International Journal for Parasitology

Volume 35, Issue 6, Pages 583-701 (May 2005)

1. Editorial board
Page co2

2. Biolistic transformation of Schistosoma mansoni with 5 flanking regions of

two peptidase genes promotes tissue-specific expression
Pages 583-589
Volker Wippersteg, Mohammed Sajid, Deirdre Walshe, Dustin Khiem, Jason P.
Salter, James H. McKerrow, Christoph G. Grevelding and Conor R. Caffrey

3. Opisthorchis viverrini antigen induces the expression of Toll-like receptor 2 in

macrophage RAW cell line
Pages 591-596
Somchai Pinlaor, Saeko Tada-Oikawa, Yusuke Hiraku, Porntip Pinlaor, Ning Ma,
Paiboon Sithithaworn and Shosuke Kawanishi

4. Insights into unique physiological features of neutral lipids in Apicomplexa:

from storage to potential mediation in parasite metabolic activities
Pages 597-615
Isabelle Coppens and Ole Vielemeyer

5. Annexin-like alpha giardins: a new cytoskeletal gene family in Giardia

lamblia
Pages 617-626
Malin E.-L. Weiland, Andrew G. McArthur, Hilary G. Morrison, Mitchell L. Sogin
and Staffan G. Svrd

6. Hsp90 is essential in the filarial nematode Brugia pahangi

Pages 627-636
Eileen Devaney, Kerry O'Neill, William Harnett, Luke Whitesell and Jane H.
Kinnaird

7. Differential effects of polyamine derivative compounds against Leishmania

infantum promastigotes and axenic amastigotes
Pages 637-646
J. Tavares, A. Ouaissi, P.K.T. Lin, A. Toms and A. Cordeiro-da-Silva
8. Patterns of host specificity and transmission among parasites of wild
primates
Pages 647-657
Amy B. Pedersen, Sonia Altizer, Mary Poss, Andrew A. Cunningham and Charles
L. Nunn

9. Redescription of Besnoitia bennetti (Protozoa: Apicomplexa) from the donkey

(Equus asinus)
Pages 659-672
J.P. Dubey, C. Sreekumar, T. Donovan, M. Rozmanec, B.M. Rosenthal, M.C.B.
Vianna, W.P. Davis and J.S. Belden

10. Clams (Corbicula fluminea) as bioindicators of fecal contamination with

Cryptosporidium and Giardia spp. in freshwater ecosystems in California
Pages 673-684
Woutrina A. Miller, Edward R. Atwill, Ian A. Gardner, Melissa A. Miller,
Heather M. Fritz, Ronald P. Hedrick, Ann C. Melli, Nicole M. Barnes and Patricia
A. Conrad

11. Developmental expression and molecular analysis of two Meloidogyne

incognita pectate lyase genes
Pages 685-692
Guozhong Huang, Ruihua Dong, Rex Allen, Eric L. Davis, Thomas J. Baum and
Richard S. Hussey

12. Echinococcus shiquicus n. sp., a taeniid cestode from Tibetan fox and plateau
pika in China
Pages 693-701
Ning Xiao, Jiamin Qiu, Minoru Nakao, Tiaoying Li, Wen Yang, Xingwang Chen,
Peter M. Schantz, Philip S. Craig and Akira Ito
 Editor-in-Chief
Nicholas Sangster, Faculty of Veterinary Science, Building B14, University of Sydney, NSW 2006, Australia
Fax: +61 (0) 2 9036 9485; E-mail: IJP@vetp.usyd.edu.au

Deputy Editor
John T. Ellis, Department of Cell and Molecular Biology,
University of Technology, Sydney, NSW 2065, Australia
Editorial Assistant
Maria Meuleman, Faculty of Veterinary Science, Building B14, University of Sydney, NSW 2006, Australia
Fax: +61 (0) 2 9036 9485; E-mail: IJP@vetp.usyd.edu.au
Specialist Editors
Peter F. BillingsleyUniversity of Aberdeen, UK James W. KazuraCase Western Reserve University,
Paul J. BrindleyTulane University Health Sciences USA
Center, USA Nirbhay KumarJohns Hopkins University, USA
Graham H. CoombsUniversity of Glasgow, UK Armand M. KurisUniversity of California, USA
John P. DaltonUniversity of Technology, Sydney, David S. LindsayVirginia Tech, USA
Australia D. Timothy J. LittlewoodThe Natural History
Jean E. FeaginSeattle Biomedical Research Institute, Museum, UK
USA Aaron G. MauleQueens University Belfast, UK
Robin B. GasserUniversity of Melbourne, Australia David A. MorrisonNational Veterinary Institute,
Timothy G. GearyPfizer Animal Health, USA Sweden
Matthias GreinerDanish Veterinary Institute, Yukifumi NawaUniversity of Miyazaki, Japan
Denmark Edoardo PozioIstituto Superiore di Sanit, Italy
Emanuela HandmanWalter and Eliza Hall Institute of R. Mark SandemanLa Trobe University, Australia
Medical Research, Australia Alan L. ScottJohns Hopkins University, USA
Wayne R. HeinAgResearch Limited, New Zealand Arne SkorpingUniversity of Bergen, Norway
Andrew HemphillUniversity of Berne, Switzerland Dominique SoldatiUniversity of Geneva, Switzerland
Nicholas H. HuntUniversity of Sydney, Australia Mary M. StevensonMcGill University Health Centre
Al IvensSanger Institute, UK Research Institute, Canada
Mark C. JenkinsAgricultural Research Service, Louis M. WeissAlbert Einstein College of Medicine,
USDA, USA USA
John T. JonesScottish Crop Research Institute,
Scotland

Copyright 2005 Australian Society for Parasitology Inc.

 International Journal for Parasitology 35 (2005) 583589
www.parasitology-online.com

Rapid communication
Biolistic transformation of Schistosoma mansoni with 5 0 flanking regions
of two peptidase genes promotes tissue-specific expression*
Volker Wipperstega, Mohammed Sajidb, Deirdre Walsheb, Dustin Khiemb, Jason P. Salterb,
James H. McKerrowb, Christoph G. Greveldinga, Conor R. Caffreyb,*
a
Insitute for Genetics, Heinrich-Heine-University, Universitatsstrae 1, D-40225 Dusseldorf, Germany
b
Sandler Center for Basic Research in Parasitic Diseases, Box 0511, University of California San Francisco, San Francisco, CA 94143, USA
Received 24 November 2004; received in revised form 10 February 2005; accepted 14 February 2005

Abstract
The gene-regulatory elements controlling peptidase expression in Schistosoma mansoni are unknown. A genomic DNA library was
constructed from which 5 0 flanking fragments of the cathepsins F (SmCF; 649 bp) and B2 (SmCB2; 810 bp) peptidase genes were isolated.
These were cloned into a GFP-expression vector for biolistic transformation of schistosomes. Both fragments promoted expression of GFP
that was localised in the gut for SmCF and tegument for SmCB2, consistent with previous immunochemical data. Promoter-deletion of the
SmCF gene indicated the importance of one or more transcription factor binding sites in the first 169 bp for both GFP-expression and its
tissue specificity.
q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: Gene promoter; GFP; Peptidase; Schistosoma; Transcription factor

Adults of the digenean trematode Schistosoma mansoni
express peptidases for their metabolic needs, including
nutrition and reproduction. Alimentary digestion of the
blood meal is facilitated by a number of gut-associated
peptidases including the Clan CA cysteine peptidases
(http://merops.sanger.ac.uk/for terminology) S. mansoni
cathepsin B1 (SmCB1, aka Sm31) and SmCF (aka
SmCL1, Caffrey et al., 2004 for a review). The potential
of gut-associated peptidases as targets for diagnosis,
vaccination and chemotherapy has spurred the molecular
and biochemical characterisation of these molecules
(Caffrey et al., 2004). In addition, cysteine peptidases
have been localised in adult tissues other than the gut, such
as SmCB2, which is found in the tegument (Caffrey et al.,
2002).
*
Note: Nucleotide sequence data reported in this paper have been
submitted to the GenBankTM, EMBL and DDBJ databases with the
accession numbers AJ786389 and AJ786390 for the 5 0 -flanking gene
regions of SmCF and SmCB2, respectively.
* Corresponding author. Tel.: C1 415 502 6866; fax: C1 415 514 3165.
E-mail address: caffrey@cgl.ucsf.edu (C.R. Caffrey).
0020-7519/$30.00 q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijpara.2005.02.002
The recent development of a reliable biolistic transform-
ation system for adult S. mansoni (Wippersteg et al., 2002b)
provides an opportunity to study more fundamental aspects
of peptidase gene expression including the spatial and
temporal regulation of expression conferred by cognate
gene promoter elements. Thus, the tissue-specific
expression of the endoplasmic recticular cysteine peptidase,
ER60, has been demonstrated by placing the requisite 5 0
regulatory region into a plasmid construct upstream of the
gene for green fluorescent protein (GFP; Wippersteg et al.,
2002a).
Using biolistic transfer, we wished to determine whether
the 5 0 flanking region of S. mansoni peptidase genes could
promote GFP expression and, if so, whether such expression
is restricted in a tissue-specific manner. To accomplish this,
we focused on the genes for SmCF and SmCB2 that are
localised in the gut (Brady et al., 1999) and tegument
(Caffrey et al., 2002), respectively. We isolated and
sequenced 5 0 flanking region fragments of both peptidase
genes and sub-cloned them into a plasmid construct
containing the GFP reporter gene (Wippersteg et al.,
2002b). Finally, we used promoter-deletion to identify
584 V. Wippersteg et al. / International Journal for Parasitology 35 (2005) 583589
the sub-region within the SmCF flanking gene sequence that
was required for GFP-expression.
A Liberian strain of S. mansoni was maintained in the
snail Biomphalaria glabrata and in Syrian golden hamsters
(Mesocricetus auratus). Adult worms were obtained by
perfusion 49 days after infection with 1500 cercariae and
washed three times in M199 medium (Invitrogen) sup-
plemented with 10% FCS, 1% HEPES (1 M, pH 7.4) and
1% ABAM (Antibiotics/Antimycotics; Sigma). The worms
were then maintained at 37 8C and 5% CO2 in the
supplemented M199 medium (Wippersteg et al., 2003).
To construct a genomic (g)DNA library, gDNA (10 mg)
from adult mixed sex S. mansoni was purified (Qiagen
DNeasy kit) and digested overnight at 37 8C with the
following restriction endonucleases (Roche); BamH I, EcoR
I, Hind III, Kpn I, Not I, Pst I, Sac I, Sal I, Spe I, Xba I, Xho I
and Xma I. Reaction volumes for gDNA digests were 100 ml
containing 10 U of each endonuclease. pBluescript II SK
(K) (3 mg; Invitrogen) was similarly digested in a final
volume of 30 ml with 3 U of each restriction enzyme.
Digested genomic and vector DNA were pooled, purified
using the QiaQuick gel extraction kit (Qiagen), and eluted in
40 ml water. The DNA pool was ligated for 8 h at room
temperature using T4 DNA ligase (10 U; Roche) and the
appropriate compensating volume of 10! ligation buffer. A
further 10 U of ligase and the appropriate volume of 10!
ligation buffer were then added and the reaction allowed to
continue overnight at room temperature. Ligation reactions
were stored at K20 8C.
Rapid amplification of genomic ends (RAGE) was
employed to amplify, clone and sequence upstream regions
of the open reading frames (ORF) of SmCB2 and SmCF.
Primary PCR reactions were carried out in a total volume of
25 ml that contained 1 ml ligated gDNA (w0.1 mg), 2.5 U
Taq DNA Polymerase (New England Biolabs) and 25 pmol
of each primer. Thermocycling conditions were as follows:
45 s at 94 8C, 1 min at 60 8C (range 5065 8C) and 3 min at
72 8C for 30 cycles. Secondary (full nested) PCR reactions
were completed in a total volume of 50 ml and contained
0.5 ml of the primary PCR product, 2.5 U Taq DNA
Polymerase and 50 pmol of each primer. Thermocycling
conditions were as above.
Sense primers targeting the vector for the primary and
secondary PCRs were 5 0 -O3 0 ACAGGAAACAGCTAT-
GACCATG and 5 0 -O3 0 GCGCAATTAACCCTCAC-
TAAAG, respectively. Primary and secondary antisense
primers targeted the 5 0 ends of the ORF of SmCB2
(Accession no. AJ312106) and SmCF (U07345), as follows:
SmCB2 primary 5 0 -O3 0 ACG TTT ATG TCG TCT AGC
ATC AA and secondary 5 0 -O3 0 CAT TTA AAG TCC CAT
AAC TCA GTA; SmCF primary 5 0 -O3 0 TCT CAT CTT
CTG TCT CAT GAT ACT and secondary 5 0 -O3 0 GTT
TTC TAT ATT TTA GCT TGA ATT GT. Secondary PCR
reactions were analysed by 1% agarose gel electrophoresis
and staining with ethidium bromide. Amplicons greater than
500 bp were excised, purified (gel extraction kit; Qiagen)
and combined with the vector pCR 2.1-TOPO according to
the manufacturers instructions (Invitrogen). Transform-
ations of chemically competent One Shot TOP10 cells
(Invitrogen) were performed according to the manufac-
turers instructions. Colonies returned on Luria Broth (LB)
agar plates under ampicillin selection were expanded in
3 ml LB-ampicillin broth. Plasmid DNA was isolated
(QiaPrep spin Mini-Prep Kit; Qiagen) and bidirectional
sequencing performed at the University of California San
Francisco Biomolecular Resource Center using the M13
reverse and M13 forward (K20) primers.
To construct vectors for transformation, two recombinant
pCR 2.1-TOPO clones, containing 810 and 649 bp upstream
regions of the SmCB2 and SmCF genes, respectively, were
used as templates for PCR. Reactions included 1 U Taq
DNA polymerase and 25 pmol of the primer pairs, CB2f and
CB2r or CFf and CFr, selective for the upstream regions of
the SmCB2 and SmCF genes, respectively. Sense and
antisense primers included restriction endonucleases sites
for EcoR I and Nco I, respectively, for ligation into the heat
shock protein (HSP) 70-GFP-HSP70 plasmid previously
described (Wippersteg et al., 2002b). Primers were as
follows: CB2f 5 0 -O3 0 GCGCGGAATTCATCTCTA-
TACCTGGCAAGTAATTGTCATGT, CB2r 5 0 -O3 0
GCGATCCATGGTTTTATTACATAATAAACAATTA-
GAATATT, CFf 5 0 -O3 0 GCGCGGAATTCAATATC-
GAAAAACACAAAGTCCACCTAT and CFr 5 0 -O3 0
GCGATCCATGGTCTCGGGATAGCAGATCGACAAC
TTTCAAG. The purified amplicons were sequentially
digested with EcoR I and Nco I. For the HSP70-GFP-
HSP70 plasmid, restriction digestion with the same
endonucleases removes the HSP70 5 0 -flanking region for
replacement by the 5 0 regions of either SmCB2 or SmCF.
Ligation reactions were transformed into One Shot TOP10
cells. Bacterial propagation, isolation and sequencing of
plasmid DNA were as described above. Large-scale plasmid
DNA was then prepared (Maxi-Prep kit; Qiagen).
For promoter deletion analysis, three progressively
truncated amplicons of SmCF, (designated SmCFA, B and
C) and incorporating 5 0 EcoR I and 3 0 Nco I restriction
enzyme sites were prepared by PCR using the SmCF-GFP-
HSP70 plasmid as template. Sense primers were: CFf(a)
(489 bp amplicon), 5 0 -O3 0 GTATGATTTCTCTCTCTG
TTTTG; CFf(b) (329 bp amplicon) 5 0 -O3 0 CAGATATGT
AGCATTTGACGCTA and CFf(c) (169 bp amplicon) 5 0 -O
3 0 GCACACTTGGGACCTTGCTTGT. The same antisense
primer, CFr (above), was used for all PCRs. After 1%
agarose gel electrophoresis and gel extraction, amplicons
were purified and ligated into the GFP-HSP70 plasmid for
eventual large-scale preparation. All reaction conditions
were as described above.
As microprojectiles, 0.6 mm gold particles were coated
with purified recombinant vector DNA (containing either
the SmCB2 or SmCF 5 0 flanking region) as described
(Wippersteg et al., 2002b). Particle bombardment, incorpor-
ating a total of 5 mg of plasmid DNA, was performed with
 V. Wippersteg et al. / International Journal for Parasitology 35 (2005) 583589 585

the PDS 1000/HE system (Bio-Rad) under 1550 psi of
helium gas pressure and a chamber vacuum of 3850 mmHg
with the target distance stage set to position one (3 cm).
Control bombardments used gold particles without DNA.
For each bombardment, approximately 25 males per 6 cm
culture dish were cultured for 1 day after perfusion.
Following gene transfer, the worms were washed once
with medium and cultured for 2448 h before microscopy.
GFP reporter gene activity was detected with a Leica TCS
NT confocal laser scanning microscope at an excitation
wavelength of 488 nm.
Total RNA from bombarded worms was isolated using
the RNA II Isolation Kit from Machery and Nagel (Duren,
Germany). The amount of RNA recovered was determined
spectrometrically. Reverse transcriptase PCR (RT-PCR)
was done stepwise in separate reactions. For reverse
transcriptase, about 500 ng S. mansoni total RNA was
denatured for 5 min at 70 8C and transcribed into cDNA for
60 min at 42 8C using 200 U M-MuLV reverse transcriptase
(New England Biolabs), 40 U RNAsin (Promega), 10%
glycerol and 1 mM of a GFP-selective primer (GFP-3:
ATCCTTATTTGTATAGTTCATCC). PCR amplifications,

Fig. 1. The 5 0 flanking region of the SmCF gene promotes expression of GFP that is restricted to the parasite gut. Adult male Schistosoma mansoni were
examined by confocal laser scanning microscopy 2448 h after biolistic transformation with a GFP-HSP70 plasmid construct containing a 649 bp 5 0 flanking
fragment of the SmCF gene. Fluorescence and bright field images are presented for the anterior (panel A) and posterior (panel B) regions of the gut.
Magnifications are indicated and includes a higher magnification (630!; within the red box) of the anterior bifurcated gut and its vesicular contents.
586 V. Wippersteg et al. / International Journal for Parasitology 35 (2005) 583589
Fig. 2. The 5 0 flanking region of the SmCB2 gene promotes expression of GFP that is restricted to the parasite tegument. Adult male Schistosoma mansoni were
examined by con-focal laser scanning microscopy 2448 h after biolistic transformation with a GFP-HSP70 plasmid construct containing a 810 bp 5 0 flanking
fragment of the SmCB2 gene. Fluorescence (A) and bright field (B) images are presented in which the lateral surface tegument and tubercles are clearly visible.
The magnification (200!) is indicated.
to yield a 420 bp amplicon, were performed with the primer
combination GFP-3 and GFP-5 (GGGAACTACAAGA-
CACGTGC) in a total volume of 25 ml as described
elsewhere in detail (Wippersteg et al., 2002b). Negative
control reactions were performed without reverse transcrip-
tase for each worm population or without template. As a
positive control, 1 ng of GFP-plasmid DNA was used as
template.
After biolistic transformation and a 2448 h incubation
period in medium, worms exhibited a wide distribution of
gold particles by bright field microscopy. However, upon
viewing at 488 nm, a distinct GFP-derived fluorescence was
localised in the gut after transformation with the 5 0
regulatory region of SmCF (Fig. 1). Fluorescence was
seen throughout the length of the gut including the anterior
(Fig. 1A) and posterior (Fig. 1B) regions. By contrast,
transformation with the SmCB2 construct yielded fluor-
escence that was solely localised in the tegument, including
the tegumental tubercles (Fig. 2). Control transformations
using gold particles alone yielded no fluorescence (not
shown) in agreement with previously published data
(Wippersteg et al., 2002b).
To examine which sub-region(s) of the cloned SmCF 5 0
gene flanking sequence (649 bp) is essential for GFP
expression, three promoter deletion constructs, SmCFA,
SmCFB and SmCFC, were prepared. These and the full-
length construct were used for biolistic experiments. After a
2448 h incubation period, GFP was only detected in
schistosomes transformed with the full length construct
(data not shown). To confirm this result at the molecular
level, total RNA was extracted from schistosomes bom-
barded with each of the promoter deletion constructs and
used for comparative RT-PCR analyses employing GFP-
specific primers. As shown in Fig. 3, the expected amplicon
of 420 bp was only observed for worms transformed with
the full-length construct. These findings indicate the
absolute requirement of the first 169 bp of the cloned
SmCF gene 5 0 regulatory region for both transcription of
GFP and its tissue-specificity (see Fig. 4).
Putative transcription factor binding sites (TFBS) were
identified using the MatInspector software (Quandt et al.,
1995; http://www.genomatix.de/index.html) setting the core
and matrix similarities to 1.0 and 0.8, respectively. A
relatively simple profile of TFBS for SmCF was identified
with the majority of sites being distributed within that 5 0
169 bp region required for expression of GFP (Fig. 4). These
sites included Oct (K625 with respect to the start ATG),
GATA (K585), PPAR/RXR (K548), NF-AT(K543),
STAT (K517) and AP1 (K511). Further, downstream,
two NF-Y elements (aka CCAAT-boxes) at K459 and
K353 were identified, as were GATA (K330) and TATA-
box elements (K243).
Fig. 3. Deletion of the 5 0 flanking region of the SmCF gene leads to loss of
the GFP reporter transcript. RNA was extracted from male Schistosoma
mansoni worms transfected with various deletion constructs of the SmCF 5 0
flanking gene region. Reverse transcriptase PCR was performed with equal
amounts of total RNA using primers selective for GFP that amplify a single
product of 420 bp. (M) molecular ladder, (1) full-length SmCF, (2) deletion
A, (3) deletion B, (4) deletion C, (5) worms transfected with gold particles
without DNA, (6) PCR without template, (7) PCR positive control with
plasmid as template. Only in lanes 1 and 7 is the expected 420 bp amplicon
visible. Other control reactions without the reverse transcriptase step
yielded no amplification product (data not shown). Identical results were
obtained in a second independent experiment.
 V. Wippersteg et al. / International Journal for Parasitology 35 (2005) 583589 587

Fig. 4. Nucleotide sequence and putative transcription factor binding sites in the 5 0 flanking region of the SmCF gene. Transcription factor binding sites (shaded
areas) were identified using the MatInspector software (http://www.genomatix.de/index.html). Core regions of transcription factor binding sites are underlined.
Negative numbers are used to mark their position upstream of the translation initiation codon ATG (marked in bold type face). The vertical lines with arrows
termed A, B and C indicate the three deletions of the SmCF 5 0 flanking region used to biolistically transform Schistosoma mansoni.

For SmCB2, MatInspector analysis only identified TFBS
that have a more general function in transcription such as
Yin-Yang, NF-Y and AP4 (Mantovani, 1999; Usheva and
Shenk, 1996; Aranburu et al., 2001; data not shown).
Therefore, promoter deletion analysis was not considered
further at this stage.
In spite of the considerable biochemical and functional
characterisation of schistosome peptidases (Caffrey et al.,
2004), the gene regulatory elements determining their
spatial and temporal expression are unknown. However,
the recent availability of a biolistic transformation system
for adult S. mansoni, incorporating plasmid constructs
with gene-specific flanking elements upstream of GFP
(Wippersteg et al., 2002a,b, 2003), provides one means by
which gene-regulation can be studied. Using biolistics,
therefore, our data demonstrate that the 5 0 genomic
fragments for two S. mansoni peptidase genes are sufficient
to promote expression of GFP. Moreover, expression is
tissue-specific, in a manner consistent with the previous
immunochemical localisations of SmCF and SmCB2 to the
gut (Brady et al., 1999) and tegument (Caffrey et al., 2002),
respectively. For promoter-deletions, we focused on SmCF
because of the higher number of candidate TFBS identified
that were considered necessary for tissue-specificity (dis-
cussed below). All three SmCF deletions failed to promote
expression of GFP as visualised by microscopy. Further,
these observations were confirmed by RT-PCR selective for
GFP. The results, therefore, demonstrate the necessity of
one or more TFBS within the first 169 bp of the cloned
SmCF gene 5 0 flanking region for directing both gene
expression and its restriction to the gut.
Prior to biolistic transformation and in the absence of
extensive archived genomic sequence information, it had
been necessary to characterise the requisite 5 0 flanking
sequences. This was accomplished by constructing a gDNA
library and performing RAGE using nested PCRs with
vector-specific forward and gene-ORF-specific reverse
primers. Towards the end of this study, however, consider-
ably more genomic sequence information had been deposited
in the relevant databases. We, therefore, compared our clones
against the genomic database available at The Wellcome
Trust Sanger Institute S. mansoni Blast server (http://www.
sanger.ac.uk/cgi-bin/blast/submitblast/s_mansoni). At the
time of writing, for SmCF, the latest contig with a greater
than 99% identity was 333748.c005310846.Contig1 cover-
ing all 649 bp of our cloned 5 0 UTR together with at least
12 bp of the SmCF ORF. For SmCB2, the respective contig
was 333748.c003608915.Contig1 covering 506 bp of the 5 0
end of our cloned UTR and also the clone SMECS44TR
covering 245 bp plus 144 bp of the SmCB2 ORF. Thus, our
588 V. Wippersteg et al. / International Journal for Parasitology 35 (2005) 583589
clones are consistent with the archived sequence
information.
The MatInspector software allowed for both the
identification of putative TFBS in the SmCF 5 0 upstream
fragment and some interpretation of the results obtained
with the promoter deletion experiments. Interestingly, most
of the putative TFBS (six of 10 sites) were located within
the critical 5 0 flanking 169 bp region (see Fig. 4). K625 to
K227 upstream of the ATG start codon.
Of particular interest is the identification of a palin-
dromic STAT motif (TT(N)5AA) at position K517. STAT
elements are also present in the cathepsin L gene promoters
of Fasciola gigantica (Grams et al., 2001; accession no.
AB010923) and Onchocerca volvulus (Hashmi et al., 2002;
AF442768) at similar positions upstream of the ATG start
codon (K543 and K668, respectively). This suggests that
STATs may function in the spatial regulation of cathepsin
L-like genes because the Fasciola ortholog is also expressed
in the gut (Grams et al., 2001) and a Caenorhabditis elegans
line carrying a heterologous O. volvulus cathepsin L gene
promoter linked to a lacZ reporter gene revealed an
expression pattern that was restricted to gut cells (Hashmi
et al., 2002). In mammals, STAT proteins are a family of
cytoplasmic transcription factors that are involved in signal
transduction cascades induced by cytokines, hormones and
growth factors (Decker, 1999). For example, upon stimu-
lation of human cells by interferon-gamma, binding of
STAT-1 to the promoter of the cysteine peptidase gene
caspase-11 is required for its expression (Schauvliege et al.,
2002).
Closely associated with the STAT element are putative
NF-AT (K543) and AP-1 (K511) binding sites. These
often occur in tandem to cooperatively bind transcription
factors involved in tissue-specific gene regulation (Rao et
al., 1997). Further, composite binding sites that would allow
the cooperative binding and dimerisation of PPARs
(peroxisome proliferator-activated receptors) and RXR (9-
cis retinoic acid receptor) were detected at K548. This
heterodimer would then bind PPREs (PP response elements)
to control tissue-specific and developmentally regulate
transcription (Qi et al., 2000).
Finally, within the first 169 bp region of the SmCF 5 0
flanking fragment, are TFBS for Oct-1/2 (K625) and
GATA-binding proteins (K585). Both are involved in the
tissue-specific expression of a number of genes in
organisms such as insects (Harshman and James, 1998)
and C. elegans. In C. elegans, the importance of GATA
elements in gut development has been demonstrated
(Maduro and Rothman, 2002). Further, the number and
placement of GATA elements are crucial in the gut-
specific regulation of a C. elegans cathepsin B gene
(Britton et al., 1998).
The present biolistic and promoter-deletion data provide
a basis for further enquiry of those gene regulatory elements
directing the temporal and spatial expression of peptidases
in schistosomes. Such studies will be greatly aided by the
future availability of the entire S. mansoni genome
sequence. An improved understanding of the function and
regulation of peptidase genes could contribute to the design
of specific intervention strategies.
Acknowledgements
This work was financially supported by the Sandler
Family Supporting Foundation, NIAID grant AI053247 and
the Deutsche Forschungsgemeinschaft (Grant-No.: GR
1549/1-3). We thank Dr Ben Kelly of the Sandler Center
for his critical reading of the manuscript.
References
Aranburu, A., Carlsson, R., Persson, C., Leanderson, T., 2001. Transcrip-
tion factor AP-4 is a ligand for immunoglobulin-kappa promoter E-box
elements. Biochem. J. 354, 431438.
Brady, C.P., Dowd, A.J., Brindley, P.J., Ryan, T., Day, S.R., Dalton, J.P.,
1999. Recombinant expression and localization of Schistosoma
mansoni cathepsin L1 support its role in the degradation of host
hemoglobin. Infect. Immun. 67, 368374.
Britton, C., McKerrow, J.H., Johnstone, I.L., 1998. Regulation of the
Caenorhabditis elegans gut cysteine protease gene cpr-1: requirement
for GATA motifs. J. Mol. Biol. 283, 1527.
Caffrey, C.R., Salter, J.P., Lucas, K.D., Khiem, D., Hsieh, I., Lim, K.C.,
Ruppel, A., McKerrow, J.H., Sajid, M., 2002. SmCB2, a novel
tegumental cathepsin B from adult Schistosoma mansoni. Mol.
Biochem. Parasitol. 121, 4961.
Caffrey, C.R., McKerrow, J.H., Salter, J.P., Sajid, M., 2004. Blood n guts:
an update on schistosome digestive peptidases. Trends Parasitol. 20,
241248.
Decker, T., 1999. STATs as essential intracellular mediators of cytokine
response. Cell. Mol. Life Sci. 55, 15051508.
Grams, R., Vichasri-Grams, S., Sobhon, P., Upatham, E.S., Viyanant, V.,
2001. Molecular cloning and characterization of cathepsin L encoding
genes from Fasciola gigantica. Parasitol. Int. 50, 105114.
Harshman, L.G., James, A.A., 1998. Differential gene expression in insects:
transcriptional control. Annu. Rev. Entomol. 43, 671700.
Hashmi, S., Britton, C., Liu, J., Guiliano, D.B., Oksov, Y., Lustigman, S.,
2002. Cathepsin L is essential for embryogenesis and development of
Caenorhabditis elegans. J. Biol. Chem. 277, 34773486.
Maduro, M.F., Rothman, J.H., 2002. Making worm guts: the gene
regulatory network of the Caenorhabditis elegans endoderm. Dev.
Biol. 246, 6885.
Mantovani, R., 1999. The molecular biology of the CCAAT-binding factor
NF-Y. Gene 239, 1527.
Qi, C., Zhu, Y., Reddy, J.K., 2000. Peroxisome proliferator-activated
receptors, coactivators, and downstream targets. Cell. Biochem.
Biophys. 32, 187204.
Quandt, K., Frech, K., Karas, H., Wingender, E., Werner, T., 1995. MatInd
and MatInspector: new fast and versatile tools for detection of
consensus matches in nucleotide sequence data. Nucleic Acids Res.
23, 48784884.
Rao, A., Luo, C., Hogan, P.G., 1997. Transcription factors of the
NF-AT family: regulation and function. Annu. Rev. Immunol. 15,
707747.
Schauvliege, R., Vanrobaeys, J., Schotte, P., Beyaert, R., 2002.
Caspase-11 gene expression in response to lipopolysaccharide and
 V. Wippersteg et al. / International Journal for Parasitology 35 (2005) 583589
interferon-gamma requires nuclear factor-kappa B and signal
transducer and activator of transcription (STAT) 1. J. Biol. Chem.
277, 4162441630.
Usheva, A., Shenk, T., 1996. YY1 transcriptional initiator: protein
interactions and association with a DNA site containing unpaired
strands. Proc. Natl Acad. Sci. USA 93, 1357113576.
Wippersteg, V., Kapp, K., Kunz, W., Grevelding, C.G., 2002a. Character-
isation of the cysteine protease ER60 in transgenic Schistosoma
mansoni larvae. Int. J. Parasitol. 32, 12191224.
589
Wippersteg, V., Kapp, K., Kunz, W., Jackstadt, P., Zahner, H.,
Grevelding, C.G., 2002b. HSP70-controlled GFP expression in
transiently transformed schistosomes. Mol. Biochem. Parasitol. 120,
141150.
Wippersteg, V., Ribeiro, F., Liedtke, S., Kusel, J.R., Grevelding, C.G.,
2003. The uptake of Texas Red-BSA in the excretory system of
schistosomes and its colocalisation with ER60 promoter-induced
GFP in transiently transformed adult males. Int. J. Parasitol. 33,
11391143.
 International Journal for Parasitology 35 (2005) 591596
www.parasitology-online.com

Rapid Communication
Opisthorchis viverrini antigen induces the expression of Toll-like
receptor 2 in macrophage RAW cell line
Somchai Pinlaora,e, Saeko Tada-Oikawab, Yusuke Hirakub, Porntip Pinlaorc,e,
Ning Mad, Paiboon Sithithaworna,e, Shosuke Kawanishib,*
a
Department of Parasitology, Khon Kaen University, Khon Kaen 40002, Thailand
b
Department of Environmental and Molecular Medicine, Mie University School of Medicine, 2-174, Edobashi, Tsu, Mie 514-8507, Japan
c
Department of Clinical Microbiology, Faculty of Associated Medical Science, Khon Kaen University, Khon Kaen 40002, Thailand
d
Department of Anatomy, Mie University School of Medicine, Mie 514-8507, Japan
e
Liver Fluke and Cholangiocarcinoma Research Center, Faculty of Medicine, Khon Kaen University, Khon Kaen 40002, Thailand
Received 28 December 2004; received in revised form 17 February 2005; accepted 18 February 2005

Abstract

Opisthorchis viverrini infection induces inflammation in and around the bile duct, leading to cholangiocarcinoma in humans. To examine
the mechanism of O. viverrini-induced inflammatory response, we assessed the expression of Toll-like receptors (TLRs) in RAW 264.7
macrophage cell line treated with an extract of O. viverrini antigen. Flow cytometry and immunocytochemistry showed that O. viverrini
antigen induced the expression of TLR2 but not TLR4. Western blotting and immunocytochemistry revealed that nuclear factor-kB (NF-kB),
inducible nitric oxide synthase (iNOS) and cyclooxygenase-2 (COX-2) were expressed in RAW 264.7 cells treated with O. viverrini antigen
in a dose-dependent manner. These results suggest that O. viverrini induces inflammatory response through TLR2-mediated pathway leading
to NF-kB-mediated expression of iNOS and COX-2.
q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: Opisthorchis viverrini; Toll-like receptor 2; Inducible nitric oxide synthase; Nuclear factor-kB; Cyclooxygenase-2; RAW 264.7 macrophage
cell line

Opisthorchis viverrini infection is a major risk factor of
cholangiocarcinoma (CCA) development in humans (IARC,
1994). Nitric oxide (NO) produced by inducible nitric oxide
synthase (iNOS) has been implicated in inflammation-
mediated carcinogenesis (Hussain et al., 2003) including
O. viverrini-caused CCA (Haswell-Elkins et al., 1994).
Recently, we have demonstrated that chronic inflammation
triggered by repeated O. viverrini infection mediates iNOS-
dependent DNA damage in intrahepatic bile duct epithelium
and inflammatory cells of hamsters, which may play an
important role in CCA development (Pinlaor et al., 2003,
2004a). However, the underlining mechanism of
O. viverrini-induced inflammatory response remains to be
clarified.
* Corresponding author. Tel./fax: C81 59 231 5011.
E-mail address: kawanisi@doc.medic.mie-u.ac.jp (S. Kawanishi).
0020-7519/$30.00 q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijpara.2005.02.003
Sripa and Kaewkes (2000) observed O. viverrini antigen
in bile ducts of the liver, liver cells, Kupffer cells,
macrophages, epithelioid and giant cells in the egg
granuloma. The presence of the antigens was associated
with heavy inflammatory cell infiltration, particularly with
mononuclear cells. Importantly, Akai et al. demonstrated
that antibody level against O. viverrini antigen was
positively associated with the severity of hepatobiliary
disease and CCA (Akai et al., 1994). Thus, local parasite-
specific immune responses induced by O. viverrini antigens
would play a role in the pathogenesis of opisthorchiasis
associated with CCA development.
Toll-like receptors (TLR) are an important membrane
receptor family, which actively participates in the stimu-
lation of the innate immune response. To date, 10 TLR
homologs have been found in humans. It has been reported
that TLR2 and TLR4 contribute to immune response to
parasitic infection including protozoa (Kropf et al., 2004)
592 S. Pinlaor et al. / International Journal for Parasitology 35 (2005) 591596

and helminth (van der Kleij et al., 2002). TLRs activate
homologous signal transduction pathways leading to
nuclear localization of NF-kB/Rel-type transcription
factors. NF-kB is a key player in inflammation that regulates
expression of various genes involved in controlling
inflammatory response such as proinflammatory mediator,
iNOS and cyclooxygenese-2 (COX-2) expression (Balkwill
and Coussens, 2004; Surh et al., 2001). In addition, COX-2
is involved in carcinogenesis through the induction of
inflammatory process (Surh et al., 2001). The expression of
COX-2 is increased in tumor tissue of CCA patients (Endo
et al., 2002). Relevantly, NF-kB functions as a tumor
promoter in inflammation-associated cancer (Pikarsky et al.,
2004).
In the present study, we investigated expression of TLRs,
NF-kB, iNOS and COX-2 in RAW 264.7 macrophage cell
line treated with O. viverrini antigen using flow cytometry
and immunofluorescence techniques. To confirm the
expression of these molecules, we also performed Western
blotting.
O. viverrini crude antigen was prepared as described
previously (Pinlaor et al., 2004b) with a minor modification.
Four months after hamsters were infected with 50
O. viverrini, adult worms were collected. Following the
addition of protease inhibitors (0.1 mM phenylmethyl
sulphonyl fluoride, 0.1 mM L-1-tosylamine-2-phenylethyl
chloromethyl ketone and 0.1 mM trans-epoxysuccinyl-
L -leucylamido-(4-guanidino)butane (E64)), the worms
were homogenized with 10 passes of a microhomogenizer
with a Teflon-coated pestle. The homogenate was frozen in
liquid nitrogen and then thawed at 37 8C. The homogenate
was centrifuged at 12,000 g at 4 8C for 30 min. The
supernatant containing crude somatic antigen was stored
at K80 8C until use. The protein concentration in the
supernatant was determined using the Coomassiew Protein
Assay Reagent Kit (Pierce Biotechnology, Rockford, USA).
RAW 264.7 mouse macrophage cell line was cultured
in high glucose Dulbeccos modified Eagles medium
(Gibco/BRL, New York, NY, USA) containing 10% heat-
inactivated fetal bovine serum and 100 mg/l kanamycin.
The cells (1!106 cells) were detached from culture dishes
by vigorous pipetting, and were centrifuged and resus-
pended in 1 ml fresh medium. Cells (1!106 cells/ml) were
incubated with crude O. viverrini antigen at 37 8C for
indicated durations.
For analysis of TLR expression, O. viverrini antigen-
treated RAW 264.7 cells were treated with 1 mg/ml of rabbit
polyclonal anti-TLR2 antibody or mouse monoclonal anti-
TLR4 antibody (Santa Cruz Biotechnology, Santa Cruz,
CA, USA) and then incubated with 1 mg/ml of Alexa

Fig. 1. TLR expression induced by Opisthorchis viverrini antigen. (A) Immunocytochemical analysis of TLR2 expression in RAW 264.7 cells treated with
O. viverrini antigen. RAW 264.7 cells were treated with O. viverrini antigen for 12 h at 37 8C, and then treated with anti-TLR2 antibody overnight at room
temperature. The cells were incubated with Alexa 594 anti-rabbit IgG antibody for 3 h and then viewed under a fluorescent microscope. Scale barZ10 mm. (B)
Flow cytometric fluorescence distribution of RAW 264.7 cells treated with O. viverrini antigen. RAW 264.7 cells were treated with 100 mg/ml of O. viverrini
antigen for 12 h at 37 8C, and then incubated with anti-TLR2 or anti-TLR4 antibody for 30 min. The cells were then incubated with Alexa 488 anti-rabbit IgG
or anti-mouse IgG antibody for 30 min and then analyzed with a flow cytometer. Open peaks, nontreated control cells; shaded peaks, O. viverrini antigen-
treated cells.
 S. Pinlaor et al. / International Journal for Parasitology 35 (2005) 591596 593

488-labeled goat antibody against mouse IgG or against
rabbit IgG (Molecular Probes, Eugene, OR, USA). The cells
were resuspended in PBS, and then analyzed on a flow
cytometer (FACScan; Becton Dickinson).
Localization of NF-kB and iNOS expression was
assessed as described previously (Pinlaor et al., 2004a)
with a minor modification. RAW 264.7 cells treated with
O. viverrini antigen were smeared on a glass slide. After
drying at 60 8C for 2 h, cells were fixed with 4%
formaldehyde in PBS for 20 min at room temperature.
The cells were treated with 0.1% Triton XK1 100 for 15 min
and then incubated with 1% skim milk for 30 min.
For simultaneous analysis of iNOS and NF-kB
expression, the cells were incubated with the primary
antibodies, rabbit polyclonal anti-iNOS antibody (1:300,
Calbiochem-Novabiochem Corporation, San Diego, USA)
and mouse monoclonal anti-NF-kB antibody (1:300, Santa
Cruz Biotechnology, Inc, USA) overnight at room
temperature. Then, the cells were incubated with Alexa
594-labeled goat antibody against rabbit IgG and Alexa
488-labeled goat antibody against mouse IgG (1:400,
Molecular Probes) for 3 h. The cells were examined under
an inverted Laser Scan Microscope (LSM 410, Zeiss,
Gottingen, Germany).
For analysis of the expression of TLR2, TLR4 and
COX-2, the cells were incubated with the primary
antibody [rabbit polyclonal anti-TLR2 antibody, mouse
monoclonal anti-TLR4 antibody (1 mg/ml, Santa Cruz
Biotechnology) or rabbit polyclonal anti-COX-2 antibody
(1:400, Oxford Biomedical Research, Oxford, Michigan,
USA)] overnight at room temperature. Subsequently, the
cells were incubated with Alexa 594-labeled goat
antibody against rabbit IgG or Alexa 488-labeled goat
antibody against mouse IgG (1:400, Molecular Probes).
The cells were examined under an inverted Laser Scan
Microscope as described above.

Fig. 2. Expression of NF-kB and iNOS induced by Opisthorchis viverrini antigen. (A) Immunocytochemical analysis of iNOS and NF-kB expression in RAW
264.7 cells treated with O. viverrini antigen. RAW 264.7 cells were treated with indicated concentration of O. viverrini antigen for 12 h, and then incubated
with anti-iNOS and anti-NF-kB antibodies followed by Alexa 594 anti-rabbit IgG or Alexa 488 anti-mouse IgG antibodies for 3 h and then viewed under an
inverted Laser Scan Microscope. iNOS expression (red color) is observed in the cytoplasm, whereas NF-kB (green color) is expressed in both nucleus and
cytoplasm. Scale barZ10 mm. (B, C) Western blotting analysis of NF-kB and iNOS expression in RAW 264.7 cells treated with O. viverrini antigen. Proteins
extracted from O. viverrini antigen-treated RAW 264.7 cells were immunoblotted and probed with mouse monoclonal anti-NF-kB antibody (B) or rabbit
polyclonal anti-iNOS antibody (C). (For interpretation of the reference to colour in this legend, the reader is referred to the web version of this article.)
594 S. Pinlaor et al. / International Journal for Parasitology 35 (2005) 591596

RAW 264.7 cells treated with crude O. viverrini antigen
were solubilized in sample buffer and boiled for 5 min.
Samples were separated by 420% SDS-PAGE and blotted
onto polyvinylidene difluoride (PVDF) membranes. The
membranes were incubated with 5% skim milk in Tris-
buffered saline containing 0.1% Tween 20 (TBS, pH 7.5).
The membranes were then incubated with the primary
antibody [mouse monoclonal anti-NF-kB antibody
(1:1000), rabbit polyclonal anti-iNOS antibody (1:2000)
or rabbit polyclonal anti-COX-2 antibody (1:1000)]. The
membranes were washed in TBS and then incubated with
horseradish peroxidase (HRP)-conjugated anti-mouse IgG
antibody (1:2000, Santa Cruz Biotechnology) or anti-rabbit
IgG antibody (1:2000, New England Biolabs, Beverly, MA,
USA). After extensive washings in TBS, enhanced chemi-
luminescence assay was performed using an ECL detection
reagent (Amersham) and positive bands were detected on
X-ray films.
Fig. 1 shows the expression of TLRs in RAW 264.7 cells
treated with O. viverrini antigen. Immunocytochemistry
revealed that O. viverrini antigen increased the expression
of TLR2 dose-dependently (Fig. 1A), but not TLR4 (data
not shown). TLR2 was not expressed in nontreated control
cells (Fig. 1A). Flow cytometry confirmed that O. viverrini
antigen increased the fluorescence intensity of TLR2, but
not TLR4 (Fig. 1B).
Fig. 2 shows the expression of NF-kB and iNOS in RAW
264.7 cells treated with O. viverrini antigen. iNOS
expression (red color) was observed in the cytoplasm,
whereas NF-kB expression (green color) was observed in
both nucleus and cytoplasm (Fig. 2A). O. viverrini antigen
increased expression of NF-kB and iNOS dose-dependently.
Western blotting showed that the expression of NF-kB
(approximately 65 kDa, Fig. 2B) and iNOS (approximately
130 kDa, Fig. 2C) were increased dose-dependently in
RAW 264.7 cells treated with O. viverrini antigen.
Fig. 3 shows COX-2 expression in RAW 264.7 cells
treated with O. viverrini antigen. The immunoreactivity of
COX-2 increased in the cytoplasm depending on the
concentration of O. viverrini antigen (Fig. 3A). COX-2
expression was not observed in nontreated control cells
(Fig. 3A). Western blotting showed that COX-2 protein
(approximately 70 kDa) was increased dose-dependently in
RAW 264.7 cells treated with O. viverrini antigen (Fig. 3B).
This is the first study showing that O. viverrini antigen
induces the expression of TLR2, which may participate in NF-
kB-mediated expression of iNOS and COX-2 in RAW 264.7
cells. The expression of these proteins was dependent on the
concentration of O. viverrini antigen. TLR family recognizes
conserved organism structures and activates signaling
pathways (Barton and Medzhitov, 2003). Relevantly, it was
reported that in schistosome infection, purified schistosomal
antigen activates TLR2 in dendritic cells (van der Kleij et al.,
2002). O. viverrini may mediate TLR2-dependent
pathway through binding its lipoproteins as demonstrated
using bacterial lipoproteins (Aliprantis et al., 1999).
Our data can reasonably explain the clinical data showing
that parasite-specific antibody levels in human hepatobili-
ary disease were associated with O. viverrini infection
(Haswell-Elkins et al., 1991). On the other hand, TLR4

Fig. 3. COX-2 expression induced by Opisthorchis viverrini antigen. (A) Immunocytochemical analysis of COX-2 expression in RAW 264.7 cells. RAW 264.7
cells were treated with indicated concentration of O. viverrini antigen for 12 h, and then incubated with anti-COX-2 antibody followed by Alexa 594 anti-rabbit
IgG antibody and then viewed under an inverted Laser Scan Microscope. COX-2 expression was observed in the cytoplasm dose-dependently, but not in
untreated cells. Scale barZ10 mm. (B) Western blotting analysis of COX-2 expression in RAW 264.7 cells treated with O. viverrini antigen. Proteins extracted
from RAW 264.7 cells treated with O. viverrini antigen were immunoblotted and probed with anti-COX-2 antibody.
 S. Pinlaor et al. / International Journal for Parasitology 35 (2005) 591596
contributes to efficient control of infection with protozoa
such as Leishmania major (Kropf et al., 2004), and
lipopolysaccharide, a major component of the outer
surface of Gram-negative bacteria (Andonegui et al.,
2003). In the present study, TLR4 expression was not
increased in O. viverrini antigen-treated cultured cells. It
has been reported that constitutively active TLR4 induces
NF-kB activation and subsequent production of inflamma-
tory cytokines (Medzhitov et al., 1997). Thus, TLR4 may
also participate in pathogenesis of O. viverrini infection.
Our results showed that NF-kB was expressed in both
the nucleus and cytoplasm, suggesting that O. viverrini
antigen induces not only NF-kB expression, but also its
nuclear translocation and activation. NF-kB regulates the
expression of iNOS and COX-2 (Surh et al., 2001;
Balkwill and Coussens, 2004). We have previously
reported that O. viverrini infection strongly induces
iNOS in bile duct epithelial cells and inflammatory
cells (Pinlaor et al., 2004a). This finding can be
explained by assuming that O. viverrini antigen mediates
NF-kB-mediated iNOS expression depending on TLR,
since it has been reported that O. viverrini antigen is
present in both epithelial bile duct and inflammatory cells
(Sripa and Kaewkes, 2000). Relevantly, Harada et al.
have reported that NF-kB is expressed through TLR and
related molecules in cultured biliary epithelial cells
treated with bacterial LPS (Harada et al., 2003). COX-
2 expression induces inflammatory cell infiltration (Surh
et al., 2001) and participates in carcinogenesis (Oshima
et al., 2004). Moreover, COX-2 expression is increased
in tumor tissue of cholangiocarcinoma patients (Endo
et al., 2002). Thus, COX-2 may participate in the
pathogenesis of O. viverrini infection through an
additional inflammatory response.
We have recently demonstrated that O. viverrini infection
induces inflammation-mediated DNA damage through NO
production via iNOS expression (Pinlaor et al., 2004a). DNA
damage was observed in bile duct epithelial cells and
inflammatory cells in the liver of hamsters with repeated O.
viverrini infection and was associated with the progression of
pathological changes (Pinlaor et al., 2004a). In the present
study, in vitro experiments have shown that O. viverrini
antigen induces the expression of TLR2, NF-kB, iNOS and
COX-2 in macrophage cell line. It is reasonably considered
that TLR2 expression mediates NF-kB expression and its
nuclear translocation, leading to expression of iNOS and
COX-2. In conclusion, the pathogenesis of O. viverrini
infection, including carcinogenesis, may be mediated by
inflammatory response triggered by its antigen via TLR2-
dependent pathway.
Acknowledgements
This work was supported by Khon Kaen University,
Research Fund and Faculty of Medicine, Khon Kaen
595
University, Thailand, and Grants-in-Aid for Scientific
Research from the Ministry of Education, Science, Sports
and Culture of Japan.
References
Akai, P.S., Pungpak, S., Chaicumpa, W., Viroj, K., Bunnag, D.,
Befus, A.D., 1994. Serum antibody response to Opisthorchis viverrini
antigen as a marker for opisthorchiasis-associated cholangiocarcinoma.
Trans. R. Soc. Trop. Med. Hyg. 88, 471474.
Aliprantis, A.O., Yang, R.B., Mark, M.R., Suggett, S., Devaux, B.,
Radolf, J.D., Klimpel, G.R., Godowski, P., Zychlinsky, A., 1999. Cell
activation and apoptosis by bacterial lipoproteins through toll-like
receptor-2. Science 285, 736739.
Andonegui, G., Bonder, C.S., Green, F., Mullaly, S.C., Zbytnuik, L.,
Raharjo, E., Kubes, P., 2003. Endothelium-derived Toll-like receptor-4
is the key molecule in LPS-induced neutrophil sequestration into lungs.
J. Clin. Invest. 111, 10111020.
Balkwill, F., Coussens, L.M., 2004. Cancer: an inflammatory link. Nature
431, 405406.
Barton, G.M., Medzhitov, R., 2003. Toll-like receptor signaling pathways.
Science 300, 15241525.
Endo, K., Yoon, B.I., Pairojkul, C., Demetris, A.J., Sirica, A.E., 2002.
ERBB-2 overexpression and cyclooxygenase-2 up-regulation in human
cholangiocarcinoma and risk conditions. Hepatology 36, 439450.
Harada, K., Ohira, S., Isse, K., Ozaki, S., Zen, Y., Sato, Y., Nakanuma, Y.,
2003. Lipopolysaccharide activates nuclear factor-kB through toll-like
receptors and related molecules in cultured biliary epithelial cells. Lab.
Invest. 83, 16571667.
Haswell-Elkins, M.R., Sithithaworn, P., Mairiang, E., Elkins, D.B.,
Wongratanacheewin, S., Kaewkes, S., Mairiang, P., 1991. Immune
responsiveness and parasite-specific antibody levels in human hepato-
biliary disease associated with Opisthorchis viverrini infection. Clin.
Exp. Immunol. 84, 213218.
Haswell-Elkins, M.R., Satarug, S., Tsuda, M., Mairiang, E., Esumi, H.,
Sithithaworn, P., Mairiang, P., Saitoh, M., Yongvanit, P., Elkins, D.B.,
1994. Liver fluke infection and cholangiocarcinoma: model of
endogenous nitric oxide and extragastric nitrosation in human
carcinogenesis. Mutat. Res. 305, 241252.
Hussain, S.P., Hofseth, L.J., Harris, C.C., 2003. Radical causes of cancer.
Nat. Rev. Cancer 3, 276285.
IARC Working Group, 1994. Infection with liver flukes (Opisthorchis
viverrini, Opisthorchis felineus and Clonorchis sinensis), In: IARC
Monographs on the Evaluation of Carcinogenic Risks to Humans, vol.
61. IARC Lyon pp. 121175.
Kropf, P., Freudenberg, M.A., Modolell, M., Price, H.P., Herath, S.,
Antoniazi, S., Galanos, C., Smith, D.F., Muller, I., 2004. Toll-like
receptor 4 contributes to efficient control of infection with the protozoan
parasite Leishmania major. Infect. Immun. 72, 19201928.
Medzhitov, R., Preston-Hurlburt, P., Janeway Jr., C.A., 1997. A human
homologue of the Drosophila Toll protein signals activation of adaptive
immunity. Nature 388, 394397.
Oshima, H., Oshima, M., Inaba, K., Taketo, M.M., 2004. Hyperplastic
gastric tumors induced by activated macrophages in COX-2/mPGES-1
transgenic mice. EMBO J. 23, 16691678.
Pikarsky, E., Porat, R.M., Stein, I., Abramovitch, R., Amit, S., Kasem, S.,
Gutkovich-Pyest, E., Urieli-Shoval, S., Galun, E., Ben-Neriah, Y.,
2004. NF-kB functions as a tumour promoter in inflammation-
associated cancer. Nature 431, 461466.
Pinlaor, S., Yongvanit, P., Hiraku, Y., Ma, N., Semba, R., Oikawa, S.,
Murata, M., Sripa, B., Sithithaworn, P., Kawanishi, S., 2003. 8-Nitrogua-
nine formation in the liver of hamsters infected with Opisthorchis
viverrini. Biochem. Biophys. Res. Commun. 309, 567571.
Pinlaor, S., Ma, N., Hiraku, Y., Yongvanit, P., Semba, R., Oikawa, S.,
Murata, M., Sripa, B., Sithithaworn, P., Kawanishi, S., 2004a. Repeated
596 S. Pinlaor et al. / International Journal for Parasitology 35 (2005) 591596
infection with Opisthorchis viverrini induces accumulation of 8-
nitroguanine and 8-oxo-7,8-dihydro-2 0 -deoxyguanine in the bile duct
of hamsters via inducible nitric oxide synthase. Carcinogenesis 25,
15351542.
Pinlaor, S., Sripa, B., Sithithaworn, P., Yongvanit, P., 2004b.
Hepatobiliary changes, antibody response, and alteration of liver
enzymes in hamsters re-infected with Opisthorchis viverrini. Exp.
Parasitol. 108, 3239.
Sripa, B., Kaewkes, S., 2000. Localisation of parasite antigens and
inflammatory responses in experimental opisthorchiasis. Int.
J. Parasitol. 30, 735740.
Surh, Y.J., Chun, K.S., Cha, H.H., Han, S.S., Keum, Y.S., Park, K.K.,
Lee, S.S., 2001. Molecular mechanisms underlying chemopreventive
activities of anti-inflammatory phytochemicals: down-regulation of
COX-2 and iNOS through suppression of NF-kB activation. Mutat. Res.
480-481, 243268.
van der Kleij, D., Latz, E., Brouwers, J.F., Kruize, Y.C., Schmitz, M., Kurt-
Jones, E.A., Espevik, T., de Jong, E.C., Kapsenberg, M.L.,
Golenbock, D.T., Tielens, A.G., Yazdanbakhsh, M., 2002. A novel
host-parasite lipid cross-talk. Schistosomal lyso-phosphatidylserine
activates toll-like receptor 2 and affects immune polarization. J. Biol.
Chem. 277, 4812248129.
 International Journal for Parasitology 35 (2005) 597615
www.parasitology-online.com

Invited review
Insights into unique physiological features of neutral lipids
in Apicomplexa: from storage to potential mediation
in parasite metabolic activities
Isabelle Coppensa,*, Ole Vielemeyerb
a
Department of Molecular Microbiology and Immunology, Johns Hopkins Bloomberg School of Public Health,
615 N. Wolfe Street, Baltimore, MD 21205-2223, USA
b
Laboratoires Constant Burg, Institut Curie, UMR 144/CNRS/IC, 26, rue dUlm, 75248 Paris, Cedex 05, France
Received 9 June 2004; received in revised form 5 January 2005; accepted 13 January 2005

Abstract
The fast intracellular multiplication of apicomplexan parasites including Toxoplasma and Plasmodium, requires large amounts of lipids
necessary for the membrane biogenesis of new progenies. Hence, the study of lipids is fundamental in order to understand the biology and
pathogenesis of these deadly organisms. Much has been reported on the importance of polar lipids, e.g. phospholipids in Plasmodium.
Comparatively, little attention has been paid to the metabolism of neutral lipids, including sterols, steryl esters and acylglycerols. In
eukaryotic cells, free sterols are membrane components whereas steryl esters and acylglycerols are stored in cytosolic lipid inclusions. The
first part of this review describes the recent advances in neutral lipid synthesis and storage in Toxoplasma and Plasmodium. New potential
pharmacological targets in the pathways producing neutral lipids are outlined. In addition to lipid bodies, Apicomplexa contain unique
secretory organelles involved in parasite invasion named rhoptries. These compartments appear to sequester most of the cholesterol found in
the exocytic pathway. The second part of the review focuses on rhoptry cholesterol and its potential roles in the biogenesis, structural
organisation and function of these unique organelles among eukaryotes.
q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: Toxoplasma; Plasmodium; Lipid bodies; Rhoptries; Multivesicular bodies; Cholesterol; Cholesteryl esters; Triacylglycerol

1. Apicomplexa and lipid bodies
1.1. Lipid bodies in eukaryotic cells are more than simple
energy storage structures
The formation of intracellular lipid particles occurs at
some point in the life cycle of nearly all organisms,
including plants, mammals, non-mammalian cells, algae
and yeast as well as in some prokaryotes (reviewed in
Zweytick et al., 2000). These structures are known as lipid
bodies, lipid droplets (in adipocytes), adiposomes or oil
bodies (in plants). Lipid bodies can be defined as a major
form of macromolecular lipid assembly in biological
systems. In contrast to the bilayer membrane of organelles,
* Corresponding author. Tel.: C1 443 287 1589; fax: C1 410 955 0105.
E-mail address: icoppens@jhsph.edu (I. Coppens).
0020-7519/$30.00 q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijpara.2005.01.009
lipid bodies are surrounded by only one monolayer of
amphipathic phospholipids, glycolipids and/or sterols that
encircles a hydrophobic core of neutral lipids, such as steryl
esters, diacylglycerol (DAG) and triacylglycerol (TAG). In
mammalian cells, the major enzymes involved in lipid
esterification are members of the membrane bound O-acyl
transferase (MBOAT) family (Hofmann, 2000) that include
acyl-CoA:cholesterol acyltransferase (ACAT) and acyl-
CoA:diacylglycerol acyltransferase (DGAT), producing
cholesteryl ester and TAG, respectively (Murphy, 2001).
Two mammalian enzymes ACAT1 and ACAT2, utilising
fatty acyl-CoA and cholesterol to produce cholesteryl esters,
have been identified (summarised in Buhman et al., 2000;
Chang et al., 2001). They are mainly localised within the
endoplasmic reticulum, and both proteins display
similar enzymologic properties with broad acyl-CoA
specificity. The glycerol-3-phosphate pathway, also known
598 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615
as the Kennedy pathway is the major route for de novo TAG
biosynthesis in all TAG-accumulating organisms (Lehner
and Kuksis, 1996). This pathway involves the stepwise
acylation of glycerol-3-phosphate and/or dihydroxyacetone
phosphate to phosphatidic acid, which in turn is hydrolysed
to DAG. DGAT catalyses the final and rate-limiting step,
the transfer of the acyl group from acyl-CoA to DAG to
form TAG (Lehner and Kuksis, 1996). Two different DGAT
have been identified in eukaryotes. DGAT1 is phylogeneti-
cally related to ACAT family members (Lehner and Kuksis,
1996; Cases et al., 1998; Oelkers et al., 1998; Bouvier-Nave
et al., 2000; Farese et al., 2000; Sorger and Daum, 2002)
while DGAT2 is not related to any known enzymes (Cases
et al., 2001; Lardizabal et al., 2001). In mammals, DGAT1
is solely a microsomal membrane-protein (Farese et al.,
2000) while in plants and yeast, this enzyme has a dual
localisation and is found both on the endoplasmic reticulum
and lipid (oil) bodies (Bouvier-Nave et al., 2000; Sorger and
Daum, 2002). DGAT1 and DGAT2 have similar maximal
capacities of TAG synthesis and share, like their ACAT
counterparts, broad acyl-CoA specificities (Cases et al.,
2001). In addition to the reaction catalysed by DGAT, other
mechanisms for TAG synthesis have been observed in
plants and yeast (Dahlqvist et al., 2000). Those include an
acyl-CoA-independent pathway involving a phospholipid:
diacylglycerol acyltransferase, which is distantly related to
the mammalian enzyme lecithin:cholesterol acyltransferase
(Oelkers et al., 2000). Recently, a new type of DGAT, the
bifunctional wax ester synthase/DGAT, has been identified
in some bacteria and plants (Kalscheuer and Steinbuchel,
2003).
In multicellular organisms, DGAT displays a variety of
physiological functions, including lipoprotein assembly,
regulation of plasma TAG concentration or cytosolic DAG
levels. As an allosteric activator of the protein kinase C,
DAG links extracellular signals to intracellular events,
resulting in important biological processes like cell
proliferation and differentiation. In addition to its messenger
properties, DAG is also a central metabolite in de novo
biosynthesis of phospholipids (reviewed in van Blitterswijk
and Houssa, 2000). In unicellular microorganisms, TAG can
function as a reservoir for metabolic energy and stored TAG
may be used as a fatty acid source for phospholipid
biosynthesis. Neither neutral lipid biosynthesis nor lipid
bodies are essential for yeast growth (Sandager et al., 2002)
but accumulation of lipid bodies containing TAG appears to
be specifically induced in response to metabolic stress or
environmental changes (e.g. osmotic stress, nitrogen
depletion; Murphy, 1990, 2001; van Blitterswijk and
Houssa, 2000).
There is now increasing evidence that most lipid bodies
contain different populations of proteins that are more or
less tightly bound to their surfaces. In mammalian cells,
proteomic analysis of lipid body-associated proteins
identified structural proteins; multiple enzymes involved
in the synthesis, storage, utilisation (e.g. perilipin,
adipophilin; reviewed in Fujimoto et al., 2004), and
degradation of cholesteryl esters and TAG; different Rab
GTPases involved in regulating membrane traffic; signaling
molecules; and proteins found in caveolae and lipid rafts
(Liu et al., 2004; Fujimoto et al., 2004). Higher eukaryotic
cells and yeast lipid bodies have strikingly similar
proteomes that emphasise common functions in lipid
metabolism (Athenstaedt et al., 1999). In plants, the major
oil body-associated proteins, oleosins function as stabilisers
of TAG. These proteins prevent subcellular oil bodies from
coalescence, maintaining them as individual structures in
plant seeds even after a long period of storage (reviewed in
Zweytick et al., 2000).
It is well known that lipid bodies store neutral lipids
utilisable for energy (e.g. fatty acids as respiratory
substrates), for membrane biogenesis (e.g. fatty acids,
cholesterol) and/or for the formation of specific lipophilic
components, (e.g. steroid hormones). However, the con-
stellation of proteins associated with the lipid droplets, as
listed above, indicates that these compartments are complex
and metabolically active. In fact, they may be a nodal point
for multiple intersecting membrane pathways, and be
directly involved in membrane lipid recycling. In plants,
for instance, another role of oil bodies might be to channel
away toxic fatty acids from membranes by incorporating
them in TAG (Ohlrogge and Jaworski, 1997). Yeast lipid
body particles may serve as a degradation compartment for
enzymes that are no longer needed or have been produced in
excess (Lum and Wright, 1995).
There are several models describing the biogenesis of
intracellular lipid bodies (Fig. 1). First, the budding model
proposes that droplets of neutral lipids are formed within the
bilayer of the endoplasmic reticulum, and then become free
cytoplasmic particles, shielded by a phospholipid mono-
layer after budding of these specific microdomains from the
endoplasmic reticulum. The second vesicle flux model
hypothesises a mechanism of translocation of vesicles from
the endoplasmic reticulum, which contain either neutral
lipids or proteins, followed by their fusion to produce
mature lipid bodies. As a third alternative, the aggregation
model suggests that specific lipid particle proteins form
aggregates in the cytosol and encapsulate neutral lipids that
pinched off from the endoplasmic reticulum (summarised in
Murphy and Vance, 1999; Zweytick et al., 2000).
Of importance, some of the most widespread human
diseases such as atheroma, steatosis, obesity, and certain
cancer types (Heid et al., 1998; Murphy and Vance, 1999)
are linked to malfunction of lipid body metabolism. In
addition, there is a series of recent findings demonstrating a
mistargeting of unexpected proteins to the surface of lipid
droplets under pathological conditions. Examples include
the a-synuclein as a major component of the pathologic
lesions characteristic of the Parkinsons disease (Cole et al.,
2002) and the Nir2 protein involved in retinal degeneration
(Litvak et al., 2002). This further substantiates the emerging
 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615 599

A BUDDING

mature LB

cytosol

lumen ER DGAT ACAT

B VESICLE FLUX

mature LB

cytosol

lumen ER DGAT ACAT

C AGGREGATION

mature LB

cytosol
Fig. 2. Ultrastructure of neutral lipid stores in intracellular apicomplexan
parasites. (A). Toxoplasma gondii inside a parasitophorous vacuole (PV)
lumen ER formed during the invasion of a human fibroblast. The parasites exhibit
DGAT ACAT
several lipid inclusions (arrows) in their cytoplasm, which are morpho-
logically comparable with lipid bodies present in the host cells. For this
neutral lipids LB-associated protein
image, malachite green was added to the fixative solution in order to
stabilise the lipid elements and to enhance the staining of lipid-containing
Fig. 1. Schematic depiction of three models for lipid body biogenesis structures (Pourcho et al., 1978). Under such conditions, lipid structures in
(adapted from Zweytick et al., 2000). See text (Section 1.1) for details. ER, host cell and parasites alike appear compact, with a deeply stained matrix
endoplasmic reticulum; LB, lipid body. and an irregular border with small protruding granules but no discernable
membrane. Toxoplasma lipid bodies contain cholesteryl esters and
central role of lipid bodies and their proteins in many human triacylglycerol (TAG). (B). Intravacuolar schizonts of Plasmodium
physiological ailments. falciparum inside a human red blood cell (RBC). As the parasites mature
within erythrocytes, the number and size of lipid bodies (arrows) increase.
The phenomenon coincides with the production of TAG in this parasite. Rh,
1.2. Morphological observations illustrate that protozoa are rhoptries; A, apicoplast; Mt, mitochondrion; Mi, microneme; ER,
endoplasmic reticulum; N, nucleus. Bars are 1 mm.
also fat accumulators

Surprisingly, there is a paucity of information about
neutral lipid metabolism and storage in ancient eukaryotes.
In many comprehensive reviews on lipid bodies, no
reference is made to lipid bodies in lower eukaryotes
(Murphy and Vance, 1999; Zweytick et al., 2000; Murphy,
2001). However, microscopy studies performed on different
species of protozoa illustrate the presence of cytoplasmic
lipid structures of 0.11.5 mm in diameter and devoid of
membranous profiles. These compartments morphologically
resemble lipid bodies in higher eukaryotic cells (Fig. 2(A)).
Such presumed lipid inclusions have been identified in
various apicomplexan parasites such as in the malarial agent
600 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615
Plasmodium falciparum (Fig. 2(B); Nawabi et al., 2003;
Palacpac et al., 2004a,b; Vielemeyer et al., 2004), Toxo-
plasma gondii the etiologic agent of toxoplasmosis
(Fig. 2(A); Lindsay et al., 1993; Dubey et al., 1998; Coppens
et al., 2000; Sonda et al., 2001; Charron and Sibley, 2002;
Quittnat et al., 2004; Nishikawa et al., in press), Neospora
caninum responsible for neosporosis in animals (Lindsay
et al., 1993), Sarcocystis neurona causing equine myeloen-
cephalitis (Speer and Dubey, 2001), Perkinsus marinus
infecting oysters (Chu et al., 2000), Isospora suis causing
porcine enteric diseases (Lindsay et al., 1991), Pterospora
floridiensis living in the body cavity of polychaeta (Landers,
2002), Caryospora bigenetica infecting snakes and mam-
mals (Sundermann and Lindsay, 1989), Eimeria nieschulzi
infecting rodents (Sibert and Speer, 1980), as well as in
different mammalian flagellates including trypanosomatids
(Soares et al., 1987; Figueiredo et al., 1994), phytoflagellates
(Klut et al., 1988), and in the ciliate Paramecium primaurelia
(Ramoino et al., 2002).
Recently, several investigator teams have tackled the
task of analysing the composition and biogenesis of lipid
bodies in some apicomplexan parasites at the molecular
level. They have furthermore attempted to determine the
role of neutral lipids in the pathogenesis of these parasites of
great medical importance. Below we summarise the unique
features of neutral lipid metabolism in P. falciparum and
T. gondii and highlight the common themes in lipid body
formation conserved from Apicomplexa to mammalian
cells.
1.3. Toxoplasma is competent to synthesise cholesteryl
esters using host mammalian cholesterol, but Plasmodium
does not produce any steryl esters
In mammalian cells, cholesterol synthesised de novo and
acquired exogenously via the low-density lipoprotein (LDL)
pathway is esterified by fatty acids and stored in lipid
bodies. Toxoplasma contains cholesterol (Foussard et al.,
1991) but lacks the biosynthetic machinery to produce
sterols (Coppens et al., 2000) as do many other protozoa
(Furlong, 1989). Instead it diverts LDL-derived cholesterol
from host mammalian cells by an unknown mechanism.
Cultivation of intracellular T. gondii in the presence of LDL
results in a remarkable increase in the number of lipid
bodies and synthesis of cholesteryl esters in a LDL
concentration-dependent manner (Nishikawa et al., in
press). This clearly indicates that akin to mammalian
cells, this parasite has developed a mechanism for
cholesterol sensing, and following a massive delivery of
exogenous cholesterol, it can expand its cholesterol storage
pools through esterification of cholesterol-derived LDL
(Charron and Sibley, 2002; Nishikawa et al., in press).
Lipoprotein depletion, on the contrary, causes a progressive
consumption of material stored in parasites lipid bodies,
and after 2 days of LDL starvation, lipid dye-positive
structures can no longer be found in the cytoplasm of
the parasite.
While cholesteryl oleate is the main ester synthesised in
mammalian cells (Yang et al., 1997), Toxoplasma prefer-
entially utilises palmitate to esterify cholesterol (Nishikawa
et al., in press). In addition, cultivation of intracellular
parasites with excess palmitate results in a significant
increase in cholesteryl ester synthesis and formation of lipid
bodies. This preference might be due to a relatively
facilitated uptake of exogenous palmitate in comparison
with other fatty acids, and/or a selective biosynthesis of this
fatty acid by the parasite. An explanation for the difference
in fatty acid specificity between host cell and Toxoplasma
enzymes may be related to a slight difference in the
composition of the respective membrane regions where the
ACAT are anchored (see below). The accessibility of
coenzyme A to the active site of the acyltransferase may
depend on the length and degree of desaturation of the acyl
chains. In Toxoplasma, the function of esterification and
storage of cholesterol molecules may be physiologically
important since the parasite has to rely on an external source
for this lipid. In fact, during its lifecycle T. gondii
experiences a quiescent encysted stage and a brief
extracellular phase that requires metabolic adaptations,
such as the accumulation of a lipid reserve.
In stark contrast to Toxoplasma, no erythrocytic stages of
Plasmodium show any evidence for steryl ester synthesis, as
shown in metabolic labeling experiments (Vial et al., 1984;
Nawabi et al., 2003; Palacpac et al., 2004a,b; Vielemeyer
et al., 2004). This observation is corroborated by the
insensitivity of Plasmodium to potent inhibitors of steryl
esterification, the lack of influence of exogenous cholesterol
on lipid body formation (Vielemeyer et al., 2004), and
finally the absence of genes in the malarial genome (http://
plasmodb.org/) encoding enzymes involved in sterol
esterification. Stores of cholesteryl esters may not be
required for the replication of erythrocytic stages of
Plasmodium since plasmodial membranes contain only
traces of cholesterol (Wunderlich et al., 1991; Nawabi et al.,
2003; Palacpac et al., 2004a,b).
1.4. In Apicomplexa, TAG synthesis occurs via the glycerol-
3-phosphate pathway and is developmentally regulated
in Plasmodium
Early studies document the presence of TAG in some
plasmodial species and an increase of TAG level in the
mature stages of erythrocytic Plasmodium grown in vivo
and in vitro (Beach et al., 1977; Vial et al., 1982a,b). More
recently, metabolic labeling studies using either saturated or
unsaturated fatty acids, or DAG reveal an efficient
incorporation of these lipid precursors into the TAG fraction
of P. falciparum (Mitamura et al., 2000; Nawabi et al.,
2003; Palacpac et al., 2004a,b; Vielemeyer et al., 2004). The
accumulation of TAG increases with parasite maturation
and seems strikingly more pronounced in mature
 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615
trophozoites (26 h p.i.) compared to schizonts (38 h p.i.),
indicating that TAG stores must be essential during the last
steps of the intraerythrocytic development. The production
of TAG in Plasmodium correlates with the biogenesis of
cytoplasmic lipid bodies, which become relatively more
abundant and larger as the parasites mature (Fig. 2(B)).
Forty-two hours p.i., TAG degradation ensues, and the
resulting breakdown productsthe free fatty acidsare
released into the medium during merozoite egress from
erythrocytes (48 h p.i.; Palacpac et al., 2004a,b). Such TAG
catabolism implies the existence of functional (phospho)-
lipases and TAG lipase during the erythrocytic development
of these parasites (Athenstaedt et al., 1999), for which a
candidate gene is present in the Plasmodium database,
during the erythrocytic development of these parasites.
Incubation with trifluoperazine, a modulator of fatty acid
incorporation into TAG, results in the disappearance of
plasmodial lipid bodies (Palacpac et al., 2004a,b). This
strongly suggests that a de novo TAG biosynthetic pathway
via phosphatidic acid contributes to lipid body formation in
Plasmodium. Incubation of Plasmodium-infected erythro-
cytes with monoacylglycerol in the presence of oleate
results in the production of DAG as a consequence of
monoacylglycerol acylation (our personal data). Thus, as
described in higher eukaryotic cells (Chen and Farese,
2000), the monoacylglycerol pathway leading to TAG
production through a DGAT-mediated pathway is present in
Plasmodium.
What can be the physiological relevance of synthesis and
accumulation of TAG in Plasmodium? As suggested for
Toxoplasma (Charron and Sibley, 2002), lipid bodies in
Plasmodium bloodstages might serve as an intracellular
compartment to divert and concentrate host cell lipids for
the biogenesis of parasite membranes, and possibly as a
place for lipid metabolism during proliferation. Indeed,
P. falciparum is capable of conversion of TAG to essential
phospholipids (Vial et al., 2003). However, TAG as a
reservoir of fatty acids for phospholipid synthesis seems less
probable in Plasmodium, since no lipid bodies are
detectable in ring stage parasites, which are rapidly growing
and expanding to form large trophozoites, thus requiring a
massive supply of lipids for membrane biogenesis. Instead,
lipid stores are abundant in late trophozoites and schizonts,
where most of the new membranes are already formed.
Another possibility is that TAG may serve as a reservoir for
fatty acids for oxidative catabolism to generate ATP.
However, this is likewise improbable because this parasite
seems to have an extremely low capacity for b-oxidation of
fatty acids based on chromatography assays (Holz, 1977;
Palacpac et al., 2004). This observation is supported by the
lack of classical enzymes required for fatty acid catabolism
in plasmodial genomes (Gardner et al., 2002). Nevertheless,
this latter statement deserves deeper investigation by using,
for example, integrated computational genomics to corro-
borate the absence of any functional pathways linked
to fatty acid oxidation in malarial parasites. Plasmodial
601
TAG could serve as a source of fatty acyl chains for the
synthesis of glycerophospholipids or glycosylphosphatidy-
linositol (GPI). This seems also unlikely, though, because
the production rate and concentration of these two lipids are
significantly higher, when compared with TAG (Vial et al.,
1982a; Naik et al., 2000; Palacpac et al., 2004a,b).
Alternatively, Mitamura and colleagues have evoked the
possibility that TAG metabolism can be regarded as a
unique dynamic cellular event that facilitates Plasmodium
egress from the host erythrocyte. According to this
hypothesis, a sudden massive release of fatty acids from
TAG stores in the mature stages would destabilise and
rupture both, the parasitophorous vacuole membrane (PVM)
and the erythrocyte plasma membrane, facilitating parasite
release and dissemination. Nonetheless, this scenario
implies a selective targeting of parasite fatty acids to the
PVM and host cell membranes, excluding parasite internal
membranes, which is difficult to conceive in the context of
our current knowledge on lipid trafficking.
Metabolic labeling experiments using fatty acids or DAG
performed on intracellular Toxoplasma demonstrate that
this parasite is also able to take up both lipid precursors and
incorporate them into a TAG fraction (Quittnat et al., 2004).
Oleate can be incorporated into Toxoplasma DAG, reveal-
ing that DAG is the acyl acceptor (Cases et al., 1998). As
monitored for cholesteryl ester synthesis, T. gondii also
incorporates preferentially palmitate into TAG, as com-
pared to other fatty acids. Stored TAG may be a reservoir of
fatty acids utilisable for phospholipid biosynthesis and/or
exploitable as respiratory substrates in Toxoplasma,
although a mechanism of oxidative degradation of fatty
acids remains to be demonstrated in this parasite.
1.5. Toxoplasma has one DGAT and at least two isoforms
of ACAT while the genome of Plasmodium contains only
one member of the MBOAT family, probably involved
in TAG synthesis
The genome of Toxoplasma (http://toxodb.org/) contains
at least two genes encoding enzymes of the MBOAT family.
Based on analysis of the predicted protein sequences and the
conserved domains characterising the members of the
MBOAT family, one Toxoplasma candidate has the major
hallmarks of a DGAT1 enzyme (Fig. 3(A); Oelkers et al.,
1998; Bouvier-Nave et al., 2000), while the other candidate
shares the characteristics of mammalian ACAT1
(Fig. 3(B)).
Similarly to their mammalian counterparts (Cases et al.,
1998), the Toxoplasma-related DGAT1 and ACAT1 protein
sequences, named TgDGAT1 or TgACAT1, respectively,
have several membrane spanning domains as well as
potential sites for N-glycosylation and tyrosine phosphoryl-
ation, though the significance of the latter features is still
unknown (Quittnat et al., 2004; Nishikawa et al., in press).
The genes of TgDGAT1 and TgACAT1 contain
multiple exons (Quittnat et al., 2004), as described for
602 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615
Fig. 3. Presence of conserved signatures characterising MBOAT, ACAT
and DGAT in the predicted protein sequences of TgACAT1, TgDGAT1 and
PfDGAT1. (A). Sequence comparisons between TgDGAT1, PfDGAT1 and
Homo sapiens DGAT1 (GenBank accession number L21934; Oelkers et al.,
1998) show the presence of an invariant histidine positioned within a long
hydrophobic region that characterises the superfamily of MBOAT
(Hofmann, 2000; Cases et al., 1998); a remarkable motif thought to
include the putative fatty acid binding site that is shared between all ACAT
and DGAT1 (Bouvier-Nave et al., 2000); and two unique signatures of
DGAT, one of them corresponding to the DAG/phorbol ester binding site
located downstream of the a putative fatty acid binding site signature. (B).
Sequence comparisons between TgACAT1 and Homo sapiens ACAT1
(accession number L21934; Chang et al., 1993) reveal the motif
characterising the superfamily of MBOAT (see above); the putative fatty
acid binding site; and the potential cholesterol binding site including a
serine surrounded by an aromatic and/or basic amino acids (Guo et al.,
2001). All these possible functional domains are consistent with the
substrate utilisation properties of the parasite acyltransferases (Quittnat
et al., 2004; Nishikawa et al., in press).
the mammalian homologues (Bouvier-Nave et al., 2000).
Interestingly, two isoforms of TgACAT1, designated
TgACAT1a and TgACAT1b derived from two full-length
cDNA have been demonstrated in Toxoplamsa (Nishikawa
et al., in press). TgACAT1a harbors a long hydrophilic
serine-rich region at the N-terminus, which is missing in
TgACAT1b (Quittnat et al., 2004). TgDGAT, TgACAT1a
and TgACAT1b are all integral membrane proteins
localised to the parasite cortical and perinuclear endoplas-
mic reticulum, and have never been detected at the surface
of lipid bodies as found in plants or yeast (Sorger and Daum,
2002).
The transformation of a Saccharomyces cerevisiae
mutant strain lacking neutral lipid production and so devoid
of lipid bodies (quadruple knock-out are1D,are2D,dga1D,
lro1D; Sandager et al., 2002) with a full length TgDGAT1
construct results in a significant increase of TAG synthesis.
Furthermore, this TAG synthesis in mutant yeast correlates
with formation of cytosolic lipid inclusions, resembling
lipid bodies present in Toxoplasma (Quittnat et al., 2004).
ACAT-deficient mammalian cells, when transfected with
either TgACAT1a or TgACAT1b are restored in their
capacity to synthesise cholesteryl ester (Quittnat et al.,
2004). Mutagenesis of the conserved fatty acid and
cholesterol binding sites of TgACAT1a abolishes the
ACAT activity. When the mutant yeast strain is transformed
with TgACAT1a in the absence of cholesterol, ergosteryl
esters are produced, although neither ergosterol nor
ergosteryl esters have ever been detected in Toxoplasma.
The addition of cholesterol to the culture medium induces
the synthesis of cholesteryl esters by the transformed mutant
yeast strain. This demonstrates that unlike in higher
eukaryotes or in yeast, where one particular steryl ester is
predominantly synthesised (Yang et al., 1997), the ester-
ification reaction in Toxoplasma is not specific to changes in
the sterol side chain since both cholesteryl and ergosteryl
esters are produced by the heterogeneously expressed
Toxoplasma ACAT.
In higher eukaryotic cells and yeast, ACAT and DGAT
are encoded by two related groups of genes and their
predicted sequences share highly homologous domains
(Bouvier-Nave et al., 2000), suggesting that DGAT and
ACAT genes might have arisen from a common ancestor,
having displayed both functions. The divergence of
Apicomplexa as well as dinoflagellates and ciliates is
believed to have occurred more than one billion years ago,
possibly before the separation of the three multicellular
kingdoms of animals, plants and fungi (Escalante and
Ayala, 1995). However, when heterogeneously expressed,
TgDGAT1 cannot synthesise steryl esters just as TgACAT1
shows no activity of TAG production (Quittnat et al., 2004;
Nishikawa et al., in press). This observation, thus assigns
clearly separate enzymatic function to the two proteins
present in lower eukaryotes.
In Toxoplasma, both LDL-derived cholesterol and free
fatty acids can serve as ACAT activators. The TgACAT1
mRNA expression, cellular cholesterol esterification and
lipid droplet biogenesis appear to be coordinately regulated.
Such a mechanism of transcriptional regulation of
TgACAT1 suggests a unique sterol regulatory-like
element present within the TgACAT1 gene promotors.
 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615 603

In mammalian cells, by contrast, ACAT1 is regulated

primarily by post-translational mechanisms (Chang et al.,
1997). Only the expression level of one of the four human
ACAT1 mRNAs changes due to the exogenous supply of
cholesterol. Free fatty acids can increase mammalian
ACAT1 mRNA levels of two on the four transcripts in a
cell type specific manner and with some acyl-CoA
preferences (Seo et al., 2001).
Only one gene with significant homology to known
members of the MBOAT family is present in the
P. falciparum genome. This protein has numerous potential
hydrophobic domains and is localised to the endoplasmic
reticulum. Phylogenetic analysis shows that the parasite
predicted protein sequence possesses the major hallmarks of
the DGAT1 family (Fig. 3(A)), with the highest degree of
homology with plant DGAT1. Recently, a direct role for
PfDGAT1 in plasmodial TAG synthesis has been assigned
to the parasite after transfection of mammalian cells with a
synthetic PfDGAT1 gene, which showed an increase in
DGAT activity compared to non transfected cells (Palacpac
et al., 2004). In addition, this PfDGAT1 immunoreacts
specifically with antibodies against TgDGAT1 and its
transcription peak occurs during the late trophozoite stage,
which coincides with maximal TAG synthesis and lipid
body formation (Vielemeyer et al., 2004).
Alternatively TAG synthesis may also occur through a
non-DGAT pathway in Plasmodium as identified in yeast.
Indeed, a phospholipid:diacylglycerol acyltransferase hom-
Fig. 4. Ultrastructure of neutral lipid deposits inside the parasitophorous
ologue, which is distantly related to the mammalian enzyme vacuole (PV) of Toxoplasma and the digestive vacuole of Plasmodium. (A).
lecithin:cholesterol acyltransferase (Dahlqvist et al., 2000; Intravacuolar Toxoplasma gondii in a human fibroblast incubated in the
Sorger and Daum, 2003) does exist in the Plasmodium presence of excess fatty acids. The arrows pinpoint large deposits of
genome, although it does not have a significant percentage malachite green-positive material in the lumen of the PV, which faces host
of identity with the yeast enzyme and its peak of organelles associated with the parasitophorous vacuole membrane (PVM).
This observation might substantiate the hypothesis on a biochemical
transcription occurs not in the late trophozoite but the ring exchange between host mitochondrial-associated membranes and the PVM
stage (Bozdech et al., 2003). However, experimental to deliver lipids to the parasites. (B). Intravacuolar trophozoite of
evidence reveals that disruption of the PfDGAT1 gene is Plasmodium falciparum inside a human red blood cell (RBC). The
deleterious in Plasmodium, leading to the presumption that presence of the malachite green-positive material closely associated with
PfDGAT1 plays a fundamental role in the intraerythrocytic the digestive vacuole suggests that neutral lipids may contribute to the
catalysis of hemozoin formation. Indeed, arrows clearly show a close
proliferation of P. falciparum and that it likely acts as the association of neutral lipid deposits with hemozoin crystals. A, apicoplast;
major contributor for TAG biosynthesis. Being the only DV, digestive vacuole; Mt, mitochondrion; ER, endoplasmic reticulum;
predicted plasmodial MBOAT, PfDGAT1 may also partici- LB, lipid bodies; N, nucleus. Bars are 0.5 mm.
pate in other aspects of lipid metabolism, e.g. modification
of the fatty acid composition of phospholipids. prominently decorates cytosolic lipid bodies and to a lesser
extent the endoplasmic reticulum, the PVM and plasma
1.6. In addition to the cytoplasm, apicomplexan parasites membrane (Sonda et al., 2001; Charron and Sibley, 2002;
accumulate neutral lipid components in unusual Quittnat et al., 2004; Nishikawa et al., in press).
compartments Surprisingly, the Toxoplasma vacuole appears to be
intensely labeled with Nile Red upon incubation with excess
The location of neutral lipid components can been finely fatty acids, thus revealing neutral lipid deposits in the
determined by exploiting the fluorescence properties of Nile vacuolar space (Quittnat et al., 2004). Ultrastructural studies
Red, known to label membranes in addition to neutral lipid confirm large and abundant malachite green-positive
particles (Greenspan et al., 1985), or by using the malachite structures in the lumen of the PV, mainly located between
green aldehyde fixative technique, which is efficient to parasites and host mitochondria or endoplasmic reticulum
stabilise lipid elements that are otherwise extracted associated with the PVM (Fig. 4(A); Quittnat et al., 2004).
from cells during the preparations of samples for electron Shortly after parasite invasion, the PVM physically interacts
microscopy (Pourcho et al., 1978). In Toxoplasma, Nile Red with the host cell mitochondria and endoplasmic reticulum,
604 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615
owing to the anchoring of a secreted parasite protein into the
PVM and the membranes of both mitochondria and
endoplasmic reticulum (Sinai and Joiner, 2001). The close
apposition of these host cell organelles with the PVM raises
the possibility that a mechanism of direct interorganellar
exchange (e.g. lipids) exists between host organelles and the
PV. In mammalian cells, lipids manufactured in the
endoplasmic reticulum transit from this organelle to
mitochondria via specialised regions named the mitochon-
drial-associated membrane (MAM), involving translocators
that deliver phospholipids to the mitochondrial membranes
for further modifications (Shiao et al., 1995). By analogy,
lipid mobilisation to Toxoplasma may be facilitated by the
physical association between host cell organelles and
specific regions of the PVM. If so, the specific recruitment
of host organelles mediated by invading parasites may be an
important adaptation for lipid acquisition that facilitates
intracellular Toxoplasma growth. The large lipid deposits
visible in the lumen of the vacuole of intravacuolar parasites
upon excess fatty acids may also be a direct consequence of
parasite ACAT overexpression, ensuing neutral lipid
secretion, as observed with mammalian apolipoprotein B
from hepatocytes (Spady et al., 2000) and lipoprotein-like
particles from astrocytes (Mutka et al., 2004).
In Plasmodium, Nile Red identifies cytoplasmic lipid
bodies as well as a population of intensely stained particles
of a few hundred nanometers in size that are closely
associated with the digestive vacuole and are composed of
DAG and TAG (Fig. 4(B); Jackson et al., 2004). These
particles may represent neutral lipid storage structures for
lipid intermediates that are generated during digestion of
phospholipids in the digestive vacuole. Indeed, there must
be a large turnover of phospholipids during the growth of
Plasmodium as it ingests the host haemoglobin through the
cytostome. The endocytic compartments that are formed are
surrounded by a double membrane originating from the
parasite plasma membrane and the PVM. The outer
membrane of the vesicles fuses with the digestive vacuole
membrane, whereas the inner membrane is thought to
be degraded by phospholipases (Yayon et al., 1984;
Slomianny, 1990). Subsequently, the phospholipid break-
down products may assemble into TAG.
The close association of neutral lipids with the digestive
vacuole suggests that they may function in processes
occurring within this organelle. The digestive vacuole is
the site of degradation of haemoglobin, which is endocy-
tosed from the host cell cytosol. The digestion process
results in the production of potentially toxic free haem. An
important pathway for the detoxification of haem is the
formation of crystals of haemozoin, the characteristic
malarial pigment present within the digestive vacuole.
Previous studies report that monoacylglycerol, DAG and
TAG suspensions promote b-haematin (the synthetic
equivalent of haemozoin) formation in vitro (Fitch et al.,
1999; Papalexi et al., 2001; Pandey et al., 2003; Jackson
et al., 2004). This observation raises the possibility that
the neutral lipid deposits associated with the parasite
digestive vacuole may represent the site of haemozin
formation in vivo and that their lipid content assist in the
crystallisation of haemozoin. In this context, it can be
assumed that aggregates of DAG and other TAG precursors
accumulate in the digestive vacuole before incorporation
into neutral lipid bodies and play a role in haem
detoxification during the early trophozoite stage of infec-
tion. It has been reported that plasmodial histidine-rich
proteins contribute to the catalysis of haemozoin formation
in vivo (Sullivan et al., 1996; Francis et al., 1997); however,
the levels of these proteins in the digestive vacuole are too
low to provide sufficient catalytic activity of haemoglobin
(Papalexis et al., 2001; Akompong et al., 2002).
Apicomplexa harbor a unique plastid-like organelle
named the apicoplast, which probably is derived from the
engulfment of a red alga in ancient times (reviewed in
Waller and McFadden, 2005). Like plant chloroplasts,
apicoplasts are semi-autonomous with their own genome
and expression machinery. Apicoplasts possess two primary
metabolic pathways for lipid synthesis: the type II pathway
for de novo fatty acid synthesis, which provides the large
bulk of fatty acids for apicomplexan parasites, and the
1-deoxy-D-xylulose 5-phosphate (DOXP) pathway produ-
cing isoprenoids, which largely replaces the mevalonate
pathway which has a very low activity in these parasites. In
contrast to Toxoplasma and Plasmodium, plants also
synthesise TAG in chloroplasts using a DGAT1 protein
associated with chloroplast membranes; in plant chloro-
plasts, the specific role of TAG appears to be the
sequestration of fatty acids released from the catabolism
of thylakoid galactolipids (Kaup et al., 2002). So far no
stores of acylglycerols have been identified in apicoplasts.
1.7. ACAT and DGAT as pharmacological targets
in Apicomplexa
It has been extensively shown for several protozoan
parasites that lipid biosynthetic pathways are essential for
their development and that these pathways represent valid
targets for chemotherapeutic intervention (reviewed in Vial
et al., 2003; Mitamura and Palacpac, 2003). Interfering with
host cholesterol acquisition by Toxoplasma impairs parasite
growth (Coppens et al., 2000). Since the parasite does not
have the capability to synthesise sterols de novo, impairing
cholesterol storage may be also detrimental for parasite
survival in host cells. In mammalian cells, perturbation of
the balance between cholesteryl esters and free cholesterol
using pharmacological inhibitors of ACAT leads to the
accumulation of excess amounts of free intracellular
cholesterol. This, in turn results in cell toxicity, and once
a critical mass of cholesterol is reached, free cholesterol
crystallisation occurs within the membrane bilayer
(Kellner-Weibel et al., 1998). When added to the
extracellular milieu, cholesterol acceptors (e.g. serum,
high-density lipoproteins, apolipoprotein E, apolipoprotein
 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615
A-I/phospholipid particles) can desorbs cholesterol from
membranes and prevent a nucleation event from occurring.
Incubation of intracellular Toxoplasma in the presence of
selected ACAT inhibitors at non-toxic concentrations for
mammalian cells, leads to a rapid rupture of parasite plasma
membrane (Nishikawa et al., in press). This cytopathic
effect is irreversible and likely linked to a massive
incorporation of cholesterol into the parasite plasma
membrane, provoking a complete destabilisation of the
lipid bilayer. The higher sensitivity of Toxoplasma towards
cholesterol esterification inhibitors compared to mammalian
cells may be explained by the absence of physiological
acceptors of cholesterol in the parasitophorous vacuolar
space. In addition, Toxoplasma lacks caveolin and caveolae
(our unpublished data), which in mammalian cells transfer
cholesterol from the endoplasmic reticulum to the plasma
membrane, and so mediate the efflux of free cholesterol
(reviewed in Batetta et al., 2003). Therefore, the process of
cholesterol esterification may be a potential drug target,
since pharmacological reduction of cholesteryl esters results
in a toxic buildup of free cholesterol in Toxoplasma.
Both TgDGAT1 and PfDGAT1 mRNA are encoded by a
single copy gene and disruption of the gene by knockout
strategy results in the generation of non viable parasites (our
personal data; Palacpac et al., 2004a,b), suggesting that the
production of TAG is vital for Toxoplasma and Plasmodium
and relies largely on their respective DGAT1, as described
in higher eukaryotic cells. In these parasites, the step of
DAG esterification may thus also be exploitable for the
pharmacological reduction of the TAG levels. In plasmodial
parasites, the potential role of neutral lipids in haemozoin
formation may lead to the exploitation of these lipid-rich
aggregates in the digestive vacuole to interfere with
haemoglobin detoxification as anti-malarial targets.
2. Apicomplexa and rhoptries
2.1. Rhoptries have unique morphological
and compositional features
In apicomplexan parasites, rhoptries are the largest apical
organelles extending from the nuclear area through the conoid
to the anterior plasma membrane. They vary in number
depending on the life-cycle stage and the species (reviewed in
Blackman and Bannister, 2001). Toxoplasma contains 812
rhoptries, while Plasmodium blood and insect stages have only
two. Furthermore, mature rhoptries vary considerably in size
and shape, ranging from pear-like structures in Plasmodium to
long club-shaped organelles with a bulbous base and a
narrowing apical peduncle in Toxoplasma.
Rhoptries are membrane-bound organelles containing
densely packaged granular material. In the intraerythrocytic
stages of Plasmodium, maturing rhoptries form
two distinctive regions, an apical reticular zone with
electron-lucent areas separated by cords of granular
605
material, and a basal region with a more homogenously
granular pattern (Bannister et al., 2000). Rhoptries in
Toxoplasma are somewhat different as they contain a
homogenous electron-dense apical region and a more
heterogeneous basal portion with a honeycomb-like appear-
ance, resembling internal lamellar structures (Fig. 5(A)).
These remarkable ultrastructural distinctions between the
duct and bulb, found in both parasites, might be ascribed to
their differing polypeptide and lipid compositions, although
the mechanism of such sub-compartmentalisation within
rhoptries is still unclear (reviewed in Ngo et al., 2004).
All rhoptry proteins identified so far are synthesised as
inactive prepro-proteins (Ngo et al., 2004). The signal
sequence is cleaved in the endoplasmic reticulum, while
the pro-domain trimming occurs within the rhoptry
compartment, which is acidic (Shaw et al., 1998) and
contains various proteases. Toxoplasma and Plasmodium
rhoptry proteins do not share any significant protein
sequence homologies (detailed in Preiser et al., 2000).
However, many of them in both parasites have predicted
membrane-spanning domains and are found inserted into
host cell membranes, suggesting a common role for
rhoptry proteins (summarised in Sam-Yellowe, 1996). In
fact, such transmembrane proteins are implicated in the
invasion process, first promoting parasite adhesion to the
host cell and subsequently expanding the PVM. In
addition, some rhoptry proteins are involved in the host
cell remodeling after completion of invasion. Fascinating
examples include the host mitochondria and endoplasmic
reticulum recruitment to the PV of Toxoplasma (Sinai
and Joiner, 2001), possibly for nutrient scavenging from
these organelles, and the structural modifications of the
plasma membrane of Plasmodium-infected erythrocytes
for promoting the cytoadhesion of infected cells to
endothelial cells (Allred and Al-Khedery, 2004; Ling
et al., 2004).
More detailed studies reveal that Toxoplasma rhoptries
contain lipids, including large amounts of phosphatidyl-
choline and cholesterol (Fig. 5(B); Foussard et al., 1991;
Coppens and Joiner, 2003). Morphological observations
demonstrate a preponderance of cholesterol at the basal
bulbous portions (Coppens and Joiner, 2003).
2.2. Rhoptries are analogous to secretory lysosomes and are
derived from multivesicular endosomal precursors
Because of the striking morphology of rhoptries and their
complex composition, the biogenesis of these organelles has
been a long-standing puzzle. Indeed, the lack of early
organelle markers has hindered the identification of rhoptry
precursors, called immature, pre- or proto-rhoptries. Fol-
lowing the attachment of Plasmodium to the erythrocyte
surface, rhoptries discharge their contents into the host
plasma membrane, then they disappear after parasite
internalisation and are formed again de novo at the end of
each erythrocytic cycle (Bannister et al., 1986). During this
606 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615

Fig. 5. Fine detailed analysis of the structure of the rhoptry compartment in Toxoplasma gondii and comparison with mammalian multivesicular bodies (MVB).
(A). Electron micrograph illustrating the intraluminal granular material in the bulbous base of rhoptries, reminiscent of lamellar sheets found in secretory
lysosomes. Bar is 0.5 mm. (B). Fluorescence microscopy picture showing the pattern of filipin labeling associated with an intravacuolar parasite 20-min after
invasion of a human fibroblast. Clearly the apical rhoptries and the parasitophorous vacuole membrane (PVM) are filipin-stained, and thus contain sterol
(modified from Coppens and Joiner, 2003). (C). Serial sections of electron micrographs showing a rhoptry compartment containing internal vesicles. This
organelle shares morphological similarities with MVB in higher eukaryotic cells, and may correspond to a multivesicular endosomal precursor of rhoptries. Bars
are 0.2 mm. (D). Electron micrograph representing a typical MVB in a human fibroblast containing lamellar sheets and multiple small vesicles presumably
originating from the invagination and pinching off of the limiting membrane into the luminal space. Bar is 0.2 mm. Rh, rhoptry; DG, dense granule; Mi microneme.

process, rhoptries first form as spheroidal structures that maturing, and that they are formed by sequential fusion of
grow progressively by fusion of small vesicles around their post-Golgi vesicles (Jaikaria et al., 1993; Bannister et al.,
margins. Microscopic observation suggests that rhoptry 2000). Recent studies emphasise that Plasmodium rhoptry
proteins travel through the secretory pathway while proteins are selectively concentrated in distinctive lipid-rich
 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615
microdomains of the endoplasmic reticulum and Golgi
(Topolska et al., 2004). These microdomains first bud to
form non-pedunculated condensing vesicles and later
mature into rhoptries.
There are multiple lines of evidence suggesting that
rhoptries are analogous to lysosomal secretory organelles
(Ngo et al., 2004). This hypothesis is based on the arguments
that follow. In mammalian cells, secretory lysosomes release
their content upon stimulation using a unique molecular
machinery (reviewed in Blott and Griffiths, 2002). Similarly,
the discharge of the rhoptry content is regulated, although the
signal that triggers rhoptry secretion at the time of invasion
remains to be identified. Mature secretory lysosomes have
diverse types of structures, containing either dense cores or
multilamellar sheaths, which are morphologically compar-
able to the granular material accumulated inside the basal
portion of rhoptries. Both secretory lysosomes and rhoptries
have a low luminal pH and contain mature forms of
lysosomal acid proteases.
Furthermore, mammalian secretory lysosomes are pro-
ducts of both biosynthetic and endocytic pathways. Their
biogenesis initially involves small multivesicular bodies
(MVB), in which the inner vesicles will progressively
increase in size and number (reviewed in Blott and Griffiths,
2002). The formation of secretory lysosomes is dependent
on adaptor complexes and sorting of transmembrane cargo
containing cytoplasmic-tail tyrosine- and dileucine-target-
ing signals. The internal membrane structures stored in the
rhoptry lumen (Fig. 5(A) and (C)) are reminiscent of the
intraluminal vesicles found in MVB and secretory lyso-
somes (Fig. 5(D)), which are released into the extracellular
milieu as exosomes (Denzer et al., 2000). Indeed, upon host
cell invasion, such rhoptry-derived exosomal vesicles from
Toxoplasma (Hakansson et al., 2001) or Plasmodium
(Nichols et al., 1983; Stewart et al., 1986) contribute to
parasite lipids and proteins of the PVM.
In mammalian cells, MVB constitute common stations in
the biosynthetic and endocytic pathways and function as
spatial and temporal intermediates between early endosomes
and lysosomes (Denzer et al., 2000; Raposo and Marks,
2002). Their internal vesicles originate from inward
invaginations of the endosomal limiting membrane, enwrap-
ping cytosolic material (van Deurs et al., 1993). During this
process, selected membrane proteins are sequestered within
these vesicles, whereas others remain at the limiting
membrane. The sorting of some proteins to the internal
vesicles is clearly determined by their association with
specific lipid microdomains and/or recognition of sorting
signals of pre-proteins by membrane receptors. Apicomplexa
also utilise conserved endocytic proteins (e.g. rab proteins;
Chaturvedi et al., 1999; Robibaro et al., 2002; Stedman et al.,
2003; Quevillon et al., 2003) and coat proteins (e.g. adaptor-
complex 1; Ngo et al., 2003) at the trans-Golgi network
(TGN) and/or early endosomes to regulate protein trafficking
to parasite compartments, including rhoptries. The delivery
of pre-rhoptry proteins to mature rhoptries is mediated by
607
cytoplasmic-tail tyrosine- and dileucine-targeting signals
that bind to the tyrosine-binding pocket of the adaptor-
complex 1 (Hoppe et al., 2000; Ngo et al., 2003).
There is emerging evidence that multivesicular endo-
somes may intersect the rhoptry biogenesis pathway in
Apicomplexa (Fig. 6; Yang et al., 2004). The inner vesicles of
multivesicular precursors containing rhoptry proteins appear
to be delivered to mature rhoptries along a pathway involving
the protein-specific chaperone, vacuolar protein sorting 4
(VPS4; Babst et al., 1998; Beyer et al., 2003). Subsequently,
the MVB morphology disappears and large electron-lucent
vacuoles appear in the parasite cytoplasm. In yeast, the
members of the class E of the VPS family may promote the
formation of luminal membranes within MVB. Yeast mutant
strains lacking any of the class E genes accumulate an
exaggerated prevacuolar/late endosome compartment that
amasses large multilamellar cisternal structures and contains
both recycling TGN proteins and vacuolar proteins (Bishop
and Woodman, 2000). The same phenomenon can be
observed in mammalian cells deleted of genes homologous
to the class E (Fujita et al., 2003).
In an attempt to visualise multivesicular endosomes
containing rhoptry material in Plasmodium or Toxoplasma
as rhoptry precursors, parasite VPS4 orthologs containing
dominant-negative mutations have been expressed in both
parasites (Fig. 6; Yang et al., 2004). As observed in other
eukaryotic cells, the expression of the VPS4 mutant leads to
the accumulation of large swollen endosomal compart-
ments, where the mutant VPS4 gene product, rhoptry
proteins, endocytic markers (e.g. Rab5) as well as
cholesterol accumulate. This observation demonstrates
that VPS4 is a key regulatory protein in the membrane
dynamics of endosomes in Apicomplexa and provides
strong support that these parasites use multivesicular
endosomes as intermediates at the cross-road of the
endocytic and exocytic pathways for rhoptry formation.
2.3. Potential roles of rhoptry cholesterol in apicomplexan
cell biology
Electron and epifluorescence microscopy studies using
filipin, a cytomarker forming specific complexes with sterols,
demonstrate the presence of sterol-filipin complexes at the
apical pole of Toxoplasma (Fig. 5(B); Coppens and Joiner,
2003). Likewise, lipid analysis of rhoptry-enriched fractions
reveals a relatively high level of free cholesterol. The molar
cholesterol to phospholipid ratio in rhoptry organelles ranges
from 1.4 to 1.5 (Foussard et al., 1991; Coppens and Joiner,
2003). This stands in sharp contrast to the amounts of
cholesterol found in mammalian plasma membranes contain-
ing the majority of cell cholesterol, which range between 0.5
and 1.0 (reviewed in Mason et al., 2003).
Nevertheless, a variety of organised cholesterol-rich
domains exist in biological membranes, which are essential
for normal cell functions (summarised in Addadi et al.,
2003). For example, sphingolipids and cholesterol
608 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615

Fig. 6. Hypothetical model of rhoptry biogenesis in apicomplexan parasites. (A). Rhoptries are formed by contributions from both secretory and endocytic
pathways. Biogenesis of rhoptries is mediated by adaptor-based vesicular trafficking (AP-1), most closely related to the sorting mechanism for secretory
lysosomes. The cargo of rhoptry proteins from the trans-Golgi network (TGN) transits the endosomal pathway through a multivesicular body (MVB) en route to
the immature rhoptry. Vacuolar protein sorting 4 (VPS4) may act as a regulator of membrane dynamics in the rhoptry endosomal pathway and/or a mediator of
cholesterol transport. Cholesterol may be involved in MVB organisation and operate to sort and transport rhoptry proteins to the mature rhoptries. (B). Evidence
for multivesicular endosomes (MVE) intersecting the rhoptry biogenesis pathway is derived from observations that Toxoplasma expressing parasite negative-
dominant mutant of VPS4 exhibit arrested and enlarged MVE (adapted from Yang et al., 2004), where cholesterol, markers for early endosomes and rhoptry
proteins lacking targeting signals accumulate. AP-1, adaptor complex-1; Go, Golgi; N, nucleus; Mt, mitochondrion; Rh, rhoptry; A, apicoplast. Bar is 0.25 mm.

dynamically cluster to form rafts that move within the lipid
bilayer as distinct units; the cholesterol/phospholipid (C/PL)
molar ratio is 1.2 in lipid rafts. Another remarkable motif of
membrane cholesterol enrichment is observed in the plasma
membrane of eye lens fiber cells, in which the C/PL molar
ratio reaches values as high as 3.7 (Jacob et al., 2001). Lipid
organisation of the lens membranes consists of
the coexistence of distinct sterol-poor and -rich regions, in
which sterol monomers are arranged in a tail-to-tail
orientation (Mason et al., 2003). The C/PL molar ratio is
lowered down to 1.6 in the cataractous lens membranes
(Jacob et al., 2001), suggesting that the maintenance of
distinct cholesterol-enriched domains is necessary for the
normal function of lens fiber cells.
It can be assumed that cholesterol associated with
rhoptries is relevant for rhoptry function, and therefore,
 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615
may play a key role in the pathogenesis of Apicomplexa. So
far it has not been determined whether cholesterol is highly
concentrated at the level of the limiting membrane and/or is
distributed towards the intraluminal vesicles of rhoptries.
Likewise there is no evidence for a specific organisation of
cholesterol into discrete domains in rhoptry membranes.
Although the functional significance of rhoptry cholesterol
is still enigmatic, several functions for this lipid can be
proposed, based on recent data on rhoptry composition,
biogenesis and role in parasite invasion.
2.3.1. Role in rhoptry structure
It is well established that cholesterol is the appropriate
molecule to maintain a delicate balance between membrane
rigidity (e.g. to allow large cell volumes) and membrane
softness and fluidity (e.g. to allow membrane-embedded
proteins to function properly; Bretscher and Munro, 1993).
Cholesterol in amounts up to 3.0 mol% softens the lipid
bilayer, and at concentrations higher than 4.0 mol% it leads
to rigification, which is responsible for mechanical coher-
ence of eukaryotic plasma membranes (reviewed in
Rukmini et al., 2001). The plasma membrane, which
contains more than 85% of total membrane cholesterol, is
more rigid than any organellar membranes (Lancrajan et al.,
2001). Acute cholesterol depletion causes accumulation of
elastic flat-coated membranes and a corresponding decrease
in stiff deep-coated pits (Subtil et al., 1999). In nervous
tissues, the formation of cholesterol-rich domains induced
by a myristoylated neuronal protein, transforms a smooth
cell surface into convoluted structures (Maekawa et al.,
1999; Epand et al., 2001). The extraordinarily high
cholesterol content within the plasma membrane of eye
lens fiber cells contributes to lens-cell rigidity and structural
order. Indeed, the lens fiber cells are elongated, ordered in
compact concentric lamellae, which minimises the extra-
cellular space, a potential light scattering region detrimental
for lens transparency (Mason et al., 2003). In Apicomplexa,
cholesterol may be present in large excess within the
limiting membrane of rhoptries. Serving as a building
material, this lipid may play a structural role that contributes
to the well-defined morphology of rhoptries, entailing the
stiffness of the organelle.
2.3.2. Role in rhoptry protein sorting and MVB formation
In addition to proteinaceous coats, lipids as the major
components of membranes play active roles in regulating
membrane dynamics (reviewed in Gu et al., 2001; Miwako
et al., 2001). Both, cholesterolprotein interactions and
cholesterol-rich microdomains seem to be necessary in
order to induce vesicle curvature and to assemble vesicle-
specific proteins and lipids, respectively (Rukmini et al.,
2001; Huttner and Zimmerberg, 2001). In mammalian cells,
internal vesicles of MVB are also enriched in cholesterol
(Mobius et al., 2003). This lipid is clearly implicated in
MVB reorganisation that drives the sorting/transport of
materials destined for the Golgi out of the pathways towards
609
lysosomes (Miwako et al., 2001). It is conceivable that in
Apicomplexa, cholesterol might modulate the MVB traffic
connected to mature rhoptries.
2.3.3. Role in formation of signaling platforms
In general, rafts represent a type of membrane domain
wherein lipids dynamically associate with each other, to
form platforms important for membrane protein sorting and
assembly of signaling complexes (Simons and Toomre,
2000). In mammalian cells, lipid rafts have been identified
in different cell compartments (reviewed in Helms and
Zurzolo, 2004). These include the endoplasmic reticulum
(role in stabilisation of the association of GPI-proteins with
the endoplasmic reticulum membranes), early Golgi cister-
nae (role in maintenance of Golgi structure and function),
the TGN (role in sorting of apical cargo proteins through
specific interaction with apical targeting signals), the
plasma membrane (after internalisation and fusion with
the early endosomal system, role in raft-associated protein
sorting towards the Golgi via late endosomes or the plasma
membrane via the recycling compartment), early endo-
somes (role in protein sorting towards the Golgi via late
endosomes or recycling endosomes), internal vesicles of
late endosomal MVB (role in protein sorting towards
lysosomes or the plasma membrane), and consequently
exosomes (participation in vesicle formation and structure;
de Gassart et al., 2003).
Cholesterol plays a central role in lipid raft organisation
(Schroeder et al., 1998). Cellular cholesterol depletion or
abnormal cholesterol accumulation in lipid raft-associated
compartments results in mistrafficking of raft components
(Pagano, 2003). In one of several models concerning the
nature of rafts (reviewed in Simons and Ikonen, 1997; Pralle
et al., 2000; Edidin, 2003; Zurzolo et al., 2003), it has been
suggested that rafts are constructed of lipid shells. This
model is based on interactions between lipids and proteins
(e.g. caveolins, flotillins, stomatins). These proteins pene-
trate the lipid bilayer and allow intimate contact with
membrane lipids, which creates and stabilises liquid ordered
phases in biological membranes (reviewed in Helms and
Zurzolo, 2004).
The existence of endoplasmic reticulum- and Golgi-
localised rafts has been suggested in Plasmodium (Topolska
et al., 2004). These lipid raft analogs contain a GPI-
anchored protein, the rhoptry-associated membrane antigen
(RAMA), which can form complexes with rhoptry
proteins. Domains within parasite Golgi membranes
enriched in RAMA bud to form vesicles that later generate
proto-rhoptries. More interestingly, the nascent PVM of
Plasmodium contains non-caveolar cholesterol-rich deter-
gent-resistant membrane (DRM) rafts from both, parasite
and host cell origin (Haldar et al., 2001; Murphy et al.,
2004). A Plasmodium stomatin secreted from rhoptries and
an erythrocyte stomatin-like protein-2 are strong candidates
for the assembly of these lipid rafts (Hiller et al., 2003). In
addition, DRM rafts are also present on the parasite plasma
610 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615
membrane, suggesting that establishment of malarial
erythrocyte infection is linked to multiple DRM rafts. The
association of stomatin with rhoptry raises the possibility
that lipid rafts are present on the membranes of inner
vesicles inside rhoptries before incorporation into the PVM.
Wherever its localisation, stomatin in Plasmodium and in
Toxoplasma where it is also present in the genome, may
promote nucleation or stabilisation of raft domains in
membranes of Apicomplexa, as reported for stomatin in
various species.
2.3.4. Role in PVM formation
Studies on Plasmodium invasion suggest that the
vacuolar membrane is derived mostly from lipids of
rhoptries (Dluzewski et al., 1995; Bannister and Mitchell,
1989). In contrast, Ward et al. (1993) suggest that the PVM
comes mainly from the erythrocyte bilayer. Upon Toxo-
plasma entry, the plasma membrane of the host cell
contributes the bulk of the lipids and selected proteins that
are required for the formation of the PVM (Suss-Toby et al.,
1996). Thereafter, Toxoplasma rhoptry-derived secretory
vesicles may supply critical parasite proteins and lipids for
the nascent formation of the hybrid PVM (Hakansson et al.,
2001). The PVM surrounding Toxoplasma contains choles-
terol at the time of invasion (Fig. 5(B); Coppens and Joiner,
2003). It has been speculated that rhoptry cholesterol is
essential for the formation of the nascent PVM. When
rhoptry cholesterol is depleted, however, parasite invasion is
not impaired. In contrast, depletion of cholesterol from the
host cell plasma membrane blocks parasite organelle
exocytosis and invasion (Coppens and Joiner, 2003). The
available data are not in favour of a substantial role of
rhoptry cholesterol as extracellular effector at the time of
invasion. Nevertheless, cholesterol molecules may be
discharged into the PVM throughout the parasites intra-
cellular lifecycle. This process could endow the PVM with
properties favourable for parasite survival by giving it a lipid
composition different from that of the host cell membranes.
2.3.5. Role in rhoptry exocytosis
Recent studies suggest that cholesterol and sphingolipid-
rich microdomains in the plasma membrane play an
essential role in regulated exocytosis pathways (summarised
in Salaun et al., 2004). The association of soluble
N-ethylmaleimide-sensitive factor-attachment protein
receptor (SNARE) mediators of membrane fusion with
lipid rafts act to concentrate the SNARE regulatory proteins,
such as the N-ethylmaleimide-sensitive factor attachment
protein (SNAP) at defined sites of the plasma membrane.
Cholesterol depletion inhibits regulated exocytosis,
suggesting that SNARE-dependent exocytosis occurs in
cholesterol-rich domains in the plasma membrane.
The soluble N-ethylmaleimide-sensitive factor (NSF)/
SNAP/SNARE machinery is present and functional in
Apicomplexa and has been shown to participate in the
discharge of secretory organelles in Toxoplasma, as shown
for dense granules (Chaturvedi et al., 1999) and in the export
of vesicular structures into the erythrocyte infected by
Plasmodium (Hayashi et al., 2001). A direct role of
cholesterol in the release of rhoptry proteins has been
investigated using parasites depleted in rhoptry cholesterol,
but data show that rhoptry discharge seems to be preserved
under such conditions (Coppens and Joiner, 2003). This
may suggest that rhoptry secretion does not rely on a direct
fusion mechanism between the rhoptry membrane and the
plasma membrane. The orientation of rhoptries with their
peduncle pointed toward the top suggests that rhoptries have
a stable connection with the apical structures, or even a
permanent opening with the outside, regulated by unknown
cholesterol-independent mechanisms.
2.3.6. Role in modulation of rhoptry protein functions
Because cholesterol affects membrane properties, it can
also manipulate the behavior and functions of proteins
residing in membranes. In mammalian cells, a variety of
integral membrane proteins, including ion channels, mem-
brane receptors and enzymes, are sensitive to physical
changes in the surrounding lipid bilayer (OConnell et al.,
2004; Harder, 2004). Some proteins can bind cholesterol
directly via a sterol-sensing domain; as a result, they become
either activated or inactivated (Kuwabara and Labouesse,
2002). Alternatively, it has been shown that the ordered
arrangement of membrane cholesterol in lens fiber cells
interfere with the ability of extrinsic cataractogenic proteins
to aggregate at the cell surface (Mason et al., 2003).
The unique environment of the rhoptry contents includ-
ing both protein and lipid components, which assemble to
form membrane-like structures, leads to the assumption that
specific biomolecular interactions between proteins and
cholesterol must be prevailing in these organelles. The
identification of rhoptry proteins that can bind to choles-
terol, resulting in their functional activation or inhibition,
will shed some light on the involvement of rhoptry
cholesterol in pathogenesis of apicomplexan parasites.
3. Concluding remarks
On one hand, Toxoplasma and Plasmodium share with
mammalian cells the same typical organelles and conserved
metabolic pathways found in any eukaryotic cells. How-
ever, several studies have revealed that these parasites often
employ them in unconventional ways. The presence of lipid
bodies in the cytoplasm of apicomplexan parasites is to
some extent not surprising, considering the important role of
neutral lipids in membrane lipid metabolism reported for
many organisms. Nonetheless, further research on lipid
body composition, organisation, function and biogenesis in
Apicomplexa will identify novel and unexpected features.
For example, Toxoplasma and Plasmodium genome data-
bases do not reveal any sequences homologous to lipid-
body-associated proteins found in mammalian cells, plants
 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615
Table 1
Questions for future research
On lipid bodies
What are the molecular mechanisms for host cholesterol transport to the
parasitophorous vacuole of Toxoplasma?
What are the carriers involved in host cholesterol translocation across
the Toxoplasma plasma membrane and trafficking between parasite
organelles?
How do nascent lipid bodies bud off from parasite membranes? Is it
from the endoplasmic reticulum?
How is the mature size of lipid bodies determined?
What are the triggers of cholesteryl esters and triacylglycerol
remobilisation from lipid bodies?
What is the nature of the parasite specific lipid body-associated
proteins?
Is the neutral lipid biosynthesis coordinated with the synthesis of lipid
body-associated proteins?
Do the cytosolic lipid bodies exchange their lipid cargo with
organelles?
Is lipid storage in lipid bodies essential or dispensable for Apicomplexa
development?
What are the common themes in lipid body biogenesis between higher
and lower eukaryotes?
On rhoptries
What is the mechanism underlying cholesterol tropism and
concentration into rhoptries?
How can Apicomplexa regulate the cholesterol content associated with
rhoptries?
Is cholesterol delivery to the rhoptries coordinated with the synthesis of
specific rhoptry proteins?
What are the potential rhoptry proteins that can bind cholesterol and
what is the nature of their physical interactions with cholesterol?
What is the fine organisation of membrane cholesterol in rhoptries?
What is the physiological relevance of cholesterol accumulation in
rhoptries?
If secreted into the parasite environment, what is the fate of rhoptry
cholesterol?
Do rhoptries play a role in cholesterol exchange between parasite
organelles?
or yeast, although such regulatory molecules are likely
present at the surface of their lipid bodies. On the other
hand, Toxoplasma and Plasmodium have unusual organelles
among eukaryotes, exemplified by rhoptries exclusively
present in the Apicomplexa phylum. Rhoptries have begun
to emerge from relative obscurity with the recognition that
they are complex proteolipid organelles playing dynamic
roles in parasite invasion, and perhaps in cholesterol
metabolism. From the cell biologys point of view, the
original features provided by the study of rhoptries may
offer a unique perspective to elucidate basic concepts in
mammalian cell biology, such as the formation of MVB,
exosomal secretion, and cholesterol homeostasis. From a
therapeutical point of view, new information on lipid bodies
or rhoptries in Apicomplexa could pinpoint potential drug
targets exploitable to cure infection caused by these deadly
parasites. Although current studies on lipid bodies and
rhoptries in apicomplexan parasites leave innumerable
questions (see Table 1), the increasing sophistication of
molecular and biochemical techniques and a greater
611
awareness of comparative studies between higher and
lower eukaryotes makes us optimistic that many of these
questions will be soon answered.
Acknowledgements
We are grateful to members of our laboratory, labora-
tories of Keith Joiner (Yale University, CT) and Vern
Carruthers (Johns Hopkins University, MD) as well as to our
collaborators who have been studying the fascinating
aspects of neutral lipid metabolism in apicomplexan
parasites. This work was supported by a grant from the
American Heart Association (SDG 0230079N) to IC.
References
Addadi, L., Geva, M., Kruth, H.S., 2003. Structural information about
organized cholesterol domains from specific antibody recognition.
Biochim. Biophys. Acta 1610, 208216.
Akompong, T., Kadekoppala, M., Harrison, T., Oksman, A.,
Goldberg, D.E., Fujioka, H., Samuel, B.U., Sullivan, D., Haldar, K.,
2002. Trans-expression of a Plasmodium falciparum histidine-rich
protein II (HRPII) reveals sorting of soluble proteins in the periphery of
the host erythrocyte and disrupts transport to the malarial food vacuole.
J. Biol. Chem. 277, 2892328933.
Allred, D.R., Al-Khedery, B., 2004. Antigenic variation and cytoadhesion
in Babesia bovis and Plasmodium falciparum: different logics achieve
the same goal. Mol. Biochem. Parasitol. 134, 2735.
Athenstaedt, K., Zweytick, D., Jandrositz, A., Kohlwein, S.D., Daum, G.,
1999. Identification and characterisation of major lipid particle proteins
of the yeast Saccharomyces cerevisiae. J. Bacteriol. 181, 64416448.
Babst, M., Wendland, B., Estepa, E.J., Emr, S.D., 1998. The Vps4p AAA
ATPase regulates membrane association of a Vps protein complex
required for normal endosome function. Eur. Mol. Biol. Org. J. 17,
29822993.
Bannister, L.H., Mitchell, G.H., Butcher, G.A., Dennis, E.D., 1986.
Lamellar membranes associated with rhoptries in erythrocytic
merozoites of Plasmodium knowlesi: a clue to the mechanism of
invasion. Parasitology 92, 291303.
Bannister, L.H., Mitchell, G.H., 1989. The fine structure of secretion by
Plasmodium knowlesi merozoites during red cell invasion. J. Protozool.
36, 362367.
Bannister, L.H., Hopkins, J.M., Fowler, R.E., Krishna, S., Mitchell, G.H.,
2000. Ultrastructure of rhoptry development in Plasmodium falciparum
erythrocytic schizonts. Parasitology 121, 273287.
Batetta, B., Mulas, M.F., Sanna, F., Putzolu, M., Bonatesta, R.R., Gasperi-
Campani, A., Roncuzzi, L., Baiocchi, D., Dessi, S., 2003. Role of
cholesterol ester pathway in the control of cell cycle in human aortic
smooth muscle cells. Fed. Am. Soc. Exp. Biol. J. 17, 746748.
Beach, D.H., Sherman, I.W., Holz Jr.., G.G., 1977. Lipids of Plasmodium
lophurae, and of erythrocytes and plasma of normal and P. lophurae-
infected Pekin ducklings. J. Parasitol. 63, 6275.
Beyer, A., Scheuring, S., Muller, S., Mincheva, A., Lichter, P., Kohrer, K.,
2003. Comparative sequence and expression analyses of four
mammalian VPS4 genes. Gene 305, 4759.
Bishop, N., Woodman, P., 2000. ATPase-defective mammalian VPS4
localises to aberrant endosomes and impairs cholesterol trafficking.
Mol. Biol. Cell 11, 227239.
Blackman, M.J., Bannister, L.H., 2001. Apical organelles of Apicomplexa:
biology and isolation by subcellular fractionation. Mol. Biochem.
Parasitol. 117, 1125.
612 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615
Blott, E.J., Griffiths, G.M., 2002. Secretory lysosomes. Nature Rev. Mol.
Cell Biol. 3, 122131.
Bouvier-Nave, P., Benveniste, P., Oelkers, P., Sturley, S.L., Schaller, H.,
2000. Expression in yeast and tobacco of plant cDNAs encoding acyl
CoA:diacylglycerol acyltransferase. Eur. J. Biochem. 267, 8596.
Bozdech, Z., Llinas, M., Pulliam, B.L., Wong, E.D., Zhu, J., DeRisi, J.L.,
2003. The transcriptome of the intraerythrocytic developmental cycle of
Plasmodium falciparum,. PLoS Biol. 1, 85100.
Bretscher, M.S., Munro, S., 1993. Cholesterol and the golgi apparatus.
Science 261, 12801281.
Buhman, K.F., Accad, M., Farese, R.V., 2000. Mammalian acyl-
3CoA:cholesterol acyltransferases. Biochim. Biophys. Acta 1529,
142154.
Cases, S., Smith, S.J., Zheng, Y.W., Myers, H.M., Lear, S.R., Sande, E.,
Novak, S., Collins, C., Welch, C.B., Lusis, A.J., Erickson, S.K., Farese
Jr., R.V., 1998. Identification of a gene encoding an acyl CoA:dia-
cylglycerol acyltransferase, a key enzyme in triacylglycerol synthesis.
Proc. Natl Acad. Sci. USA 95, 1301813023.
Cases, S., Stone, S.J., Zhou, P., Yen, E., Tow, B., Lardizabal, K.D.,
Voelker, T., Farese Jr.., R.V., 2001. Cloning of DGAT2, a second
mammalian diacylglycerol acyltransferase, and related family mem-
bers. J. Biol. Chem. 276, 3887038876.
Chang, C.C., Huh, H.Y., Cadigan, K.M., Chang, T.Y., 1993. Molecular
cloning and functional expression of human acyl-coenzyme A:choles-
terol acyltransferase cDNA in mutant Chinese hamster ovary cells.
J. Biol. Chem. 268, 2074720755.
Chang, T.Y., Chang, C.C., Cheng, D., 1997. Acyl-coenzyme A:cholesterol
acyltransferase. Annu. Rev. Biochem. 66, 613638.
Chang, T.Y., Chang, C.C., Lin, S., Yu, C., Li, B.L., Miyazaki, A., 2001.
Roles of acyl-coenzyme A:cholesterol acyltransferase-1 and -2. Curr.
Opin. Lipidol. 12, 289296.
Charron, A.J., Sibley, L.D., 2002. Host cells: mobilizable lipid resources
for the intracellular parasite Toxoplasma gondii. J. Cell Sci. 115, 3049
3059.
Chaturvedi, S., Qi, H., Coleman, D., Rodriguez, A., Hanson, P.I.,
Striepen, B., Roos, D.S., Joiner, K.A., 1999. Constitutive calcium-
independent release of Toxoplasma gondii dense granules occurs
through the NSF/SNAP/SNARE/Rab machinery. J. Biol. Chem. 274,
24242431.
Chen, H.C., Farese Jr.., R.V., 2000. DGAT and triglyceride synthesis:
a new target for obesity treatment?. Trends Cardiovasc. Med. 10,
188192.
Chu, F.E., Soudant, P., Volety, A.K., Huang, Y., 2000. Uptake and
interconversion of fluorescent lipid analogs in the protozoan parasite,
perkinsus marinus, of the oyster, crassostrea virginica. Exp. Parasitol.
95, 240251.
Cole, N.B., Murphy, D.D., Grider, T., Rueter, S., Brasaemle, D.,
Nussbaum, R.L., 2002. Lipid droplet binding and oligomerization
properties of the parkinsons disease protein a-synuclein. J. Biol. Chem.
277, 63446352.
Coppens, I., Joiner, K.A., 2003. Host but not parasite cholesterol controls
Toxoplasma cell entry by modulating organelle discharge. Mol. Biol.
Cell 14, 38043820.
Coppens, I., Sinai, A.P., Joiner, K.A., 2000. Toxoplasma gondii exploits
host LDL receptor-mediated endocytosis for cholesterol acquisition.
J. Cell Biol. 149, 167180.
Dahlqvist, A., Stahl, U., Lenman, M., Banas, A., Lee, M., Sandager, L.,
Ronne, H., Stymne, S., 2000. Phospholipid:diacylglycerol acyltransfer-
ase: an enzyme that catalyzes the acyl-CoA-independent formation
of triacylglycerol in yeast and plants. Proc. Natl Acad. Sci. USA 97,
64876492.
de Gassart, A., Geminard, C., Fevrier, B., Raposo, G., Vidal, M., 2003.
Lipid raft-associated protein sorting in exosomes. Blood 102,
43364344.
Denzer, K., Kleijmeer, M.J., Heijnen, H.F., Stoorvogel, W., Geuze, H.J.,
2000. Exosome: from internal vesicle of the multivesicular body to
intercellular signaling device. J. Cell Sci. 19, 33653374.
Dluzewski, A.R., Zicha, D., Dunn, G.A., Gratzer, W.B., 1995. Origins of
the parasitophorous vacuole membrane of the malaria parasite: surface
area of the parasitized red cell. Eur. J. Cell. Biol. 68, 446449.
Dubey, J.P., Lindsay, D.S., Speer, C.A., 1998. Structures of Toxoplasma
gondii tachyzoites, bradyzoites, and sporozoites and biology and
development of tissue cysts. Clin. Microbiol. Rev. 11, 267299.
Edidin, M., 2003. The state of lipid rafts: from model membranes to cells.
Annu. Rev. Biophys. Biomol. Struct. 32, 257283.
Epand, R.M., Maekawa, S., Yip, C.M., Epand, R.F., 2001. Protein-induced
formation of cholesterol-rich domains. Biochemistry 40, 1051410521.
Escalante, A.A., Ayala, F.J., 1995. Evolutionary origin of Plasmodium and
other Apicomplexa based on rRNA genes. Proc. Natl Acad. Sci. USA
92, 57935797.
Farese Jr., R.V., Cases, S., Smith, S.J., 2000. Triglyceride synthesis:
insights from the cloning of diacylglycerol acyltransferase. Curr. Opin.
Lipidol. 11, 229234.
Figueiredo, R.C., Steindel, M., Soares, M.J., 1994. The reservosomes of
epimastigote forms of Trypanosoma cruzi: occurrence during in vitro
cultivation. Parasitol. Res. 80, 517522.
Fitch, C.D., Cai, G.Z., Chen, Y.F., Shoemaker, J.D., 1999. Involvement of
lipids in ferriprotoporphyrin IX polymerization in malaria. Biochim.
Biophys. Acta 1454, 3137.
Foussard, F., Leriche, M.A., Dubremetz, J.F., 1991. Characterization of
the lipid content of Toxoplasma gondii rhoptries. Parasitology 102,
367370.
Francis, S.E., Sullivan Jr., D.J., Goldberg, D.E., 1997. Hemoglobin
metabolism in the malaria parasite Plasmodium falciparum. Annu.
Rev. Microbiol. 51, 97123.
Fujimoto, Y., Itabe, H., Sakai, J., Makita, M., Noda, J., Mori, M.,
Higashi, Y., Kojima, S., Takano, T., 2004. Identification of major
proteins in the lipid droplet-enriched fraction isolated from the human
hepatocyte cell line HuH7. Biochim. Biophys. Acta 1644, 4759.
Fujita, H., Yamanaka, M., Imamura, K., Tanaka, Y., Nara, A.,
Yoshimori, T., Yokota, S., Himeno, M., 2003. A dominant negative
form of the AAA ATPase SKD1/VPS4 impairs membrane trafficking
out of endosomal/lysosomal compartments: class E vps phenotype in
mammalian cells. J. Cell Sci. 116, 401414.
Furlong, S.T., 1989. Sterols of parasitic protozoa and helminths. Exp.
Parasitol. 68, 482485.
Gardner, M.J., Shallom, S.J., Carlton, J.M., Salzberg, S.L., Nene, V.,
Shoaibi, A., Ciecko, A., Lynn, J., Rizzo, M., Weaver, B., Jarrahi, B.,
Brenner, M., Parvizi, B., Tallon, L., Moazzez, A., Granger, D., Fujii, C.,
Hansen, C., Pederson, J., Feldblyum, T., Peterson, J., Suh, B., Angiuoli,
S., Pertea, M., Allen, J., Selengut, J., White, O., Cummings, L.M.,
Smith, H.O., Adams, M.D., Venter, J.C., Carucci, D.J., Hoffman, S.L.,
Fraser, C.M., 2002. Sequence of Plasmodium falciparum chromosomes
2, 10, 11 and 14. Nature 419, 531534.
Greenspan, P., Mayer, E.P., Fowler, S.D., 1985. Nile Red: a
selective fluorescent stain for intracellular lipid droplets. J. Cell Biol.
100, 965973.
Gu, F., Crump, C.M., Thomas, G., 2001. Trans-golgi network sorting. Cell.
Mol. Life Sci. 58, 10671084.
Guo, Z., Cromley, D., Billheimer, J.T., Sturley, S.L., 2001. Identification of
potential substrate-binding sites in yeast and human acyl-CoA sterol
acyltransferases by mutagenesis of conserved sequences. J. Lipid Res.
42, 12821291.
Hakansson, S., Charron, A.J., Sibley, L.D., 2001. Toxoplasma evacuoles: a
two-step process of secretion and fusion forms the parasitophorous
vacuole. Eur. Mol. Biol. Org. J. 20, 31323144.
Haldar, K., Samuel, B.U., Mohandas, N., Harrison, T., Hiller, N.L., 2001.
Transport mechanisms in Plasmodium-infected erythrocytes: lipid rafts
and a tubovesicular network. Int. J. Parasitol. 31, 13931401.
Harder, T., 2004. Lipid raft domains and protein networks in T-cell receptor
signal transduction. Curr. Opin. Immunol. 16, 353359.
Hayashi, M., Taniguchi, S., Ishizuka, Y., Kim, H.S., Wataya, Y.,
Yamamoto, A., Moriyama, Y., 2001. A homologue of N-ethylmalei-
mide-sensitive factor in the malaria parasite Plasmodium falciparum is
 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615
exported and localised in vesicular structures in the cytoplasm of
infected erythrocytes in the brefeldin A-sensitive pathway. J. Biol.
Chem. 276, 1524915255.
Heid, H.W., Moll, R., Schwetlick, I., Rackwitz, H.R., Keenan, T.W., 1998.
Adipophilin is a specific marker of lipid accumulation in diverse cell
types and diseases. Cell Tissue Res. 294, 309321.
Helms, J.B., Zurzolo, C., 2004. Lipids as targeting signals: lipid rafts and
intracellular trafficking. Traffic 5, 247254.
Hiller, N.L., Akompong, T., Morrow, J.S., Holder, A.A., Haldar, K., 2003.
Identification of a stomatin orthologue in vacuoles induced in human
erythrocytes by malaria parasites. J. Biol. Chem. 278, 4841348421.
Hofmann, K., 2000. A superfamily of membrane-bound o-acyltransferases
with implications for wnt signaling. Trends Biochem. Sci. 25, 111112.
Holz Jr.., G.G., 1977. Lipids and the malarial parasite. World Health Organ
55, 237248 (Bull).
Hoppe, H.C., Ngo, H.M., Yang, M., Joiner, K.A., 2000. Targeting to
rhoptry organelles of Toxoplasma gondii involves evolutionarily
conserved mechanisms. Nature Cell Biol. 2, 449456.
Huttner, W.B., Zimmerberg, J., 2001. Implications of lipid microdomains
for membrane curvature, budding and fission. Curr. Opin. Cell Biol. 13,
478484.
Jackson, K.E., Klonis, N., Ferguson, D.J., Adisa, A., Dogovski, C.,
Tilley, L., 2004. Food vacuole-associated lipid bodies and hetero-
geneous lipid environments in the malaria parasite, Plasmodium
falciparum. Mol. Microbiol. 54, 109122.
Jacob, R.F., Cenedella, R.J., Mason, R.P., 2001. Evidence for distinct
cholesterol domains in fiber cell membranes from cataractous human
lenses. J. Biol. Chem. 276, 1357313578.
Jaikaria, N.S., Rozario, C., Ridley, R.G., Perkins, M.E., 1993. Biogenesis of
rhoptry organelles in Plasmodium falciparum. Mol. Biochem. Parasitol.
57, 269279.
Kalscheuer, R., Steinbuchel, A., 2003. A novel bifunctional wax ester
synthase/acyl-CoA:diacylglycerol acyltransferase mediates wax ester
and triacylglycerol biosynthesis in Acinetobacter calcoaceticus ADP1.
J. Biol. Chem. 278, 80758082.
Kaup, M.T., Froese, C.D., Thompson, J.E., 2002. A role for diacylglycerol
acyltransferase during leaf senescence. Plant Physiol. 129, 16161626.
Kellner-Weibel, G., Jerome, W.G., Small, D.M., Warner, G.J.,
Stoltenborg, J.K., Kearney, M.A., Corjay, M.H., Phillips, M.C.,
Rothblat, G.H., 1998. Effects of intracellular free cholesterol accumu-
lation on macrophage viability: a model for foam cell death.
Arterioscler. Thromb. Vasc. Biol. 18, 423431.
Klut, M.E., Bisalputra, T., Antia, N.J., 1988. The use of fluorochromes in
the cytochemical characterization of some phytoflagellates. Histochem.
J. 20, 3540.
Kuwabara, P.E., Labouesse, M., 2002. The sterol-sensing domain: multiple
families, a unique role?. Trends Genet. 18, 193201.
Lancrajan, I., Diehl, H.A., Socaciu, C., Engelke, M., Zorn-Kruppa, M.,
2001. Carotenoid incorporation into natural membranes from
artificial carriers: liposomes and beta-cyclodextrins. Chem. Phys.
Lipids 112, 110.
Landers, S.C., 2002. The fine structure of the gamont of Pterospora
floridiensis (Apicomplexa: Eugregarinida). J. Eukaryot. Microbiol. 49,
220226.
Lardizabal, K., Mai, J.T., Wagner, N.W., Wyrick, A., Voelker, T.,
Hawkins, D.J., 2001. DGAT2 is a new diacylglycerol acyltransferase
gene family: purification, cloning, and expression in insect cells of two
polypeptides from Mortierella ramanniana with diacylglycerol acyl-
transferase activity. J. Biol. Chem. 276, 3886238869.
Lehner, R., Kuksis, A., 1996. Biosynthesis of triacylglycerols. Prog. Lipid
Res. 35, 169201.
Lindsay, D.S., Blagburn, B.L., Toivio-Kinnucan, M., 1991. Ultrastructure
of developing Isospora suis in cultured cells. Am. J. Vet. Res. 52,
471473.
Lindsay, D.S., Speer, C.A., Toivio-Kinnucan, M.A., Dubey, J.P.,
Blagburn, B.L., 1993. Use of infected cultured cells to
613
compare ultrastructural features of Neospora caninum from dogs and
Toxoplasma gondii. Am. J. Vet. Res. 54, 103106.
Ling, I.T., Florens, L., Dluzewski, A.R., Kaneko, O., Grainger, M., Yim
Lim, B.Y., Tsuboi, T., Hopkins, J.M., Johnson, J.R., Torii, M.,
Bannister, L.H., Yates 3rd., J.R., Holder, A.A., Mattei, D., 2004. The
Plasmodium falciparum clag9 gene encodes a rhoptry protein that is
transferred to the host erythrocyte upon invasion. Mol. Microbiol. 52,
107118.
Litvak, V., Shaul, Y.D., Shulewitz, M., Amarilio, R., Carmon, S., Lev, S.,
2002. Targeting of Nir2 to lipid droplets is regulated by a specific
threonine residue within its PI-transfer domain. Curr. Biol. 12, 1513
1518.
Liu, P., Ying, Y., Zhao, Y., Mundy, D.I., Zhu, M., Anderson, R.G., 2004.
Chinese hamster ovary K2 cell lipid droplets appear to be metabolic
organelles involved in membrane traffic. J. Biol. Chem. 279,
37873792.
Lum, P.Y., Wright, R., 1995. Degradation of HMG-CoA reductase-induced
membranes in the fission yeast, Schizosaccharomyces pombe. J. Cell
Biol. 131, 8194.
Maekawa, S., Sato, C., Kitajima, C.K., Funatsu, N., Kumanogoh, H.,
Sokawa, Y., 1999. Cholesterol-dependent localisation of NAP-22 on
a neuronal membrane microdomain (raft). J. Biol. Chem. 274,
2136921374.
Mason, P.R., Tulenko, T.N., Jacob, R.F., 2003. Direct evidence for
cholesterol crystalline domains in biological membranes: role in human
pathology. Biochim. Biophysica. Acta 1610, 198207.
Mitamura, T., Palacpac, N.M., 2003. Lipid metabolism in Plasmodium
falciparum-infected erythrocytes: possible new targets for malaria
chemotherapy. Microbes Infect. 5, 545552.
Mitamura, T., Hanada, K., Ko-Mitamura, E.P., Nishijima, M., Horii, T.,
2000. Serum factors governing intraerythrocytic development and
cell cycle progression of Plasmodium falciparum. Parasitol. Int. 49,
219229.
Miwako, I., Yamamoto, A., Kitamura, T., Nagayama, K., Ohashi, M., 2001.
Cholesterol requirement for cation-independent mannose 6-phosphate
receptor exit from multivesicular late endosomes to the Golgi. J. Cell
Sci. 114, 17651776.
Mobius, W., van Donselaar, E., Ohno-Iwashita, Y., Shimada, Y.,
Heijnen, H.F., Slot, J.W., Geuze, H.J., 2003. Recycling compartments
and the internal vesicles of multivesicular bodies harbor most of the
cholesterol found in the endocytic pathway. Traffic 4, 222231.
Murphy, D.J., 1990. Storage lipid bodies in plants and other organisms.
Prog. Lipid Res. 29, 299324.
Murphy, D.J., 2001. The biogenesis and functions of lipid bodies in
animals, plants and microorganisms. Prog. Lipid Res. 40, 325438.
Murphy, D.J., Vance, J., 1999. Mechanisms of lipid-body formation.
Trends Biochem. Sci. 24, 109115.
Murphy, S.C., Samuel, B.U., Harrison, T., Speicher, K.D., Speicher, D.W.,
Reid, M.E., Prohaska, R., Low, P.S., Tanner, M.J., Mohandas, N.,
Haldar, K., 2004. Erythrocyte detergent-resistant membrane proteins:
their characterization and selective uptake during malarial infection.
Blood 103, 19201928.
Mutka, A.L., Lusa, S., Linder, M.D., Jokitalo, E., Kopra, O.,
Jauhiainen, M., Ikonen, E., 2004. Secretion of sterols and the NPC2
protein from primary astrocytes. J. Biol. Chem. 279, 4865448662.
Naik, R.S., Davidson, E.A., Gowda, D.C., 2000. Developmental stage-
specific biosynthesis of glycosylphosphatidylinositol anchors in
intraerythrocytic Plasmodium falciparum and its inhibition in a novel
manner by mannosamine. J. Biol. Chem. 275, 2450624511.
Nawabi, P., Lykidis, A., Ji, D., Haldar, K., 2003. Neutral-lipid analysis
reveals elevation of acylglycerols and lack of cholesterol esters
in Plasmodium falciparum-infected erythrocytes. Eukaryot. Cell 2,
11281131.
Ngo, H.M., Yang, M., Pypaert, M., Hoppe, H.C., Joiner, K.A., 2003.
Biogenesis of secretory organelles in an apicomplexan parasite is
dependent on the AP-1 adaptin. J. Biol. Chem. 278, 53435352.
614 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615
Ngo, H.M., Yang, M., Joiner, K.A., 2004. Are rhoptries in Apicomplexan
parasites secretory granules or secretory lysosomal granules? Mol.
Microbiol. 52, 15311541.
Nichols, B.A., Chiappino, M.L., OConnor, G.R., 1983. Secretion from the
rhoptries of Toxoplasma gondii during host-cell invasion. J. Ultrastruct.
Res. 83, 8598.
Nishikawa, Y., Quittnat, F., Stedman, T.T., Voelker, D.R., Choi, J.Y., Zahn,
M., Yang, M., Joiner, K.A., Coppens, I, in press. Host cell lipids control
cholesteryl ester synthesis and storage in intracellular Toxoplasma.
Cell. Microbiol, in press.
OConnell, K.M., Martens, J.R., Tamkun, M.M., 2004. Localization of ion
channels to lipid raft domains within the cardiovascular system. Trends
Cardiovasc. Med. 14, 3742.
Oelkers, P., Behari, A., Cromley, D., Billheimer, J.T., Sturley, S.L., 1998.
Characterization of two human genes encoding acyl coenzyme
A:cholesterol acyltransferase-related enzymes. J. Biol. Chem. 273,
2676526771.
Oelkers, P., Tinkelenberg, A., Erdeniz, N., Cromley, D., Billheimer, J.T.,
Sturley, S.L., 2000. A lecithin cholesterol acyltransferase-like gene
mediates diacylglycerol esterification in yeast. J. Biol. Chem. 275,
1560915612.
Ohlrogge, J.B., Jaworski, J.G., 1997. Regulation of fatty acid synthesis.
Annu. Rev. Plant. Physiol. Plant Mol. Biol. 48, 109136.
Pagano, R.E., 2003. Endocytic trafficking of glycosphingolipids in
sphingolipid storage diseases. Philos. Trans. R. Soc. Lond. B Biol.
Sci. 358, 885891.
Palacpac, N.M., Hiramine, Y., Mi-Ichi, F., Torii, M., Kita, K.,
Hiramatsu, R., Horii, T., Mitamura, T., 2004a. Developmental-stage-
specific triacylglycerol biosynthesis, degradation and trafficking as lipid
bodies in Plasmodium falciparum-infected erythrocytes. J. Cell Sci.
117, 14691480.
Palacpac, N.M., Hiramine, Y., Shintaro Seto, S., Hiramatsu, R., Horii, T.,
Mitamura, T., 2004b. Evidence that Plasmodium falciparum diacyl-
glycerol acyltransferase is essential for intraerythrocytic proliferation.
Biochem. Biophys. Res. Commun. 321, 10621068.
Pandey, A.V., Babbarwal, V.K., Okoyeh, J.N., Joshi, R.M., Puri, S.K.,
Singh, R.L., Chauhan, V.S., 2003. Hemozoin formation in malaria: a
two-step process involving histidine-rich proteins and lipids. Biochem.
Biophys. Res. Commun. 308, 736743.
Papalexis, V., Siomos, M.A., Campanale, N., Guo, X., Kocak, G.,
Foley, M., Tilley, L., 2001. Histidine-rich protein 2 of the malaria
parasite, Plasmodium falciparum, is involved in detoxification of the
by-products of haemoglobin degradation. Mol. Biochem. Parasitol. 115,
7786.
Pourcho, R.G., Bernstein, M.H., Gould, S.F., 1978. Malachite green:
applications in electron microscopy. Stain Technol. 53, 2935.
Pralle, A., Keller, P., Florin, E.L., Simons, K., Horber, J.K., 2000.
Sphingolipid-cholesterol rafts diffuse as small entities in the plasma
membrane of mammalian cells. J. Cell Biol. 148, 9971008.
Preiser, P., Kaviratne, M., Khan, S., Bannister, L., Jarra, W., 2000. The
apical organelles of malaria merozoites: host cell selection, invasion,
host immunity and immune evasion. Microbes Infect. 2, 14611477.
Quevillon, E., Spielmann, T., Brahimi, K., Chattopadhyay, D.,
Yeramian, E., Langsley, G., 2003. The Plasmodium falciparum family
of Rab GTPases. Gene 306, 1325.
Quittnat, F., Nishikawa, Y., Stedman, T.T., Voelker, D.R., Choi, J.Y.,
Zahn, M.M., Murphy, R.C., Barkley, R.M., Pypaert, M., Joiner, K.A.,
Coppens, I., 2004. On the biogenesis of lipid bodies in ancient
eukaryotes: synthesis of triacylglycerols by a Toxoplasma DGAT1-
related enzyme. Mol. Biochem. Parasitol. 138, 107122.
Ramoino, P., Fronte, P., Fato, M., Beltrame, F., Diaspro, A., 2002. Mapping
cholesteryl ester analogue uptake and intracellular flow in Paramecium
by confocal fluorescence microscopy. J. Microsc. 208, 167176.
Raposo, G., Marks, M.S., 2002. The dark side of lysosome-related
organelles: specialization of the endocytic pathway for melanosome
biogenesis. Traffic 3, 237248.
Robibaro, B., Stedman, T.T., Coppens, I., Ngo, H.M., Pypaert, M.,
Bivona, T., Nam, H.W., Joiner, K.A., 2002. Toxoplasma gondii Rab5
enhances cholesterol acquisition from host cells. Cell. Microbiol. 4,
139152.
Rukmini, R., Rawat, S.S., Biswas, S.C., Chattopadhyay, A., 2001.
Cholesterol organization in membranes at low concentrations: effects
of curvature stress and membrane thickness. Biophys. J. 81, 21222134.
Salaun, C., James, D.J., Chamberlain, L.H., 2004. Lipid rafts and the
regulation of exocytosis. Traffic 5, 255264.
Sam-Yellowe, T.Y., 1996. Rhoptry organelles of the apicomplexa: their
role in host cell invasion and intracellular survival. Parasitol. Today 12,
308316.
Sandager, L., Gustavsson, M.H., Stahl, U., Dahlqvist, A., Wiberg, E.,
Banas, A., Lenman, M., Ronne, H., Stymne, S., 2002. Storage lipid
synthesis is non-essential in yeast. J. Biol. Chem. 277, 64786482.
Schroeder, R.J., Ahmed, S.N., Zhu, Y., London, E., Brown, D.A., 1998.
Cholesterol and sphingolipid enhance the triton X-100 insolubility of
glycosylphosphatidylinositol-anchored proteins by promoting the
formation of detergent-insoluble ordered membrane domains. J. Biol.
Chem. 273, 11501157.
Seo, T., Oelkers, P.M., Giattina, M.R., Worgall, T.S., Sturley, S.L.,
Deckelbaum, R.J., 2001. Differential modulation of ACAT1 and
ACAT2 transcription and activity by long chain free fatty acids in
cultured cells. Biochemistry 40, 47564762.
Shaw, M.K., Roos, D.S., Tilney, L.G., 1998. Acidic compartments and
rhoptry formation in Toxoplasma gondii. Parasitology 117, 435443.
Shiao, Y.J., Lupo, G., Vance, J.E., 1995. Evidence that phosphatidylserine
is imported into mitochondria via a mitochondria-associated membrane
and that the majority of mitochondrial phosphatidylethanolamine is
derived from decarboxylation of phosphatidylserine. J. Biol. Chem.
270, 1119011198.
Sibert, G.J., Speer, C.A., 1980. Fine structure of zygotes and oocysts of
Eimeria nieschulzi. J. Protozool. 27, 374379.
Simons, K., Ikonen, E., 1997. Functional rafts in cell membranes. Nature
387, 569572.
Simons, K., Toomre, D., 2000. Lipid rafts and signal transduction. Nat.
Rev. Mol. Cell Biol. 1, 3139.
Sinai, A.P., Joiner, K.A., 2001. The Toxoplasma gondii protein ROP2
mediates host organelle association with the parasitophorous vacuole
membrane. J. Cell Biol. 154, 95108.
Slomianny, C., 1990. Three-dimensional reconstruction of the feeding
process of the malaria parasite. Blood Cells 16, 369378.
Soares, M.J., De Souza, M.F., De Souza, W., 1987. Ultrastructural
visualization of lipids in trypanosomatids. J. Protozool. 34, 199203.
Sonda, S., Ting, L.M., Novak, S., Kim, K., Maher, J.J., Farese Jr., R.V.,
Ernst, J.D., 2001. Cholesterol esterification by host and parasite is
essential for optimal proliferation of Toxoplasma gondii. J. Biol. Chem.
276, 3443434440.
Sorger, D., Daum, G., 2002. Synthesis of triacylglycerols by the acyl-
coenzyme A:diacyl-glycerol acyltransferase Dga1p in lipid particles of
the yeast Saccharomyces cerevisiae. J. Bacteriol. 184, 519524.
Sorger, D., Daum, G., 2003. Triacylglycerol biosynthesis in yeast. Appl.
Microbiol. Biotechnol. 61, 289299.
Spady, D.K., Willard, M.N., Meidell, R.S., 2000. Role of acyl-coenzyme
A:cholesterol acyltransferase-1 in the control of hepatic very low
density lipoprotein secretion and low density lipoprotein receptor
expression in the mouse and hamster. J. Biol. Chem. 275, 2700527012.
Speer, C.A., Dubey, J.P., 2001. Ultrastructure of schizonts and merozoites
of Sarcocystis neurona. Vet. Parasitol. 95, 263271.
Stedman, T.T., Sussmann, A.R., Joiner, K.A., 2003. Toxoplasma gondii
Rab6 mediates a retrograde pathway for sorting of constitutively
secreted proteins to the Golgi complex. J. Biol. Chem. 278, 54335443.
Stewart, M.J., Schulman, S., Vanderberg, J.P., 1986. Rhoptry secretion of
membranous whorls by Plasmodium falciparum merozoites. Am.
J. Trop. Med. Hyg. 35, 3744.
 I. Coppens, O. Vielemeyer / International Journal for Parasitology 35 (2005) 597615
Subtil, A., Gaidarov, I., Kobylarz, K., Lampson, M.A., Keen, J.H.,
McGraw, T.E., 1999. Acute cholesterol depletion inhibits clathrin-
coated pit budding. Proc. Natl Acad. Sci. USA 96, 67756780.
Sullivan, D.J., Gluzman, I.Y., Goldberg, D.E, 1996. Plasmodium hemozoin
formation mediated by histidine-rich proteins. Science 271, 219222.
Sundermann, C.A., Lindsay, D.S., 1989. Ultrastructure of in vivo-produced
caryocysts containing the coccidian Caryospora bigenetica (Apicom-
plexa: Eimeriidae). J. Protozool. 36, 8186.
Suss-Toby, E., Zimmerberg, J., Ward, G.E., 1996. Toxoplasma invasion:
the parasitophorous vacuole is formed from host cell plasma membrane
and pinches off via a fission pore. Proc. Natl Acad. Sci. USA 93, 8413
8418.
Topolska, A.E., Lidgett, A., Truman, D., Fujioka, H., Coppel, R.L., 2004.
Characterization of a membrane-associated rhoptry protein of Plasmo-
dium falciparum. J. Biol. Chem. 279, 46484656.
van Blitterswijk, W.J., Houssa, B., 2000. Properties and functions of
diacylglycerol kinases. Cell Signal. 12, 595605.
van Deurs, B., Holm, P.K., Kayser, L., Sandvig, K., Hansen, S.H., 1993.
Multivesicular bodies in HEp-2 cells are maturing endosomes. Eur.
J. Cell Biol. 61, 208224.
Vial, H.J., Thuet, M.J., Broussal, J.L., Philippot, J.R., 1982a. Phospholipid
biosynthesis by Plasmodium knowlesi-infected erythrocytes: the
incorporation of phospohlipid precursors and the identification of
previously undetected metabolic pathways. J. Parasitol. 68, 379391.
Vial, H.J., Thuet, M.J., Philippot, J.R., 1982b. Phospholipid biosynthesis
in synchronous Plasmodium falciparum cultures. J. Protozool. 29,
258263.
Vial, H.J., Philippot, J.R., Wallach, D.F., 1984. A reevaluation of the status
of cholesterol in erythrocytes infected by Plasmodium knowlesi and P.
falciparum. Mol. Biochem. Parasitol. 13, 5365.
615
Vial, H.J., Eldin, P., Tielens, A.G., van Hellemond, J.J., 2003.
Phospholipids in parasitic protozoa. Mol. Biochem. Parasitol. 126,
143154.
Vielemeyer, O., McIntosh, M.T., Joiner, K.A., Coppens, I., 2004. Neutral
lipid synthesis and storage in the intraerythrocytic stages of Plasmo-
dium falciparum. Mol. Biochem. Parasitol. 135, 197209.
Waller, R.F., McFadden, G.I., 2005. The apicoplast: a review of the derived
plastid of apicomplexan parasites. Curr. Issues Mol. Biol. 7, 5779.
Ward, G.E., Miller, L.H., Dvorak, J.A., 1993. The origin of parasitophorous
vacuole membrane lipids in malaria-infected erythrocytes. J. Cell Sci.
106, 237248.
Wunderlich, F., Fiebig, S., Vial, H., Kleinig, H., 1991. Distinct lipid
compositions of parasite and host cell plasma membranes from
Plasmodium chabaudi-infected erythrocytes. Mol. Biochem. Parasitol.
44, 271277.
Yang, H., Cromley, D., Wang, H., Billheimer, J.T., Sturley, S.L., 1997.
Functional expression of a cDNA to human acyl-coenzyme A:choles-
terol acyltransferase in yeast. Species-dependent substrate specificity
and inhibitor sensitivity. J. Biol. Chem. 272, 39803985.
Yang, M., Coppens, I., Wormsley, S., Baevova, P., Hoppe, H.C.,
Joiner, K.A., 2004. Rhoptry biogenesis in apicomplexan parasites
involves a multivesicular body intermediate. J. Cell Sci. 117, 3831
3838.
Yayon, A., Timberg, R., Friedman, S., Ginsburg, H., 1984. Effects of
chloroquine on the feeding mechanism of the intraerythrocytic human
malarial parasite Plasmodium falciparum. J. Protozool. 31, 367372.
Zurzolo, C., van Meer, G., Mayor, S., 2003. The order of rafts. Conference
on microdomains, lipid rafts and caveolae. Eur. Mol. Biol. Org. J. 4,
11171121.
Zweytick, D., Athenstaedt, K., Daum, G., 2000. Intracellular lipid particles
of eukaryotic cells. Biochim. Biophys. Acta 1469, 101120.
 International Journal for Parasitology 35 (2005) 617626
www.parasitology-online.com

Annexin-like alpha giardins: a new cytoskeletal gene family

in Giardia lamblia*
Malin E.-L. Weilanda, Andrew G. McArthurb, Hilary G. Morrisonb,
Mitchell L. Soginb, Staffan G. Svarda,c,*
a
Microbiology and Tumor Biology Center, Karolinska Institutet, SE-171 77 Stockholm, Sweden
b
Marine Biological Laboratory, Woods Hole, MA, USA
c
Department of Cell and Molecular Biology, Uppsala University, BMC Box 596, SE-751 24 Uppsala, Sweden
Received 28 October 2004; received in revised form 10 December 2004; accepted 14 December 2004

Abstract
Through a genome survey and phylogenetic analysis, we have identified and sequenced 14 new coding regions for alpha-giardins in
Giardia lamblia. These proteins are related to annexins and comprise a multi-gene family with 21 members. Many alpha giardins are highly
expressed proteins that are very immunogenic during acute giardiasis in humans. However, little is known about the function of these
proteins. By using PCR with different combinations of gene-specific primers, we demonstrated that several of the genes localised to the same
chromosomal fragment. These data point towards a molecular evolution through gene duplication and subsequent functional divergence.
Semi-quantitative reverse transcriptase-PCR analysis of the Giardia life cycle revealed large differences in mRNA expression levels of the
alpha giardins. Epitope tagging of the alpha-giardins localised them to different cytoskeletal components, such as the flagella and the
adhesive disc, but also to the plasma membrane. These localisation experiments suggest alpha-giardins play a role in cell motility, attachment
and membrane stability.
q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: Giardia; Parasite; Annexin; Cytoskeleton; Flagella

1. Introduction intestine. To ensure transmission and survival outside

Giardia lamblia, a commonly found intestinal protozoan
parasite, occurs throughout the world and triggers a form of
diarrhea called giardiasis. Giardia is thought to be an early
diverging eukaryote and has a simple life cycle that
alternates between the cyst stage and the trophozoite
stage. Infection starts through ingestion of the infectious
cysts. The cysts passage through the stomach triggers
excystation and the cysts transform into trophozoites that
will attach, divide and colonise the upper part of the small
*
Nucleotide sequence data reported in this paper are available in the
GenBanke, under the accession numbers AY781314AY781334.
* Corresponding author. Address: Microbiology and Tumor Biology
Center, Karolinska Institutet, SE-171 77 Stockholm, Sweden. Tel.: C46 18
4714558; fax: C46 18 530396.
E-mail address: staffan.svard@icm.uu.se (S.G. Svard).
0020-7519/$30.00 q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijpara.2004.12.009
the host, Giardia undergoes encystation into highly
resistant, dormant cysts, which are shed to the environment
through feces, thereby completing the life cycle.
The genome sequencing of G. lamblia is nearly
complete, with greater than 99% of Giardias coding
capacity represented in contigs or unplaced reads available
on GiardiaDB (www.mbl.edu/Giardia) (McArthur et al.,
2000). This enables researchers to compare specific genes
between early and late diverging eukaryotes. We have
recently characterised some of the immunodominant
proteins during acute human giardiasis (Palm et al., 2003).
One of these proteins was alpha-1 giardin, a protein that we
recently characterised as an annexin with glycosaminogly-
can-binding activity (Weiland et al., 2003).
Crossley and Holberton originally described giardins as
cytoskeletal proteins unique to Giardia (Crossley and
Holberton, 1983a,b). The giardins are associated with
618 M.E.-L. Weiland et al. / International Journal for Parasitology 35 (2005) 617626
the cytoskeleton and belong to at least three separate gene
families: the alpha giardinsannexin homologs (Morgan and
Fernandez, 1995, 1997), beta giardin, a striated fibre (SF)-
assemblin homolog (Weber et al., 1993) and gamma
giardin, a protein without notable homologs (Nohria et al.,
1992). Up until now, published reports identified six alpha
giardins including, alpha-1 and -2 giardin (Peattie et al.,
1989; Alonso and Peattie, 1992), alpha-3 (Weiland et al.,
2003), as well as alpha-7.1, -7.2 and -7.3 giardin (Palm
et al., 2003). These genes share an ancestry with human
annexins (Morgan and Fernandez, 1995; Szkodowska et al.,
2002).
Annexins are Ca2C and phospholipid-binding proteins
forming an evolutionary conserved multigene family
(Gerke and Moss, 2002). The name annexin derives
from terms meaning bring/hold together and describes
the principal property of binding to and possibly holding
together membranes. Today, more than 160 unique
annexin proteins are known from more than 65 different
eukaryotic species (Gerke and Moss, 2002). Annexins
have a wide range of biological functions. Despite
detailed biochemical and structural knowledge, a clear
physiological function for most annexins remains
unknown.
In Giardia, there is a strong link between the cytoske-
leton and virulence, since Giardia, by aid of its cytoskele-
ton, has to move within the intestine and attach to the
epithelium to avoid being swept away (Elmendorf et al.,
2003). Novel cytoskeletal proteins might be potential drug
targets for treatment of giardiasis. During Giardia differen-
tiation, there are extensive, highly organised rearrangements
of the cytoskeleton and membranous compartments (Que
et al., 1996). Annexins interact with the cytoskeleton, in
particular F-actin, suggesting their participation in such
rearrangements (Gerke and Moss, 2002). In this study, we
characterise 14 new members belonging to the alpha giardin
multigene family of annexin-like proteins from Giardia, and
Table 1
PCR primers (5 0 3 0 ) used for cloning and sequencing of alpha-giardins (see text for numbering)
Primer Forward primer
A4 TGC GGG TCT GGA AGA ACT TCC G
A5 CCT GTA AAC CTT CGA ACC CTT CGC
A6 CAC GGT CAT AAT ACT TTG GAT AAT
A8 TAT ACG TGC GCA ACC CGA TGA CCT
A9 CTT TTG TGA ACC CTT TTC ATT AAA TC
A10 TAT CGG CCA GAG ACA TCG TCC GCC
A11 GCA TTT GAA TAA GGT GTG ACT GTT
A12 GGT GCG CCG GAT TCA GTC TAG
A13 GCA TTT AAC AGG TGT GAA TTA ATA
A15 GTC AGT AAA TGT GTG GGA AGT CAC
A16 GAA GCC ACG CAG CCC CTG AAA ACA
A17 TAG TAA AGG CGA CTG TAG TTT TGA C
A18 CGA CTC AGT TGA GTG TAG GAC ATA
A19 GTT GTG CCG CAT CAA CTC CCG
explore their role in regulation of membrane-cytoskeleton
dynamics.
2. Materials and methods
2.1. Reagents, cell culture and differentiation
Unless otherwise indicated, reagents were obtained from
Sigma Chemical Co, St Louis, MO. All restriction enzymes
used were obtained from Invitrogen. Giardia lamblia strain
WB (ATCC 50803) clone C6 or A11 trophozoites were
grown vegetatively and encysted as described (Bernander et
al., 2001). Excystation was carried out by a two-step method
as described (Boucher and Gillin, 1990).
2.2. Cloning and sequencing of genes belonging to alpha
giardin gene family
PCR primers (Table 1) were designed to amplify
complete giardin genes from G. lamblia strain WB clone
C6 genomic DNA, using Ready To Go PCR beads
(Amersham Pharmacia Biotech) or Platinum Taq DNA
polymerase (Invitrogen). PCR conditions were as
described by each manufacturer, with 35 cycles at 95 8C
for 30 s, 55 8C for 30 s and 72 8C for 1 min. The amplified
gene products were run on a 1% agarose gel, cut out, and
gel-extracted using a QIAquick Gel Extraction Kit
(Qiagen). Purified gene products were cloned into the
TOPO TA cloning vector for DNA sequencing (Invitro-
gen). Sequencing of the PCR amplified genes was
performed using an ABI PRISM Big Dye Terminator
version 3.0 cycle sequencing kit (Applied Biosystems)
with T3 and T7 primers on an ABI 377 automatic DNA
sequencer.
Reverse primer
GAC CAT GCC TTT ACT CAT GAG ATG
CAT TTT CTA AAT GTA AGC GCT CAG AA
CCT TTG GCT CAC ACA ACT GGA GAT
GCT AGT AGA TAA GCT GGG CCA AAT G
TAT TAA TTC ACA CCT GTT AAA TGC
GTG ACT TCC CAC ACA TTT ACT GAC
TAA ATT GAA GAG TGA CAC GCG GCC
CTC GTA GTC AGA CAT CTG CCT GAA G
TTT AAA GTC CGT CAG TTT ACC CTG
AAC AGT CAC ACC TTA TTC AAA TGC
GCG TCG GAC AGG CTC CGC AGC AGT T
TGG CCT ACT ATA TTC CCG TAT GAC
ATG CAT TTA GCG CAA TCA CTT G
TCA GTC GCC GCG GGG AGT CGA G
 M.E.-L. Weiland et al. / International Journal for Parasitology 35 (2005) 617626
2.3. Phylogenetic analysis
To identify proteins homologous to already known alpha
giardins (alpha-1, -2, -3, -7.1, -7.2 and -7.3 giardin) we
performed a survey of the G. lamblia genome database
(McArthur et al., 2000). We identified 14 additional genes
belonging to the annexin-like alpha giardin gene family
(alpha-4 to -6, -8 to -13 and -15 to -19 giardin). In order to
clarify the nomenclature prior to phylogenetic analysis on
all members of the annexin-like alpha giardin gene family,
we renamed the already identified annexin XXI (Morgan
and Fernandez, 1997) to alpha-14 giardin.
All giardin amino acid sequences were aligned using
ClustalX 1.81 (Thompson et al., 1997), with manual
correction using MacClade 4.0fc4 (Maddison and Maddison,
1989). Gaps and highly divergent or ambiguous regions of
the alignments were excluded from analyses. A number of
vertebrate annexin sequences served as outgroups, based on
their similarity to giardins in BLAST searches (Altschul and
Koonin, 1998). The alignment is available upon request.
Phylogenetic relationships of the giardin sequences
were assessed using a Bayesian statistical procedure,
as implemented by the computer program Mr Bayes
(Huelsenbeck and Ronquist, 2001). Mr Bayes performs a
Metropolis-coupled Markov chain Monte Carlo (MC3)
estimation of posterior probabilities (Shoemaker et al.,
1999; Lewis and Swofford, 2001; Huelsenbeck et al., 2002).
We performed MC3 estimation of posterior probabilities
using non-informative prior probabilities, the JTTCICG
(Jones et al., 1992) substitution model with inclusion of
unequal amino acid frequencies, and four incrementally
heated Markov chains with different random starting trees.
The Markov chains were run for 10,000,000 generations,
with sampling of topologies every 100 generations.
Posterior probabilities of topologies, clades, and parameters
were estimated from the sampled topologies after removal
of MC3 burn-in.
2.4. RNA extraction and DNAse I treatment
Giardia lamblia strain WB clone A11 cultures of 500 ml
trophozoites, 4 h encysted cells, 10 h encysted cells and
cysts (water treated, 24 h encysted cells) were put on ice
20 min before harvesting by centrifugation at 2500 rpm for
5 min. Cells were washed two to three times in PBS before
homogenising the resulting cell pellet in 1 ml Trizol
(Invitrogen) or 1 ml Trizol plus glass beads for cysts. The
homogenates, were snap frozen to K80 8C and stored until
use. Total cellular RNA was extracted, precipitated and
washed according to the manufacturer (Invitrogen). RNA
from cysts was extracted after shaking the homogenates
with glass beads for 60 s at maximum speed in bead beater
(Techtum Lab). The RNA pellets were dissolved in 43 ml
ddH2O, 5 ml of One Phor All (OPA) buffer (Amersham
Pharmacia Biotech), 1 ml DNAse I (Stratagene) and 1 ml
RNaseOUT (Invitrogen) and DNAse I treated at 37 8C for
619
30 min followed by enzyme inactivation for 5 min at 95 8C.
The RNA was precipitated by adding 160 ml pure ethanol
and incubated at K20 8C overnight, before being dissolved
in ddH2O and quantified by measuring OD260.
2.5. Reverse transciptase polymerase chain reaction
(RT-PCR)
Complementary DNA was prepared from 5 mg DNAse I
treated total parasite RNA by Superscript II RNase H-
Reverse Transcriptase (Invitrogen) according to the manu-
facturer. Briefly, 5 mg total RNA was mixed with 2 ml oligo
dT primer, 2 ml dNTP and ddH2O, and incubated for 5 min
at 65 8C. After addition of 2 ml RNaseOUT (Invitrogen),
4 ml 0.1 M DTT and 8 ml first strand buffer, the reaction
mixture was incubated for 2 min at 42 8C. Two microliters
of Superscript II were added and incubation continued at
42 8C for 50 min. The reaction was inactivated for 15 min at
70 8C and stored at K20 8C. As a control all reactions were
made with or without Superscript II to check for genomic
DNA contamination.
2.6. Semi-quantitative expression analysis of alpha giardins
To remove RNA from the synthesised cDNA, heating at
95 8C for 5 min inactivated the Superscript II enzyme and
digestion with 0.5 ml RNace-It (Stratagene) at 37 8C for
30 min degraded residual RNA. A final incubation at 95 8C
for 5 min inactivated the enzyme in RNace-It. Gene-specific
primers were designed for semi-quantitative PCR (Table 2).
The semi-quantitative analysis was done by comparing the
alpha giardin gene expression to the expression of beta
giardin, which was used as a standard in a reverse
transcriptase PCR (RT-PCR) experiment). All PCR reac-
tions were run in triplicate for 2530 cycles with
denaturation at 95 8C for 30 s, annealing at 55 8C for 30 s
and extension at 72 8C for 1 min. Amplified gene products
were run on a 1% agarose gel and analysed with the program
Quantity One Volume Quick Analysis (Biorad). The PCR
fragments of the alpha giardin genes 8, 12, 15 and 16 had
similar size as beta giardin and were therefore amplified
separately, although analysed on a gel containing beta
giardin in separate lanes.
2.7. Construction of AU-1 tagged alpha giardins
The promoter and coding sequence of each alpha giardin
gene was amplified from G. lamblia strain WB clone six
genomic DNA as an EcoRI/MluI fragment (Table 3). The
alpha-2 giardin promoter and coding sequence was cut out
from 2 mg of the AU-1 expression vector (Weiland et al.,
2003) with EcoRI and MluI and run on a 1% agarose gel.
The cut vector fragment was gel extracted using a QIAquick
Gel Extraction Kit (Qiagen) and ligated with the purified
EcoRI/MluI cut PCR fragment of the alpha giardin genes at
16 8C overnight. Ligations were transformed into
620 M.E.-L. Weiland et al. / International Journal for Parasitology 35 (2005) 617626

Table 2
PCR primers (5 0 3 0 ) used for semi-quantitative RT-PCR analysis of alpha giardins (see text for numbering)

Primer Forward primer Reverse primer

A1 exp ACG AGG TCC AGA TCG CGT TCA TAG ATC TCT CGA TGT CTG TGC CCA TGT C
A2 exp ATG CCA AGG ACG AGG CCC AGA TCG CAC CTT TCG ACG TCG GCT TTC ATA C
A3 exp GGT CGT AGC GGA CCT GAA TGC CGC TCT TCA ACC GCA TCC TTA AAG GCG
A4 exp CAT TCC AGC CTA TCA CGC AAA C TCG CAA AGC CAG TGG TGT G
A5 exp CTC AGA TAT GCT CGG ATC TCA AGG GGC CAT TCG CTC GTC AAA GCA GCG
A6 exp TCC AAT GAG CAG ATG ATA GCA GG TCG CTG ACC GAT GTG GTT CC
A7.1 exp AGG AGC GTC GCC GAC TTC GAC GAC AGG ACC TTG CAC TGC TCC GCC TGA
A7.2 exp AGG AGC GTC GCC GAC TTC GAC GAT GCG TAG AGG ACC TTG CAC TGC TCT
A7.3 exp GCG AGG AGC GTC GCC GAC TTC AAT TTC CTG ATG TCC TTG GCT AAG TTC
A8k exp AGA GAC GCA TTG GTG GAC CTT G TTG TTT GTG TGC TCC TCG TAA GC
A9 exp GCA CCG AGA CAA GGT CGT AAT C CGT AGT TGG ACG CAA TAA GGG TC
A10 exp ATT TAT CAA GGT TGC AGA CGA GCT G CTC GCC GAA ATT GCT CTT GAA CAC
A11 exp TCA AGA AGT GCA TCA AGA ATG GCC CAG AAC ATG AGG CAT GCA TCG CGG
A12k exp CGT CAA AGC GAC ATT CAT TCA CG TTC CTG TGG TAG CAC TGG GTT C
A13 exp CAA AGG GAC AAG GTC CTT CCG CTG CGC AGA TCT GCT GCT AGT GAT CC
A14 exp GCG GTT GTG CAG AAG GTT GTG G CAA CTC GAG AGT CAT CGG CCT G
A15k exp AGA TGT CTC ACG ACG CAG GAT G TCC ACA GGA GGC AAA GTT TCG C
A16k exp GAA GCA CGG ATG TTA GAC TTA GCG TGC CCT GAG CGT TTT GGA TTC
A17 exp GGA GAT GAG GTT GGC CTG GCT GTG GCT AAG ATC CTT GTC AGT AAC AGC
A18 exp ATG CAT TTA GCG CAA TCA CTT G CTA TAT CTT CCA GAC TGT TGC
B exp CGA CGA CCT CAC CCG CAG TGC GAC GTT CTG GAG ATT TGT CTC AAC G

Escherichia coli Top10 for identification of complete
constructs. Trophozoites were transfected using electro-
poration of 20 mg pure plasmid DNA and the plasmids were
maintained through puromycin selection (Knodler et al.,
1999). Cells were processed for immunofluorescence
analysis as below.
(1:50, Dako A/S). Slides were re-washed in PBS and
mounted in Vectashield for digital confocal microscoscopy
(Nikon Eclipse E800, Nikon Instruments). Processing of
images was performed using Openlab 3.11 software
(Improvision, Heidelberg, Germany).

2.8. Immunofluorescence and alpha giardin localisation 3. Results

AU-1 transfected trophozoites were put onto slides and
allowed to dry. Slides were either air dried or fixed in 100%
acetone on ice for 20 min, dried, and stored at room
temperature until use. After rinsing the slides in PBS, 20 ml
monoclonal AU-1 antibody (Covance), diluted 1:300 in
PBS, was added and incubated for 1 h at room temperature
in a moisture chamber. Slides were washed with 10 droplets
of PBS before addition of the secondary fluorescein
isothiocyanate (FITC) conjugated goat anti-mouse antibody
3.1. Cloning, alignment and phylogenetic analysis
All new giardin sequences were deposited in GenBank
under accession numbers AY781314AY781334. After
exclusion of poorly aligned regions, the final amino acid
alignment included 271 characters, all of which were
parsimony informative. The alignment verified the con-
served annexin and alpha-giardin specific motifs identified
by Szkodowska et al. (2002). A chi-square test indicated

Table 3
Primers (5 0 3 0 ) used to construct AU-1 epitope tagged alpha giardins (see text for numbering)

Primer Forward primer Reverse primer

A3-AU1 CCG AAT TCA AGA TTC AAG CGG ACC GTG GCT TC CCA CGC GTA CCT CTC CAA AGG ACC TTG AGA AC
A5-AU1 CCG AAT TCC TGT AAA CCT TCG AAC CCT TCG CCA CGC GTG CAT CTG TGA CGC GCC ATA GCG T
A6-AU1 CCG AAT TCA CGG TCA TAA TAC TTT GGA TAA CCA CGC GTA CAT AAC ACG CCA GAG GAT ATT AA
A7.1-AU1 CCG AAT TCT GTG TGG AAA CAT AGA CCC TC CCA CGC GTC ATG ACG TGC CAG AGG ACG CAG
A7.2-AU1 CCG AAT TCT GTG TGG AGA CAT AGA CCC CT CCA CGC GTC ATG ACG TGC CAG AGT ACG CAG C
A7.3-AU1 CCG AAT TCA CGG ACT GGA GCG TCA GTT GTG C CCA CGC GTC ATG ACG TGC CAG AGG ACG CAG C
A8-AU1 CCG AAT TCT ATA CGT GCG CAA CCC GAT GAC C CCA CGC GTA GTG ATG TTC TCG TGT TTT TCT GC
A9-AU1 CCG AAT TCT TGT GAA CCC TTT TCA TTA AAT C CCA CGC GTG TTT GCG CCT TCG GTG TCA ACA CG
A10-AU1 CCG AAT TCT ATC GGC CAG AGA CAT CGT CC CCA CGC GTT AGA ACC CCC CAC ATG GTA CAA C
A15-AU1 CCG AAT TCG TCA GTA AAT GTG TGG GAA GT CCA CGC GTT GGG CAC AGG GAA GGG ATG C
A16-AU1 CCG AAT TCG AAG CCA CGC AGC CCC TGA AA CCA CGC GTG GCG CCG CCG TCC ACA AGC TC
A17-AU1 CCG AAT TCT AGT AAA GGC GAC TGT AGT TT CCA GCC GTG CCT CGG GAG CCC AAA GCT CG
 M.E.-L. Weiland et al. / International Journal for Parasitology 35 (2005) 617626 621

1.00 alpha-1 giardin supercontig 707

0.95 alpha-2 giardin
0.80 alpha-5 giardin

supercontig 791
0.99 alpha-3 giardin
0.99
alpha-6 giardin
0.67 alpha-9 giardin

0.97 alpha-4 giardin

0.61
alpha-8 giardin
alpha-13 giardin

0.93 alpha-7.1 giardin

1.00 alpha-7.2 giardin supercontig 699
0.99
alpha-7.3 giardin
0.91
alpha-10 giardin

supercontig 710
0.91 alpha-11 giardin

1.00 alpha-12 giardin

0.52 alpha-15 giardin
alpha-17 giardin
0.51
alpha-18 giardin
alpha-16 giardin

1.00 0.96 alpha-14 giardin

alpha-19 giardin

1.00 annexin A4 Human

supercontig 727
0.94 annexin IV Cow
annexin A5 Zebrafish supercontig 699
1.00
annexin max2 Medaka
annexin max3 Medaka
0.1

Fig. 1. The annexingiardin phylogenetic tree and mapping of the genes within the Giardia genome. Bayesian posterior probabilities are superimposed upon
the tree. The tree is rooted using vertebrate annexin sequences. Horizontal branch lengths are representative of evolutionary change. The giardins were mapped
to supercontigs (scaffolds) in the genome assembly giardia12 (http://www.mbl.edu/Giardia).

that the amino acid composition of alpha-16 giardin was 3.2. Alpha giardin gene distribution
significantly different from the other sequences. This could
bias its phylogenetic placement. The length of the MC3 Giardia lamblia has a very compact genome of approxi-
burn-in was 5000 of 10,000,000 generations, resulting in a mately 12 Mbp, distributed over five chromosomes. It has
sample of 99,950 trees for estimation of posterior few introns and most genes have short 5 0 - and 3 0 -UTRs
probabilities. Replicate analyses provided very similar (Adam, 2001). To understand how the alpha giardin gene
estimates of posterior probabilities, providing evidence that family has evolved, we looked at the distribution of the alpha
convergence was obtained and mixing was adequate. giardin genes in the genome by using PCR with different
Overall resolution of giardin phylogeny was high, as combinations of gene-specific primers and by mapping their
indicated by posterior probabilities (Fig. 1). Two major placement within the current Giardia genome sequence
clades of giardins were found, each reflective of gene assembly. As mentioned above, alpha giardins predomi-
duplications within restricted regions of the genome nantly localised to two regions of the Giardia genome
(supercontigs 710 and 791). Although the support of both assembly version giardia12 (http://www.mbl.edu/giardia).
of these clades was low (posterior probabilities of 0.52 and Alpha-1, -2, -4, -5, -6, -8, -9, and -13 giardin are all located
0.61, respectively), the co-mapping of their members within a w80 kb region of the genome, with spatial clustering
provides independent support for their monophyly. As of three groups of giardins within w5 kb sub-regions (Fig. 2).
such, we hypothesise that the majority of alpha-giardin In a separate region of the assembly, alpha-7.1, -7.2, -10, -11,
diversity is the product of gene duplication events in two -12, -15, -16, -17, and -18 are all located within a w330 kb
regions of the genome. A third clade (alpha-14 giardinC stretch of the genome, with all but alpha-7.1 and -7.2
alpha-19 giardin) represents the product of an earlier gene clustered within a w25 kb sub-region (Fig. 2). Three alpha
duplication event and shows no co-mapping of its two giardins (alpha-7.3, -12, and -14 giardin) co-mapped to
members. Notably, phylogenetic analysis confirmed the supercontig 699 but violations of several assembly criteria
common ancestry of alpha-7.1, -7.2 and -7.3 giardin. suggest there may be incorrect links between contigs in this
However, alpha-7.3 giardin did not co-map with alpha-7.1 scaffold. Preliminary evidence suggests that alpha-14 giardin
and -7.2 giardin. As discussed below, this may be an belongs in a different part of the genome while alpha-7.3
artifact in the current genome assembly. and -12 giardin may be linked close together within
622 M.E.-L. Weiland et al. / International Journal for Parasitology 35 (2005) 617626

Super contig 791

200 K 225 K 250 K 300 K

Contig 7702
Contig 6504

205 K 210 K
250 K 255 K
Alpha-2
Alpha-5 Alpha-13
Alpha-6 Alpha-9
Alpha-4
272 K 278 K

Alpha-8 Alpha-1

Super contig 710

200 K 300 K 400 K 500 K 600 K

Contig 3370 Contig 3313 Contig 3545 Contig 7532

Contig 4154

200 K 205 K 340 K 342 K 502 K 510 K 520 K

Alpha-7.1 Alpha-7.2 Alpha-10 Alpha-16

Alpha-15 Alpha-17
Alpha-11 Alpha-18

Fig. 2. Mapping of alpha-giardins to the giardia12 genome assembly. Arrows for alpha-giardins represent orientation of genes relative to the assembly
consensus sequences (contigs).

supercontig 710. This would be in agreement with the analysis and to get a semi-quantitative measurement of the
phylogenetic results. Confidence in this result will have to mRNA levels of the different alpha giardins (Table 4 and
await further editing of the genome assembly. Fig. 3). Most alpha giardins were constantly expressed
The pattern of physical clustering of alpha giardins throughout encystation when compared to beta giardin, in
within the Giardia genome assembly strongly correlates a single tube RT-PCR reaction. Only one reaction gave an
with phylogenetic relationships of the encoded proteins. If extra band, due to non-specific amplification (alpha-7.3
the corrections to supercontig 699 above withstand scrutiny, giardin, Fig. 3). Alpha-4, -6, -9 and -18 giardin showed
only alpha-3 giardins physical placement does not reflect
Table 4
its phylogenetic ancestry, suggestive of gene duplication via Semi-quantitative PCR comparison of alpha giardin gene expression to beta
a different mechanism from the other alpha-giardins. giardin gene expression (standardZ1.0)
Examination of both supercontigs 710 and 791 found no
Giardin gene Trophozoites 4 h encysting 10 h encysting Cysts
evidence of repeated sequences of other possible sequence
signatures to suggest mechanisms for the gene diversifica- Alpha-1 1.02 1.06 1.26 1.73
Alpha-2 0.54 0.50 0.83 0.73
tion of alpha-giardins. It should be noted that two partial Alpha-3 0.51 0.73 0.77 0.76
copies of alpha-9 giardin and one partial copy of alpha-13 Alpha-4 0.13 0.13 0.22 0.32
giardin were found within the Giardia genome. These occur Alpha-5 !0.10 0.29 0.21 0.22
near the end of problematic assembly contigs and may Alpha-6 0.20 0.17 0.23 0.40
simply be artifactual. Alpha-7.1 0.16 0.15 0.24 0.17
Alpha-7.2 0.38 0.50 0.62 0.56
Alpha-7.3 0.61 0.69 1.05 0.80
3.3. Expression of alpha giardins
Alpha-8 !0.10 !0.10 !0.10 !0.10
Alpha-9 0.17 0.24 0.37 0.47
Annexins are frequently described as being ubiquitous, Alpha-10 !0.10 !0.10 !0.10 0.13
since any given cell appears to express a range of different Alpha-11 1.17 1.11 1.45 1.04
annexins (Gerke and Moss, 2002). The expression of Alpha-12 !0.10 !0.10 !0.10 !0.10
individual annexins have been correlated to cell prolifer- Alpha-13 0.11 !0.10 !0.10 0.12
Alpha-14 !0.10 !0.10 !0.10 !0.10
ation and/or differentiation. To investigate the gene
Alpha-15 !0.10 !0.10 !0.10 !0.10
expression of the alpha giardins, we performed a reverse Alpha-16 !0.10 !0.10 !0.10 !0.10
transcriptase PCR (RT-PCR) analysis to determine the Alpha-17 !0.10 !0.10 !0.10 !0.10
alpha giardin gene-specific mRNA levels in trophozoites, Alpha-18 0.44 0.26 0.57 0.78
encysting cells and cysts. Beta giardin, a protein thought Alpha-giardin genes with expression values below 0.10 have been excluded
to have a relative constant gene expression throughout the due to low expression. Standard deviation between experiment means is 0.
parasite life cycle, was used to standardise the RT-PCR 12. An experiment was not performed for alpha-19 giardin.
 M.E.-L. Weiland et al. / International Journal for Parasitology 35 (2005) 617626 623

Fig. 3. Gene specific reverse transcriptase PCR analysis of alpha giardins from different timepoints of the Giardia life cycle. Time points are trophozoites (T),
4 h encystation (4), 10 h encystation (10) and cysts (C). Arrows indicate alpha- and beta-giardin reverse transcriptase PCR products. The DNA ladder is
included to indicate fragment size.

a tendency toward higher expression during late encysta- 4. Discussion

tion and in cysts (Table 4). Alpha-1 and -11 giardin were
the only genes with higher expression levels than beta Giardia is the simplest organism known to express
giardin. Within the alpha giardin gene family, these two annexins. Given that G. lamblia may represent one of the
genes are the most abundantly expressed, followed by earliest diverging eukaryotes, the occurrence of as many as
alpha-2, -3, -7.2, -7.3 and -18 giardin. This was also seen 21 annexin homologs is surprising. Several crown-group
in two-dimensional protein gels (Palm et al., 2003; Svard, eukaryotes have no or only two to four annexin homologs
unpublished data). The rest of the alpha giardins are all e.g. Saccharomyces cerevisiae, Dictyostelium discoideum,
expressed at low levels compared to beta giardin. Neurospora crassa and Caenorhabditis elegans (Gerke and
Moss, 2002). The alpha giardin gene family, like other
annexins, has probably undergone expansion by multiple
3.4. Localisation of AU-1 epitope tagged alpha giardins gene duplications, accompanied by selection based on
functions related to the specific needs of the parasite (Morgan
Reports of the cellular localisation of the alpha and Fernandez, 1997). Notably, annexin XXI (here alpha-14
giardins have been contradictory, as reviewed in giardin), identified and characterised by Szkodowska et al.
Elmendorf et al. (2003). In a previous attempt to clarify (2002), is as distant from alpha-1 and -2 giardin as it is from
this issue we used two different approaches to localise human annexin. This diversification of alpha-giardins is
alpha-1 giardin; confocal immunofluorescence with a suggestive of functional importance in the parasite.
peptide antibody and transfection of Giardia trophozoites
with epitope tagged (AU-1) alpha-1 giardin (Weiland
et al., 2003). In this study we used the second approach Table 5
Localisation of AU-1 epitope tagged alpha gardins
to localise the remaining members in the alpha giardin
gene family. All Giardia proteins were expressed with AU-1 tagged giardin protein Main localisation
their own promoter thereby reducing the risk of Alpha-1 Plasma membrane
mislocalisation due to unnatural over expression. The Alpha-2 Plasma membrane, flagella
results from the localisation experiments are summarised Alpha-3 Disc edge
Alpha-4 Lethal to the parasite
in Table 5 and Fig. 4AG. Most alpha giardins show a Alpha-5 Ventral flagella, diffuse disc
plasma membrane association and/or localisation to Alpha-6 Spotty cytoplasmatic
different flagella. A few alpha giardins (alpha-3, -5, Alpha-7.1 Lethal to the parasite
and -17 giardin, Fig. 4A, B and G) also localised to the Alpha-7.2 Plasma membrane, cyst wall
associated
adhesive disc. Alpha-15 and -16 giardin showed a patchy Alpha-7.3 Plasma membrane, cytoplasmatic
distribution along the plasma membrane, resembling Alpha-8 Lethal to the parasite
vesicles (Fig. 4E). Interestingly, transfection of alpha- Alpha-9 Flagella mainly ventral
7.2 giardin disturbed the cell division and caused Alpha-10 Ventral flagella
Alpha-11 Lethal to the parasite
parasites to encyst. In the cysts, alpha-7.2 giardin
Alpha-14 All flagella, weak median body
localised in or close to the cyst wall (Fig. 4D). Alpha-15 Spotty cytoplasmatic close to pm
Transfection experiments with constructs encoding Alpha-16 Spotty cytoplasmatic close to pm
alpha-4, -7.1, -8 and -11 were lethal to the parasite Alpha-17 Disc, flagella mainly ventral
and therefore another approach to localise these proteins Experiments were not performed for alpha-12, -13, -18, and -19 giardin (see
is needed. Fig. 4 for sample images).
624 M.E.-L. Weiland et al. / International Journal for Parasitology 35 (2005) 617626
Fig. 4. Localisation of AU-1 epitope tagged alpha giardins, using confocal
immunofluorescence photomicrographs of transformed parasites probed
with a monoclonal anti-AU-1 antibody (green). Bar is 5 mm. (A) epitope
tagged alpha-3 giardin; (B) epitope tagged alpha-5 giardin; (C and D)
epitope tagged alpha-7.2 giardin; (E) epitope tagged alpha-9 giardin; (F)
epitope tagged alpha-15 giardin; (G) epitope tagged alpha-17 giardin.
The evolutionary distance between different annexin sub-
families in plants, protists and animals tells us that there has
been extensive divergence within each of these separate
kingdoms (Morgan and Fernandez, 1997). In G. lamblia, the
existence of multiple alpha giardins with unique 5 0 coding
and regulatory regions has probably facilitated their
adaptation to cell specific needs. That epitope tagging and
over-expression of some proteins were lethal to the parasite,
with impaired cell division and/or differentiation, supports
their putative functional importance. Several of the alpha
giardins are among the most highly expressed proteins in
Giardia. For example, alpha-1 and -11 giardin show high
levels of transcription of mRNA (Table 4) as well as the
corresponding protein expression observed on 2D protein
gels (Palm et al., 2003). There are large differences in the
mRNA levels between different alpha giardins (Table 4), yet
over-expression of weakly and highly expressed alpha
giardins is lethal. This illustrates the need for tight
transcriptional regulation of these genes.
During encystation, a marked expansion of the rough
endoplasmatic reticulum and Golgi takes place, facilitating
a rapid secretion and synthesis of the cyst wall proteins
(Lujan et al., 1995; Mowatt et al., 1995; Sun et al., 2003).
Surprisingly, few genes involved in vesicular transportation
are up-regulated during encystation (Svard et al., 1999;
Marti and Hehl, 2003). Thus, even if the alpha giardins are
constantly expressed, they may participate in the encysta-
tion process. The fact that alpha-7.2 giardin localised in
close association with the cyst wall and that alpha-15 and
-16 giardin showed a patchy distribution along the plasma
membrane, resembling vesicles, may be indicative of their
participation in the encystation process. Annexins have been
shown to participate in vesicle transport (Gerke and Moss,
2002) and it is possible that alpha giardins are involved in
vesicle formation, vesicle stabilisation and/or vesicle
transportation along the cytoskeleton (Fig. 4).
De novo synthesis of proteins would also be expected
during excystation, when cytoskeletal structures need to
be rebuilt and the cell resumes its vegetative growth.
Up-regulation of genes has been seen early in excystation
(Hetsko et al., 1998), but it is a rapid process that may not
rely entirely on new gene expression (Cardenas et al., 1998;
Hetsko et al., 1998). One major challenge for the re-
emerging excyzoite is to undergo its first round of
cytokinesis, re-assemble the flagella and adhesive disc to
recover motility, and attach before being swept down with
the intestinal flow (Reiner et al., 2003). Giardia has solved
this challenge by a conservative approach, simply breaking
the disk and flagella into several large fragments, storing
these in the cyst cytoplasm and then re-assembling the
fragments during excystation (Elmendorf et al., 2003).
The excystation process is triggered by changes in Ca2C
concentration in the cell (Reiner et al., 2003) and Ca2C is
known to influence cell mobility and microtubule stability,
two important factors during membrane rearrangements in
excystation. A calcium responsive Giardia calmodulin
(CaM) localised to the basal bodies may function as a
control center to coordinate the assembly of new flagella
during this process (Reiner et al., 2003). Since alpha
giardins are calcium-binding proteins (Bauer et al., 1999)
that are associated with the cytoskeleton, including the
adhesive disc and flagella, they may participate in the highly
dynamic process of excystation.
Another possible function of alpha giardins may be to
stabilise the newly formed plasma membrane of the
emerging excyzoites. Giardia lives in the extremely hostile
environment of the gut, which has high concentrations of
bile salts. In this environment, the parasite requires a very
stable cytoskeleton in order to resist membrane leakage and
structural collapse. Several studies have shown that Giardia
tubulin is highly modified to maximise its stability
(Elmendorf et al., 2003) and alpha giardins might help in
 M.E.-L. Weiland et al. / International Journal for Parasitology 35 (2005) 617626
stabilising the cytoskeleton by cross-linking the plasma
membrane to the underlying microtubules/microfilaments.
Actin or actin-associated proteins would be expected
underneath the dorsal plasma membrane, since this convex
side implies cytoskeletal support (Elmendorf et al., 2003).
Researchers have tried to find homologs to microfilament
supportive proteins, such as alpha-actinin, vinculin and
tropomyosin, in the Giardia genome without success
(Elmendorf et al., 2003). This could be due to the extreme
divergence of Giardia proteins in comparison to homologs
in other organisms, but it is also possible that Giardia lacks
these proteins. Several alpha giardins were located close to
the plasma membrane (Table 5) and therefore the alpha
giardin gene family might be the functional homolog of the,
so far, missing actin-associated proteins in Giardia.
Many alpha giardins localise to the flagella (alpha-2, -5,
-9, -10, -14 and -17 giardin), thereby indicating their role in
cell motility. The flagella of Giardia, in addition to
enabling the parasite to swim, have been implicated in a
number of different processes. First, it may function in
attachment through the hydrodynamic model presented by
Holberton (1974). Second, in detachment, by aid of the
ventraldorsal beating of the caudal flagella, it helps the
parasite to bend sharply upward dorsally, thereby freeing
the parasite (Erlandsen and Feely, 1984; Ghosh et al.,
2001). The third proposed function for the flagella is in food
uptake, where the beating of the ventral flagella would
create a negative pressure in the marginal groove, thereby
sweeping nutrients under the parasite (Ghosh et al., 2001).
The more precise function of the ventral flagella could be
understood by using the antisense RNA technique (Touz
et al., 2002) to knock out the alpha giardins specifically
localised to the ventral flagella (alpha-5, -9, -10 and -17
giardin). The study of particular flagella may also help
bring light to the events of Giardia cell division, which is
not clearly understood (Elmendorf et al., 2003).
Similar to other annexins, one possible function for the
alpha giardins could be to serve as lectins, by interacting
with glycosaminoglycans on the epithelial cell lining,
mediating the intriguing initial contact with the microvilli
(Weiland et al., 2003), or through direct interaction with the
microfilament system of the host.
Several benzimidazoles are highly effective anti-giardia-
sis drugs in humans, pets and livestock (Mank and Zaat,
2001). Ultra-structural observations of cells treated with
these drugs implicate the microtubular cytoskeleton as the
drug target (Oxberry et al., 1994). Treating trophozoites
with albendazole causes profound structural effects, with
fracture of the ventral disc (Chavez et al., 1992a,b)
indicating the importance of the cytoskeleton. Similarly,
the alpha giardin gene family, with its 21 members, might
serve as interesting new drug targets. So far, no alpha
giardin has been crystallised and the structure model of
alpha-1 giardin that exists is based upon the structure of
human annexinV (Bauer et al., 1999). A crystallised alpha
giardin would help to identify putative functional domains
625
shared by the alpha giardins and in the effective develop-
ment of drugs affecting the alpha giardins. An effective drug
against the unique alpha giardins might be selective for the
parasite cytoskeleton but not the cytoskeleton of host cells.
Acknowledgements
The authors are grateful to Katarina Roxstrom-Lindquist
for critical commentary on the manuscript. SGS and ME-
LW were supported by grants from the Swedish Natural
Science Research Council and/or the Swedish Medical
Research Council. ME-LW is supported by a grant from the
I&V programme of the Strategic Foundation and by a grant
from Karolinska Institutet. AGM was supported by NIH
grant AI51089 and the Marine Biological Laboratorys
Program in Global Infectious Diseases, funded by the
Ellison Medical Foundation. Computational resources were
provided by the Josephine Bay Paul Center for Comparative
Molecular Biology and Evolution (Marine Biological
Laboratory) through funds provided by the W.M. Keck
Foundation and the G. Unger Vetlesen Foundation. The
Giardia genome project is supported by the National
Institute of Allergy and Infectious Diseases (Grant numbers
AI43273 and AI42488), the G. Unger Vetlesen Foundation,
and LI-COR Biotechnology.
References
Adam, R.D., 2001. Biology of Giardia lamblia. Clin. Microbiol. Rev.
14 (3), 447475.
Alonso, R.A., Peattie, D.A., 1992. Nucleotide sequence of a second alpha
giardin gene and molecular analysis of the alpha giardin genes and
transcripts in Giardia lamblia. Mol. Biochem. Parasitol. 50 (1), 95104.
Altschul, S.F., Koonin, E.V., 1998. Iterated profile searches with PSI-
BLASTa tool for discovery in protein databases. Trends Biochem.
Sci. 23 (11), 444447.
Bauer, B., Engelbrecht, S., Bakker-Grunwald, T., Scholze, H., 1999.
Functional identification of alpha 1-giardin as an annexin of Giardia
lamblia. FEMS Microbiol. Lett. 173 (1), 147153.
Bernander, R., Palm, J.E., Svard, S.G., 2001. Genome ploidy in different
stages of the Giardia lamblia life cycle. Cell Microbiol. 3 (1), 5562.
Boucher, S.E., Gillin, F.D., 1990. Excystation of in vitro-derived Giardia
lamblia cysts. Infect. Immun. 58 (11), 35163522.
Cardenas, M.E., Sanfridson, A., Cutler, N.S., Heitman, J., 1998. Signal-
transduction cascades as targets for therapeutic intervention by natural
products. Trends Biotechnol. 16 (10), 427433.
Chavez, B., Cedillo-Rivera, R., Martinez-Palomo, A., 1992a. Giardia
lamblia: ultrastructural study of the in vitro effect of benzimidazoles.
J. Protozool. 39 (4), 510515.
Chavez, B., Espinosa-Cantellano, M., Cedillo Rivera, R., Ramirez, A.,
Martinez-Palomo, A., 1992b. Effects of albendazole on Entamoeba
histolytica and Giardia lamblia trophozoites. Arch. Med. Res. 23 (2),
6367.
Crossley, R., Holberton, D.V., 1983a. Characterization of proteins from the
cytoskeleton of Giardia lamblia. J. Cell Sci. 59, 81103.
Crossley, R., Holberton, D.V., 1983b. Selective extraction with Sarkosyl
and repolymerization in vitro of cytoskeleton proteins from Giardia.
J. Cell Sci. 62, 419438.
626 M.E.-L. Weiland et al. / International Journal for Parasitology 35 (2005) 617626
Elmendorf, H.G., Dawson, S.C., McCaffery, J.M., 2003. The cytoskeleton
of Giardia lamblia. Int. J. Parasitol. 33 (1), 328.
Erlandsen, S.L., Feely, D.E., 1984. Trophozoite motility and the
mechanism of attachment. In: Erlandsen, S.L., Meyer, E.A. (Eds.),
Giardia and Giardiasis: Biology, Pathogenesis and Epidemiology.
Plenum Press, New York, pp. 3363.
Gerke, V., Moss, S.E., 2002. Annexins: from structure to function. Physiol.
Rev. 82 (2), 331371.
Ghosh, S., Frisardi, M., Rogers, R., Samuelson, J., 2001. How Giardia swim
and divide. Infect. Immun. 69 (12), 78667872.
Hetsko, M.L., McCaffery, J.M., Svard, S.G., Meng, T.C., Que, X.,
Gillin, F.D., 1998. Cellular and transcriptional changes during
excystation of Giardia lamblia in vitro. Exp. Parasitol. 88 (3), 172183.
Holberton, D.V., 1974. Attachment of Giardiaa hydrodynamic model
based on flagellar activity. J. Exp. Biol. 60 (1), 207221.
Huelsenbeck, J.P., Ronquist, F., 2001. MRBAYES: Bayesian inference of
phylogenetic trees. Bioinformatics 17 (8), 754755.
Huelsenbeck, J.P., Larget, B., Miller, R.E., Ronquist, F., 2002. Potential
applications and pitfalls of Bayesian inference of phylogeny. Syst. Biol.
51 (5), 673688.
Jones, D.T., Taylor, W.R., Thornton, J.M., 1992. The rapid generation of
mutation data matrices from protein sequences. Comput. Appl. Biosci.
8 (3), 275282.
Knodler, L.A., Svard, S.G., Silberman, J.D., Davids, B.J., Gillin, F.D.,
1999. Developmental gene regulation in Giardia lamblia: first evidence
for an encystation-specific promoter and differential 5 0 mRNA
processing. Mol. Microbiol. 34 (2), 327340.
Lewis, P.O., Swofford, D.L., 2001. Back to the future: Bayesian inference
arrives in phylogenetics. Trends Ecol. Evol. 16 (11), 600601.
Lujan, H.D., Mowatt, M.R., Conrad, J.T., Bowers, B., Nash, T.E., 1995.
Identification of a novel Giardia lamblia cyst wall protein with leucine-
rich repeats. Implications for secretory granule formation and protein
assembly into the cyst wall. J. Biol. Chem. 270 (49), 2930729313.
Maddison, W.P., Maddison, D.R., 1989. Interactive analysis of phylogeny
and character evolution using the computer program MacClade. Folia
Primatol. (Basel) 53 (14), 190202.
Mank, T.G., Zaat, J.O., 2001. Diagnostic advantages and therapeutic
options for giardiasis. Expert Opin. Investig. Drugs 10 (8), 15131519.
Marti, M., Hehl, A.B., 2003. Encystation-specific vesicles in Giardia: a
primordial Golgi or just another secretory compartment? Trends
Parasitol. 19 (10), 440446.
McArthur, A.G., Morrison, H.G., Nixon, J.E., Passamaneck, N.Q., Kim, U.,
Hinkle, G., Crocker, M.K., Holder, M.E., Farr, R., Reich, C.I.,
Olsen, G.E., Aley, S.B., Adam, R.D., Gillin, F.D., Sogin, M.L., 2000.
The Giardia genome project database. FEMS Microbiol. Lett. 189 (2),
271273.
Morgan, R.O., Fernandez, M.P., 1995. Molecular phylogeny of annexins
and identification of a primitive homologue in Giardia lamblia. Mol.
Biol. Evol. 12 (6), 967979.
Morgan, R.O., Fernandez, M.P, 1997. Distinct annexin subfamilies in
plants and protists diverged prior to animal annexins and from a
common ancestor. J. Mol. Evol. 44 (2), 178188.
Mowatt, M.R., Lujan, H.D., Cotten, D.B., Bowers, B., Yee, J., Nash, T.E.,
Stibbs, H.H., 1995. Developmentally regulated expression of a Giardia
lamblia cyst wall protein gene. Mol. Microbiol. 15 (5), 955963.
Nohria, A., Alonso, R.A., Peattie, D.A., 1992. Identification and
characterization of gamma-giardin and the gamma-giardin gene from
Giardia lamblia. Mol. Biochem. Parasitol. 56 (1), 2737.
Oxberry, M.E., Thompson, R.C., Reynoldson, J.A., 1994. Evaluation of
the effects of albendazole and metronidazole on the ultrastructure
of Giardia duodenalis, Trichomonas vaginalis and Spironucleus
muris using transmission electron microscopy. Int. J. Parasitol. 24 (5),
695703.
Palm, J.E., Weiland, M.E., Griffiths, W.J., Ljungstrom, I., Svard, S.G.,
2003. Identification of immunoreactive proteins during acute human
giardiasis. J. Infect. Dis. 187 (12), 18491859.
Peattie, D.A., Alonso, R.A., Hein, A., Caulfield, J.P., 1989. Ultrastructural
localization of giardins to the edges of disk microribbons of Giarida
lamblia and the nucleotide and deduced protein sequence of alpha
giardin. J. Cell Biol. 109 (5), 23232335.
Que, X., Svard, S.G., Meng, T.C., Hetsko, M.L., Aley, S.B., Gillin, F.D.,
1996. Developmentally regulated transcripts and evidence of differen-
tial mRNA processing in Giardia lamblia. Mol. Biochem. Parasitol. 81
(1), 101110.
Reiner, D.S., Hetsko, M.L., Meszaros, J.G., Sun, C.H., Morrison, H.G.,
Brunton, L.L., Gillin, F.D., 2003. Calcium signaling in excystation of
the early diverging eukaryote, Giardia lamblia. J. Biol. Chem. 278 (4),
25332540.
Shoemaker, J.S., Painter, I.S., Weir, B.S., 1999. Bayesian statistics in
genetics: a guide for the uninitiated. Trends Genet. 15 (9),
354358.
Sun, C.H., McCaffery, J.M., Reiner, D.S., Gillin, F.D., 2003. Mining the
Giardia lamblia genome for new cyst wall proteins. J. Biol. Chem. 278
(24), 2170121708.
Svard, S.G., Rafferty, C., McCaffery, J.M., Smith, M.W., Reiner, D.S.,
Gillin, F.D., 1999. A signal recognition particle receptor gene from the
early-diverging eukaryote, Giardia lamblia. Mol. Biochem. Parasitol.
98 (2), 253264.
Szkodowska, A., Muller, M.C., Linke, C., Scholze, H., 2002. Annexin
XXI(ANX21) of Giardia lamblia has sequence motifs uniquely sdhared
by giardial annexins and is specifically localized in the flagella. J. Biol.
Chem. 277 (28), 2570325706.
Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, D.G.,
1997. The CLUSTAL_X windows interface: flexible strategies for
multiple sequence alignment aided by quality analysis tools. Nucleic
Acids Res. 25 (24), 48764882.
Touz, M.C., Gottig, N., Nash, T.E., Lujan, H.D., 2002. Identification and
characterization of a novel secretory granule calcium-binding protein
from the early branching eukaryote Giardia lamblia. J. Biol. Chem. 277
(52), 5055750563.
Weber, K., Geisler, N., Plessmann, U., Bremerich, A., Lechtreck, K.F.,
Melkonian, M., 1993. SF-assemblin, the structural protein of the 2-
nm filaments from striated microtubule associated fibers of algal
flagellar roots, forms a segmented coiled coil. J. Cell Biol. 121 (4),
837845.
Weiland, M.E., Palm, J.E., Griffiths, W.J., McCaffery, J.M., Svard, S.G.,
2003. Characterisation of alpha-1 giardin: an immunodominant
Giardia lamblia annexin with glycosaminoglycan-binding activity.
Int. J. Parasitol. 33 (12), 13411351.
 International Journal for Parasitology 35 (2005) 627636
www.parasitology-online.com

Hsp90 is essential in the filarial nematode Brugia pahangi

Eileen Devaneya,*, Kerry ONeilla, William Harnettb, Luke Whitesellc, Jane H. Kinnairda
a
Parasitology Group, Division of Infection and Immunity, Institute of Comparative Medicine, University of Glasgow, Bearsden Road, Glasgow G61 1QH, UK
b
Department of Immunology, University of Strathclyde, Glasgow G4 0NR, UK
c
Steele Memorial Childrens Research Center, University of Arizona, Tuscon, AZ 85724, USA
Received 16 December 2004; received in revised form 19 January 2005; accepted 19 January 2005

Abstract
The development of a compound with activity against filarial nematodes (a macrofilaricide) has been a long-standing goal of the World
Health Organization. However, adult filariae have proved remarkably difficult to kill. To some extent this reflects a lack of understanding of
key pathways and processes in filarial nematodes that may be suitable targets for chemotherapy. In this paper we show that geldanamycin
(GA), a specific inhibitor of the activity of the heat shock protein 90 (Hsp90) family, kills adult worms and microfilariae (Mf) of Brugia
pahangi at nanomolar concentrations. In addition, release of Mf from adult worms is inhibited within 24 h of exposure to GA and is not
recoverable, demonstrating that GA effectively sterilises the worm. Similar results were obtained with a second filarial worm
Acanthocheilonema viteae. In contrast GA has no effect on the free-living nematode Caenorhabditis elegans despite a high degree of
conservation between the nematode Hsp90 sequences. In keeping with these findings, Brugia Hsp90 binds GA in a solid phase pull-down
assay while the binding of C. elegans Hsp90 to immobilised GA is undetectable. In other eukaryotes, GA is known to bind in the N-terminal
ATP pocket of Hsp90, disrupting its interactions with client proteins which are then targeted for degradation via the proteasome pathway.
Thus, Hsp90 or some of its client proteins may provide novel targets for the chemotherapy of filarial infection.
q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: Hsp90; Filarial worms; Geldanamycin; Macrofilaricide

1. Introduction differ in their dependence on Hsp90 activity (Xu and

Heat shock proteins (HSPs) are best characterised for
their role in cell survival during periods of stress, where
their ability to bind denatured or misfolded proteins is
essential for survival (Morimoto, 1998). Heat shock proteins
also function during normal growth and division as
chaperones for protein folding and transport. Heat shock
protein 90 (Hsp90) is unique amongst the family of HSPs
because of the specific nature of the proteins with which it
interacts under non-stress conditions. These include a range
of signalling and receptor molecules with important roles in
cell division, cell cycle and apoptosis (Aligue et al., 1994;
de Carcer et al., 2001). This interaction is highly specific, in
that closely related proteins such as p60v-src and p60c-src
* Corresponding author. Tel.: C44 141 330 6925; fax: C44 141 330
5603.
E-mail address: e.devaney@vet.gla.ac.uk (E. Devaney).
Lindquist, 1993). This phenomenon seems to reflect
intrinsic differences in the sequence of the kinases, their
ability to fold and their requirement for various co-
chaperones to attain stability (Xu et al., 1999). As knockout
of hsp90 is lethal in eukaryotes, drugs that inhibit Hsp90
interactions have been widely used to explore the con-
sequences of inactivating Hsp90. The most widely used of
these compounds is geldanamycin (GA), a naturally
occurring benzoquinone ansamycin, which binds in the
N-terminal ATP pocket of Hsp90 (Grenert et al., 1997).
Heat shock protein 90 functions as part of a multi-protein
complex that promotes the folding and stability of client
proteins. Geldanamycin binding alters the conformation of
Hsp90 in such a way as to destabilise client proteins, which
are then targeted for degradation via the proteasome
(Whitesell et al., 1998). Consequently in the presence of
GA, the steady state level of many Hsp90 client proteins is
reduced to undetectable levels (Bagatell et al., 2001).

0020-7519/$30.00 q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijpara.2005.01.007
628 E. Devaney et al. / International Journal for Parasitology 35 (2005) 627636
Because of its interactions with various oncogenic proteins
and steroid hormone receptors, inhibition of Hsp90 using a
derivative of GA, 17-allylaminogeldanamycin (17-AAG) is
in Phases I/II clinical trials for the chemotherapy of certain
tumours (Workman, 2003). Additionally many screens are
underway for novel compounds that inhibit Hsp90 activity
(Neckers, 2003; Chiosis et al., 2003).
As in other eukaryotes, hsp90 (daf-21) is required for
normal development in the free-living model nematode
Caenorhabditis elegans (Birnby et al., 2000). An hsp90 null
(nr2081) arrests at the L2/L3 stages, while worms expres-
sing hsp90 with a single point mutation (p673) are dauer
constitutive (Daf-c), i.e. form developmentally arrested
dauer larvae under favourable growth conditions. In
addition to its role in larval development, RNAi studies
have demonstrated a requirement for Hsp90 in adult worms.
Injection of dsRNA results in an embryonic lethal
phenotype in the progeny and the cessation of egg
production in hermaphrodite worms (Piano et al., 2000),
the severity of the hsp90 (RNAi) phenotype presumably
reflecting knock down of both maternal and zygotic RNA.
Despite these findings, exposure to GA causes no dis-
cernible phenotype in C. elegans (David et al., 2003).
Indeed Hsp90 in soluble extracts of C. elegans fails to bind
GA immobilised on a solid support, in contrast to
mammalian Hsp90. This observation was proposed to
represent an example of adaptive evolution, as C. elegans
and the Streptomyces species that synthesise GA inhabit the
same ecological niche of the soil (David et al., 2003).
In order to determine whether the resistance of
C. elegans Hsp90 to GA is a feature of other nematodes,
we investigated the effect of GA on the parasitic filarial
nematode Brugia pahangi. Two species of lymphatic
dwelling filarial worms are important human pathogens,
the related Brugia malayi and Wuchereria bancrofti. These
parasites have a two-host life cycle with adult worms living
in the afferent vessels of the lymphatics where the females
produce an abundance of first stage larvae or microfilariae
(Mf) that circulate in the blood. Microfilariae are ingested
by a feeding mosquito in which they develop to the L3, the
infective form for the mammalian host. An estimated 120 m
individuals in tropical regions of the world are infected with
lymphatic filarial parasites with an additional 1.2 billion at
risk of infection (Michael and Bundy, 1997). Here we show
that Hsp90 from B. pahangi is exquisitely sensitive to GA;
at nanomolar concentrations GA is lethal for both Mf and
adults of Brugia while exposure of adult female worms also
results in the almost immediate cessation of Mf production.
In contrast to C. elegans, Brugia Hsp90 binds GA in a solid
phase pull-down assay. Thus, despite the extremely high
degree of conservation between Hsp90 from the two
nematodes, their sensitivity to GA is markedly different.
These studies confirm that minor changes in amino acid
sequence are sufficient to alter the affinity of Hsp90 for GA
and suggest that compounds that target Hsp90 or some of its
client proteins may have clinical efficacy against filarial
nematodes in vivo.
2. Material and methods
2.1. Maintenance of B. pahangi, Acanthocheilonema viteae
and C. elegans
B. pahangi was maintained by serial passage in
mosquitoes (Aedes aegypti) and Mongolian jirds (Meriones
ungiculatus). Adult worms or Mf of B. pahangi were
obtained by peritoneal lavage of jirds infected at least 3
months previously. Parasites were washed in Hanks
Balanced Salt Solution (HBSS, Invitrogen) at 37 8C and
Mf were purified from host cells exactly as described
previously (Devaney et al., 1992). Third stage larvae were
obtained from infected mosquitoes using standard pro-
cedures (Devaney and Jecock, 1991). Microfilariae and
adult worms of A. viteae were harvested from infected jirds
(Meriones libycus) as described previously (Steptek et al.,
2004). For the experiments in which parasites were
subjected to different temperatures, they were incubated in
water baths at 28, 37 or 41 8C in RPMI 1640 plus 25 mM
Hepes, 5 mM glutamine, 100 m/ml of penicillin, 100 mg/ml
of streptomycin and 5% heat inactivated FCS (all Invitro-
gen), as described previously (Devaney et al., 1992).
Worms of the N2 strain of C. elegans were maintained on
agar plates at a temperature of 21 8C with OP50 as a
bacterial food source (Lewis and Fleming, 1995). Worms
were washed off plates with M9 buffer and collected by
centrifugation, following which they were washed exten-
sively in M9. Worms were heat shocked at 33 or 35 8C and
either washed off plates immediately or allowed to recover
for 1 h at 21 8C.
2.2. Western blot analysis of Hsp90 expression in different
life cycle stages
Worms were extracted directly into SDS-PAGE sample
cocktail by boiling for 5 min. Aliquots of worm extract
containing equivalent amounts of protein (as assessed by
Comassie blue staining of replicate gels) were run on 10%
SDS-polyacrylamide gels and blotted onto nitro-cellulose
paper using standard methods. Blots were probed with an
antiserum raised against recombinant B. pahangi Hsp90 . A
714 bp construct expressing the C-terminal 238 amino acid
residues of Hsp90 was obtained by PCR amplification from
a cDNA template using primers 5 0 -CTGGTGGATCCAGG-
GAAGCTGTCGCC-3 0 and 5 0 - CCATGGTACCTTAAT-
CAACTTCTTCCATCC-3 0 (restriction sites underlined).
PCR conditions were one cycle of 94 8C for 1 min, 55 8C for
1 min and 72 8C for 3 min, followed by 29 cycles of 94 8C
for 1 min, 65 8C for 1 min and 72 8C for 3 min. The
amplified product was digested and ligated into the BamH I
and Kpn I sites of plasmid pQE30 (Qiagen), which contains
 E. Devaney et al. / International Journal for Parasitology 35 (2005) 627636
a 6! Histidine tag, and transformed into the host strain
M15 [pREP4] following the manufacturers protocol.
Three plasmids containing inserts of the correct size were
fully sequenced and one was used for production of the
28 kDa recombinant protein, largely according to the
manufacturers recommendations. Briefly, the protein was
induced with 1 mM IPTG for 4 h, and the cells resuspended
in lysis buffer containing 1 mg/ml lysozyme prior to
sonication. The fusion protein was purified from the soluble
supernatant on a Ni-agarose affinity column and eluted with
a 0.10.5 M imidazole gradient. The fractions were
analysed by SDS-PAGE and those containing the recombi-
nant protein were dialysed against PBS and used to
immunise rabbits.
Two rabbits were immunised with approximately 130 mg
recombinant protein emulsified in Freunds incomplete
adjuvant followed by two boosts. The resulting antibody
was used at 1:10001:2000 dilution (see Section 3 for details)
followed by a 1:20,000 dilution of HRP-labelled anti-rabbit
IgG (Sigma). Bound antibody was detected by chemilumi-
nescence using the Super Signalw West Pico Kit (Pierce).
2.3. Treatment of B. pahangi with GA
A single female worm or two male worms or approxi-
mately 2!103 purified Mf were cultured in a 2.0 ml volume
of RPMI 1640 containing 25 mM Hepes, 5 mM glutamine,
100 u/ml of penicillin and 100 mg/ml of streptomycin and
5% heat inactivated FCS (all Invitrogen) in 24 well tissue
culture plates. GA (Sigma) was dissolved in DMSO to give
a stock of 10 mM which was then diluted in culture medium
to the appropriate concentration (1.0 mM maximum).
Controls included worms incubated with the appropriate
volume of carrier DMSO or cultured in medium alone. Six
replicate wells containing adult worms or duplicate wells for
Mf were set up for each concentration of drug or control.
Worms were incubated at 37 8C in an atmosphere of 5%
CO2 in air and monitored daily by microscopy for viability
and Mf output. In some experiments, adult female worms
were removed from the drug after 24 h exposure and
incubated for a further 7 days in medium without drug. For
experiments involving adult A. viteae, individual female
worms, which are significantly larger than Brugia, were
incubated in 10 ml medium plus/minus drug or DMSO.
The release of Mf from adult female worms was assessed
at appropriate times (see Section 3 for details). Adult female
worms were removed to fresh wells containing drug or
medium alone and Mf in the spent medium collected by
centrifugation. Microfilariae were counted using an inverted
microscope and results presented as the mean outputGSD
of six individual female worms. Statistical significance
between groups was calculated using the Mann Whitney
test, with P values !0.05 being considered significant.
Adult female worms were dissected at various times post-
exposure to GA (2 days onwards). Worms were placed on
glass slides in a drop of HBSS and cut into four sections
629
using a sterile scalpel. The contents of the female
reproductive system were extruded and examined using a
Zeiss AxioScope 2 with an Openlab imaging system
2.4. Preparation of GA matrix and pull-down assays
One mg GA was derivatised exactly as described
(Whitesell et al., 1994) using 1,6, hexanediamine to yield
17-hexamethylenediamine-17 demethoxygeldanamycin.
This compound was re-dissolved in DMSO and incubated
with 400 ml of washed Affi-Gel 10 beads (Bio-Rad) for 4 h
in the dark. Control beads were reacted with DMSO alone.
Beads were washed and then rocked o/n in buffer to block
residual binding sites.
Lysates of B. pahangi adult worms or mixed stage
C. elegans were prepared in ice-cold TNES buffer (50 mM
TrisHCl, pH 7.4, 100 mM NaCl, 2 mM EDTA, 1% NP-40)
with protease inhibitors added just prior to use (Complete
Mini Protease Inhibitor Cocktail, Roche Diagnostics). A
pellet of C. elegans worms was ground to a fine powder in a
pestle and mortar containing liquid Nitrogen. Ground
worms were re-suspended in 1.0 ml of TNES, following
which the mortar was washed out with a further 0.5 ml of
buffer and the extract incubated on ice for 15 min with
intermittent vortexing. Insoluble material was pelleted at
14,000 g for 15 min at 4 8C, after which the supernatant was
removed and assayed for protein concentration using the
Bio-Rad protein assay. Adult B. pahangi were ground in a
glass homogeniser in ice-cold TNES and then processed as
described for C. elegans.
For the pull-down, 300500 mg of total protein in a
volume of 300 ml was mixed with 25 ml packed volume of
GA or control beads for 2 h at 4 8C. In some experiments,
soluble GA to a final concentration of 20 mM, or an
equivalent volume of DMSO, was added to the incubation
mix. The beads were washed by centrifugation with TNES
and bound protein eluted by boiling in SDS-PAGE sample
cocktail. Aliquots of the input fraction, and protein eluted
from control and GA beads were run on a 10% gel and
analysed by western blotting with the anti-Hsp90 antiserum,
as described previously.
3. Results
3.1. The sequence of Brugia and C. elegans Hsp90 is highly
conserved
The molecular architecture and sequence of Hsp90
proteins is similar in most eukaryotes, with a modular
structure consisting of the N-terminal domain with a well-
characterised ATP-binding site in which GA binds
(Grenert et al., 1997), a highly charged linker domain and
the C-terminus containing the dimerization domain (Minami
et al., 1994). More recent analysis suggests that the
C-terminal domain contains an additional ATP-binding
630 E. Devaney et al. / International Journal for Parasitology 35 (2005) 627636

site, although GA appears not to bind at this site (Csermely et
al., 1998; Soti et al., 2002). Comparison of the predicted
amino acid sequence of B. pahangi and C. elegans Hsp90
reveals a high degree of conservation over the
whole sequence (87% identical, 92% similar overall).
The N-terminal domain is particularly well conserved
between the two nematodes with a minimal number of
substitutions (see Fig. 1). In comparison with human and
Brugia Hsp90 both of which are susceptible to GA, there are a
few amino acid residues that differ only in C. elegans. Most
notably, the charged linker region of C. elegans Hsp90
contains fewer amino acids than Brugia and human Hsp90. In
addition, there are 15 other substitutions in C. elegans Hsp90
which are conserved between human and B. pahangi
Hsp90, all of which are confined to the linker region or
the C-terminus. Some are conservative substitutions, while
others are not and potentially could contribute to confor-
mational differences between B. pahangi and C. elegans

Fig. 1. Hsp90 is conserved in Brugia pahangi and Caenorhabditis elegans. Figure shows a Clustal-W alignment of Hsp90 from B. pahangi (Bp, Accession
number AJ005784), C. elegans (Ce, Accession number Z75530) and human Hsp90a (Hs, Accession number NM005348). Identical amino acids are in bold and
boxed, similar amino acids are in bold. **Denotes the start and the end of the N-terminal ATP-binding domain. Brugia pahangi Hsp90 is 80% identical and
87% similar to human Hsp90 and 87% identical, 92% similar to C. elegans Hsp90; C. elegans Hsp90 is 78% identical and 84% similar to human Hsp90.
 E. Devaney et al. / International Journal for Parasitology 35 (2005) 627636 631

Hsp90 (e.g. E to S at position 219 of C. elegans Hsp90, N to D

at 238, S to K at 483, A to K at 536 and T to E at 683). While
many organisms contain two hsp90 genes (a and b), the
C. elegans genome contains only one hsp90 (daf-21).
Previous studies demonstrated that the B. pahangi hsp90
gene cloned in this laboratory was single copy as determined
by Southern blotting (Thompson et al., 2001), a finding
supported by examination of the B. malayi genome (http://
www.tigr.org/tdb/e2k1/bma1/).

3.2. The expression of Hsp90 in Brugia and C. elegans is

constitutive and not significantly altered by heat shock

Analysis of proteins extracted from different life cycle

stages of both nematodes demonstrates that Hsp90 is
constitutively expressed, a band of w83 kDa being detected
in all life cycle stages of Brugia by western blot using an
antibody raised to recombinant B. pahangi Hsp90 (Fig. 2a).
As the L3 and the Mf of Brugia are subject to a heat shock as
they are transmitted between mammalian and mosquito
hosts (2837 8C), we investigated whether Hsp90 was up-
regulated co-incident with this transition or following
exposure to a more profound heat shock (41 8C). No
significant increase in the level of Hsp90 was observed in
any life cycle stage of Brugia exposed to elevated
temperatures (see Fig. 2b for Mf). These results correlate
well with previous analysis of Brugia Mf and adult worms,
where a minimal induction of hsp90 mRNA levels (1.21.5-
fold) was observed following 2 h of heat shock (Thompson
et al., 2001). However, in that study there was a significant
decrease in hsp90 mRNA in Mf exposed to 28 8C, while
levels of Hsp90 protein do not vary over a short exposure to
lowered temperature (see Fig. 2b). The difference in mRNA
and protein profiles at 28 8C presumably reflects the
abundance and stability of Hsp90.
A similar analysis was undertaken with C. elegans, in
which L4 worms were exposed to a range of temperatures
and Hsp90 levels investigated by western blot using the
cross-reactive Brugia antibody, which detects a protein of
w80 kDa in all life cycle stages of C. elegans (data not
shown). As with Brugia, Hsp90 levels did not differ
significantly following a short exposure to elevated
temperature (Fig. 2c). Worms were allowed to recover
from heat shock for varying periods of time up to 7.5 h, but
no significant induction of Hsp90 was observed (data not
shown). These results suggest that Hsp90 is not a major
heat-inducible protein in C. elegans or Brugia.
3.3. GA inhibits Mf release from B. pahangi and has
macrofilaricidal activity
An intriguing feature of C. elegans Hsp90 is its apparent
resistance to GA. To examine whether this phenomenon is
common to other nematodes, adult female worms of
B. pahangi were incubated in medium containing 1 mM
GA (a concentration known to inhibit mammalian Hsp90
Fig. 2. Levels of Hsp90 are not significantly affected by temperature in
Brugia pahangi or Caenorhabditis elegans. (a) Different life cycle stages of
B. pahangi were lysed in 2!SDS-PAGE sample cocktail and analysed on a
10% SDS-polyacrylamide gel. Equal amounts of protein (as assessed by
Coomassie blue staining) of Mf (lane 1), L3 (lane 2), adult females (lane 3)
or adult males (lane 4) were analysed. Proteins were transferred to NCP
using standard methods, blocked o/n in 5% dried milk in PBS/0.5% Tween
20 and then the blot probed with 1:2000 dilution of the anti-Hsp90 Ab.
Bound Ab was detected using a 1:20,000 dilution of anti-rabbit IgG
conjugated to HRP. The blot was developed using the Pierce Super Signal
West Pico chemiluminescence kit. (b) Mf were incubated for 60 min at
28 8C (lane 1), 37 8C (lane 2) or 41 8C (lane 3) in tissue culture medium
(RPMI 1640 plus 25 mM Hepes, 5 mM glutamine, 100 u/ml of penicillin
and 100 mg/ml of streptomycin and 5% heat inactivated FCS) then lysed
and analysed exactly as described above. (c) Mixed stages of C. elegans
were cultured at 21 8C, or exposed to 33 or 35 8C for 1 h, followed by a 1 h
recovery at 21 8C. Worms were lysed directly into sample cocktail and
analysed on a 10% SDS-PAGE gel as described above.
activity) and their viability and Mf output assessed. GA had
an immediate and profound effect on Mf release. Within
24 h of exposure to the drug, there was a significant decrease
in the numbers of Mf released from adult worms (see
Fig. 3a, P!0.0001, GA vs DMSO or medium alone control)
and by 48 h, Mf release had ceased altogether. Similar
results were obtained at 500 nM GA, but not at 100 nM GA.
To determine whether the inhibition of Mf production
was reversible, worms exposed to 1.0 mM GA for 24 h were
washed and then incubated for a further 48 h in medium
without drug. Fig. 3b shows that adult worms fail to recover
the capacity to release Mf when removed from the drug. In
some experiments, incubations were continued for up to 7
632 E. Devaney et al. / International Journal for Parasitology 35 (2005) 627636
(a) 3000
2500
2000
Mf numbers
1500
1000
500
0 3.4. GA has direct microfilaricidal activity against Brugia
GA DMSO Control
(b) 3000
2500
2000
Mf numbers
1500
1000
500
0 62.5 nM. The effect of GA was progressive with time and at
GA DMSO Control
Fig. 3. The release of microfilariae does not recover following 24 h
exposure to geldanamycin. Six adult female Brugia pahangi were
incubated individually in 2.0 ml of tissue culture medium containing
1.0 mM GA, an equivalent volume of DMSO or medium alone for 24 h (a).
Worms were removed to fresh wells containing 2.0 ml of tissue culture
medium alone for a further 48 h and Mf output assessed (b). Graphs show
Mf output as the meanGSD of six worms per group. These results were
representative of four separate experiments.
days, but on no occasion did Mf release recover. Similar
results were obtained at 500 nM GA.
Although Mf output was significantly different within
24 h of exposure to GA, there was no obvious effect on adult
worms at this time. However, by 72 h the female worms
exposed to GA appeared flaccid and became progressively
more lethargic until, by days 78, 100% were dead. As the
effect of GA on adult female worms might simply reflect the
retention of Mf within the uterus, male worms were exposed
to 1.0 mM and 500 nM GA. Male worms were more
sensitive to GA than females with motility being severely
compromised by 48 h of exposure. By days 56, 100% of
adult males were dead following exposure to both
concentrations of drug. Thus GA appears to have a direct
macrofilaricidal effect on Brugia.
In order to investigate how GA might affect Mf release
from adult worms, the contents of the female reproductive
system were examined following exposure to the drug. The
uterine contents of untreated worms or worms exposed to
DMSO alone revealed a mass of active, wriggling Mf
(Fig. 4a), while in the GA treated worms, the mature Mf
were elongate and motionless (Fig. 4c). Released from
the confines of the uterus, the Mf did not regain their
motility and were clearly dead. Examination of other parts
of the reproductive system, revealed an effect on developing
embryos, which were dead and vacuolated in the treated
worms (see Fig. 4d) compared to embryos from control
worms which were undergoing normal morphogenesis
(Fig. 4b). These observations suggest that GA kills Mf
and embryonic stages in utero.
Microfilariae normally exhibit a vigorous thrashing
activity, but those produced by adult females in cultures
containing 1.0 mM or 500 nM GA were noticeably lethargic
and died over a period of 34 days. To further investigate
this effect, Mf purified from the peritoneal cavity of infected
animals were cultured in GA concentrations ranging from
1 mM to 1.0 nM. At high concentrations of drug (1 mM to
250 nM), the effects of GA were visible within 48 h of
exposure, with Mf appearing to be paralysed by the drug. By
day 5, 100% of Mf were dead at concentrations down to
lower concentrations (31.25 and 15.625 nM) the drug
affected the motility of Mf but did not kill them. No effect
was observed at concentrations !15 nM over the time scale
tested (up to 10 days).
3.5. A filarial species lacking Wolbachia is also susceptible
to GA
Most species of pathogenic filarial worms are known to
contain endosymbiotic bacteria related to rickettsia, and it is
now well documented that treatment with antibiotics such as
tetracycline affects worm viability probably by targeting the
Wolbachia endosymbiont (Hoerauf et al., 2002; Chirgwin
et al., 2003). To determine whether the activity of GA was
directed against Brugia Hsp90 or the Wolbachia homologue
of Hsp90, Mf and adult worms of the Wolbachia-free
species A. viteae were cultured with GA. As observed with
B. pahangi, Mf of A. viteae were killed by exposure to GA at
nanomolar concentrations. By day 4 of culture, 100% of Mf
were dead at 1.0 mM and 500 nM GA, with 50% mortality at
250 and 125 nM. However, as the Mf of A. viteae survive
less well in vitro than do B. pahangi Mf, the experiments
were not continued beyond day 56. Adult worms of
A. viteae were also affected by exposure to GA with 100%
dying by 96 h at 1.0 mM and 500 nM GA. Similarly Mf
output was significantly reduced after 24 h of exposure to
1.0 mM and 500 nM GA (see Table 1), and was reduced to
almost zero after 72 h exposure to drug. Although a trend
was noted in these experiments (as in those with Brugia) for
an increased release of Mf in the DMSO controls compared
to medium only, the difference was not significant in any
experiment.
 E. Devaney et al. / International Journal for Parasitology 35 (2005) 627636 633

Fig. 4. Geldanamycin is lethal to microfilariae in utero. Adult female worms exposed to medium alone (a and b) or to 1.0 mM GA for 48 h (c and d) were
dissected on glass slides in a drop of HBSS. The contents of the reproductive system were examined under the !40 lens of a Zeiss AxioScope 2 microscope
and photographed using an Openlab imaging system. Panel a shows live coiled Mf in utero in a control worm; panel b shows live embryos from a control worm;
panel c shows dead Mf in utero following 48 h exposure to GA; panel d shows dead vacuolated embryos at the comma stage of development following 48 h
exposure to GA. Scale bar equals 20 mm.

3.6. Brugia Hsp90 binds GA in a solid phase
pull-down assay
The effect of GA on filarial worms contrasts with its lack
of activity on the free-living C. elegans, which is not
susceptible to GA. To determine whether there were
corresponding differences in the affinity of worm Hsp90
for GA, soluble extracts of adult B. pahangi or mixed stage
C. elegans were incubated with GA immobilised on a solid
support. Proteins that bound to the GA matrix were analysed
by western blotting using the antiserum to B. pahangi
Hsp90. These experiments demonstrated that B. pahangi
Hsp90 bound the GA beads (see Fig. 5, lane 5) while, under
the conditions tested, it was not possible to detect binding of
C. elegans Hsp90 (lane 3). Furthermore the binding of
B. pahangi Hsp90 was significantly reduced by the addition
of 20 mM GA compared to addition of the DMSO carrier
(see Fig. 5, lane 7), demonstrating the specificity of the
interaction. While the antibody used in these studies detects
Table 1
The release of microfilariae from adult Acanthocheilonema viteae cultured in the presence or absence of geldanamycin
Brugia Hsp90 with a higher affinity than C. elegans Hsp90,
this does not explain the difference in binding to
immobilised GA. Prolonged exposure of autoradiographs
failed to detect a specific signal with the C. elegans extract.
4. Discussion
Using a specific inhibitor of Hsp90 activity, we show that
Hsp90 is essential in the filarial nematode B. pahangi, as it is
in the free-living nematode C. elegans. Indeed the results of
inhibiting Hsp90 in B. pahangi using GA bear a striking
similarity to those described for depletion of hsp90 by RNAi
in C. elegans (Piano et al., 2000). In both nematodes
inhibition/depletion of Hsp90 results in cessation of egg
production and embryonic death. Treatment of adult female
Brugia with GA results in the retention of Mf within the
adult female, an effect similar to the egl (egg-laying
defective) phenotype in C. elegans that causes bagging

Treatment Control DMSO GA 500 nM GA 1.0 mM

1
Mf output 24 h 3980G2552 6325G2914 1260*G597 608*G505
Mf out put 72 h2 17,270G9972 21,575G10,223 300*G711 42*G80

Control vs DMSO, not significant at either time point. *Significantly different at P!0.05.
a
Mf release over the first 24 h of culture.
b
Cumulative Mf release over the subsequent 48 h.
634 E. Devaney et al. / International Journal for Parasitology 35 (2005) 627636
Fig. 5. Brugia Hsp90 binds immobilised geldanamycin. Lane 1, detergent
extract of mixed stage Caenorhabditis elegans or Brugia pahangi adult
worms (lane 2); pull-down using the GA matrix and C. elegans (lane 3), or
Brugia extract (lane 5); control pull-down using Affi-gel beads alone for
C. elegans (lane 4) and for Brugia (lane 6). Lane 7 shows a pull-down with
the Brugia input incubated with the GA matrix in the presence of 20 mM
GA or an equivalent volume of DMSO (lane 8). Samples were run on 10%
SDS-polyacrylamide gels, blotted and blots probed with a 1:1000 dilution
of rabbit anti-Hsp90 Ab, except for lanes 1 and 2, which were probed with a
1:2000 dilution of anti-Hsp90 antibody. These results were representative
of three separate experiments.
in adult hermaphrodites and compromises viability. How-
ever, exposure of adult male Brugia to GA was also lethal
suggesting that the drug is directly macrofilaricidal in
addition to its effect on Mf release.
By analogy with its effects on mammalian cells, GA
presumably kills filarial parasites by disrupting interactions
between Hsp90 and key client proteins required for
viability. Although many targets of Hsp90 have been
identified in mammalian cells and yeast (Richter and
Buchner, 2001; Young et al., 2001), few interactions have
been defined in nematodes. Until recently the only known
target of Hsp90 in C. elegans was UNC-45 a protein
involved in muscle assembly (Barral et al., 2002). As with
many Hsp90 co-chaperones (Young et al., 1998), UNC-45
contains a TPR (tetratricopeptide repeat) domain that binds
the C-terminus of Hsp90. The unc-45 loss of function
phenotype is embryonic lethal (Venolia et al., 1990, 1999),
with embryos arresting at the two-fold stage while animals
bearing temperature sensitive alleles of unc-45 are paral-
ysed at the restrictive temperature. However, the large-scale
interactome projects underway with C. elegans have
identified a variety of proteins that interact with Hsp90,
including kinases and nuclear hormone receptors, additional
TPR proteins and components of the TGF-b signalling
pathway (Li et al., 2004; Tewari et al., 2004). Two of the
proteins identified in these screens, DAF-1 and DAF-4, are
the type 1 and type-2 TGF-b receptors, which together form
a complex that is activated by the DAF-7 ligand (Tewari
et al., 2004). Under normal growth conditions DAF-7 is
expressed in the ASI neurone, but is down-regulated when
worms are exposed to dauer-inducing conditions (Ren et al.,
1996). By analogy with its function in other eukaryotes,
Hsp90 may stabilise the TGF-b receptor complex in
C. elegans; thus the Daf-c phenotype of worms expressing
hsp90 with a single point mutation may stem from reduced
levels of Hsp90 (Yue et al., 1999) or from alterations in the
binding properties of the mutant protein with consequent
defects in TGF-b signalling. These studies demonstrate that
the known interactions of Hsp90 in C. elegans are in
keeping with those described in other organisms and thus
may be similarly conserved in other nematodes. In this
respect it is interesting to note that several genes with
homology to components of the TGF-b signalling pathway
have been cloned from Brugia (Gomez-Escobar et al., 1997,
2000).
Hsp90 was not significantly induced by heat shock in
either C. elegans or B. pahangi. As the experiments with
Brugia were of necessity carried out under in vitro
conditions, it is possible that the worms were already
stressed and that further expression of Hsp90 was not
observed following the application of a heat shock.
However, previous studies on a small HSP of Brugia are
not consistent with this hypothesis. The small HSPs are
heat-inducible in Brugia but are not expressed when adult
worms are cultured under similar conditions to those used in
the present study (Devaney et al., 1992). Likewise, our
previous studies demonstrated that knock-down of the heat
shock transcription factor in C. elegans by RNAi severely
compromised the heat-induced expression of the small HSP,
Hsp16-2, but not that of Hsp90 (Walker et al., 2003). Taken
together these results suggest that Hsp90 is not a major heat-
inducible protein in nematodes.
The differing affinity of Hsp90 from C. elegans and
Brugia to GA is intriguing, given the degree of conservation
between the two nematode proteins. These results were
substantiated by the pull-down assay in which C. elegans
Hsp90 failed to bind a GA matrix in contrast to Brugia
Hsp90. Previous studies in which C. elegans Hsp90 was
expressed in yeast indicated that the free-living nematode
protein was relatively resistant to GA, requiring four-fold
more GA to inhibit growth compared to yeast expressing
human Hsp90b (Piper et al., 2003a). Point mutations in
yeast hsp90 have been shown to significantly affect GA
binding (Piper et al., 2003b) although none of the critical
residues so far identified in yeast differ in C. elegans (data
not shown). However, it is noteworthy that substitution of
amino acid residues outside the N-terminal ATP-binding
pocket of yeast Hsp90 can affect GA binding, e.g. A587T
increases the affinity of Hsp90 for GA (Piper et al., 2003b).
Comparison of the C. elegans and Brugia sequences
highlights a few differences for future study, none of
which are in the N-terminus. Recent studies have confirmed
an additional ATP-binding domain at the C-terminus of
Hsp90 (Soti et al., 2002) and the two nucleotide-binding
domains appear to interact in such a way that occupancy of
one ATP-binding pocket influences nucleotide binding in
the second (Soti et al., 2002). However, domain swap
experiments with C. elegans Hsp90 suggested that the
diminished affinity for GA mapped to the N-terminus
(David et al., 2003). Consequently, a proper understanding
of the differing affinities of Brugia and C. elegans Hsp90 for
GA will require detailed structural and mutational studies,
an analysis that will be facilitated by comparison with
additional sequences from other parasitic nematodes.
B. pahangi, in common with most other filarial species,
contains a Wolbachia endosymbiont, which appears to be
required for viability (Chirgwin et al., 2003). The genome of
the Wolbachia from B. malayi has recently been sequenced
 E. Devaney et al. / International Journal for Parasitology 35 (2005) 627636
and contains a homologue of the bacterial hsp90 gene, HtpG
(Foster et al., 2004). In contrast to eukaryotic hsp90,
Eschereria coli htpG is not essential and cannot comp-
lement yeast deficient in hsp90 (Palmer et al., 1995).
However, HtpG is GA-sensitive (Panaretou et al., 1998). To
determine whether the effect of GA on Brugia reflected
inhibition of filarial or Wolbachia Hsp90, Mf and adult
worms of the Wolbachia-free species A. viteae were
exposed to the drug. However, A. viteae adults and Mf
were equally susceptible to GA, suggesting that filarial
Hsp90 is the target of GA.
These results demonstrate that inhibition of Hsp90
activity may represent a novel target for chemotherapy of
filariasis. Given its poor solubility and problems with
toxicity, GA itself would not be acceptable for the
chemotherapy of filariasis. However, considerable effort is
being invested in identifying, designing and refining
compounds that target Hsp90 (Workman, 2003; Neckers,
2003; Chiosis et al., 2003). For example, 17-AAG is
entering Phase II trials for various malignancies, while a
more recent derivative 17-(Dimethylaminoethylamino)-17-
Demethoxygeldanamycin (17-DMAG) is beginning Phase I
testing (reviewed in Bagatell and Whitesell, 2004). This
compound has the advantage of being water-soluble, orally
bioavailable and less hepato-toxic than 17-AAG (Egorin
et al., 1998; Kaur et al., 2004). Although the degree of
similarity between the human and parasite molecules may
yet rule out Hsp90 as a target, it remains possible that some
of the differences in Hsp90 sequence may confer differential
sensitivity to GA or related compounds. Furthermore,
defining the proteins with which Hsp90 complexes in
Brugia and which are essential for filarial viability may
provide additional novel targets for the chemotherapy of
filariasis.
Acknowledgements
We would like to acknowledge the assistance of
Margaret McFadyen and Caitlan Egan for the maintenance
of the B. pahangi and A. viteae life cycles, respectively. Part
of this work was supported by a MRC grant. JK was
supported by the Wellcome Trust.
References
Aligue, R., Akhavan-Niak, H., Russell, P., 1994. A role for Hsp90 in cell
cycle control: Wee 1 tyrosine kinase activity requires interaction with
Hsp90. Eur. Med. Biol. Org. J. 13, 60996106.
Bagatell, R., Whitesell, L., 2004. Altered Hsp90 function in cancer: A
unique therapeutic opportunity. Mol. Cancer Ther. 3, 10211030.
Bagatell, R., Khan, O., Paine-Murrieta, G., Taylor, C.W., Akinaga, S.,
Whitesell, L., 2001. Destabilisation of steroid hormone receptors by
heat shock protein 90-binding drugs: a ligand independent approach to
hormonal therapy of breast cancer. Clin. Cancer Res. 7, 20762084.
635
Barral, J.M., Hutagalung, A.H., Brinker, A., Hartl, F.U., Epstein, H.F.,
2002. Role of the myosin assembly protein UNC-45 as a molecular
chaperone for myosin. Science 295, 669671.
Birnby, D.A., Link, E.M., Vowels, J.J., Tian, H., Colacurcio, P.L.,
Thomas, J.H., 2000. A transmembrane guanylyl cyclase (DAF-11)
and Hsp90 (DAF-21) regulate a common set of chemosensory
behaviours in Caenorhabditis elegans. Genetics 155, 85104.
Chiosis, G., Lucas, B., Huezo, H., Solit, D., Basso, A., Rosen, N., 2003.
Development of purine-scaffold small molecule inhibitors of Hsp90.
Curr. Cancer Drug Targets 3, 371376.
Chirgwin, S.R., Nowling, J.M., Coleman, S.U., Klei, T.R., 2003. Brugia
pahangi and Wolbachia: the kinetics of bacteria elimination, worm
viability, and host responses following tetracycline treatment. Exp.
Parasitol. 103, 1626.
Csermely, P., Schnaider, T., Soti, C., Prohaszka, Z., Nardai, G., 1998. The
90-kDa molecular chaperone family: structure, function, and clinical
applications. A comprehensive review. Pharmacol. Ther. 79, 129168.
David, C.L., Smith, H.E., Raynes, D.A., Pulcini, E.J., Whitesell, L., 2003.
Expression of a unique drug-resistant Hsp90 ortholog by the nematode
C. elegans. Cell Stress Chaperones 8, 93104.
de Carcer, G., do Carmo Avides, M., Lallena, M.J., Glover, D.M.,
Gonzalez, C., 2001. Requirement of Hsp90 for centrosomal function
reflects its regulation of Polo kinase stability. Eur. Mol. Biol. Org. J. 20,
28782884.
Devaney, E., Jecock, R.M., 1991. The expression of the Mr 30,000
antigen in the third stage larvae of Brugia pahangi. Parasite
Immunol. 13, 7587.
Devaney, E., Egan, A., Lewis, E., Warbrick, E.V., Jecock, R.M., 1992. The
expression of small heat shock proteins in the microfilaria of Brugia
pahangi and their possible role in development. Mol. Biochem.
Parasitol. 56, 209218.
Egorin, M.J., Rosen, D.M., Wolff, J.H., Callery, P.S., Musser, S.M.,
Eiseman, J.L., 1998. Metabolism of 17-(allylamino)-17-demethoxy-
geldanamycin (NSC 330507) by murine and human hepatic prep-
arations. Cancer Res. 58, 23852396.
Foster, J.M., Kumar, S., Ganatra, M.B., Kamal, I.H., Ware, J., Ingram, J.,
Pope-Chappell, J., Guiliano, D., Whitton, C., Daub, J., Blaxter, M.L.,
Slatko, B.E., 2004. Construction of bacterial artificial chromosome
libraries from the parasitic nematode Brugia malayi and physical
mapping of the genome of its Wolbachia endosymbiont. Int. J. Parasitol.
34, 733746.
Gomez-Escobar, N., van den Biggelaar, A., Maizels, R.M., 1997. A
member of the TGF-b receptor gene family in the parasitic nematode
Brugia pahangi. Genetics 199, 101109.
Gomez-Escobar, N., Gregory, W.F., Maizels, R.M., 2000. Identification of
tgh-2, a filarial nematode homolog of Caenorhabditis elegans daf-7 and
human transforming growth factor b, expressed in microfilarial and
adult stages of Brugia malayi. Infect. Immun. 68, 64026410.
Grenert, J.P., Sullivan, W.P., Fadden, P., Haystead, T.A., Clark, J.,
Mimnaugh, E., Krutzsch, H., Ochel, H.J., Schulte, T.W., Sausville, E.,
Neckers, L.M., Toft, D.O., 1997. The amino-terminal domain of heat
shock protein 90 (Hsp90) that binds geldanamycin is an ATP/ADP
switch domain that regulates Hsp90 conformation. J. Biol. Chem. 272,
2384323850.
Hoerauf, A., Adjei, O., Buttner, D.W., 2002. Antibiotics for the treatment of
onchocerciasis and other filarial infections. Curr. Opin. Investig. Drugs
3, 533537.
Kaur, G., Belotti, D., Burger, A.M., Fisher-Nielson, K., Borsotti, P.,
Riccardi, E., Thillainathan, J., Hollingshead, M., Sausville, E.A.,
Giavazzi, R., 2004. Antiangiogenic properties of 17-(dimethylami-
noethylamino)-17-demethoxygeldanamycin: an orally bioavailable
heat shock protein 90 modulator. Clin. Cancer Res. 10, 48134821.
Lewis, J.A., Fleming, J.T., 1995. Basic culture methods. Methods Cell Biol.
48, 329.
Li, S., Armstrong, C.M., Bertin, N., Ge, H., Milstein, S., Boxem, M.,
Vidalain, P.O., Han, J.D., Chesneau, A., Hao, T., Goldberg, D.S., Li, N.,
636 E. Devaney et al. / International Journal for Parasitology 35 (2005) 627636
Martinez, M., Rual, J.F., Lamesch, P., Xu, L., Tewari, M., Wong, S.L.,
Zhang, L.V., Berriz, G.F., Jacotot, L., 2004. A map of the interactome
network of the metazoan C. elegans. Science 303, 540543.
Michael, E., Bundy, D.A., 1997. Global mapping of lymphatic filariasis.
Parasitol. Today 13, 472476.
Minami, Y., Kimura, Y., Kawasaki, H., Suzuki, K., Yahar, I., 1994. The
carboxy terminal region of mammalian HSP90 is required for its
dimerization and function in vivo. Mol. Cell Biol. 14, 14591464.
Morimoto, R.I., 1998. Regulation of the heat shock transcriptional
response: cross talk between a family of heat shock factors, molecular
chaperones, and negative regulators. Genes Dev. 12, 37883796.
Neckers, L., 2003. Development of small molecule Hsp90 inhibitors:
utilizing both forward and reverse chemical genomics for drug
identification. Curr. Med. Chem. 10, 733739.
Palmer, G., Louvian, J.F., Tibbetts, R., Engman, D.M., Picard, D., 1995.
Trypanosma cruzi heat-shock protein 90 can functionally complement
yeast. Mol. Biochem. Parasitol. 70, 199202.
Panaretou, B., Prodromou, C., Roe, S.M., OBrien, R., Ladbury, J.E.,
Piper, P.W., Pearl, L.H., 1998. ATP binding and hydrolysis are essential
to the function of the Hsp90 molecular chaperone in vivo. Eur. Mol.
Biol. Org. J. 17, 48294836.
Piano, F., Schetter, A.J., Mangone, M., Stein, L., Kemphues, K.J., 2000.
RNAi analysis of genes expressed in the ovary of C. elegans. Curr. Biol.
10, 16191622.
Piper, P.W., Panaretou, B., Millson, S.H., Trumana, A., Mollapour, M.,
Pearl, L.H., Prodromou, C., 2003a. Yeast is selectively hypersensitised
to heat shock protein 90 (Hsp90)-targetting drugs with heterologous
expression of the human Hsp90b, a property that can be exploited in
screens for new Hsp90 chaperone inhibitors. Genetics 302, 165170.
Piper, P.W., Millson, S.H., Mollapour, M., Panaretou, B., Siligardi, G.,
Pearl, L.H., Prodromou, C., 2003b. Sensitivity to Hsp90-targeting drugs
can arise with mutation to the Hsp90 chaperone, cochaperones and
plasma membrane ATP binding cassette transporters of yeast. Eur.
J. Biochem. 270, 46894695.
Ren, P., Lim, C.S., Johnsen, R., Albert, P.S., Pilgrim, D., Riddle, D.L.,
1996. Control of C. elegans larval development by neuronal expression
of a TGF-b homolog. Science 274, 13891391.
Richter, K., Buchner, J., 2001. Hsp90: chaperoning signal transduction.
J. Cell Physiol. 188, 281290.
Soti, C., Racz, A., Csermely, P., 2002. A nucleotide-dependent molecular
switch controls ATP binding at the C-terminal domain of Hsp90. N-
terminal nucleotide binding unmasks a C-terminal binding pocket.
J. Biol. Chem. 277, 70667075.
Steptek, G., Houston, K.M., Goodridge, H.S., Devaney, E., Harnett, W.,
2004. Stage-specific and species-specific differences in the production
of the mRNA and protein for the filarial nematode secreted product, ES-
62. Parasitology 128, 9198.
Tewari, M., Hu, P.J., Ahn, J.S., Ayivi-Guedehoussou, N.,
Vidalain, P.O., Li, S., Milstein, S., Armstrong, C.M., Boxem, M.,
Butler, M.D., Busiguina, S., Rual, J.F., Ibarrola, N., Chaklos, S.T.,
Bertin, N., Vaglio, P., Edgley, M.L., King, K.V., Albert, P.S.,
Vandenhaute, J., Pandey, A., Riddle, D.L., Ruvkun, G., Vidal, M.,
2004. Systematic interactome mapping and genetic perturbation
analysis of a C. elegans TGF-beta signaling network. Mol Cell 13,
469482.
Thompson, F.J., Cockroft, A.C., Wheatley, I., Britton, C., Devaney, E.,
2001. Heat shock and developmental expression of hsp83 in the filarial
nematode Brugia pahangi. Eur. J. Biochem. 268, 58085815.
Venolia, L., Waterston, R.H., 1990. The unc-45 gene of Caenorhabditis
elegans is an essential muscle-affecting gene with maternal expression.
Genetics 126, 345353.
Venolia, L., Ao, W., Kim, S., Kim, C., Pilgrim, D., 1999. unc-45 gene of
Caenorhabditis elegans encodes a muscle-specific tetratricopeptide
repeat-containing protein. Cell Motil. Cytoskeleton 42, 163177.
Walker, G.A., Thompson, F.J., Brawley, A., Scanlon, T., Devaney, E.,
2003. Heat shock factor functions at the convergence of the stress
response and developmental pathways in Caenorhabditis elegans. Fedr.
Am. Soc. Exp. Biol. J. 2003;, 171960171962.
Whitesell, L., Mimnaugh, E.G., De Costa, B., Myers, C.E., Neckers, L.M.,
1994. Inhibition of heat shock protein HSP90-pp60v-src heteroprotein
complex formation by benzoquinone ansamycins: essential role for
stress proteins in oncogenic transformation. Proc. Natl Acad. Sci. USA
91, 83248328.
Whitesell, L., Sutphin, P.D., Pulcini, E.J., Martinez, J.D., Cook, P.H., 1998.
The physical association of multiple molecular chaperone proteins with
mutant p53 is altered by geldanamycin, an Hsp90-binding agent. Mol.
Cell Biol. 18, 15171524.
Workman, P., 2003. Overview: translating Hsp90 biology into Hsp90
drugs. Curr. Cancer Drug Targets 3, 297300.
Xu, Y., Lindquist, S., 1993. Heat-shock protein Hsp90 governs the
activity of pp60v-src kinase. Proc. Natl Acad. Sci. USA 90, 7074
7078.
Xu, Y., Singer, M.A., Lindquist, S., 1999. Maturation of the tyrosine kinase
c-src as a kinase and as a substrate depends upon the molecular
chaperone Hsp90. Proc. Natl Acad. Sci. USA 96, 109114.
Young, J.C., Obermann, W.M.J., Hartl, F.U., 1998. Specific binding of
tetratricopeptid repeat proteins to the C-terminal 12 kDa domain of
hsp90. J. Biol. Chem. 273, 1800718010.
Young, J.C., Moarefi, I., Hartl, F.U., 2001. Hsp90: a specialized but
essential protein-folding tool. J. Cell Biol. 154, 267273.
Yue, L., Karr, T.L., Nathan, D.F., Swift, H., Srinivasian, S., Lindquist, S.,
1999. Genetic analysis of viable Hsp90 alleles reveals a critical role in
Drosophila spermatogenesis. Genetics 151, 10651079.
 International Journal for Parasitology 35 (2005) 637646
www.parasitology-online.com

Differential effects of polyamine derivative compounds against

Leishmania infantum promastigotes and axenic amastigotes
J. Tavaresa,b, A. Ouaissic, P.K.T. Lind, A. Tomasb,e, A. Cordeiro-da-Silvaa,b,*
a
Laboratorio de Bioqumica, Faculdade de Farmacia da Universidade do Porto, Rua Anibal Cunha, 164, 4050-047 Porto, Portugal
b
Instituto de Biologia Molecular e Celular, Universidade do Porto, Porto, Portugal
c
IRD UR008 Pathogenie des Trypanosomatides, Centre IRD de Montpellier, Montpellier, France
d
School of Life Sciences, The Robert Gordon University, Aberdeen, UK
e
ICBAS, Instituto de Ciencias Biomedicas Abel Salazar da Universidade do Porto, Porto, Portugal
Received 19 November 2004; received in revised form 21 January 2005; accepted 21 January 2005

Abstract
The natural polyamines are ubiquitous polycationic compounds that play important biological functions in cell growth and differentiation.
In the case of protozoan species that are causative agents of important human diseases such as Leishmaniasis, an exogenous supply of
polyamines supports parasite proliferation. In the present study, we have investigated the effect of three polyamine derivatives, (namely bis-
naphthalimidopropyl putrescine (BNIPPut), spermidine (BNIPSpd) and spermine (BNIPSpm)), on the proliferative stages of Leishmania
infantum, the causative agent of visceral leishmaniasis in the Mediterranean basin. A significant reduction of promastigotes and axenic
amastigotes growth was observed in the presence of increasing concentrations of the drugs, although the mechanisms leading to the parasite
growth arrest seems to be different. Indeed, by using a number of biochemical approaches to analyse the alterations that occurred during early
stages of parasite-drug interaction (i.e. membrane phosphatidylserine exposure measured by annexin V binding, DNA fragmentation,
deoxynucleotidyltranferase-mediated dUTP end labelin (TUNEL), mitochondrial transmembrane potential loss), we showed that the drugs
had the capacity to induce the death of promastigotes by a mechanism that shares many features with metazoan apoptosis. Surprisingly, the
amastigotes did not behave in a similar way to promastigotes. The drug inhibitory effect on amastigotes growth and the absence of propidium
iodide labelling may suggest that the compounds are acting as cytostatic substances. Although, the mechanisms of action of these compounds
have yet to be elucidated, the above data show for the first time that polyamine derivatives may act differentially on the Leishmania parasite
stages. Further chemical modifications are needed to make the polyamine derivatives as well as other analogues able to target the amastigote
stage of the parasite.
q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: Polyamines derivative compounds; Apoptosis-like death; Leishmania

1. Introduction donovani infantum, especially affects children and South

In humans, parasites from the Leishmania donovani
complex cause visceral leishmaniasis or Kala-azar (Zucke-
man and Lainson, 1977). In Africa, L. donovani also infects
adults but frequently presents a high resistance to antimonial
treatment. Mediterranean Kala-azar, caused by Leishmania
* Corresponding author. Address: Laboratorio de Bioqumica, Faculdade
de Farmacia da Universidade do Porto, Rua Anibal Cunha, 164, 4050-047
Porto, Portugal. Tel.: C351 22 2078906/07; fax: C351 22 2003977.
E-mail address: cordeiro@mail.ff.up.pt (A. Cordeiro-da-Silva).
0020-7519/$30.00 q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijpara.2005.01.008
American visceral leishmaniasis caused by Leishmania
donovani chagasi is a disease of both adults and children.
Geographical and clinical data have been used in the past to
classify the organisms causing this visceral disease.
Chemotherapy is limited to the use of pentavalent
antimonials such as sodium stibogluconate (Pentostan),
N-methylglucamine (Glucantime), amphotericin B or penta-
midine (Murry, 2001). However, due to the prolonged
duration of therapy, causing adverse reactions and resistance
to these compounds, the discovery of new antileishmanial
compounds is required. The concept of drug resistance in
638 J. Tavares et al. / International Journal for Parasitology 35 (2005) 637646
Leishmaniasis is not straightforward; sensitivity to drugs has
to be evaluated carefully and considered in relation to the
differences in intrinsic drug sensitivity between species and
zoonotic area. Treatment efficacy is also compromised in case
of immunosuppression, in particular due to HIV co-infection.
This can correspond to the exacerbation of the disease or
emergence from latent infection due to the depletion of
immune capacity and consequently unsuccessful standard
chemotherapy (Berhe et al., 1999; Croft and Coombs, 2003;
Watkins, 2003). Therefore, a search for new chemotherapeutic
agents acting on parasite development is still warranted.
The natural polyamines, spermidine and spermine, and
their precursor diamine putrescine, are present in most
eukaryotic cells and have an important role in cell
proliferation and differentiation (Muller et al., 2001).
In trypanosomatid protozoa, polyamines have an additional
role participating in the endogenous redox equilibrium
through the compound (N1, N8-bis (glutathionyl) spermi-
dine), named trypanothione T(S)2, which is maintained in
its reduced dithiol form, T(SH)2, by trypanothione
reductase (TR) (Fairlamb and Cerami, 1992). T(S)2 is the
major redox reactive metabolite in trypanosomatids, this
molecule and the enzymes involved in its metabolism are
good drug targets (Fairlamb and Cerami, 1992; Barrett et al.,
1999). Spermidine could be an essential polyamine needs
for the maintenance of normal proliferation of trypanoso-
matid protozoa. This role is supported by the sensitivity of
L. donovani cell lines deficient in the spermidine synthase
gene (Roberts et al., 2001). Inhibition of proliferation occurs
when the endogenous spermidine level was reduced to one
third or less of its normal value, independently of the
intracellular concentrations of putrescine (Gonzalez et al.,
2001).
The intracellular concentrations of polyamines in
mammalian cells are regulated by feedback mechanisms
and involve multiple routes of synthesis and interconversion
(Muller et al., 2001). Furthermore, targeting parasitic
protozoan polyamines and their associated enzymes, like
TR, ornithine decarboxylase, have become attractive targets
for antiparasite therapy (Bonnet, 1997; Tovar et al., 1998;
Muller et al., 2001). Therefore, it is reasonable to assume
that selective interference with the parasitic polyamine
metabolism will lead to an alteration of natural defence
mechanisms of the parasite.
In fact, interference with the usual functions of
polyamine has been one strategy in the search for effective
antitumour drugs. The use of terminally alkylated poly-
amine analogues, able to mimic the natural polyamines in
their self-regulatory role, but unable to act as substitutes for
polyamines in their cell growth regulatory functions, has
been shown to induce the growth inhibitory effects on
several cancer cell lines (Ha et al., 1997; Dai et al., 1999;
Davidson et al., 2000; Pavlov et al., 2002). In some cases,
evidence reported showed that the polyamine analogues
induced cell death via apoptosis seen by the fact that the
dying cells displayed cell shrinkage, nuclear condensation
and fragmentation, though failing to obtain clear evidence
of the DNA ladder, typical of apoptosis (McCloskey et al.,
1995; Pavlov et al., 2002).
In this paper, we present the results of a comparative
study of the effects of polyamine derivative compounds,
assigned as BNIPPut, BNIPSpd and BNIPSpm, on the
proliferative stages of Leishmania infantum (promastigotes
and amastigotes). We show that the three compounds are
potent antiproliferative agents against both forms of the
parasite. However, analyses of the drug-induced alterations
leading to parasite growth arrest revealed a killing process
sharing many features with metazoan apoptosis only in the
case of promastigotes whereas the activity against the
amastigotes is likely due to a cytostatic effect.
2. Materials and methods
2.1. Polyamine compounds
The polyamine derivative compounds assigned as bis-
naphthalimidopropyl putrescine (BNIPPut), spermidine
(BNIPSpd), and spermine (BNIPSpm) were synthesised as
described previously (Lin and Pavlov, 2000). Stock
solutions of BNIPPut, BNIPSpd and BNIPSpm were
prepared in 20% dimethylsulfoxide (DMSO) and stored at
4 8C. Working solutions were freshly diluted with culture
medium until the desired final concentrations for the
different assays were reached. The final concentrations of
DMSO (0.02%) used did not interfere with any of the
biological activities tested in this work.
2.2. Parasites
Leishmania infantum (clone MHOM/MA671TMA-
P263) promastigotes were grown at 27 8C in RPMI medium
(Gibco) supplemented with 10% heat inactivated fetal
bovine serum (FBS-Gibco), 2 mM L-glutamine (Gibco),
20 mM Hepes (Gibco), 100 U/ml penicillin (Gibco) and
100 mg/ml streptomycin (Gibco) (Cordeiro-da-Silva et al.,
2003). The parasites (106/ml) in the logarithmic phase
(2 days of culture) were incubated with a serial range of
concentrations of each drug for 5 days at 28 8C. The growth
of parasites was determined by three methods: (i) the
counting in Newbauer chamber and by flow cytometry; (ii)
by optical density at 600 nm; (iii) by methylthiazoletetra-
zolium (MTT) assay. The percentage of growth inhibition
was calculated as (1-growth rate of the experimental
culture/growth rate of the control culture)!100. The IC50,
that is the concentration of the drug required to inhibit the
growth by 50%, was determined by linear regression
analysis.
Leishmania infantum axenic amastigote forms were
grown at 37 8C with 5% CO2 in a cell free medium called
MAA/20 (medium for axenically grown amastigotes)
(Sereno and Lemesre, 1997). MAA/20 consisted of
 J. Tavares et al. / International Journal for Parasitology 35 (2005) 637646
modified medium 199 (GIBCO) with Hanks balanced salt
solution supplemented with 0.5% trypto-casein (Oxoid),
15 mM D8-glucose (SIGMA), 5 mM glutamine (GIBCO),
4 mM NaHCO 3 (Sigma), 0.023 mM bovine hemin
(FLUKA), and 25 mM HEPES to a final pH of 6.5 and
supplemented with 20% heat inactivated FBS (Gibco). The
parasites (2!105/ml) were incubated with a serial range of
concentrations of each drug for 5 days at 37 8C with 5%
CO2. The growth of parasites was determined by the same
methods described above.
To monitor parasite death, short incubation periods were
used. Promastigotes (106/ml) or amastigotes (106/ml) were
treated with 10 mM of each drug for times ranging from 6 to
24 h. Different approaches were developed to explore
the mechanisms of drug induced cell growth inhibition
(see below).
2.3. Flow cytometric analysis of external
phosphatidylserine exposure
Promastigotes or axenic amastigotes (1!106/ml) were
incubated with 10 mM of polyamine derivative compounds,
for 22 h. Parasites were stained with FITC-conjugated
annexin V, for 15 min, at room temperature, in a Ca2C
enriched binding buffer (apoptosis detection kit, R&D
Systems). Then, propidium iodide (PI) was added to exclude
the necrotic cells with disrupted plasma membrane
permeability, and the parasites were analysed by flow
cytometry (FACSCalibur, BD Biosciences). Promastigotes
exposed to UV for 15 min were used as a positive control.
Axenic amastigotes were also treated with 4 mM of
staurosporine during 22 h and used as a positive control.
2.4. In situ TUNEL assay
DNA fragmentation was analysed in situ using a
fluorescent detection system (In Situ Cell Death Detection
Kit, Fluorescein-Roche). Slides containing promastigotes
treated for 6, 12, and 24 h with polyamine derivative
compounds and untreated promastigotes or amastigotes
were fixed for 1 h with 4% paraformaldehyde (Merck) in
PBS, washed with PBS, and stored at K20 8C until used.
The TUNEL assay was conducted following the manufac-
turers instructions. The parasites were observed under a
fluorescent microscope (Axioskop-Carl Zeiss, Germany) at
400! magnification and images captured with a digital
camera (Spot 2-Diagnostic Instruments, USA) and the
software Spot 3.1 (Diagnostic Instruments, USA).
2.5. Flow cytometric analysis of DNA content
The DNA content of 1!106/ml promastigotes or
amastigotes treated and untreated with 10 mM of either of
the three drugs during 24 h, was determined using
propidium iodide after cell permeabilisation with 2% of
saponin (Sigma) in PBS. Cells were analysed using FLH-3
639
detector and the Cell Quest software from a FACSCalibur
(BD Biosciences).
2.6. Immunofluorescence assay
After 24 h treatment with 10 mM of polyamine derivative
compounds, L. infantum promastigotes were fixed with 4%
paraformaldehyde in PBS for 20 min at room temperature.
After several washes, the parasites were permeabilised with
0.1% (v/v) Triton X-100 in PBS. Parasites were then
incubated with a rabbit immune serum to a L. infantum
recombinant mitochondrial protein belonging to the perox-
iredoxin family (anti-LimTXNPx antibodies) (Castro et al.,
2002) diluted 1:1000 in PBS containing 1% bovine serum
albumin (PBS-BSA). The secondary antibody used was
Alexa Fluor 488 goat anti-rabbit IgG diluted 1:2000 in PBS-
BSA (Molecular Probes). Washed parasites were mounted
in Vectashield (Vector Laboratories) and analysed with a
fluorescent microscope (Axioskop-Carl Zeiss, Germany) at
1000! magnification and images captured with a digital
camera (Spot 2-Diagnostic Instruments, USA) and the
software Spot 3.1 (Diagnostic Instruments, USA).
A negative control with untreated parasites was used.
2.7. Measurement of mitochondrial membrane potential in
live Leishmania promastigotes or amastigotes
Tetramethylrhodamine ethylester perchlorate (TMRE,
Molecular Probes) is a cationic lipophilic dye that
accumulates in the negatively charged mithocondrial
matrix according to the Nernst equation potential
(Ehrenberg et al., 1988). A stock of TMRE was prepared
at 4 mg/ml in DMSO and stored at K20 8C. For the
determination of mitochondrial membrane potential,
L. infantum promastigotes or amastigotes were incubated
with 10 mM of BNIPPut, BNIPSpd or BNIPSpm for 6,
12 or 24 h, washed once in PBS and resuspended at
2!106 cells/ml in PBS containing 100 nM TMRE. Cells
were incubated for 15 and 30 min at room temperature
and analysed by flow cytometry on the FL2-H channel.
The PI at 4 mg/ml was used before the last acquisition to
exclude dead cells. Propidium iodide (PI) fluorescence
was recorded on FL3-H channel. As a positive control,
cells already labeled with TMRE for 30 min were treated
during 15 min with 200 mM final concentration of
carbonyl cyanide m-clorophenylhydrazone (CCCP,
Sigma) during 15 min which depolarises mitochondria
by abolishing the proton gradient across the inner
mitochondrial membrane (Scaduto and Grotyohann,
1999).
2.8. Statistical analysis
The data were analysed using the Students t-test.
Probability values of ! 0.01 or 0.05 were considered
significant with 99 or 95% of confidence, respectively.
640 J. Tavares et al. / International Journal for Parasitology 35 (2005) 637646

3. Results

3.1. The polyamine derivative drugs control the growth

of Leishmania infantum parasites

Treatment of L. infantum promastigotes and axenic

amastigote forms with BNIPPut, BNIPSpd or BNIPSpm at
concentrations ranging from 0.125 to 25 mM (Fig. 1),
resulted in a dose dependent inhibition of parasite growth,
determined by quantification in a hemocytometer chamber
(Fig. 2). It can be seen that at R12 mM all the three drugs
completely blocked both promastigote and amastigote
multiplication. However, below the above-mentioned con-
centration (down to 8 mM), BNIPPut and BNIPSpd were
able to arrest promastigote growth. In the case of
promastigotes, the calculated IC50GSD values for BNIP-
Put, BNIPSpd and BNIPSpm were 1.30G0.24, 1.9G0.33,
and 7.56G0.04 mM, respectively. The BNIPPut seems to be
the most effective compound compared to BNIPSpd and
BNIPSpm (PZ0.0635; P!0.01). The effect of the three
drugs on parasite axenic amastigote forms allowed us to
determine the following IC50GSD values: 1.68G0.06,
4.56G0.23, and 12.28G1.30 mM, for BNIPPut, BNIPSpd
and BNIPSpm, respectively (Fig. 2). Although the IC50
values appear higher than those determined in the case of
promastigotes, the BNIPPut remained the most effective
drug (P!0.01) when comparing BNIPPut IC50 value to
those obtained with BNIPSpd and BNIPSpm, respectively.
Three other independent methods, counting in flow
Fig. 2. Effect of different drugs on parasite growth in vitro. Representative
cytometry, MTT assay and optical density determinations, inhibition growth curves of promastigotes and axenic amastigotes.
were used with similar results (data not shown). Promastigote (closed square) and amastigote (open square) forms were
incubated with a concentration range of 0.12525 mM of bis-naphthalimi-
dopropyl putrescine (BNIPPut) (A), spermidine (BNIPSpd) (B), and
3.2. Polyamine derivative drugs induce apoptosis-like DNA
spermine (BNIPSpm) (C). Percentages of promastigote and amastigote
fragmentation in Leishmania infantum promastigotes growth inhibition were determined after 5 days by counting using a
Newbauer chamber under light microscopy. Each point represents the mean
In order to determine whether the observed effect was of two replicates G standard deviation. The data shown is representative of
due to a leishmanicidal activity and to establish the type at least three independent experiments. The growth rate of the promastigote
and amastigote parasite forms without drug treatment was 24 and 100 times
related to the initial concentration, respectively.

of death that the drugs induced, promastigotes or

Fig. 1. Chemical structures of polyamine derivative compounds. Bis-
naphthalimidopropyl putrescine (BNIPPut) (A), spermidine (BNIPSpd) (B)
and spermine (BNIPSpm) (C).
amastigotes were incubated with 10 mM of each of the
three compounds for 6, 12 and 24 h at 27 or 37 8C,
respectively. The changes occurring in the nuclear
material was evaluated in situ by the TUNEL assay
which detects the free ends of DNA after breakage. As
shown in Fig. 3, in contrast to the parasites incubated
without the drug where fewer labeled nuclei could be
seen in Fig. 3A all promastigotes samples treated with
either of the three drugs showed bright fluorescent nuclei
indicating fragmention of DNA (Fig. 3BD). However,
examination of a number of slides showed slightly
different patterns of labelling. Indeed, the BNIPSpd
induced fluorescent nuclei started to be visible only
after 6 h of incubation. Moreover, only moderate
 J. Tavares et al. / International Journal for Parasitology 35 (2005) 637646 641

Fig. 3. In situ analysis of Leishmania infantum promastigotes DNA fragmentation. Tunel assay (TUNEL, FITC) and phase contrast, were performed on
logarithmic phase promastigotes. Untreated promastigotes (A) and parasites treated with 10 mM of the polyamine derivative compounds bis-
naphthalimidopropyl putrescine (BNIPPut) (B), spermidine (BNIPSpd) (C) and spermine (BNIPSpm) (D) for 6, 12, and 24 h, were submitted to the
TUNEL assay as described under Section 2. The parasites were examined under a fluorescent microscope at 400! magnification. The data are representative
of three independent experiments.

fluorescent reactions could be seen in the case of
BNIPPut (Fig. 3B) when compared to BNIPSpd
(Fig. 3B) or BNIPSpm (Fig. 3C). As a positive control,
parasite samples treated with DNAse showed that almost
all the nuclei and kinetoplasts were fluorescent (data not
shown).
However, when studying the effect of the drugs on
amastigotes, although some positive signals could be
observed on some parasites after 6 and 12 h of
incubation with either of the three drugs, no positive
signals could be seen after 24 h of incubation (data not
shown).
Since previous studies have shown that apoptosis-like
death in trypanosomes and Leishmania parasites (Ouaissi,
2003) is associated with DNA fragmentation in nucleosome
sized DNA fragments, we examined the occurrence of such
phenomenon in genomic DNA samples from drug-treated
parasites. We were unable to demonstrate any typical
oligonucleosomal fragmentation in either promatigote or
amastigote forms (data not shown).
We further examined by flow cytometry analysis after
cell permeabilisation, the labeling with PI in order to
quantify the percentage of pseudohypodiploid cells. The
amount of bound dye is correlated with the DNA content
in a given cell, and DNA fragmentation in apoptotic cells
translates into a fluorescence intensity lower than that of
G1 cells (sub-G1 peak) (Nicoletti et al., 1991). After
24 h of promastigote incubation with 10 mM BNIPPut,
BNIPSpd or BNIPSpm, the cells were found in the sub-
G1 peak region, 34.82, 36.88, and 34.13%, respectively
(Fig. 4BD) compared with 7.64% in control cells
(Fig. 4A), suggesting, therefore, that BNIPPut, BNIPSpd
and BNIPSpm induced parasite DNA degradation. When
the same approach was applied to the drug treated
amastigotes, we could not observe similar profiles as
those obtained in the case of promastigotes. Indeed, after
24 h of amastigotes incubation with 10 mM BNIPPut,
BNIPSpd (Fig. 4F) or BNIPSpm, the percentages of
parasites found in the sub-G1 peak were 4.23, 5.55 and
3.81, respectively. These values are comparable with that
642 J. Tavares et al. / International Journal for Parasitology 35 (2005) 637646

Fig. 4. Flow cytometry analysis of Leishmania infantum DNA content. Promastigotes or axenic amastigotes were incubated for 24 h in culture medium alone
(A, E) or with 10 mM of bis-naphthalimidopropyl putrescine (BNIPPut) (B), spermidine (BNIPSpd) (C, F) or spermine (BNIPSpm) (D). DNA content
degradation was assessed by flow cytometry after cell permeabilisation and propidium iodide staining. Promastigotes incubated with: culture medium alone
(A), BNIPPut (B), BNIPSpd (C) and BNIPSpm (D). Amastigotes incubated with culture medium alone (E) or culture medium plus BNIPSpd (F). The profile
obtained with BNIPSpd is comparable to those obtained with BNIPPut and BNIPSpm (data not shown). The data shown are representative of three independent
experiments.

observed in the case of amastigotes incubated in the
culture medium alone (Fig. 4E).
3.3. Quantification of phosphatidylserine externalisation in
Leishmania infantum promastigotes and amastigotes
treated with polyamine derivative compounds
In mammalian cell apoptosis, a ubiquitous alteration is
the translocation of phosphatidylserine from the inner side
to the outer layer of the plasma membrane. Annexin V,
a Ca2C dependent phospholipid binding protein with
affinity for phosphatidylserine, is routinely used in a
fluorescein conjugated form to label externalisation of
phosphatidylserine. Since annexin V-FITC can also label
necrotic cells following the loss of membrane integrity,
simultaneous addition of PI, which does not permeate cells
with an intact plasma membrane, allows discrimination
between apoptotic cells (annexin V positive and PI
negative), necrotic cells (annexin V positive and PI positive)
and live cells (annexin V negative and PI negative) (Vernes
et al., 1995). Leishmania infantum promastigotes or
amastigotes were incubated with 10 mM of each compound
for 22 h at 27 or 37 8C, respectively and processed for
phosphatidylserine externalisation measurement as indi-
cated under Section 2. As shown in Fig. 5, increased
expression of phosphatidylserine on the surface of drug-
treated promastigotes was observed. Indeed, the percentages
of Annexin positive parasites in the presence of the drug
increased dramatically from 35.97 to 61.36% (Fig. 5CE)
when compared to non-treated cells (1.72%, Fig. 5A).
As a positive control UV exposed promastigotes showed
16.57% Annexin V positive cells (Fig. 5B).
In contrast to promastigotes, BNIPPut treated amasti-
gotes showed no significant increase of Annexin-V positive
cells (Fig. 5G) when compared to the non treated cells
(Fig. 5F). Similar profiles of Annexin-V binding were
obtained when axenic amastigotes were treated with either
BNIPSpd or BNIPSpm (data not shown). This is likely due
to the absence of apoptosis induction since amastigotes
treated with 4 mM staurosporine, a drug known to induce
apoptosis-like death in Leishmania showed around 80%
Annexin-V positive cells (Fig. 5H).
3.4. Leishmania infantum cell death is associated with the
modification of membrane mitochondrial permeability
and potential
Previous studies suggested that nuclear features of
apoptosis in metazoan cells, like condensation of nuclei
and fragmentation of DNA, are preceded by alterations in
mitochondrial structure and transmembrane potential
(Zamzami et al., 2001). Thus, we explored whether
BNIPPut, BNIPSpd and BNIPSpm treatment caused an
early response to the parasite mitochondrial membrane
potential using TMRE fluorescent probe and FACS
analysis. Thus, promastigotes or amastigotes were treated
with BNIPPut, BNIPSpd and BNIPSpm for 6, 12, and 24 h
and then incubated with the probe for 15 and 30 min.
 J. Tavares et al. / International Journal for Parasitology 35 (2005) 637646 643

Fig. 5. Flow cytometry analysis of phosphatidylserine exposure to Leismania infantum after polyamine derivative compound treatment. Promastigotes (AE)
and axenic amastigotes (FH), were incubated in the absence (A, F) or presence of 10 mM of polyamine derivative compounds: bis-naphthalimidopropyl
putrescine (BNIPPut) (C, G), spermidine (BNIPSpd) (D) and spermine (BNIPSpm) (E) for 22 h. Phosphatidylserine exposure was monitored by flow
cytometry. Dead cells were excluded by propidium iodide incorporation. Dot plots are representative of three independent assays. The promastigotes were UV
exposed for 15 min (B) and amastigotes were treated with 4 mM of staurosporine (H) during 22 h as a positive control. The profils obtained with BNIPSpd and
BNIPSpm treated amastigotes were similar to that shown for BNIPPut in G (data not shown).

Fluorescence of cells treated with TMRE and PI was
measured by FACS analysis. Although we did not observe
the modification of mitochondrial membrane potential after
6 h of drug treatment in both amastigotes and promastigotes,
upon 12 h promastigote incubation with BNIPPut, BNIPSpd
or BNIPSpm, a significant decrease in parasite mitochon-
drial membrane potential (31.18, 30.32 and 37.19%,
respectively) occurred compared to the non treated parasites
(98.14%, Fig. 6A). Moreover, after 24 h of BNIPPut,
BNIPSpd and BNIPSpm incubation, similar profiles as
those observed after 12 h treatment were recorded (data not
shown).
In contrast to the promastigotes, we were unable to detect
any modification of amastiogotes mitochondrial membrane
potential after 12 h, (Fig. 6B) or 24 h (data not shown) of
treatment with 10 mM of each of the three drugs.
In previous work, we found that the L. infantum
mitochondrial protein, LimTXNPx, co-localise with the
mitochondrial standard staining (Mito Tracker) (Castro
et al., 2002). Therefore, we further examined the mitochon-
drial integrity by using the anti-LimTXNPx as a probe to
determine the cellular distribution of LimTXNPx protein.
DAPI was also used to label the nucleus and kinetoplast
DNA. After 24 h of incubation with 10 mM of BNIPPut,
BNIPSpd or BNIPSpm, the mitochondrial protein
LimTXNPx showed cytoplasmatic localisation, whereas in
the non treated cells the LimTXNPx remained inside the
mitochondria (Fig. 7). These results reinforce the notion that
drug treatment induced the alteration of mitochondrial
membrane potential leading to apoptosis-like death of
promastigotes.
4. Discussion
The polyamine family including putrescine, spermidine
and spermine have important physiological roles (Gonzalez
et al., 2001; Muller et al., 2001). On the basis of the vast
array of biochemical functions and their role in cell growth
and differentiation, several polyamine synthetic analogues
have been used as tools for the study of the physiological
roles of natural polyamines in terms of their antiprolifera-
tive properties (Thomas and Thomas, 2001). For several
years one of us has been synthesising polyamine derivatives
644 J. Tavares et al. / International Journal for Parasitology 35 (2005) 637646

Fig. 6. Effect of bis-naphthalimidopropyl putrescine (BNIPPut), spermidine (BNIPSpd) and spermine (BNIPSpm) on the mitochondrial membrane potential as
measured by the TMRE probe. Live Leishmania promastigotes (A) or axenic amastigotes (B) were treated with 10 mM of BNIPPut (sketch line), BNIPSpd (dot
line), and BNIPSpm (sketch and dot line) for 12 h. Untreated cells (full peak) showed positive control labelling, whereas the negative control consisted of
parasites treated with the uncoupler CCCP (continuous line). The histogram shows fluorescence intensity of TMRE after 30 min of fluorescent probe time
incubation. In the part B we have an overlay effect of BNIPPut (sketch line), and BNIPSpd (dot line) with the control (full peak).

Fig. 7. Alteration of mitochondrial permeability by polyamine derivative drugs. The localisation of the mitochondrial protein LimTXNPx was analysed by
immunofluorescence assay. Leishmania infantum promastigotes were treated with culture medium alone (A) or with 10 mM of polyamine derivative
compounds: bis-naphthalimidopropyl putrescine (BNIPPut) (B), spermidine (BNIPSpd) (C) and spermine (BNIPSpm) (D) for 24 h and reacted with the rabbit
anti-LimTXNPx antibodies followed by Alexa Fluor 488 goat anti-rabbit IgG. The parasites were examined under a fluorescent microscope at 1000!
magnification. The data shown are representative of three independent experiments.
 J. Tavares et al. / International Journal for Parasitology 35 (2005) 637646
having the capacity to interfere with the growth of cancer
cells. This kind of polyamine derivative compounds showed
a high in vitro cytotoxicity against the human breast cancer
MCF-7 cell line (Pavlov et al., 2000). Given that polyamines
are known to be essential factors for the growth of the
parasites in their host and that cancer cells and Leishmania
parasites share at least one common feature, that is their
mutual capacity for rapid cell division, we attempted to
answer the following: first, whether the three polyamine
derivatives, BNIPPut, BNIPSpd and BNIPSp, could exert an
anti-parasite proliferative activity, and if so whether there is
a phenotypic variance in cell growth arrest (i.e. leishmanio-
static or leishmaniolytic) and second, the type of cell death.
Both antiproliferative and leishmanicidal effects, at least in
the case of promastigotes, were observed for concentrations
of these compounds in the micromolar range. The
reductions in the proliferation of the promastigotes as well
as the amastigotes, the vertebrate stage of the parasite, were
dose dependent for all three drugs tested. However,
comparing the structures of the three compounds, BNIPPut
with the smallest polyamine chain, was found to be the most
potent compound with IC50 values of 1.30 and 1.68 mM for
promastigotes and amastigotes, respectively. The increase
in the length of the polyamine chains leads to increased IC50
values.
Using biochemical and morphological approaches, we
showed that treatment with polyamine derivative com-
pounds induces promastigotes death sharing phenotypic
features with metazoan apoptosis (Ameisen, 2002; Zangger
et al., 2002). Although we did not observe DNA ladders, a
hallmark of metazoan apoptosis, similar observations have
been reported by other investigators who failed to
demonstrate a typical oligonucleosomal fragmentation of
DNA from staurosporine treated Leishmania major para-
sites while showing a DNA degradation (hypoploidy) by
flow cytometry (Arnoult et al., 2002). Furthermore, it is
noteworthy that although the kinetoplastid parasites may
share a death process with features reminiscent of
mammalian nucleated cells apoptosis, the pathways (induc-
tion/execution) may differ at the molecular level (Ouaissi,
2003). Moreover, similar polyamine derivative compounds
which induced apoptosis in cancer cells, failed to give DNA
laddering (Gooch and Yee, 1999; Pavlov et al., 2002),
suggesting that this may be due to the characteristics of
these compounds.
Interestingly, the TUNEL staining clearly demonstrated
that DNA fragmentation occurs in the promastigote nuclei
during polyamine drugs induced cell death before membrane
integrity is compromised, supporting the notion that the
death process observed is not necrotic. Moreover, additional
experiments performed showed that a number of features also
found in apoptosis of multicellular organisms (i.e. cell
shrinkage, phosphatidylserine exposure with preservation of
plasma membrane integrity, and modification of mitochon-
dria potential and permeability), also occurred in polyamine
derivatives treated L. infantum promastigotes. Furthermore,
645
using an antibody probe directed against a L. infantum
mitochondrial protein, LimTXNPx, we could demonstrate
that mitochondrial membrane alterations occurred during
parasite drug interaction. Surprisingly, although a growth
inhibition could be clearly demonstrated in the case of
drug treated axenic amastigotes, we were unable to
observe any of the apoptosis-like features occurring in
the drug treated promastigote forms. Since the drug
treated amastigotes did not incorporate the PI, it is
unlikely that the amastigotes were dying by necrosis. The
simplest interpretation of this phenomenon could be that
the drugs are exerting a cytostatic effect on the
amastigote forms. Another possibility could be that the
amastigotes develop some kind of mechanism allowing
them to resist the apotosis inducing effect of the drugs.
Further studies are needed to clarify this point.
Parasites have distinct polyamine metabolisms compared
to mammalian cells. Interfering with the parasite specific
polyamine metabolism pathway will have more severe
consequences for the parasite than for its host. Recent
studies on parasitic protozoa have focused on the biochemi-
cal properties and susceptibility to inhibition by enzymes
involved in the polyamine synthesis. Protozoa belonging to
the trypanosomatid family such as Leishmania, synthesise a
unique co-factor that is a conjugate of spermidine
and glutathione. This co-factor, termed trypanothione, is
required to maintain redox balance in the cell (Garforth
et al., 1997). Different inhibitors of trypanothione reductase
have antitrypanosomal activity (Bonnet et al., 2000).
Thus, the parasite specific enzymes of trypanothione
metabolism are considered attractive targets for drug
development (Krauth-Siegel and Coombs, 1999). Further
chemical modifications of the polyamine compounds is
needed to target the amastigote, the vertebrate stage of the
parasites.
In summary, our results indicate that polyamine
derivative drugs, BNIPPut, BNIPSpd and BNIPSpm induce
an apoptosis-like death in L. infantum promastigotes and
failed to exert similar effects on axenic amastigotes. Better
understanding of the mechanisms of action of these
polyamine derivative compounds together with the identi-
fication of the major pathways involved in Leishmania
apoptosis-like death is essential to the development of new
analogues which may target the amastigote, the vertebrate
stage of the parasite.
Acknowledgements
This work was supported by Fundacao para a Ciencia e
Tecnologia (FCT) grant POCT/CVT/14209/98/2001 and
FEDER, Cooperacao Luso-Francesa-Programa Pessoa
2004, grant 706 C3, INSERM and IRD UR 008. JT is
supported by a fellowship from FCT number
SFRH/BD/18137/2004.
646 J. Tavares et al. / International Journal for Parasitology 35 (2005) 637646
References
Ameisen, J.C., 2002. On the origin, evolution, and nature of programmed
cell death: a timeline of four billion years. Cell Death Differ. 9, 367
393.
Arnoult, D., Akarid, K., Grodet, A., Petit, P.X., Estaquier, J., Ameisen, J.C.,
2002. On the evolution of programmed cell death: apoptosis of the
unicellular eukaryote Leishmania major involves cysteine proteinase
activation and mitochondrion permeabilization. Cell Death Differ. 9,
6581.
Barrett, M.P., Mottram, J.C., Coombs, G.H., 1999. Recent advances in
identifying and validating drug targets in trypanosomes and leishma-
nias. Trends Microbiol. 7, 8288.
Berhe, N., Wolday, D., Hailu, A., Abraham, Y., Ali, A., Gebre-Michael, T.,
Desjeux, P., Sonnerborg, A., Akuffo, H., Britton, S., 1999. HIV viral
load and response to antileishmanial chemotherapy in co-infected
patients. AIDS 13, 19211925.
Bonnet, B., Soullez, D., Davioud-Charvet, E., Landry, V., Horvath, D.,
Sergheraert, C., 1997. New spermine and spermidine derivatives as
potent inhibitors of Trypanosoma cruzi trypanothione reductase. Biorg.
Med. Chem. 5, 12491256.
Bonnet, B., Soullez, D., Girault, S., Maes, L., Landry, V., Davioud-
Charvet, E., Sergheraert, C., 2000. Trypanothione reductase inhibi-
tion/trypanocidal activity relationships in a 1,4-bis(3-aminopropyl)pi-
perazine series. Bioorg. Med. Chem. 8, 95103.
Castro, H., Sousa, C., Santos, M., Cordeiro-da-Silva, A., Flohe, L.,
Tomas, A., 2002. Complementary antioxidant defense by cytoplasmic
and mitochondrial peroxiredoxins in Leishmania infantum. Free Radic.
Biol. Med. 33, 15521562.
Cordeiro-da-Silva, A., Cardoso, L., Araujo, N., Castro, H., Tomas, A.,
Rodrigues, M., Cabral, M., Vergnes, B., Sereno, D., Ouaissi, A., 2003.
Identification of antibodies to Leishmania silent information regulatory
2 (SIR2) protein homologue during canine natural infections:
pathological implications. Immunol. Lett. 86, 155162.
Croft, S.L., Coombs, G.H., 2003. Leishmaniasiscurrent chemotherapy
and recent advances in the search for novel drugs. Trends Parasitol. 19,
502508.
Dai, H., Kramer, D.L., Yang, C., Murti, K.G., Porter, C.W., Cleveland, J.L.,
1999. The polyamine oxidase inhibitor MDL-72,527 selectively
induces apoptosis of transformed hematopoietic cells through lysoso-
motropic effects. Cancer Res. 59, 49444954.
Davidson, N.E., Hahm, H.A., McCloskey, D.E., Woster, P.M.,
Casero, R.A., 2000. Clinical aspects of cell death in breast cancer:
the polyamine pathway as a new target for treatment. Endocrinol.
Related Cancer 6, 6973.
Ehrenberg, B., Montana, V., Wei, M.D., Wuskell, J.P., Loew, L.M., 1988.
Membrane potential can be determined in individual cells from the
nernstian distribution of cationic dyes. Biophys. J. 53, 785794.
Fairlamb, A.H., Cerami, A., 1992. Metabolism and functions of
trypanothione in the kinetoplastida. Annu. Rev. Microbiol. 46, 695
729.
Garforth, J., Yin, H., McKie, J.H., Douglas, K.T., Fairlamb, A.H., 1997.
Rational design of selective ligands for trypanothione reductase from
Trypanosoma cruzi. Structural effects on the inhibition by dibenzaze-
pines based on imipramine. J. Enzyme Inhib. 12, 161173.
Gonzalez, N.S., Huber, A., Algranati, I.D., 2001. Spermidine is essential for
normal proliferation of trypanosomatid protozoa. Fed. Eur. Biol. Soc.
Lett. 508, 323326.
Gooch, J.L., Yee, D., 1999. Strain-specific differences in formation of
apoptotic DNA leaders in MCF-7 breast cancer cells. Cancer Lett. 144,
3137.
Ha, C.H., Woster, P.M., Yager, J.D., Casero, R.A., 1997. The role of
polyamine catabolism in polyamine-induced programmed cell death.
Proc. Natl Acad. Sci. 94, 15571562.
Krauth-Siegel, R.L., Coombs, G.H., 1999. Enzymes of parasite thiol
metabolism as drug targets. Parasitol. Today 15, 404409.
Lin, P.K.T., Pavlov, V.A., 2000. The synthesis and in vitro cytotoxic studies
of novel bis-naphthalimidopropyl polyamine derivatives. Bioorg. Med.
Chem. Lett. 10, 16091612.
McCloskey, D.E., Casero, R.A., Woster, P.M., Davidson, N.E., 1995.
Induction of programmed cell death in human breast cancer cells by an
unsymetrically alkylated polyamine analogue. Cancer Res. 53, 2071
2075.
Muller, S., Coombs, G.H., Walter, R.D., 2001. Targeting polyamines of
parasitic protozoa in chemotherapy. Trends Parasitol. 17, 242249.
Murry, H.W., 2001. Clinical and experimental advances in treatment of
visceral Leishmaniasis. Antimicrob. Agents Chemother. 45, 2185
2197.
Nicoletti, I., Migliorati, G., Pagliacei, M.C., Grignani, F., Riccardi, C.,
1991. A rapid and simple method for measuring thymocyte apoptosis by
propidium iodide staining and flow cytometry. J. Immunol. Methods
139, 271279.
Ouaissi, A., 2003. Apoptosis-like death in trypanosomatids: Search for
putative pathways and genes involved. Kinetoplastid Biol. Dis. 2,
15.
Pavlov, V., Rodilla, V., Lin, P.K.T., 2000. Cytotoxicity, DNA binding and
localisation of novel bis-naphthalimidopropyl polyamine derivatives.
Chem. Biol. Interact. 137, 1524.
Pavlov, V., Rodilla, V., Lin, P.K.T., 2002. Growth, morphological and
biochemical changes in oxa-spermine derivative-treated MCF-7 human
breast cancer cells. Life Sci. 71, 11611173.
Roberts, S.C., Jiang, Y., Jardim, A., Carter, N.S., Heby, O., Ullman, B.,
2001. Genetic analysis of spermidine synthase from Leishmania
donovani. Mol. Biochem. Parasitol. 115, 217226.
Scaduto, J.r.R.C., Grotyohann, L.W., 1999. Measurement of mitochondrial
membrane potential using fluorescent rhodamine derivatives. Biophys.
J. 76, 469477.
Sereno, D., Lemesre, J.L., 1997. Axenically cultured amastigote forms as
an in vitro model for investigation of antileishmanial agents.
Antimicrob. Agents Chemother. 41, 972976.
Thomas, T., Thomas, T.J., 2001. Polyamines in cell growth and cell death:
molecular mechanisms and therapeutic applications. Cell. Mol. Life
Sci. 58, 244258.
Tovar, J., Wilkinson, S., Mottram, J.C., Fairlamb, A.H., 1998. Evidence
that trypanothione reductase is an essential enzyme in Leishmania by
targeted replacement of the tyrA gene locus. Mol. Microbiol. 29,
653660.
Vernes, I., Haanen, H., Steffens-Nakken, H., Reutelingsperger, C., 1995. A
novel assay for apoptosis. Flow cytometry detection of phosphatidyl-
serine expression on early apoptotic cells using fluorescein labelled
annexin V. J. Immunol. Mehods 184, 3951.
Watkins, B.M., 2003. Drugs for the control of parasitic diseases: current
status and development. Trends Parasitol. 19, 477478.
Zamzami, N., Maisse, C., Metivier, D., Kroemer, G., 2001. Measurement of
membrane permeability and permeability transition of mitochondria.
Methods Cell Biol. 65, 147158.
Zangger, H., Mottram, J.C., Fasel, N., 2002. Cell death in Leishmania
induced by stress and differentiation: programmed cell death or
necrosis?. Cell Death Differ. 9, 11261139.
Zuckeman, A., Lainson, R., 1977. Leishmania in Parasitic Protozoa, vol. 1.
Academic Press, New York, pp. 159163.
 International Journal for Parasitology 35 (2005) 647657
www.parasitology-online.com

Patterns of host specificity and transmission among parasites

of wild primates
Amy B. Pedersena,*, Sonia Altizerb, Mary Possc, Andrew A. Cunninghamd, Charles L. Nunne
a
Department of Biology, University of Virginia, Charlottesville, VA 22904, USA
b
Department of Environmental Studies, Emory University, 400 Dowman Drive Suite E510, Atlanta, GA, USA
c
Division of Biological Sciences, University of Montana, Missoula, MT 59812, USA
d
Institute of Zoology, Zoological Society of London, London NW1 4RY, UK
e
Department of Integrative Biology, University of California, Berkeley, CA 94720-3140, USA
Received 15 December 2004; received in revised form 18 January 2005; accepted 19 January 2005

Abstract
Multihost parasites have been implicated in the emergence of new diseases in humans and wildlife, yet little is known about factors that
influence the host range of parasites in natural populations. We used a comprehensive data set of 415 micro- and macroparasites reported
from 119 wild primate hosts to investigate broad patterns of host specificity. The majority (68%) of primate parasites were reported to infect
multiple host species, including animals from multiple families or orders. This pattern corresponds to previous studies of parasites found in
humans and domesticated animals. Within three parasite groups (viruses, protozoans and helminths), we examined parasite taxonomy and
transmission strategy in relation to measures of host specificity. Relative to other parasite groups, helminths were associated with the greatest
levels of host specificity, whereas most viruses were reported to infect hosts from multiple families or orders. Highly significant associations
between the degree of host specificity and transmission strategy arose within each parasite group, but not always in the same direction,
suggesting that unique constraints influence the host range of parasites within each taxonomic group. Finally characteristics of over 100
parasite species shared between wild primates and humans, including those recognised as emerging in humans, revealed that most of these
shared parasites were reported from multiple host orders. Furthermore, nearly all viruses that were reported to infect both humans and non-
human primates were classified as emerging in humans.
q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: Infectious disease; Multihost pathogen; Transmission strategy; Zoonosis; Wildlife; Emerging pathogens

1. Introduction Dobson and Foufopoulos, 2001; Taylor et al., 2001;

The majority of parasites examined to date, including
over 60% of human pathogens and 80% of those
infecting domesticated animals, are capable of infecting
multiple host species (Cleaveland et al., 2001; Taylor
et al., 2001; Woolhouse et al., 2001). Recent emerging
diseases in humans, including those caused by Ebola,
Influenza A, Nipah, and SARS viruses, are associated
with multihost pathogens maintained in non-human
animal populations (Murphy, 1998; Hahn et al., 2000;
* Corresponding author. Tel.: C1 434 243 5077; fax: C1 434 982 5626.
E-mail address: abp3a@virginia.edu (A.B. Pedersen).
0020-7519/$30.00 q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijpara.2005.01.005
Williams et al., 2002). Multihost pathogens also pose
problems for wildlife conservation, and have caused
population declines or high mortality among African
carnivores, harbor seals, sea otters, black-footed ferrets,
and African apes (Roelke-Parker et al., 1996; Harvell
et al., 1999; Daszak et al., 2000; Jensen et al., 2002;
Miller et al., 2002; Walsh et al., 2003). Despite their
threats to global health and conservation, factors that
influence the evolution and ecology of multihost parasites
and their distributions in wild animal populations are not
well understood (Desdevises et al., 2002), in part because
conventional studies have focused most commonly on
single host-pathogen systems (Anderson and May, 1991;
648 A.B. Pedersen et al. / International Journal for Parasitology 35 (2005) 647657
Bull, 1994; Day, 2001) or on domesticated animals
and humans (Cleaveland et al., 2001; Taylor et al.,
2001).
A variety of ecological and evolutionary factors may
influence patterns of parasite specificity. Here we focus
on two key variables. First, transmission strategies that
increase encounters with new hosts, such as transmission
by biting arthropods or through contaminated soil or
water, are predicted to increase the range of hosts that a
pathogen can infect (Woolhouse et al., 2001). Second,
greater genetic variability and more rapid generation
times might allow certain pathogens to readily exploit
new host species. For instance, parasites with high
antigenic variation should have an increased ability to
recognise host proteins or evade host immune defenses
(Bitter et al., 1998; Simon et al., 1998).
In this study, we assembled a comprehensive database
of infectious diseases reported from wild primate
populations to investigate the frequency and character-
istics of multihost pathogens. Indeed, several recent
papers have highlighted that a majority of parasites
reported to infect humans and domesticated animals can
in fact infect more than one host species (Cleaveland
et al., 2001; Taylor et al., 2001). This is not necessarily
surprising because both humans and domesticated
animals have cosmopolitan distributions, exist in large
dense populations, and can be exposed to a wide
diversity of pathogens from wildlife sources. In contrast
to these earlier papers, our study focused on parasites
reported from wild populations of non-human primates,
in part because they represent a diverse group of
mammals that have been threatened by infectious
diseases (Wallis and Lee, 1999; Walsh et al., 2003;
Leroy et al., 2004), and they have been identified as wild
reservoirs or ancestral hosts for a growing number of
human pathogens (Wolfe et al., 1998, 2004).
Using a similar database of host-pathogen records, Nunn
et al. (2003) tested a suite of socioecological factors
predicted to influence parasite species richness in wild
primates, and demonstrated that host population density and
geographic range size were positively associated with
parasite species richness in wild primates. In this paper,
we focused on traits of the parasites themselves by
examining the association between host specificity and
transmission strategy within and among multiple parasite
taxonomic groups. We predicted that parasites transmitted
by direct contact would be relatively more host specific than
parasites with greater opportunities for among-host dis-
persal. We also predicted that viruses, owing to their short
generation times and high mutation rates, would have
broader host ranges. Finally we investigated the trans-
mission and specificity of parasites reported from both
humans and wild non-human primates to ascertain the risks
that these shared parasites pose in relation to human
emerging diseases.
2. Materials and methods
2.1. Parasite records from wild primates
A database of disease-causing organisms reported from
wild primate populations was compiled from 258 published
references, including scientific papers and book sections.
We searched for reports using primate species Latin
binomials as keywords in major online reference databases
(Biological Abstracts, AGRICOLA, Medline, Web of
Science). We also searched by primate genus name,
following the taxonomic scheme of Corbet and Hill
(1991), as well as by common taxonomic variants (based
on Rowe, 1996; Groves, 2001). The vast majority of papers
and reports used in our data set were published between
1970 and 1999, with the total time frame spanning
19402002. For each parasite species reported from a wild
primate population, we recorded information on hosts,
parasites and sampling locality. We defined parasites
broadly to include viruses, protozoa, fungi, arthropods,
helminths and bacteria. Our original data set included 2173
cumulative lines of data, where each line was a record of a
parasite species reported from a wild primate population.
Nomenclature for viruses followed the International Com-
mittee on the Taxonomy of Viruses database (ITCVdb)
available online (http://www.ncbi.nlm.nih.gov/ICTVdb/
Ictv/), and nomenclature for other parasite species followed
the guidelines published by the National Center for
Biotechnology Information (NCBI). Additional details
about the database are described in Nunn et al. (2003),
Nunn and Altizer (2005), and are available on the Global
Mammal Parasite Database website (www.mammalpara-
sites.org). From our original host-parasite database, we
generated a list of parasite species reported from free-living
primate populations. To this initial list of parasite species,
we recorded whether or not the parasite has been reported to
infect humans based on records from Ashford and Crewe
(1998) and Taylor et al. (2001), and also recorded
information on transmission and specificity described below.
2.2. Scoring parasite specificity
For all parasite species, information on the taxonomic
range of primates and other species reported as hosts was
compiled by searching for published records of vertebrate
animals infected under natural conditions using two online
databases (BIOSIS and Web of Science). To assess virus
specificity, we searched the PubMed online database (in
addition to Web of Science) using ICTV virus names. We
assigned a score from 1 to 5 to each parasite based on
the taxonomic affiliation of affected hosts as follows:
1Zspecies specific; 2Zgenus specific; 3Zfamily specific;
4Zorder specific; and 5Zvertebrate specific (i.e. can infect
multiple host orders). For host groupings at the level of
primate families, we followed the taxonomy of Corbet and
Hill (1991). We based specificity scores on recorded
 A.B. Pedersen et al. / International Journal for Parasitology 35 (2005) 647657
infections from wild populations only (i.e. excluding
records of parasites from captive animals or infection
experiments), and considered only vertebrate host species in
assigning specificity scores, ignoring invertebrates that
might serve as intermediate hosts or vectors.
To each specificity score, we added a confidence score
based on the quality and quantity of published records that
provided information on affected host species. Highest
confidence scores (3) were assigned to records based on
direct isolation of pathogens or their DNA, in combination
with a sufficient number of published records (usually 5 or
more); intermediate confidence (2) was assigned to scores
based on indirect measures of pathogen presence (e.g.
serology) or a limited number of published records, and low
confidence (1) was assigned to specificity scores based on
unpublished knowledge, strong inference, or information
derived solely from our primate parasite database. Confi-
dence scores were used to evaluate the sensitivity of patterns
by analyzing the data once with all available records, and
second by excluding data with low confidence. All viral
specificity scores were assigned high or moderate confi-
dence; thus, for this group we did not repeat analyses
excluding low confidence data.
As a second measure of confidence, we examined the
sampling effort associated with each parasite species using
citation counts from three common bibliographic databases:
ISI Web of Science (WOS; www.isiwebofknowledge.com),
Biosis Previews (www.biosis.org), and Pubmed (www.
pubmed.org). We searched each database for corrected
parasite Latin binomials or ICTVdb virus names. For each
parasite species we computed average sampling effort as the
log-transformed arithmetic mean of all three citation counts
(adding one to each count first to avoid taking the log of
zero). Analysis of variance showed that categorical
measures of confidence were highly associated with the
average of citation counts (F2,409Z132.9; P!0.0001) so
that specificity scores assigned high confidence were also
generally from the best studied parasites in the scientific
literature (mean log-referencesZ2.18). Parasites with
moderate confidence scores had intermediate numbers of
citations (mean log-referencesZ1.04), and those with the
lowest confidence were generally not well represented in the
scientific literature (mean log-references 0.29).
Not surprisingly, specificity scores were associated with
both categorical and continuous measures of confidence. In
particular, the least specific parasites were associated with
the highest confidence scores, whereas genus or species
specific parasites were more likely to be assigned inter-
mediate or low confidence scores. This association was
highly significant when examined across all parasite groups
(c2Z193.6, d.f.Z8, P!0.0001), as well as within each of
the three major parasite groups: viruses (c2Z15.3, d.f.Z4,
PZ0.004), protozoa (c2Z33.3, d.f.Z8, P!0.0001), and
helminths (c2Z61.5, d.f.Z8, P!0.0001). A similar pattern
was found when we examined the association between
specificity and average citation counts (F4,407Z83.04;
649
P!0.0001). In this case, mean citation counts were lowest
among parasites assigned genus- or species-specific status
(mean log-referencesZ0.25 and 0.34, respectively). Cita-
tion counts were intermediate for parasites assigned family
specific status (meanZ0.90) and were greatest for parasites
in the order-specific and multi-order categories (meanZ
1.83 and 2.24, respectively).
2.3. Parasite transmission mode
The transmission strategy of each parasite species was
recorded as one or more of the following four categories:
close contact, non-close contact, biting arthropod vectors, or
intermediate hosts. Parasites spread by close contact were
those that were highly contagious and communicable by
close proximity or direct contact such as biting, scratching,
mating contact, or other touching. Close contact trans-
mission was further subdivided into three routessexual,
vertical (parent to offspring), or close non-sexual contact.
Non-close contact involved transmission via fomites or
contact with contaminated soil or water. Vector-borne
parasites were those spread via biting arthropods (ticks,
mites, fleas, flies, and other invertebrates), and parasites
with intermediate hosts were those characterised by the
presence of complex life cycles and/or trophic transmission.
Transmission strategies were assigned by multiple web
searches using parasite Latin binomials or ICTV virus
names. If direct information was not available for a
particular parasite species, transmission strategies were
assigned based on information from closely related parasites
(if similar transmission strategies were highly likely) or
remained unassigned.
2.4. Data analysis
We first examined the occurrence of host specificity and
major transmission strategies across all parasites from wild
primates, and then within each of the three best-represented
groups: viruses, helminths, and protozoa. Second, we tested
whether the predicted associations between transmission
and host specificity were consistent with observed patterns
within each of the major parasite groups. Finally, the
taxonomic distribution of parasites from wild primates
(arthropods, bacteria, fungi, helminths, protozoa, and
viruses) was compared with reports from earlier studies of
humans and domesticated animals (e.g. Cleaveland et al.,
2001; Taylor et al., 2001) to investigate whether the types
and characteristics of parasites from wild hosts were similar
to those reported for economically important species, and to
examine characteristics of parasites from wild primates that
were also reported as emerging in human populations.
Unless otherwise stated, we used Pearson Chi square
analysis or logistic regression (general linear model with
binomial errors) to examine associations between host
specificity, parasite taxonomy, and transmission strategy.
650 A.B. Pedersen et al. / International Journal for Parasitology 35 (2005) 647657

3. Results

A total of 415 parasite species from 119 host species

were represented in the database of parasites reported from
wild primate populations. The distribution of parasite
species was: 163 helminths (39%), 82 protozoans (20%),
82 viruses (20%), 46 arthropods (11%), 32 bacteria (8%)
and 10 fungi (2%). Thus, three major groupshelminths,
protozoa, and virusescollectively represented 79% of all
parasites reported from wild primates.

3.1. Patterns of host specificity

Of all primate parasites, the majority (68%) were

reported to infect multiple host species, and only 131
(32%) were identified as species specific. Among multi-host
parasites, 40% were reported from both primate and non-
primate hosts, and another 25% were reported from multiple
primate families (nZ284). Among the less well represented
taxonomic groups, multihost pathogens were common
70% of all arthropods, 90% of bacteria, and 90% of fungi
were reported to infect more than one host species, and in
most cases affected hosts were from multiple orders.
Host specificity varied dramatically among the three
best-represented groups of primate parasites (Fig. 1).
Viruses were dominated by multi-order parasites
(Fig. 1(a)); only 13% were categorised as species specific,
and nearly half of all primate viruses were also reported to
infect non-primate hosts. Protozoan parasites were nearly
evenly distributed among all five levels of specificity
(Fig. 1(b)). Most helminths, on the other hand, were
categorised as host specific, with 48% recorded from single
Fig. 1. Levels of host specificity among the three best-represented parasite
host species (Fig. 1(c)). These differences among the three
groups from wild primates: (a) viruses (nZ82), (b) protozoa (nZ82), and
major parasite groups were statistically significant when (c) helminths (nZ163). Differences in host specificity levels were
tested using all parasite records (c2Z76.0, d.f.Z8, significantly different among the three parasite groups shown.
P!0.0001), and remained significant when we removed
data from parasites with low confidence scores (c2Z31.7, (Fig. 2(a)) and was less common among helminths and
d.f.Z8, P!0.0001) or those with five or fewer references protozoa (c2Z33.96, d.f.Z2, P!0.0001). In fact, viruses
on average (c2Z44.22, d.f.Z8, P!0.0001). were the only pathogen group where sexual transmission was
relatively common, with 17 (20%) of all primate viruses
3.2. Transmission strategies transmitted by sexual contact. Moreover, there was a high
degree of overlap among sexual, vertical, and close non-sexual
Of the 408 parasite species for which we could identify transmission, in that 35% of primate viruses transmitted by
major transmission routes, 185 (45%) could be transmitted close non-sexual contact could also be transmitted both
by close non-sexual contact, 20 (5%) by sexual contact, 20 sexually and vertically. Vector transmission, on the other
(5%) by vertical transmission, 175 (43%) by non-close hand, was far more common among protozoa than other
contact, 132 (32%) by arthropod vectors, and 59 (14%) parasite groups (Fig. 2(b); c2Z35.0, d.f.Z2, P!0.0001).
through intermediate hosts or trophic transmission. These Finally, helminths were the only group where transmission by
categories were not mutually exclusive, with 34% (139) of intermediate hosts was common, with another 25% of
all parasite species being associated with more than one helminths transmitted by biting vectors (Fig. 2(c)).
transmission strategy (and, therefore, counted more than
once in the above percentages). 3.3. Associations between host specificity and
The importance of different transmission routes varied transmission mode
significantly among the three major parasite groups (Fig. 2).
Close contact transmission (including sexual, non-sexual Across all parasites we found a significant association
and vertical routes) was most common among viruses between transmission mode and specificity score (c2Z27.43,
 A.B. Pedersen et al. / International Journal for Parasitology 35 (2005) 647657
Fig. 2. Transmission strategies among the three best-represented parasite
groups from wild primates: (a) viruses (nZ81), (b) protozoa (nZ80), and
(c) helminths (nZ157). Close contact here includes close non-sexual
contact in addition to vertical and sexual transmission (although these latter
two modes are represented only by viruses). Because some parasites could
be transmitted by more than one strategy, we show the frequencies of
parasites with both single and combined strategies.
d.f.Z16, PZ0.037; Med/High confidence: c2Z20.35, d.f.Z
16, PZ0.205), but visual inspection of the data indicated that
different taxonomic groups exhibited contrasting patterns.
Therefore, we investigated the relationship between host
specificity and transmission within viruses, protozoa and
helminths in two ways. First, we assigned each parasite species
a single transmission strategy based on the singular and
combined modes shown in Fig. 2 (close contact, non-close
contact, closeCnon-close contact, vector, and intermediate
host). Second, we examined each route alone and contrasted
parasites capable of transmission by that route against all other
parasites.
3.3.1. Viruses
Associations between categorical measures of viral
transmission and specificity were highly significant
(nZ80; c2Z39.35, d.f.Z8, P!0.0001). Approximately
651
half of all viruses transmitted by close contact, or by a
combination of close and non-close contact, were specific at
the level of primate species, genus or family (Fig. 3(a);
Table 1a). As predicted, parasites transmitted by sexual or
vertical routes tended to be highly host specific
(c2sexual Z 36:51, d.f.Z4, P!0.0001, c2vertical Z 48:22;
d.f.Z4; P!0.0001). In contrast, nearly all vector-borne
viruses were reported from hosts from multiple orders
(Fig. 3(a)) and this result was highly significant (Table 1(a)).
All but one of these vector-borne viruses were RNA viruses.
Furthermore, we observed a major dichotomy in specificity
patterns among DNA vs. RNA viruses, in that virtually all
(86% of 57) RNA viruses in our database could infect hosts
from multiple families or orders (with the exception being
nine retroviruses), whereas specificity levels among the 25
DNA viruses were more evenly distributed among speci-
ficity categories.
3.3.2. Protozoa
Protozoa were dominated by vector-borne parasites but
also included a large number of host-specific parasites. As
with viruses, we found highly significant associations
between specificity scores and categorical transmission
strategies of protozoan parasites using all data (c2Z29.27,
d.f.Z12, PZ0.007), and when excluding specificity scores
with low confidence (c2Z28.76, d.f.Z12, PZ0.010) or for
which five or fewer citations on average were reported
(c2Z21.28, d.f.Z12, PZ0.026). In contrast to the patterns
found in viruses, however, vector borne protozoa were
primarily reported from single primate species, genera, or
families (Fig. 3(b)), and this positive association between
vector transmission and host specificity was highly
significant (Table 1(b)). By comparison, the majority of
protozoa transmitted by both close and non-close contact
were only specific at the level of host order or beyond
(Fig. 3(b)), and again, this association was highly significant
(Table 1(b)).
3.3.3. Helminths
Helminths were generally specific to single host species,
genera, or family (Fig. 3(c)), but levels of specificity
differed significantly among the four major transmission
strategies when tested using data for all helminths (c2Z
41.4, d.f.Z12, P!0.0001), and when restricting the
analysis to specificity scores with high or moderate
confidence only (c2Z31.20, d.f.Z12, PZ0.002)but
results were non-significant when performed using the
subset of parasite species for which an average of five or
more citations were reported. Helminths capable of
transmission by close contact were primarily specific to
single host species (Fig. 3(c)) and this association was
highly significant (Table 1(c)). Incidentally, all helminths
transmitted by close contact could also be transmitted by
non-close contact (Fig. 3(c)). Vector-borne helminths also
tended to be relatively host specific, and many were reported
only from single host species (Fig. 3(c); Table 1(c)).
652 A.B. Pedersen et al. / International Journal for Parasitology 35 (2005) 647657

Fig. 3. The relationship between transmission strategy and host specificity within the three best-represented groups: (a) viruses (nZ80), (b) protozoa (nZ79),
and (c) helminths (nZ153). The number of parasites transmitted by each strategy is represented by the height of each bar, and the specificity scores assigned to
individual parasite species are represented by colors within each bar, with darker colors indicating a greater level of host specificity.

A higher proportion of helminths transmitted by intermedi-
ate hosts or non-close contact alone could infect hosts from
multiple families or orders (Fig. 3(c) and Table 1(c)).
3.4. Comparison with parasites from humans and
domesticated animals
The taxonomic distribution of parasites of wild primates
(i.e. non-human primates) differed from parasites reported
from humans and domesticated species (Fig. 4(a)(d); note
that arthropods were omitted for this comparison). The most
significant differences in the taxonomic representation of
parasites occurred between humans vs. wild primates
(Fig. 4(a) and (b)), and this difference was highly significant
(c2Z318.03, d.f.Z4, P!0.0001). In particular, 87% of
parasites from wild primates were helminths, viruses and
protozoa, whereas bacteria and fungi comprised the
majority (60%) of human pathogens.
Despite these differences, there was a high degree of
overlap among parasites from humans and wild primates,
with 114 (27.5%) shared between wild primates and humans
(Table 2). Approximately half of all bacteria and fungi
reported from wild primates were also reported as human
pathogens. Even though the greatest total number of
parasites shared between the two databases involved
helminths, only 20% of all wild primate helminths were
reported to infect humans. As predicted, the vast majority
(90%) of non-human primate parasites also reported from
 A.B. Pedersen et al. / International Journal for Parasitology 35 (2005) 647657 653

Table 1
Associations between parasite transmission strategy and host specificity based on chi-square analysis of major transmission routes (presence/absence) in
relation to host specificity scores

(a) Viruses
Transmission strategy All data (nZ80)a
Dir c2 d.f. Pb
Close contact C 33.24 4 0.000
CloseCnon-close NS 4.77 4 0.372
Vector borne K 31.60 4 0.000
(b) Protozoans
Transmission strategy All data (nZ79) Excl. low confidence (nZ50)
2
Dir c d.f. P Dir c2 d.f. Pc

Close contact K 18.44 4 0.001 K 14.05 4 0.007

Non-close contact C 13.12 4 0.011 C 12.17 4 0.016
CloseCnon-close K 16.97 4 0.002 K 14.10 4 0.007
Vector borne C 16.37 4 0.003 C 14.31 4 0.006
(c) Helminths
Transmission strategy All data (nZ163) Excl. low confidence (nZ80)
2
Dir c d.f. P Dir c2 d.f. Pd

CloseCnon-close C 23.04 4 0.000 C 23.20 4 0.000

Non-close contact K 15.10 4 0.005 K 12.10 4 0.016
Vector borne C 10.91 4 0.028 NS 2.56 4 0.633
Intermediate host K 31.72 4 0.000 K 9.95 4 0.014

Each analysis was repeated using all available data and those for which high or moderate confidence was assigned to specificity scores. Symbols (C/K)
indicate the directionality of the association (i.e. whether parasites with that transmission strategy were more or less specific than parasites transmitted by other
routes). c2, Chi-square scores. d.f., Degrees of freedom, P, P value corrected by the Bonferroni method (aZ0.05/n, where n is the number of tests and a, b, c
and d refer to the Bonferroni corrected P values.
a
No low confidence scores.
b
aZ0.017.
c
aZ0.0125.
d
aZ0.0125.

humans were recorded as multihost parasites, capable of
infecting hosts from multiple families and orders.
Viruses reported from both humans and wild primates
were primarily RNA viruses (85% of shared viruses), and
over half were transmitted by arthropod vectors (Fig. 5).
Protozoa reported to infect both wild primates and humans
spanned multiple phyla and were transmitted either by a
combination of close and non-close contact (70%) or
arthropod vectors (29%; Fig. 5). Approximately half
(45%) of the helminths that overlapped with wild primate
and human hosts had complex life cycles and transmission
by intermediate hosts, and the remainder were nematodes
transmitted by non-close contact or arthropod vectors
(Fig. 5). Only 10% of the shared parasites were recorded
as being transmitted by close contact only, and all of these
were viruses.
Finally, 44.7% (51 of 114) of non-human primate
parasites reported to infect humans were classified as
emerging based on a recent analysis of risk factors
associated with human diseases (Taylor et al., 2001). Over
half of these emerging parasites were viruses; in fact, 28%
of all viruses reported from wild primates were classified
as emerging in humans (Table 2). Moreover, there was
a significant association between emerging diseases
and transmission strategy (c2Z15.37; d.f.Z4; PZ0.008).
Given that a shared parasite was transmitted by close
contact, it was highly likely to be recorded as emerging in
humans (Table 2(b)), and the same was true for parasites
transmitted by arthropod vectors. Shared parasites trans-
mitted by intermediate hosts were least likely to be classified
as emerging in humans (Table 2(b)).
4. Discussion
We found that a remarkable number of primate
pathogens (68%) were reported to infect multiple host
species. Since most of these affected hosts were from more
than one family or order, our data suggest that species-
specific parasites are an exception to the rule (Fig. 1). This
observation is consistent with patterns from studies of
pathogens in humans and domesticated animals (Cleaveland
et al., 2001; Taylor et al., 2001; Woolhouse et al., 2001).
However, it runs counter to the hypothesis that host
specificity (and, more generally, niche specialisation) offers
greater opportunity for diversification and coexistence
among parasitic organisms and their hosts (Berenbaum,
1996; McPeek, 1996; Poulin, 1998). A number of factors
654 A.B. Pedersen et al. / International Journal for Parasitology 35 (2005) 647657

Fig. 4. Taxonomic distribution of parasites from (a) free-living primates (nZ369), (b) humans (nZ1415), (c) domestic livestock (cattle, sheep, goats, pigs and
horses; nZ616) and (d) domestic carnivores (dogs and cats; nZ374). Fig. 1(b)(d) were recreated with data from Cleaveland et al. (2001). For comparison
with other host groups, 46 arthropod parasites from non-human primates were omitted.

could influence patterns of host specificity. First, taxonomic and May, 1991), and their relatively more complex life
patterns pointed to a possible role for fast replication and history strategies could limit their ability to infect or adapt
high genetic variability in facilitating infection of multiple to new host species. However, the strong associations
hosts. Second, links between transmission mode and host between specificity and parasite type could also be
specificity indicated that transmission might determine
encounter probabilities between host and parasite species. Table 2
Viruses might be dominated by generalist pathogens Characteristics of shared wild primate and human parasites (human records
taken from Ashford and Crewe, 1998; Taylor et al., 2001)
because their high mutation rates, greater antigenic
diversity, high population diversity and short generation (a) Parasite type
times (Morand et al., 1996; Whitlock, 1996; Gupta et al., Parasite type No. reported to infect No. recorded as
1998) allow them to rapidly adapt to a larger number of wild primates and emerging in humans
humans
hosts. In fact, nearly half of primate viruses could infect
hosts from multiple orders, and virtually all of these (38/39) Arthropod 4 a
Bacteria 19 12
multi-order viruses were RNA viruses, which have higher
Fungi 5 3
mutation rates than DNA viruses (Drake, 1991; Domingo Helminth 33 4
and Holland, 1997; Holmes, 2003). This taxonomic bias in Protozoa 24 6
levels of host specificity is consistent with recent analyses of Virus 30 26
zoonotic pathogens and risk of emergence in humans and (b) Transmission strategy
domesticated species, as previous studies showed that RNA Transmission strategy No. reported to infect No. recorded as
viruses of humans are more than twice as likely to be wild primates and emerging in humans
zoonotic than DNA viruses (Cleaveland et al., 2001; Taylor humans
et al., 2001).
Close contact 13 9
Relative to viruses, protozoa were evenly distributed CloseCnon-close 38 17
among the specificity classes, and nearly half of the Non-close contact 16 4
helminths reported from primates were recorded as Vector 30 19
species-specific (Fig. 1(c)). Helminths might have longer Intermediate host 15 1
generation times than most microparasites (Anderson a
Taylor et al. (2001) did not include arthropods.
 A.B. Pedersen et al. / International Journal for Parasitology 35 (2005) 647657
Fig. 5. Major transmission strategies of viruses (nZ30), protozoa (nZ24), and helminths (nZ33) reported to infect both wild non-human primates and
humans. Nearly all of these pathogens (O90%) were recorded as multi-order or multi-family generalists in the primate database.
influenced by taxonomic biases innate in applying the
concept of a species across a wide diversity of phyla.
Because this bias is potentially substantial, we restricted
most of our analyses to comparisons within the three major
parasite groups.
Clear but contrasting patterns emerged between speci-
ficity and transmission type within viruses, protozoa and
helminths. First, viruses were dominated by vector-borne
pathogens capable of infecting hosts from multiple orders,
and close contact (including sexual and vertical) transmitted
pathogens with greater levels of host specificity (Fig. 3(a)).
As predicted, most viruses capable of sexual transmission,
including SIVs and STLVs, were highly host specific,
although most of these were also capable of transmission by
close contact and vertical routes. Second, most protozoan
parasites fell into one of two transmission-specificity
categories: multihost (multi-family or order) parasites
transmitted by a combination of close and non-close
contact, and vector-borne parasites that were more host
specific (Fig. 3(b)). Helminths were by far the most
host-specific primate parasites, but were also dominated
by species transmitted by indirect routes, including non-
close contact, vectors, and intermediate hosts (Fig. 3(c)).
Specificity patterns among helminths infecting wild pri-
mates were difficult to characterise, possibly due to the
diverse groups encompassed within helminths (i.e. acantho-
cephalens, nematodes, cestodes and trematodes). Never-
theless, these tended to fall into three major groups: highly
specific parasites capable of transmission by combined
close and non-close contact, vector-borne parasites with
relatively high specificity, and parasites transmitted by
intermediate hosts that included both highly specific and
multi-order parasites (Fig. 3(c)).
Patterns reported from our data set have implications for
understanding disease risk in humans and domesticated
species, in part because wild primates are humans closest
relatives, and also because of a growing awareness that
many emerging diseases are multihost pathogens that can be
shared among humans, domesticated animals, and wildlife.
655
Given their close evolutionary relationship, the taxonomic
distribution of parasites from wild primates was surprisingly
unlike patterns from humans (Taylor et al., 2001; Cleave-
land et al., 2001), which were dominated by bacteria and
fungi (Fig. 4). Nevertheless, over 100 parasite species from
wild primates were also reported in humans (Fig. 5), and
nearly all of these could infect hosts from multiple orders or
families. Moreover, 46 wild primate parasites were also
reported as emerging in human populations, and nearly half
of these were viruses transmitted by close contact or
arthropod vectors. Further investigation should reveal which
of these pathogens are associated with primates as reservoir
hosts, and which are primarily human pathogens that spill
over into wild primates (e.g. Haydon et al., 2002).
Although we believe this study provides an important
first step in quantifying the characteristics of pathogens
from wild host species, several limitations of this dataset
should be noted. First and foremost, although we system-
atically searched for wild primate parasites, our list is
unlikely to be complete, in part because many primate
species have not been sampled adequately for parasites in
the wild (Nunn et al., 2003). Similarly, our study might be
limited by sampling biases towards particular pathogen
groups and accurate knowledge of parasite taxonomy. Thus,
one explanation for the high proportion of multi-host
parasites among viruses is that virus taxonomy is poorly
resolved. Improved information could lead to viruses
currently classified as multihost being separated into a
series of host-specific species (i.e. host races). Moreover,
biased sampling might occur for parasites among different
taxonomic groups. For example, nematologists might be
most interested in collecting and describing new helminth
species, whereas virologists might be most interested in
studying zoonotic pathogens. Although we attempted to
control for study biases in parasite sampling effort by
calculating specificity confidence scores and running the
analyses excluding those with low confidence, analyses of
confidence scores and citation counts suggest that well
studied parasites have a wider reported host range.
656 A.B. Pedersen et al. / International Journal for Parasitology 35 (2005) 647657
Overall, our results indicate that there are fundamental
differences in specificity within the taxonomic groups and
that these differences are significantly related to trans-
mission strategies, but these patterns might change as our
understanding of parasite taxonomy and sampling effort
improve.
The distribution and population biology of multi-host
pathogens is becoming increasingly important for wildlife
conservation and human health (Murray et al., 1999; Daszak
et al., 2000), and adding multiple host species to infectious
disease systems can have major consequences for disease
spread and evolution (Frank, 1993; Begon et al., 1999;
Woolhouse et al., 2001; Antonovics et al., 2002; Gandon,
2002; Holt et al., 2003; Gandon, 2004). For example, the
presence of reservoir hosts can lead to periodic pathogen
resurgence following long durations of disease-free periods
in highly susceptible host species (Cleaveland and Dye,
1995; Keeling and Gilligan, 2000; Haydon et al., 2002;
Swinton et al., 2002), and parasites in multiple host systems
can intensify disease impacts on sensitive wildlife species
(Greenman and Hudson, 2000). For other pathogens, such as
Borrelia burdorferi (the causative agent of Lyme disease), a
greater diversity of host species can reduce pathogen
impacts and prevalence (Schmidt and Ostfeld, 2001;
LoGiudice et al., 2003). The presence of multiple host
species could also affect the evolution of pathogen virulence
and the evolution of transmission routes among different
hosts (Gandon, 2004).
In summary, despite the fact that parasites are often
assumed to be under selection for specialisation on
commonly infected host species (Fry, 1996; Whitlock,
1996; Kawecki, 1998), multihost pathogens are common in
wild primate hosts. These multihost pathogens might
experience several advantages, including reduced risks of
ecological extinction, as their fates are not linked to the
success of one or a few host species (Poulin, 1998). Results
from our study provided limited support for two key
processes that might favor the occurrence of generalism,
including the ability to disperse to and infect multiple host
species. However, our results also indicate that the
evolutionary and ecological pressures influencing parasite
transmission and host specificity differ among parasite
taxonomic groups, perhaps because of their unique biology
and life cycles. A clearer understanding of the limits to and
opportunities for specialisation within each parasite group
should help to explain these patterns, as will a clearer
picture of potential sampling biases that could influence
observed records of host and pathogen species
combinations.
Acknowledgements
The authors thank Nick Vitone for assistance in
compiling records for data on host specificity of a large
number of parasite species. Helpful discussion and
comments on earlier drafts of the manuscript were provided
by Vanessa Ezenwa, Andy Fenton, Janis Antonovics,
Michael Hood, Andy Dobson, Peter Thrall, and Andy
Davis. Useful comments were provided by two anonymous
reviewers. This research was supported by funding from the
NSF (Grant #DEB-0212096 to CN and SA) and the Center
for Applied Biodiversity Science at Conservation Inter-
national. This work was also conducted as part of the
Infectious Disease and Host Behaviour Working Group
supported by the National Center for Ecological Analysis
and Synthesis (NCEAS), funded by the NSF, the University
of California, and the Santa Barbara campus.
References
Anderson, R.M., May, R.M., 1991. Infectious Diseases of Humans:
Dynamics and Control. Oxford University Press, Oxford, UK.
Antonovics, J., Hood, M., Partain, J., 2002. The ecology and genetics of a
host shift: Microbotryum as a model system. Am. Nat. 160, S40S53.
Ashford, R.W., Crewe, W., 1998. The Parasites of Homo Sapiens: An
Annotated Checklist of the Protozoa, Helminths and Arthropods for
Which We are Home. Cromwell Press, Trowbridge.
Begon, M., Hazel, S.M., Baxby, D., Brown, K., Cavanagh, R., Chantrey, J.,
Jones, T., Bennett, M., 1999. Transmission dynamics of a zoonotic
pathogen within and between wildlife host species. Proc. R. Soc. Lond.
B 266, 19391945.
Berenbaum, M.R., 1996. Introduction to the symposium: on the evolution
of specialization. Am. Nat. 148, S78S83.
Bitter, W.H.G., Kieft, R., Borst, P., 1998. The role of transferin-receptor
variation in the host range of Tyrpanosoma brucei. Nature 391,
499502.
Bull, J., 1994. Virulence. Evolution 48, 14231438.
Cleaveland, S., Dye, C., 1995. Maintenance of a microparasite infecting
several host species: rabies in the Serengeti. Parasitology 111, S33S47.
Cleaveland, S., Laurenson, M.K., Taylor, L.H., 2001. Diseases of humans
and their domestic mammals: pathogen characteristics, host range, and
the risk of emergence. Philos. Trans. R. Soc. Lond. B 356, 991999.
Corbet, G.B., Hill, J.E., 1991. A World List of Mammalian Species. Oxford
University Press, Oxford, UK.
Daszak, P., Cunningham, A.A., Hyatt, A.D., 2000. Emerging infectious
diseases of wildlifethreats to biodiversity and human health. Science
287, 443449.
Day, T., 2001. Parasite transmission modes and the evolution of virulence.
Evolution 55, 23892400.
Desdevises, Y., Morand, S., Legendre, P., 2002. Evolution and determi-
nants of host specificity in the genus Lamellodiscus (Monogenea). Biol.
J. Linnean Soc. 77, 431443.
Dobson, A., Foufopoulos, J., 2001. Emerging infectious pathogens in
wildlife. Philos. Trans. R. Soc. Lond. B 356, 10011012.
Domingo, F., Holland, J.J., 1997. RNA virus mutations and fitness for
survival. Annu. Rev. Microbiol. 51, 151178.
Drake, J.W., 1991. A constant rate of spontaneous mutation in DNA-based
microbes. Proc. Natl Acad. Sci. USA 88, 71607164.
Frank, S.A., 1993. Specificity versus dectable polymorphism in host-
parasite genetics. Proc. R. Soc. Lond. B 254, 191197.
Fry, J.D., 1996. The evolution of host specialization: are trade-offs
overrated?. Am. Nat. 148, S84S107.
Gandon, S., 2002. Local adaptation and the geometry of host-parasite
coevolution. Ecol. Lett. 5, 246256.
Gandon, S., 2004. Evolution of multihost parasite systems. Evolution 58,
455469.
 A.B. Pedersen et al. / International Journal for Parasitology 35 (2005) 647657
Greenman, J., Hudson, P., 2000. Parasite-mediated and direct competition
in a two-host shared macroparasite system. Theor. Popul. Biol. 57,
1334.
Groves, C.P., 2001. Primate Taxonomy. Smithsonian Institution Press,
Washington, DC.
Gupta, S., Ferguson, N., Anderson, R.M., 1998. Chaos, persistence, and
evolution of strain structure in antigenically diverse infectious agents.
Science 280, 912915.
Hahn, B.H., Shaw, G.M., De Cock, K.M., Sharp, P.M., 2000. AIDS as a
zoonosis: scientific and public health implications. Science 287, 607
614.
Harvell, C.D., Kim, K., Burkholder, J.M., Colwell, R.R., Epstein, P.R.,
Grimes, D.J., Hofmann, E.E., Lipp, E.K., Osterhaus, A.D.M.E.,
Overstreet, R.M., Porter, J.W., Smith, G.W., Vasta, G.R., 1999.
Emerging marine diseasesclimate links and anthropogenic factors.
Science 285, 15051510.
Haydon, D.T., Cleaveland, S., Taylor, L.H., Laurenson, M.K., 2002.
Identifying reservoirs of infection: a conceptual and practical challenge.
Emerg. Infect. Dis. 8, 14681473.
Holmes, E.C., 2003. Error thresholds and the constraints to RNA virus
evolution. Trends Microbiol. 11, 543548.
Holt, R., Dobson, A., Begon, M., Bowers, R., Schauber, E., 2003. Parasite
establishment in host communities. Ecol. Lett. 6, 837842.
Jensen, T., Bildt, M.V.D., Dietz, H.H., Andersen, T.H., Hammer, A.S.,
Kuiken, T., Osterhaus, A., 2002. Another phocine distemper outbreak in
Europe. Science 297, 209.
Kawecki, T., 1998. Red Queen meets Santa Rosalia: arms races and the
evolution of host specialization in organisms with parasitic life cycles.
Am. Nat. 152, 635651.
Keeling, M.J., Gilligan, C.A., 2000. Metapopulation dynamics of bubonic
plague. Nature 407, 903906.
Leroy, E.M., Rouquet, P., Formenty, P., Souquiere, S., Kilbourne, A.,
Froment, J.-M., Bermejo, M., Smit, S., Karesh, W., Swanepoel, R.,
Zaki, S.R., Rollin, P.E., 2004. Multiple Ebola transmission events and
rapid decline of central African wildlife. Science 303, 387390.
LoGiudice, K., Ostfeld, R.S., Schmidt, K.A., Keesing, F., 2003. The
ecology of infectious disease: effects of host diversity and community
composition on Lyme disease risk. Proc. Natl Acad. Sci. USA 100,
567571.
McPeek, M.A., 1996. Trade-offs, food web structure, and the coexistence of
habitat specialists and generalists. Am. Nat. 148, S124S138.
Miller, M.A., Gardner, I.A., Kreuder, C., Paradies, D.M., Worcester, K.R.,
Jessup, D.A., Dodd, E., Harris, M.D., Ames, J.A., Packham, A.E.,
Conrad, P.A., 2002. Coastal freshwater runoff is a risk factor for
Toxoplasma gondii infection of southern sea otters (Enhydraw lutris
nereis). Int. J. Parasitol. 32, 9971006.
Morand, S., Manning, S.D., Woolhouse, M.E.J., 1996. Parasite-host
coevolution and geographic patterns of parasite infectivity and host
susceptibility. Proc. R. Soc. Lond. B 263, 119128.
Murphy, F.A., 1998. Emerging zoonoses. Emerg. Infect. Dis. 4, 429435.
Murray, D., Kapke, C., Evermann, J., Fuller, T., 1999. Infectious disease
and the conservation of free-ranging large carnivores. Anim. Conserv.
2, 241254.
657
Nunn, C.L., Altizer, S., 2005. The global mammal parasite database: an
online resource for infectious disease records in wild primates. Evol.
Anthropol., 14, 12.
Nunn, C.L., Altizer, S., Jones, K.E., Secrest, W., 2003. Comparative tests of
parasite species richness in primates. Am. Nat. 162, 597614.
Poulin, R., 1998. Evolutionary Ecology of Parasites. Chapman and Hall,
New York.
Roelke-Parker, M.E., Munson, L., Packer, C., Kock, R., Cleaveland, S.,
Carpenter, M., Obrien, S.J., Pospischil, A., Hofmann-Lehmann, R.,
Lutz, H., Mwamengele, G.L.M., Mgasa, M.N., Machange, G.A.,
Summers, B.A., Appel, M.J.G., 1996. A canine distemper virus
epidemic in Serengeti lions (Panthera leo). Nature 376, 441445.
Rowe, N., 1996. The Pictorial Guide to the Living Primates. Pogonias
Press, East Hampton, NY.
Schmidt, K.A., Ostfeld, R.S., 2001. Biodiversity and the dilution effect in
disease ecology. Ecology 82, 609619.
Simon, J.H.M., Miller, D.L., Fouchier, R.A.M., Soares, M.A.,
Peden, K.W.C., Malim, M.H., 1998. The regulation of primate
immunodeficiency virus infectivity by Vif is cell species restricted: a
role for Vif in determing virus host range and cross-species
transmission. Eur. Med. Biol. Org. J. 17, 12591267.
Swinton, J., Woolhouse, M.E.J., Begon, M., Dobson, A., Ferroglio, E.,
Grenfell, B.T., Guberti, V., Hails, R.S., Heesterbeek, J.A.P.,
Lavanzza, A., Roberts, M.G., White, P.J., Wilson, K., 2002. Micro-
parasite transmission and persistence. In: Hudson, P., Rizzoli, A.,
Grenfell, B.T., Heesterbeek, H., Dobson, A.P. (Eds.), Ecology of
Wildlife Diseases. Oxford University Press, Oxford, UK.
Taylor, L.H., Latham, S., Woolhouse, M.E.J., 2001. Risk factors for human
disease emergence. Philos. Trans. R. Soc. Lond. B 356, 983989.
Wallis, J., Lee, D.R., 1999. Primate conservation: the prevention of disease
transmission. Int. J. Primatol. 20, 803826.
Walsh, P.D., Abernethy, K.A., Bermejo, M., Beyersk, R., De Wachter, P.,
Akou, M.E., Huljbregis, B., Mambounga, D.I., Toham, A.K.,
Kilbourn, A.M., Lahm, S.A., Latour, S., Maisels, F., Mbina, C.,
Mihindou, Y., Obiang, S.N., Effa, E.N., Starkey, M.P., Telfer, P.,
Thibault, M., Tutin, C.E.G., White, L.J.T., Wilkie, D.S., 2003.
Catastrophic ape decline in western equatorial Africa. Nature 422,
611614.
Whitlock, M.C., 1996. The Red Queen beats the jack-of-all-trades: the
limitations on the evolution of phenotypic plasticity and niche breadth.
Am. Nat. 148, S65S77.
Williams, E.S., Yuill, T., Artois, M., Fischer, J., Haigh, S.A., 2002.
Emerging infectious diseases in wildlife. Sci. Tech. Rev. Int. Off. Epiz.
21, 139157.
Wolfe, N., Escalante, A., Karesh, W., Kilbourn, A., Spielman, A., Lal, A.,
1998. Wild primate populations in emerging infectious disease
research: the missing link?. Emerg. Infect. Dis. 4, 149158.
Wolfe, N.D., Switzer, W.M., Carr, J.K., Bhullar, V.B., Shanmugam, V.,
Tamoufe, U., Prosser, A.T., Torimiro, J.N., Wright, A., Mpoudi-
Ngole, E., McCutchan, F.E., Brix, D.L., Folks, T.M., Burke, D.S.,
Heneine, W., 2004. Naturally acquired simian retrovirus infections in
central African hunters. Lancet 363, 932937.
Woolhouse, M.E.J., Taylor, L.H., Haydon, D.T., 2001. Population biology
of multi-host pathogens. Science 292, 11091112.
 International Journal for Parasitology 35 (2005) 659672
www.parasitology-online.com

Redescription of Besnoitia bennetti (Protozoa: Apicomplexa)

from the donkey (Equus asinus)*
J.P. Dubeya,*, C. Sreekumara, T. Donovanb, M. Rozmanecb, B.M. Rosenthala,
M.C.B. Viannaa, W.P. Davisc, J.S. Beldend
a
Animal Parasitic Diseases Laboratory, Agricultural Research Service, United States Department of Agriculture, Animal and Natural Resources Institute,
Beltsville Agricultural Research Center, Building 1001, BARC-East, Beltsville, MD 20705-2350, USA
b
Department of Biomedical Sciences, New York State College of Veterinary Medicine, Cornell University, Ithaca, NY 14853, USA
c
New Hampshire Veterinary Diagnostic Laboratory, University of New Hampshire, Durham, NH 03824, USA
d
14800, 46th Lane South, Wellington, FL 33414, USA
Received 23 November 2004; received in revised form 17 January 2005; accepted 19 January 2005

Abstract

Besnoitia bennetti tissue cysts were found in four naturally-infected donkeys (Equus asinus) from the USA. Infectivity of its bradyzoites
and tachyzoites to animals and cell culture was studied. The bradyzoites were not infectious to out-bred Swiss Webster mice, rabbits or
gerbils. When fed tissue cysts, cats did not excrete oocysts. However, the parasite was infectious to interferon-gamma gene knock out mice.
The parasite from tissues of two donkeys was grown successfully in bovine monocyte monolayers for the first time. Non-dividing,
uninucleate tachyzoites were approximately 6!1.5 mm in size. Longitudinally-cut bradyzoites in tissue sections measured 8.7!1.9 mm.
Ultrastructurally, tachyzoites and bradyzoites were similar to those in other Besnoitia species, and in particular to parasites described from
cattle (Besnoitia besnoiti) and reindeer (Besnoitia tarandi), in that their bradyzoites lacked enigmatic bodies. Based on comparative analysis
of three portions of nuclear ribosomal DNA (the small and large subunits and the first internal transcribed spacer) B. bennetti was found to be
more closely related to the other congeners described from ungulates. The parasite was formally redescribed and specimens deposited in the
US National Parasite Collections.
Published by Elsevier Ltd. on behalf of Australian Society for Parasitology Inc.

Keywords: Besnoitia bennetti; Donkeys; Equus asinus; USA; Mice; Gerbils; Cell culture; Ultrastructure; Sequence analysis; Taxonomy

1. Introduction from different intermediate host species have traditionally

Species of the coccidian genus Besnoitia parasitise cattle,
goats, equids, reindeer, caribou, opossums, rabbits, rodents,
and lizards (Dubey et al., 2003a). There are several
recognised and unrecognised species in the genus but the
life cycles of only three (Besnoitia darlingi, Besnoitia
wallacei, and Besnoitia oryctofelisi) are known, and
morphological differences among the species are poorly
defined (Dubey et al., 2003a). Although parasites isolated
*
Nucleotide sequence data reported in this paper are available in the
GenBank, EMBL and DDBJ databases under the accession numbers:
AY827838, 827839.
* Corresponding author. Tel.: C1 301 504 8128; fax: C1 301 504 9222.
E-mail address: jdubey@anri.barc.usda.gov (J.P. Dubey).
0020-7519/$30.00 Published by Elsevier Ltd. on behalf of Australian Society for Parasitology Inc.
doi:10.1016/j.ijpara.2005.01.004
been presumed to represent distinct parasite taxa, it will be
difficult to test their actual host specificity until natural
definitive hosts are identified.
Besnoitia bennetti was first reported by Bennett (1927,
1933) in four horses from Sudan. He believed this parasite
was a species of Sarcocystis or Globidium and that it
infected both cattle and horses. Babudieri (1932) transferred
this parasite to the genus Besnoitia. Schulz and Thorburn
(1955) found Besnoitia sp. tissue cysts in a horse from South
Africa. Pols (1960) listed other unpublished reports of
cutaneous besnoitiosis in horses in South Africa and
indicated that he could not infect two horses with Besnoitia
besnoiti of cattle. Bigalke (1970) found Besnoitia sp. in
donkeys, mules, and horses in South Africa; he was unable
to transmit the Besnoitia species from equids to cattle or
660 J.P. Dubey et al. / International Journal for Parasitology 35 (2005) 659672

from cattle to equids, indicating that equine besnoitiosis
may result from a parasite specific to equids. Furthermore,
equid Besnoitia sp. was not infective to rabbits, hamsters,
white mice, or guinea pigs. van Heerden et al. (1993)
provided the first ultrastructural description of B. bennetti
from a horse from South Africa. Lane et al. (1986) reported
B. bennetti-like tissue cysts associated with laryngeal
papillomatosis in a horse imported to England from
Argentina. However, the photomicrographs lack details
that permit unambiguous diagnosis.
The only reports of equine besnoitiosis out side of Africa
are from the USA. Terrell and Stookey (1973) found
Besnoitia in two of 15 Mexican burrows in a closed
experimental herd maintained at the US Army Research
Institute, Frederick, Maryland. Davis et al. (1997) found
Besnoitia sp. in one of eight donkeys on a farm in Montana,
USA. Recently clinical besnoitiosis was reported in 14 of 38
miniature donkeys from a herd in Michigan, USA
(Elsheikha et al., 2005). To our knowledge, besnoitiosis in
horses has not been reported in the USA.
The life cycle of B. bennetti is not yet completely known.
Only the tissue cyst and bradyzoite stages are known, and
their structures are not fully described. Thus, little basis
exists to evaluate whether the same etiological agent
establishes infection, and induces disease, in horses and
donkeys and possibly in other hosts. Here, we report the
successful isolation of B. bennetti in culture and in
immunodeficient mice, describe the structure of its
bradyzoites and tachyzoites, and define portions of its
ribosomal DNA to aid its diagnosis and comparison to
related coccidia. Specimens are also deposited in museum
collections to aid future comparison.
2. Materials and methods
2.1. Naturally-infected donkeys
2.1.1. Donkey no. 1
A two year old female (Fig. 1E) and her sibling were
purchased when seven months old by the present owner
from a commercial breeder in Texas. This donkey
developed severe pruritus and loss of hair at about 18 months
of age, whereas its sibling and two other donkeys (nos. 24),

Fig. 1. Lesions and Besnoitia bennetti tissue cysts in donkey no. 5. (A) Before onset of clinical signs at approximately one year of age. (B) Before euthanasia,
eight months after onset of clinical signs. Note marked cachexia, poor hair coat, and skin lesions (alopecia, hypotrichosis, hyper pigmentation, thickening and
crusting) over lateral neck, shoulders, carpi, stifle, hock, periocular region, around ears and muzzle. (C) Lateral aspect of hind limb. Note alopecia and
hypotrichosis, with thickened, irregular crusty hyper pigmented skin over caudal aspect of stifle and lateral hock (arrows). (D) Multifocal white pinpoint
granular structures (Besnoitia tissue cysts) within the sclera and conjunctiva (arrows) of the eye. (E) Lateral view of donkey no. 1 after treatment with SMZ and
TM. Note improvement in hair coat.
 J.P. Dubey et al. / International Journal for Parasitology 35 (2005) 659672
and seven horses on the same premises remained clinically
normal. Haematologic evaluation revealed anaemia (hae-
moglobin 9.8 d/L, red blood cells 5!106/cm3 haematocrit
28.4%), elevated creatinine phosphokinase (382 U/L),
elevated alkaline phosphatase (259 U/L). Cutaneous biopsy
revealed granulomatous dermatitis with intralesional Bes-
noitia sp. tissue cysts. Another biopsy was obtained a few
weeks later to attempt isolate the parasite in culture. The
donkey was treated with sulfamethaxzole (SMZ, 800 mg)
and trimethoprim (TM, 160 mg) tablets (five tablets daily)
for seven months. Following two months of therapy, folic
acid (10 mg) was added to the treatment regime to combat
thrombocytopenia, usually associated with sulfonamide
treatment. All lesions clinically resolved except a solitary
lesion on the bridge of the nose, which was biopsied seven
months after the initiation of treatment. Donkey nos. 24
were also treated with SMZ and TM for two months
although they were clinically normal. Blood samples were
obtained from all four donkeys at the time of the last biopsy
and a week later from the seven horses.
2.1.2. Donkey no. 5
This miniature 21month old female donkey (Fig. 1A and B)
was born in Vermont and shipped to Massachusetts. Neither
this nor any other contact animal had traveled outside of the
United States. This donkey had been clinically normal until
skin lesions were noted eight months prior to euthanasia.
Initially, skin lesions consisted of crusting on both flanks
and the perivulvar area. Leukocytosis was present (12.89 m/
mm3, range 5.012.0 m/mm3), total protein was elevated at
9.0 g/dL (range 5.78.0), and globulins were elevated at
6.0 g/dL (range 2.75.0). The donkeys hematocrit was just
below normal at 31% (3253% reference range). Right
carpal lameness was noted at this time, and the animal was
treated with SMZ (unknown dose) and tetracycline
(2500 mg, orally, twice daily for 30 days). Lyme disease,
bacterial arthritis and immune mediated arthritis were
considered unlikely due to negative Borrelia burgdorferi
antibody titers, synovial fluid culture, and anti-nuclear
antibody serological testing, respectively.
One month after the onset of clinical signs, the donkey
was suspected to have forelimb laminitis. Blood evaluation
at this time revealed that the leukocyte counts were within
normal range, protein levels were still elevated (total protein
9.5 g/dL, globulin 7.5 g/dL), and the hematocrit had
dropped further to 26.7%. Approximately three months
after the onset of clinical signs, protein levels were still
elevated (total protein 8.9 g/dL, and globulins 6.6 g/dL),
and the packed cell volume was further diminished at
25.2%. A repeat examination two months later revealed low
albumin at 1.9 g/dL (range 2.23.7). Total protein was
8.4 g/dL and globulins were 6.5 g/dL. Haematocrit was low
at 23%, and the leukocyte count was at the low end of the
normal range at 6.41 m/mm3. The skin was biopsied
approximately one month before euthanasia, which revealed
a locally extensive, chronic fibrosing perivascular to
661
interstitial lymphoplasmacytic and granulomatous super-
ficial dermatitis with intradermal Besnoitia tissue cysts. As
the skin lesions progressed, the donkey became cachectic,
and was euthanised.
2.1.3. Donkey nos. 6 and 7
These animals were from the same farm as donkey no. 5.
Donkey no. 6, an 11 year old female, had a three month
history of progressive alopecia, erythema and crusting
involving the face, neck, vulva, and lower leg regions.
Donkey no. 7, a one year old female, had a similar history of
alopecia and crusting that was confined to the muzzle and
neck regions. Cutaneous besnoitiosis was diagnosed based
on histologic examination of skin biopsy specimens from
these animals. Both animals were originally treated for
several months with SMZ, but Donkey no. 6 was eventually
euthanised following a relapse of the skin lesions. Donkey
no. 7 responded well to SMZ therapy. Treatment was
continued with another antiprotozoal, Navigator paste
(32% w/w nitazoxanide) for five days at 5 mg/kg followed
by 23 days at 10 mg/kg (Idexx Laboratories, Inc.,
Westbrook, ME). Although clinical improvement was
noted, Besnoitia tissue cysts and chronic inflammation
were observed in follow-up skin biopsy specimens.
2.2. Infection of laboratory animals
Laboratory-reared gerbils, rabbit, mice, and cats were
inoculated with B. bennetti. Animal experiments were
performed according to Animal Care Protocols of the US
Department of Agriculture.
The cats were from a parasite-free colony (Dubey, 1995).
Muscles and skin heavily infected with tissue cysts from
donkey no. 5 were fed to two cats. Faeces of cats were
examined for coccidian oocysts by sugar flotation daily for
28 days thereafter (Dubey, 1995).
The five to eight week old female gerbils (Meriones
unguiculatus) were obtained from Charles River Labora-
tory, Stoneridge, New York, USA. The 2025 g female
Swiss Webster (SW) out-bred mice were obtained from
Taconic Farms, Germantown, New York, USA. The 510
week old female interferon-gamma gene knockout (KO)
mice were obtained from Jackson Laboratories; these KO
mice are highly susceptible to intracellular parasites because
they lack the capacity to produce the cytokine interferon-
gamma, necessary for killing of intracellular parasites
(Dubey and Lindsay, 1998).
A skin biopsy from donkey no. 1 was homogenised in
phosphate buffer saline (PBS) by mortar and pestle. The
homogenate was sieved through gauze, centrifuged, resus-
pended in antibiotics (1000 units of penicillin and 100 mg of
streptomycin) and inoculated into animals (Table 1). Pieces
of skin from donkey no. 5, containing numerous tissue cysts
were homogenised in PBS, centrifuged, washed in antibiotic
solution, and inoculated into animals (Table 1). The
released bradyzoites were filtered through a 5-mm filter,
662 J.P. Dubey et al. / International Journal for Parasitology 35 (2005) 659672

Table 1
Infectivity of Besnoitia bennetti to mice and gerbils

Source of inoculum Animal species No. inoculated Inoculum Route Outcomea

Donkey no. 1 KO mice 4 Bradyzoitesb s.c. D57, D75, K79, K79
4 Bradyzoitesb i.p. D44, K44, K79, K79
5 Tachyzoitesc s.c., i.p. DK 2136
3 Tachyzoitesd i.p. D19, D 29, D49
2 Tachyzoitesc s.c DK35, K36
3 Tachyzoitesc i.p. DK21, DK30, K36
Bradyzoitesb s.c. K36
9 Tachyzoitesc i.p. K36
Gerbils 2 Bradyzoitesb s.c. K79, K79
5 Tachyzoitesc s.c., i.p. K36
Donkey no. 5 KO mice 8 Bradyzoitesb s.c. DK41, K53, K53,
K53, K45, DK49,
K48, K54
4 Bradyzoitesb i.p. D18, D19, D52, K 53
3 Bradyzoitesb oral K66
2 Cloned bradyzoites 1e i.p. K24, D40
2 Cloned bradyzoites 2e i.p. K45
3 Bradyzoitesf s.c., i.p. K76, K91, K91
2 Tachyzoitesc s.c.,i.p. DK27
Gerbils 3 Bradyzoitesb s.c., i.p., oral K66
a
D, Died; DK, Killed when sick; K, Killed when healthy. Parasites were demonstrated by histology only in the tissues of the animals indicated in bold.
b
Homogenised tissue; number of organisms in inoculum unknown.
c
1!106 tachyzoites from culture.
d
Tachyzoites from mouse; number of organisms in inoculum unknown.
e
Released from a single tissue cyst.
f
Pepsin digested tissue; number of organisms in inoculum unknown.

and inoculated on to cell cultures and in mice and gerbils
(Table 1). Pepsin digested (Dubey, 1998) material was
inoculated s.c. and i.p. into KO mice (Table 1). Addition-
ally, two individual tissue cysts were teased out from the
fascia and the surrounding tissues were removed as much as
possible. The cysts were further washed in antibiotic saline
and ruptured using a 27-gauge needle. The released
bradyzoites were inoculated i.p. separately into two groups
of KO mice.
An eight week old New Zealand White rabbit
(no. 65173) was inoculated s.c. and i.m. with 1!106 cell
culture-derived B. bennetti tachyzoites (isolate from donkey
no. 1), and 40 days later with another dose of 1!106
tachyzoites. The rabbit was bled and euthanised 68 days
after the initial inoculation.
2.3. In vitro cultivation
Bradyzoites released from tissue cysts from donkey nos.
1 and 5 were filtered through a 5-mm filter and seeded onto
culture flasks containing monolayers of African Green
Monkey (CV-1) cells, and M617 cells (Dubey et al., 2002).
Monolayers of M617 cells were also infected with the
peritoneal exudate of a KO mouse that was previously
inoculated with the bradyzoites released from a single tissue
cyst. To study parasite morphology, tachyzoites from cell
cultures were seeded on to M617 cells grown on coverslips.
Coverslips were removed at periodic intervals, fixed in
Bouins, stained with Giemsa, and examined
microscopically. Organisms were measured using the
motorised system microscope (Olympus Optical Co.,
Tokyo, Japan) with an Olympus FV camera and Olympus
Microsuitee B3SV software.
2.4. Necropsy and histopathological examination
Samples of brain, lung, heart, tongue, liver, kidney,
intestine, mesenteric lymph node, urinary bladder and limb
muscle of gerbils, rabbit, mice, and cats were fixed in 10%
neutral buffered formalin. Paraffin-embedded sections were
cut at 5 mm, and examined after staining with H&E or
periodic acid Schiff reaction counter stained with haema-
toxylin (PASH).
2.5. Transmission electron microscopy
For TEM, tissues were fixed in glutaraldehyde or in 10%
buffered neutral formalin. They were subsequently post-
fixed in osmium and processed for TEM. For the study of
tachyzoites, M617 cells infected in vitro and the skin of a
KO mouse inoculated with the isolate from donkey no. 5, 45
days p.i were used.
2.6. Immunohistochemical staining
Two types of primary antibodies were used for immuno-
histochemical staining. A polyclonal anti-B. bennetti antibody
from rabbit No. 65173 was used at 1:5001:8000 serum
 J.P. Dubey et al. / International Journal for Parasitology 35 (2005) 659672
dilutions and a bradyzoite-specific rabbit antibody (BAG-1,
also called BAG-5) directed against a heat-shock protein from
Toxoplasma gondii (supplied by McAllister et al., 1996) was
used at 1:10,000 serum dilution. Staining was performed as
described previously (Dubey and Sreekumar, 2003).
2.7. Serologic examination
Using tachyzoites grown in M617 cells and rabbit anti-
horse IgG whole molecule FITC conjugate (Sigma,
St Louis, MO, USA) at 1:50 dilution indirect fluorescent
antibody testing (IFAT) was used to detect antibodies to B.
bennetti in sera of four donkeys.
2.8. Molecular systematics
DNA from B. bennetti tachyzoites of isolates from
donkey nos. 1 and 5 was extracted by proteinase K
digestion, and purification using Qiagen DNAeasy columns
according to the manufacturers protocol. Aliquots along
with extraction negatives, were subsequently used as
templates in PCR assays to characterise a 1961 bp portion
of ribosomal DNA encompassing most of the ssrDNA as
well as the first internal transcribed spacer 1 (ITS-1), and a
1912 bp portion of the lsrDNA using previously described
methods (Dubey et al., 2004). These products were directly
sequenced on an ABI 3100 fluorescent sequencer using
BigDye 3.1 terminator chemistries. Sequence contigs were
assembled and edited using ContigExpress (VectorNTI,
Informax Inc.) and compared with other homologues by
means of Neighbor Joining trees reconstructed from Kimura
2-parameter distances from 500 bootstrap replicates of
multiple sequence alignments created using CLUSTALW,
as implemented by MEGA version 2.0 (Kumar et al., 2001).
The phylogenetic position of B. bennetti was placed in a
broad comparative framework by inclusion of other
available sequences of lsrDNA from diverse species
belonging to the Sarcocystidae; by limiting a subsequent
analysis to members of the genus Besnoitia, we were able to
more precisely define its phylogenetic position with respect
to its closest relatives using a longer portion of ungapped
alignment.
3. Results
3.1. Lesions in donkeys
Gross skin lesions in donkey no. 5 consisted of multifocal
to coalescing areas of patchy alopecia and hypotrichosis,
with thickened, irregular crusty hyperpigmented skin of the
neck, around the muzzle, eyes, base of the ears, over the
carpus, elbow, stifle, tarsus, extending over the cannon
bones to the pastern and fetlocks of all four limbs and over
the perineum (Fig. 1 (BD). Within the affected dermis and
epidermis, hundreds of pinpoint white granular structures
663
were present, ranging from 0.5 to 1.5 mm in diameter. The
subcutaneous tissues and superficial fascia under the
affected skin, as well as the scleral conjunctiva and vaginal
mucosa, contained millions of white granules ranging from
0.5 to 1.5 mm in diameter. The prescapular, mandibular, and
popliteal lymph nodes were diffusely enlarged, ranging
from 2!2!3 to 8!3!2 cm3. Lymphoid tissue surround-
ing the larynx and pharynx was noted to be hyperplastic.
The donkey as a whole was markedly cachectic, with
minimal fat stores, and serous atrophy of fat at the base of
the heart. The distal phalanges of the left and right forelimb
and right hind limb were ventrally rotated 35, 20, and 408,
respectively, from the hoof wall. At the ileocecocolic
junction, approximately 20 tapeworms (Anoplocephala
perfoliata) were present.
Histologically, in all skin sites examined, numerous
Besnoitia tissue cysts were present within the dermis,
panniculus, and underlying muscle. The skin crust consisted
of desquamated host tissues as well as numerous free
bradyzoites (Fig. 2). A moderate multifocal perivascular to
interstitial chronic fibrosing lymphoplasmacytic, histiocy-
tic, and eosinophilic dermatitis and myositis accompanied
the tissue cysts (Fig. 2). Multifocal sites of tissue cyst
rupture incited a severe acute neutrophilic, eosinophilic and
lymphoplasmacytic dermatitis. Within the superficial
dermis, moderate, multifocal perivascular lymphoplasma-
cytic infiltrates were present.
The vulvar mucosa contained large numbers of Besnoitia
tissue cysts with characteristics similar to those described
above throughout the lamina propria and muscularis. A mild
to moderate multifocal perivascular to interstitial chronic
fibrosing lymphoplasmacytic, histiocytic, and eosinophilic
vulvitis and myositis accompanied the cysts. Multifocal
sites of tissue cyst rupture created a severe acute
neutrophilic, lymphoplasmacytic, histiocytic and eosino-
philic vulvitis.
Besnoitia sp. tissue cysts were present within the sclera
and iris, the right thyroid gland and perithyroid and
perilaryngeal connective tissue and were associated with
inflammation in these tissues. Severe, diffuse, chronic
reactive lymphoid hyperplasia and moderate diffuse sinus
histiocytosis were present in all peripheral lymph nodes
examined. A moderate, multifocal to coalescing subacute to
chronic lymphoplasmacytic interstitial pneumonia with
mild multifocal neutrophilic and histiocytic alveolar septal
infiltrates, and moderate, multifocal subacute pulmonary
and pleural oedema were present. Within the colon, a
severe, focal chronic submucosal lymphoid hyperplasia was
present, with an adjacent histiocytic and lymphoplasmacytic
colitis.
The kidney contained evidence of mild multifocal acute
tubular necrosis, and multifocal moderate chronic lympho-
plasmacytic and histiocytic interstitial nephritis. Within the
brain, more commonly in the rostral than caudal aspects,
multifocal areas of lymphohistiocytic perivascular cuffing
664 J.P. Dubey et al. / International Journal for Parasitology 35 (2005) 659672

Fig. 2. Besnoitia bennetti and associated lesions in skin of donkey no. 5. (A) A crust containing desquamated host tissue and bradyzoites, attached to epidermis
(arrow). An intact tissue cyst is in the dermis (arrowheads). H&E. (B) Bradyzoites (arrows) among degenerating host tissue. H&E. (C) Section of a tissue cyst.
There are three layers, outer (1) thick layer, a middle layer (2) and a third (3) thin layer. 1 mm section stained with toluidine blue. Numerous bradyzoites
(arrowheads) are enclosed in layer 3. (D) Bradyzoites (arrows) released from a tissue cyst in a smear stained with Giemsa.

were present. The liver had a moderate, diffuse
haemosiderosis.
3.2. Structure of tissue cysts and bradyzoites
Tissue cysts were located in the skin, subcutaneous
tissue, conjunctiva, sclera, occasionally in soft tissues
(Fig. 1). They were white to glistening white to the naked
eye and embedded in host tissue. Tissue cysts were up to
400 mm long and the cyst walls were up to 40 mm thick
(Fig. 2). The thickness of the tissue cyst wall varied. Under
the light microscope three layers were recognised. The outer
layer (layer 1) consisted of hyaline connective tissue; it
stained weakly positive with PASH. The middle layer (layer
2) enclosed host cell elements including host cell nuclei; it
stained intensely with PASH. The innermost layer (layer 3,
the parasitophorous vacuolar membrane (PVM) was less
than 1 mm thick; it was PAS-negative. Numerous PAS-
positive bradyzoites were enclosed in layer 3. The
bradyzoites released from four tissue cysts were
8.7!1.9 mm (7.011.9!1.22.8, nZ100). Longitudinally
cut bradyzoites in 1-mm Toluidine blue sections from four
tissue cysts were 7.8!1.7 mm (5.810.3!1.12.3, nZ
100). Tissue cysts were not septate. Occasionally, two tissue
cysts were enclosed in the same host cell (Fig. 3), giving the
false impression of septa.
The three layers seen under light microscopy were also
recognised by TEM. The outer layer consisted of connective
tissue. The middle layer contained host cell nuclei and an
accumulation of endoplasmic reticula; filamentous exten-
sions were seen that projected towards the outer most layer.
The innermost layer consisted of a thin PVM, which was
lined by an amorphous granular layer, which varied
in thickness. Numerous bradyzoites were enclosed in
 J.P. Dubey et al. / International Journal for Parasitology 35 (2005) 659672 665

Fig. 3. TEM of two tissue cysts enclosed in the same host cell (HC), giving the false impression of a septum. Note a thin layer of host cell material (HC)
between two tissue cysts. Also note the host cell nucleus (HCN), and parasitophorous vacuolar membrane (PVM) aand electron dense layer (EDL) of the
separate tissue cysts.

the granular layer. Bradyzoites varied in size and shape
(Figs. 3 and 4), some were plump whereas others were
slender (Fig. 4). Ten longitudinally cut bradyzoites were
4.98.5.0!1.191.70 mm in size on TEM sections. Brady-
zoites contained a conoid, micronemes, rhoptries, a nucleus,
amylopectin, a mitochondrium and dense granules. Micro-
nemes were numerous (up to 100 per section) and were
located thoughout the bradyzoite but were more numerous at
the conoidal end. They were up to 570 nm long and
approximately 50 nm wide. The rhoptries extended up to
the posterior end (Fig. 4). The contents of rhoptries were
electron-dense. The bulbous end of rhoptries was up to
400 nm wide. Occasionally, the bulbous end was turned
towards the conoidal end of the bradyzoite (Fig. 4B). A
maximum of five rhoptries were seen in a given section. The
position of the nucleus in bradyzoites was central to terminal
(Fig. 4). The amylopectin granules were located throughout
the bradyzoite. Enigmatic bodies were not found in any of
the bradyzoites examined from four tissue cysts (Fig. 4).
3.3. Isolation of B. bennetti in cell culture
Tachyzoites were seen in the M617 monolayer 43
days after inoculation with bradyzoites from donkey no. 1.
666 J.P. Dubey et al. / International Journal for Parasitology 35 (2005) 659672

Fig. 4. TEM of Besnoitia bennetti bradyzoites of different shapes and sizes (A, B, C). Bar in B applies to all parts. Note conoid (C), dense granules (Dg),
microtubules (Mt), micronemes (Mi), amylopectin (A, empty spaces), rhoptries (R), and different shapes, location, and sizes of nucleus (N). Also note that the
bulbous end of one rhoptry (arrowhead in B) is pointed towards the conoidal end.

The monolayer was completely destroyed in 56 days and the
tachyzoites were continuously passaged. The growth rate of
the tachyzoites in vitro was slow and the average passage
interval was more than 60 days despite using a large
inoculum of more than 105 tachyzoites. Though occasional
dividing forms were initially noticed, the parasite never
established itself in the CV1 monolayers and the cultures
were discarded after 63 days.
The tachyzoites from the peritoneal exudate of a KO
mouse that was infected with bradyzoites from a single
tissue cyst (cloned isolate) was established in M617
monolayers and was passaged continuously. The tachy-
zoites from the first passage were cryopreserved for future
use.
In coverslip cultures, stray intracellular parasites could
be seen in the coverslips stained 4 h after inoculation. By
48 h, numerous cells with two, four and eight parasites
could be seen (Fig. 5F). Spontaneous ingress of parasites
from intact cells was noticed. Numerous plaques, sur-
rounded with parasitised cells and numerous extracellular
 J.P. Dubey et al. / International Journal for Parasitology 35 (2005) 659672 667

Fig. 5. Besnoitia bennetti stages. (AE) from KO mice and (F) from cell culture. (A and B) section of cutaneous granuloma with faintly stained tachyzoites
(arrowheads). 45 days p.i. H&E. (C) Section of eye with a tissue cyst (arrow) in periorbital granuloma. Note a thick walled tissue cyst enclosing two host cell
nuclei (small arrows) and bradyzoites (arrowheads). This was the only tissue cyst found in all KO mice inoculated with B. bennetti. 45 days p.i. (D) Three
tachyzoites with dividing nuclei (arrows) inside a neutrophil from smear of peritoneal exudate of KO mouse, 45 days p.i. Giemsa stain. (E) Lunate tachyzoites
(arrows) in and out of a macrophage from the same mouse as in Fig. 6D. Giemsa stain. (F) Two tachyzoites within one parasitophorous vacuole (arrow) in
M617 cell 48 h p.i. Giemsa stain.

tachyzoites, were observed after 27 days. The monolayer
was destroyed in 43 days.
Extracellular tachyzoites from cell culture were 5.9!
1.6 mm (5.07.3!1.12.5, nZ50).
3.4. Ultrastructure of tachyzoites
Tachyzoites were located in a parasitophorous vacuole
containing tubular network, in the host cell cytoplasm of
neutrophils, mononuclear cells, and fibroblasts in subcu-
taneous tissue of the mouse (Fig. 6). Tachyzoites divided by
endodyogeny. They contained a conoid, several micro-
nemes, rhoptries, an elongated mitochondrion, a nucleus,
but no amylopectin granules (Fig. 6). The micronemes were
located anterior to the nucleus and often arranged in rows
(Fig. 6). The number of micronemes was highly variable
from a few (Fig. 6C) to numerous (Fig. 6B). The rhoptries
were few in number and their contents were electron-dense.
3.5. Serologic examination
The IFAT titers of the donkey nos. 14 were 1:5000,
1:1600, 1:800, 1:200, respectively, suggesting exposure to
B. bennetti; donkey no. 1 had clinical besnoitiosis.
The IFAT titers of the seven horses were 1:800 in one, 1:
400 in one, and 1:200 or less in five horses.
668 J.P. Dubey et al. / International Journal for Parasitology 35 (2005) 659672

Fig. 6. TEM of Besnoitia bennetti tachyzoites. (A, B) Individual tachyzoites (arrows) in dermal tissue of a KO mouse 45 days p.i. and (C) from cell culture.
Note conoid (C), micronemes (Mi), rhoptries (R), an elongated mitochondrium (Mc), and a nucleus (N).
 J.P. Dubey et al. / International Journal for Parasitology 35 (2005) 659672 669

3.6. Immunohistochemistry
Besnoitia bennetti LSU rDNA
The innermost layer (third layer) and the enclosed 96
Besnoitia besnoiti bull strain AF076866
bradyzoites in tissue cysts stained positively with anti- Besnoitia caprae AF076867
B. bennetti antibodies up to 1:8000 serum dilution. 100

However, tachyzoites in tissue sections of KO mice were Besnoitia tarandi AY616164

stained only at 1:500 dilution. Bradyzoites, but not the 99 Besnoitia besnoiti AF076900
97
tachyzoites, were stained with BAG-1-antibodies. Besnoitia besnoiti BWB AF076869
100 Besnoitia jellisoni AF076868
3.7. Infection of laboratory animals 89
Hammondia hammondi AF101077
No oocysts were seen in faeces of cats nor were any 89 Toxoplasma gondii AF076901
100
protozoa identified in their histological sections. Hammondia heydorni AF159240
Protozoa were not identified in tissues of any of the 54
Neospora caninum AF001946
gerbils or out-bred mice (Table 1), whether inoculated with
donkey tissues or with tachyzoites derived from cell culture Isospora felis IFU85705
or from KO mice. The KO mice inoculated with bradyzoites 100 Sarcocystis singaporensis AF237617
or tachyzoites became infected but the organisms were not
Sarcocystis zamani AF237616
demonstrable in all mice (Table 1). Cutaneous granulomas
developed at the injection site in mice inoculated subcu- 99 Frenkelia glareoli AF044251
taneously, and tachyzoites were present in these lesions 77
99
Frenkelia microti AF044252
(Fig. 5B and C). The i.p. inoculated mice developed ascites 59
Sarcocystis neurona AF092927
that contained white fluid and tachyzoites were present in
the fluid. Besnoitia were not seen in liver, lungs, brain, and Sarcocystis muris AF012883
98
the heart. Only tachyzoites were identified in all infected Sarcocystis gigantea SGU85706
mice except in periorbital tissue of a KO mouse that had 99
Sarcocystis arieticanis AF076904
been inoculated subcutaneously with bradyzoites from 100
donkey no.1; a small tissue cyst was present in this mouse Sarcocystis tenella AF076899
65
(Fig. 5C). Sarcocystis cruzi AF076903
Protozoa were not identified in tissues of the rabbit
Eimeria tenella AF026388
inoculated with tachyzoites from cell culture. 0.05

Fig. 7. Placement of species of Besnoitia within a large phylogenetic

3.8. Molecular systematics
The sequences of ssrDNA of B. bennetti were compared
with those of other Besnoitia sp., as well as other
apicomplexans. The 1961 bp of small subunit rDNA
(GenBank AY827838) and ITS-1 (GenBank AY827839)
were identical to those derived from an isolate of B. bennetti
from a miniature donkey in Michigan (GenBank accession
AY665399, Elsheikha et al., 2005). The 1912 bp of large
subunit rDNA of each of two isolates described here were
identical to each other and were inferred to be closely
related to other tissue cyst forming coccidia (Fig. 7). This
sequence has been deposited as GenBank accession
AY778965.
3.9. Effect of chemotherapy
Donkey no.1 had remarkable clinical improvement after
a seven month treatment with SMZ and TM (Fig. 1E). All
tissue cysts seen in the biopsy of the only grossly visible
lesions on the muzzle were degenerated and mineralised.
context reconstructed from lsurDNA. Midpoint-rooted neighbor joining
tree reconstructed from 500 bootstrap resamplings of 503 ungapped
positions of nuclear large subunit ribosomal DNA gene using Kimura 2-
parameter distances.
3.10. Taxonomic summary
Intermediate type hosts: Equids (Equus caballus, Equus
asinus)
Definitive host: Unknown.
Locality: USA, Africa.
Specimens deposited Specimens were deposited in the
United States National Parasite Collection (USNPC),
United States Department of Agriculture, Beltsville, MD,
USA: (a) histologic sections of tissue cysts from subcu-
taneous tissue of donkey no.5 stained with H&E (USNPC
No. 95254), (b) stained with anti-B. bennetti antibodies
(USNPC No. 95255), (c) 1-mm section stained with
Toluidine blue (USNPC No. 95256) and (d) unstained
tissue in 10% buffered neutral formalin (USNPC
No. 95257), (e) histologic sections of tissues of a KO
mice experimentally-infected with cell-culture derived
670 J.P. Dubey et al. / International Journal for Parasitology 35 (2005) 659672
B. bennetti tachyzoites, 45 days p.i., stained with H&E
(USNPC No. 95258), and (f) B. bennetti tachyzoites in cell
culture preparations of M617 cells stained with Giemsa
(USNPC No.95259).
Live cultures of tachyzoites of B. Bennetti (from donkey
no.5) were deposited in the American Type Culture
Collection (ATCC), Manassas, VA 20108, USA, ATCC
Accession No. PRA-145.
4. Discussion
In general, the structure of tissue cysts of B. bennetti
found in the present study is similar to that of B. oryctofelisi.
Dubey and Lindsay (2003) described the development and
structure of B. oryctofelisi tissue cysts in experimentally-
infected animals. They provided evidence that the outer two
layers of the tissue cyst walls were of host origin, as they did
not stain with the B. oryctofelisi antibodies. However, the
inner (third) layer and the bradyzoites were stained with the
antibodies indicating parasitic origin. The immunohisto-
chemical staining results with anti-B. bennetti polyclonal
serum in the present study, confirms the findings of Dubey
and Lindsay (2003). Why B. bennetti tachyzoites in sections
of mice had a lower reactivity (positive at 1:500) to anti-
B. bennetti antibodies than bradyzoites in tissue cysts is
unexplained because the polyclonal serum was against
tachyzoites.
Until recently, enigmatic bodies, first described in
Besnoitia jellisoni by Senaud (1969), were considered
characteristic for Besnoitia spp. bradyzoites. They contain a
central core enclosed in a membrane; their function is
unknown. However, extensive studies failed to identify
enigmatic bodies in bradyzoites of B. besnoiti of cattle
(Dubey et al., 2003b), nor were they previously observed in
B. bennetti from a horse in South Africa (van Heerden et al.,
1993). No enigmatic bodies were identified in bradyzoites
of B. bennetti from donkeys we studied here, nor were they
reported in either of the two ultrastructural studies of
Canadian reindeer Besnoitia species (Glover et al., 1990;
Ayroud et al., 1995).
In B. bennetti, the rhoptries were as long as the
bradyzoites and in some, the bulbous end was directed
towards the conoidal end instead of the non-conoidal end.
This is the second report of this feature of rhoptries of
Besnoitia. We previously found this feature in rhoptries of
B. tarandi (Dubey et al., 2004). Ultrastructurally, B. bennetti
bradyzoites appeared to possess w100 micronemes. This is
in contrast to the higher number of micronemes (w300)
found in the bradyzoites of B. darlingi (Dubey et al., 2002),
B. oryctofelisi (Dubey et al., 2003a), and Besnoitia akodoni
(Dubey et al., 2003c).
Cutaneous besnoitiosis attributed to B. besnoiti causes
substantial economic loss among cattle in Africa, Israel, and
some parts of Europe (Dubey et al., 2003a). Although
besnoitiosis in goats was presumed to be caused by a distinct
taxon, Besnoitia caprae, an absence of diagnostic variation
in the ITS-1 between caprine isolates and those from cattle
and wildebeest has questioned this distinction (Ellis et al.,
2000). In order to verify the placement of the B. bennetti
isolate within a monophyletic group comprising other
named Besnoitia congeners, and in order to evaluate the
degree to which it could be differentiated from such
congeners, three portions of nuclear ribosomal DNA were
characterised and compared with other available homol-
ogues. Including this report, a total of three isolates of
B. bennetti have been characterised at a portion of small
subunit and ITS-1 rDNA comprising some 1960 base pairs.
The absolute uniformity of these isolates, originating in
Vermont, Florida, and Michigan, preliminarily suggests that
donkey parasites inducing clinical besnoitiosis are con-
specific. As this sequence differs, if only slightly, from
homologues of congeneric parasites, we conclude that this
sequence may be useful in future diagnoses. Previous
phylogenetic analyses (Dubey et al., 2004) have also
supported the monophyly of the genus Besnoitia and
indicated a particularly close relationship between parasites
of cattle, goats, caribou, and equids (B. besnoiti, B. caprae,
B. tarandi, and B. bennetti) with the exclusion of other
congeners that parasitise rodents, lagomorphs, and opos-
sums (B. jellisoni, B. akodoni, B. oryctofelisi, and
B. darlingi).
Our new analysis of lsrRNA confirms previous con-
clusions based upon other genetic loci (Dubey et al., 2004;
Elsheikha et al., 2005) that the genus Besnoitia is
monophyletic with the exclusion of other tissue cyst
forming coccidia, and that they form a sister group to the
one containing species assigned to the genera Toxoplasma,
Neospora, and Hammondia. Additionally, our analysis
indicates that the genus Besnoitia includes distinct lineages,
one of which has given rise to the parasites of ungulates,
B. besnoiti, B. tarandi, B. caprae, and B. bennetti. Whether
the parasite in horses is the same as the parasite in donkeys
requires further study; notably, B. bennetti infection has not
been reported in horses in the USA. If future studies
determine a distinct phylogenetic difference between the
parasite of horses and donkeys, a new name for the donkey
parasite would be required.
In the present study, donkey no. 1 improved clinically
and all tissue cysts were dead after a seven month treatment
with SMZ and TM. Somewhat similar results were reported
by Davis et al. (1997) in a donkey that was medicated with
SMZ for 30 days, discontinued for 30 days, and then treated
for an additional 30 days; this treatment resolved pruritic
dermatitis but results were not confirmed by biopsy.
Although these studies are uncontrolled, they do suggest
that treatment with SMZ can reduce suffering from
besnoitiosis. To our knowledge this is the first documented
killing of Besnoitia tissue cysts after treatment with any
anti-protozoal therapy.
Developing adequate laboratory models would promote
the study of the biology and epidemiology of ungulate
 J.P. Dubey et al. / International Journal for Parasitology 35 (2005) 659672
besnoitiosis. Tissue cysts of B. bennetti have not yet been
produced in laboratory animals or in cell culture. A single
tissue cyst was found in an infected KO mouse in this study.
Gerbils that are susceptible to experimental infections with
other species of Besnoitia (B. darlingi, B. besnoiti,
B. oryctofelisi, and B. jellisoni) do not become infected by
B. bennetti and B. tarandi. Limited trials have also revealed
that, whereas rabbits are refractive to infection with
B. tarandi, B. bennetti and B. darlingi, they could be
infected with B. besnoiti and B. oryctofelisi. Out-bred Swiss
Webster mice were also refractive to infection with
B. tarandi and B. bennetti. Unlike other Besnoitia species,
even KO mice were not highly susceptible to B. bennetti.
Tachyzoites of B. bennetti were found to multiply much
slower in vitro than those of B. darlingi, B. oryctofelisi,
B. akodoni and B. tarandi. The passage interval was more
than 60 days, which is considerably higher. Furthermore, we
were unable to establish B. bennetti in CV1 cells, which
support the growth of B. darlingi, B. oryctofelisi, B. akodoni
and B. tarandi.
Domestic cats serve as definitive hosts for the three
species of Besnoitia whose life cycles are known
(B. darlingi, B. oryctofelisi and B. wallacei). Results of
the present study indicate that domestic cats are not
likely to play this role for B. bennetti. Epidemiologic
evidence argues against congenital transmission of
Besnoitia species. The mode of infection in equids
remains a mystery; transmission by large members of
Felidae (cougars, lynx) remains a possibility. Glover
et al. (1990) observed an inverse correlation between
transmission of Besnoitia and exposure to biting flies,
among reindeers. The presence of numerous extracellular
bradyzoites in the dermal crust in the present study lends
credence to the hypothesis that arthropods might play a
role in the transmission of besnoitiosis. In the present
study, only three of 150 donkeys on a farm had clinical
besnoitiosis but the exposure rate was unknown. Finding
of antibodies to B. bennetti in three of three donkeys and
at least two of the seven horses without any clinical
signs suggests that Besnoitia infection may be more
common in equids in the USA than realised. However,
serological results are preliminary because we did not
have access to sera from donkeys not exposed to B.
bennetti. Besnoitiosis in donkeys causes economic losses.
Hence it is essential to have a better understanding of the
biology and life cycle of this parasite, including the
possible definitive host(s), to formulate adequate control
measures.
Acknowledgements
We would like to thank John Jenkins, Armed Forces
Institute of Pathology, Washington, D.C., for the electron
micrographs, O.C.H. Kwok, USDA, for his assistance
with the cell culture, Detiger Dunams for her assistance
671
with the genetic analysis, Drs Clive Huxtable, Ana
Alcaraz and Emily Meseck for diagnostic work and
photography, Dr Keith Beebe and Mr Martin Ross for
photographs and facilitation of the necropsy procedure of
donkey no. 5, Dr M. S. Rose for clinical information on
donkeys 6 and 7, and Ms Marcy Freedman for help with
investigations on donkey nos. 14.
References
Ayroud, M., Leighton, F.A., Tessaro, S.V., 1995. The morphology and
pathology of Besnoitia sp. in reindeer (Rangifer tarandus tarandus).
J. Wildlife Dis. 31, 319326.
Babudieri, B., 1932. Sarcosporidi e le sarcosporidiosi (Studio monogra-
fico.). Arch. Protistenkd. 76, 421580.
Bennett, S.C.J., 1927. A peculiar equine sarcosporidium in the Anglo-
Egyptian Sudan. Vet. J. 83, 297304.
Bennett, S.C.J., 1933. Globidium infections in the Sudan. J. Comp. Pathol.
Thera. 46, 115.
Bigalke, R.D., 1970. Studies on equine besnoitiosis. J. Parasitol. 56, 29.
Davis, W.P., Peters, D.F., Dunstan, R.W., 1997. Besnoitiosis in a miniature
donkey. Vet. Dermatol. 8, 139143.
Dubey, J.P., 1995. Duration of immunity to shedding of Toxoplasma gondii
oocysts by cats. J. Parasitol. 81, 410415.
Dubey, J.P., 1998. Refinement of pepsin digestion method for isolation
of Toxoplasma gondii from infected tissues. Vet. Parasitol. 74,
7577.
Dubey, J.P., Lindsay, D.S., 1998. Isolation in immunodeficient mice of
Sarcocystis neurona from opossum (Didelphis virginiana) faeces,
and its differentiation from Sarcocystis falcatula. Int. J. Parasitol.
28, 18231828.
Dubey, J.P., Lindsay, D.S., 2003. Development and ultrastructure of
Besnoitia oryctofelisi tachyzoites, tissue cysts, bradyzoites, schizonts
and merozoites. Int. J. Parasitol. 33, 807819.
Dubey, J.P., Sreekumar, C., 2003. Redescription of Hammondia hammondi
and its differentiation from Toxoplasma gondii. Int. J. Parasitol. 33,
14371453.
Dubey, J.P., Lindsay, D.S., Rosenthal, B.M., Sreekumar, C., Hill, D.E.,
Shen, S.K., Kwok, O.C.H., Rickard, L.G., Black, S.S., Rashmir-
Raven, A., 2002. Establishment of Besnoitia darlingi from opossums
(Didelphis virginiana) in experimental intermediate and definitive
hosts, propagation in cell culture, and description of ultrastructural and
genetic characteristics. Int. J. Parasitol. 32, 10531064.
Dubey, J.P., Sreekumar, C., Lindsay, D.S., Hill, D., Rosenthal, B.M.,
Venturini, L., Venturini, M.C., Greiner, E.C., 2003a. Besnoitia
oryctofelisi n. sp. (Protozoa: Apicomplexa) from domestic rabbits.
Parasitology 126, 521539.
Dubey, J.P., Shkap, V., Pipano, E., Fish, L., Fritz, D.L., 2003b.
Ultrastructure of Besnoitia besnoiti tissue cysts and bradyzoites.
J. Eukaryot. Microbiol. 50, 240244.
Dubey, J.P., Sreekumar, C., Rosenthal, B.M., Lindsay, D.S., Grisard, E.C.,
Vitor, R.W.A., 2003c. Biological and molecular characterization of
Besnoitia akodoni n. sp.(Protozoa: Apicomplexa) from the rodent
Akodon montensis in Brazil. Parassitologia 45, 6170.
Dubey, J.P., Sreekumar, C., Rosenthal, B.M., Vianna, M.C.B., Nylund, M.,
Nikander, S., Oksanen, A., 2004. Redescription of Besnoitia tarandi
(Protozoa: Apicomplexa) from the reindeer (Rangifer tarandus). Int.
J. Parasitol. 34, 12731287.
Ellis, J.T., Holmdahl, O.J.M., Ryce, C., Njenga, J.M., Harper, P.A.W.,
Morrison, D.A., 2000. Molecular phylogeny of Besnoitia and the
genetic relationships among Besnoitia of cattle, wildebeest and goats.
Protistologica 151, 329336.
672 J.P. Dubey et al. / International Journal for Parasitology 35 (2005) 659672
Elsheikha, H.M., Mackenzie C.D., Rosenthal, B.M., Marteniuk J.V.,
Steficek, B., Windsor, S., Saeed, A.M., Mansfield, L.S., An outbreak of
besnoitiosis in miniature donkeys. J. Parasitol. in press.
Glover, G.J., Swendrowski, M., Cawthorn, R.J., 1990. An epizootic of
besnoitiosis in captive caribou (Rangifer tarandus caribou), reindeer
(Rangifer tarandus tarandus), and mule deer (Odocoileus hemionus
hemionus). J. Wildlife Dis. 26, 186195.
Kumar, S., Tamura, K., Jakobsen, I.B., Nei, M., 2001. MEGA2:
molecular evolutionary genetics analysis software. Bioinformatics
17, 12441245.
Lane, J.G., Lucke, V.M., Wright, A.I., 1986. Parasitic laryngeal
papillomatosis in a horse. Vet. Rec. 119, 591593.
McAllister, M.M., Parmley, S.F., Weiss, L.M., Welch, V.J.,
McGuire, A.M., 1996. An immunohistochemical method for detecting
bradyzoite antigen (BAG5) in Toxoplasma gondii-infected tissues
cross-reacts with a Neospora caninum bradyzoite antigen. J. Parasitol.
82, 354355.
Pols, J.W., 1960. Studies on bovine besnoitiosis with special reference to
the aetiology. Onderstepoort J. Vet. Res. 28, 265356.
Schulz, K.C.A., Thorburn, J.A., 1955. Globidiosisa cause of dermatitis in
horses. J. S. Afr. Vet. Assoc. 26, 3943.
Senaud, J., 1969. Ultrastructure des formations kystiques de Besnoitia
jellisoni (Frenkel 1953) protozoaire, Toxoplasmea, parasite de la souris
(Mus musculus). Protistologica 5, 413430.
Terrell, T.G., Stookey, J.L., 1973. Besnoitia bennetti in two Mexican
burros. Vet. Pathol. 10, 177184.
van Heerden, J., Els, H.J., Raubenheimer, E.J., Williams, J.H., 1993.
Besnoitiosis in a horse. J. S. Afr. Vet. Assoc. 64, 9295.
 International Journal for Parasitology 35 (2005) 673684
www.parasitology-online.com

Clams (Corbicula fluminea) as bioindicators of fecal contamination

with Cryptosporidium and Giardia spp. in freshwater
ecosystems in California*
Woutrina A. Miller, Edward R. Atwill, Ian A. Gardner, Melissa A. Miller, Heather M. Fritz,
Ronald P. Hedrick, Ann C. Melli, Nicole M. Barnes, Patricia A. Conrad*
School of Veterinary Medicine, University of California, Davis CA 95616, USA

Received 2 November 2004; received in revised form 4 January 2005; accepted 10 January 2005

Abstract
This study evaluated clams as bioindicators of fecal protozoan contamination using three approaches: (i) clam tissue spiking experiments
to compare several detection techniques; (ii) clam tank exposure experiments to evaluate clams that had filtered Cryptosporidium oocysts
from inoculated water under a range of simulated environmental conditions; (iii) sentinel clam outplanting to assess the distribution and
magnitude of fecal contamination in three riverine systems in California. Our spiking and tank experiments showed that direct fluorescent
antibody (DFA), immunomagnetic separation (IMS) in combination with DFA, and PCR techniques could be used to detect Cryptosporidium
in clam tissues. The most analytically sensitive technique was IMS concentration with DFA detection of oocysts in clam digestive gland
tissues, which detected 10 oocysts spiked into a clam digestive gland 83% of the time. In the tank experiment, oocyst dose and clam
collection time were significant predictors for detecting Cryptosporidium parvum oocysts in clams. In the wild clam study, Cryptosporidium
and Giardia were detected in clams from all three study regions by IMS-DFA analysis of clam digestive glands, with significant variation by
sampling year and season. The presence of C. parvum DNA in clams from riverine ecosystems was confirmed with PCR and DNA sequence
analysis.
q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: Cryptosporidium; Giardia; Bivalve; Clam; Fecal pollution; Waterborne pathogens

1. Introduction cause clinical disease in human volunteers (Okhuysen

In 2003, over 1200 human cases of cryptosporidiosis
and 9100 cases of giardiasis were reported to the United
States Centers for Disease Control and Prevention (http://
wonder.cdc.gov/mmwr/mmwr_reps.asp?mmwr_tableZ
2B&mmwr_yearZ2004&mmwr_weekZ30), with many
more cases going unreported or undiagnosed. Cryptos-
poridium and Giardia are protozoan pathogens trans-
mitted by fecaloral ingestion, and as few as 10100
Cryptosporidium parvum oocysts have been shown to
*
Note: Nucleotide sequence data reported in this paper are available in
the EMBL, GenBanke and DDJB databases under the accession numbers
AY864316 and AY864317.
* Corresponding author. Tel.: C1 530 752 7210; fax: C1 530 752 3349.
E-mail address: paconrad@ucdavis.edu (P.A. Conrad).
0020-7519/$30.00 q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijpara.2005.01.002
et al., 1999). These oocysts can survive for over a year
in aquatic environments and have caused waterborne
human diarrheal outbreaks worldwide (Robertson et al.,
1992; Ong et al., 1999). Cryptosporidium and Giardia
spp. are shed by a variety of animals, including humans,
livestock, domestic animals, and wildlife (Xiao et al.,
1999; Heitman et al., 2002; McGlade et al., 2003; Zhou
et al., 2004). Some species of Cryptosporidium and
Giardia are thought to be host-specific, while other
species can infect many hosts (Monis and Thompson,
2003).
Bivalve shellfish have been used for decades as
bioindicators of aquatic contamination with heavy metals
and pesticides (OConnor, 2002). In recent years, shellfish
have also been recognised worldwide as bioindicators of
aquatic contamination with fecal origin bacteria, viruses,
674 W.A. Miller et al. / International Journal for Parasitology 35 (2005) 673684
and parasites (Fayer et al., 1998; Freire-Santos et al., 2000;
Pommepuy et al., 2004). Filter feeding shellfish such as
oysters, mussels, and clams have been shown to concentrate
Cryptosporidium and Giardia spp. in tank exposure
experiments, with oocysts detected in shellfish even when
none were detected in the surrounding waters (Atwill and
Rose, 1998; Fayer et al., 1998; Graczyk et al., 1999;
Tamburrini and Pozio, 1999). Multiple species of Cryptos-
poridium have been detected in wild shellfish, and the
oocysts were proven viable using mouse bioassays (Fayer
et al., 1998, 1999; Gomez-Bautista et al., 2000).
Methods for detecting Cryptosporidium and Giardia spp.
in bivalve shellfish include DNA amplification by PCR and
direct immunofluorescent antibody (DFA) assays (Graczyk
et al., 1998; Gomez-Couso et al., 2003). An advantage of
conventional PCR is that the DNA sequences can be
amplified and analysed to identify the parasite genotype, but
a limitation of PCR is that results are not quantitative. In
contrast, detection methods utilising DFA analysis provide
quantitative data but these methods cannot distinguish
between all parasite genotypes. Immunomagnetic separ-
ation (IMS) concentrates the parasites and thereby allows
for a larger sample volume to be analysed per test.
Immunomagnetic separation can be used in combination
with DFA or PCR detection methods (Sturbaum et al., 2002;
Ware et al., 2003). The use of IMS has increased the
minimum oocyst detection limit by 12 logs when testing
fecal samples (Pereira et al., 1999), and may be an important
tool to improve the analytic sensitivity of bivalve testing
methods.
Cryptosporidium and Giardia are endemic in animal and
human populations in California, but we have insufficient
understanding regarding the pathways and mechanisms of
how these pathogens enter waterways from the terrestrial
component of a watershed. Testing freshwater clams
(Corbicula fluminea) may be a useful approach to detect
fecal pathogens in freshwater ecosystems because these
clams filter large volumes of water (up to 2.5 L/water per h
per clam), they survive well in polluted aquatic environ-
ments, and they are easily collected and transported to the
laboratory for analysis (McMahon and Bogan, 2001). In
addition to the practicality of testing clams as indicators of
water quality, they are also of interest because both humans
and animals harvest them as a food source (unpublished
observation), suggesting that bivalves might expose con-
sumers to pathogens when eaten raw. In laboratory studies,
Cryptosporidium and Giardia have been detected in
Corbicula spp. after tank exposure to parasite-inoculated
waters (Atwill and Rose, 1998; Graczyk et al., 1998, 2003;
Izumi et al., 2004), but these clams have not been evaluated
as bioindicators of fecal contamination in natural
ecosystems.
In this study we evaluated methods for Cryptospor-
idium detection in C. fluminea, and then applied this
methodology to evaluate experimentally exposed and
sentinel C. fluminea outplanted in riverine ecosystems in
California. The study objectives were threefold: (i) to
evaluate DFA, IMS-DFA, and PCR methods for Cryptos-
poridium detection using spiked clam tissues; (ii) to
evaluate these methods on clams that had filtered oocysts
from inoculated waters under a range of simulated
environmental conditions; (iii) to evaluate outplanted
clams as bioindicators of fecal pathogen pollution in
three riverine ecosystems in California.
2. Methods
2.1. Experimental design
Tissue spiking experiments were first conducted to
determine which diagnostic methods would be sensitive
enough to detect the low numbers of Cryptosporidium
oocysts in clam (C. fluminea) tissues that might be expected
in environmental samples. The DFA, IMS-DFA, and PCR
methods were evaluated using spiked digestive
gland tissues. Two PCR methods (Morgan et al., 1997;
Xiao et al., 1999) were also evaluated on spiked hemolymph
tissues. Spiked hemolymph samples were not processed by
DFA methods because hemocytes were found to autofluor-
esce in our preliminary studies, making oocyst visualisation
difficult.
For the tissue spiking experiments, clam digestive
gland and hemolymph were inoculated with C. parvum
dilutions containing 0, 1, 10, 100, or 1000 oocysts, with
six replicates per tissue and oocyst dose. Spiked digestive
gland samples were homogenised, sieved, and centrifuged
to create a pellet that was re-suspended and split for DFA
and PCR analysis. A 10 ml digestive gland aliquot from
each of the six replicates was dried onto a slidewell for
DFA analysis, with the rest of the tissue then concentrated
with IMS followed by DFA analysis. Six hemolymph
replicates were each concentrated by centrifugation and
the cell pellet frozen for PCR analysis. Corbicula gills
were not assessed in this study due to their very small
size.
The same Cryptosporidium detection methods used in
the tissue spiking experiments were then evaluated on
clam samples obtained from a tank exposure experiment.
The experiment was designed to evaluate a range of
environmental conditions, including two water tempera-
tures (10 and 20 8C) and three oocyst concentrations (20,
200, and 2000 oocysts/L water), using two different
batches of C. parvum oocysts. During the tank exposure
experiment, clams were kept in tubs containing 10 L of
water, 80 clams, and an airstone that provided continuous
aeration and water mixing. Two clam tubs were spiked for
each treatment combination (oocyst dose, water tempera-
ture, and oocyst batch). In addition, negative control tubs
contained water and clams but no oocysts, while positive
control tubs contained water and oocysts but no clams.
At the time of oocyst inoculation, the clams were fed,
 W.A. Miller et al. / International Journal for Parasitology 35 (2005) 673684
and after a 6 h exposure the clams were moved to clean
freshwater tubs for Cryptosporidium depuration over the
next 3 weeks. Water samples (500 ml) from the inoculated
and negative control clam tanks were processed by
centrifugation for 10 min at 1000!g, followed by IMS
concentration and DFA analysis to quantitate the remain-
ing oocysts. Clams were collected at 3, 6, 9 h, 1, 3, 7, 14,
and 21 days after initial oocyst exposure. The effect of
water temperature, oocyst dose, and clam collection
timepoint were analysed with a negative binomial
regression model (Hardin and Hilbe, 2001).
The next step in evaluating clams as bioindicators of
pathogen pollution was to test clams that may have
filtered fecal pathogens from contaminated freshwater
ecosystems in California. Three freshwater study regions
were chosen based on water quality data that indicated a
history of fecal contamination problems and inclusion in
the state list of impaired water quality areas (http://www.
waterboards.ca.gov/tmdl/303d_lists.html). The San Lor-
enzo River feeds into the Monterey Bay near Santa Cruz,
CA, utilises mainly septic systems for sewage manage-
ment, and has minimal crop and animal production
industries. The Salinas River feeds into the Monterey
Bay near Moss Landing, CA, uses mainly sewer systems
for sewage management, and has significant crop and
animal production industries. Putah Creek is a freshwater
tributary that feeds into the California Delta, has some
sewer and some septic tank areas, and has significant crop
and animal production industries.
Along each riverine region, sentinel clams were out-
planted in mesh grids at an upstream and downstream site at
least 5 km apart, left undisturbed to filter the water for at
least 30 days, and collected as batches of 30 clams per site
and timepoint. Clam outplanting and collections occurred in
the wet and dry seasons of 2002 and 2003. Clams were
dissected and analysed for Cryptosporidium using IMS-
DFA on six pools of five clam digestive glands from each
batch of 30 clams. All Cryptosporidium-positive DFA slides
were scraped, washed, amplified with PCR, and the
sequences analysed to obtain parasite genotypes. Hemo-
lymph was analysed for all clams individually. The pooling
of clam digestive glands was possible because all the tissue
could be analysed using the IMS concentration procedure
before DFA detection, whereas the maximal PCR test
volume was approximately one hemolymph sample. For
each study site, information was gathered on environmental
variables that could be used in a statistical risk factor
analysis.
Giardia testing was added to the study in the second year.
Based on our Cryptosporidium spiking and tank experiment
findings that IMS-DFA of digestive gland was the most
sensitive detection method, we focused on IMS-DFA
methods for a Giardia tissue spiking experiment. In the
spiking trial, the traditional IMS method of using a full dose
of 100 ml IMS beads was compared with using a half dose of
IMS beads, a modification that would significantly decrease
675
the cost of each IMS test. The traditional method of acid
dissociation was also compared with heat dissociation,
based on a recent publication showing that heat dissociation
could improve parasite recovery efficiency in spiked water
samples (Ware et al., 2003). Finally, negative binomial
regression was used to compare the three methods, with the
best method applied to all the wild clam samples collected
in the second year of the study as described previously for
Cryptosporidium testing.
2.2. Cryptosporidium oocysts
Wild-type C. parvum genotype 2 oocysts (synonymous
with bovine genotype A), henceforth called C. parvum, were
used for sensitivity experiments and the tank exposure
experiment. Feces were obtained from infected calves from
commercial dairies near the Veterinary Medical Teaching
and Research Center, Tulare, CA. Calf feces were sieved
through a series of mesh strainers and then purified by
sucrose flotation methods (Arrowood and Sterling, 1987).
Oocyst concentrations for spiking experiments were deter-
mined with eight hemacytometer counts and confirmed with
DFA enumeration. Oocyst suspensions were kept at 4 8C
and used within 1 month of collection.
2.3. Clams
For the clam spiking and tank exposure experiments,
C. fluminea were harvested near Oakley, CA, in
cooperation with the California State Mussel Watch
Program. Clams were transported on ice and held in
freshwater flow-through tanks at the University of
California, Davis, Aquatic Pathogen Facility. Clams
were maintained for at least 3 weeks before being used
in any experiments so they could acclimatise to their
environment and depurate any residual oocysts they
might have been exposed to in the wild. The tanks were
constantly aerated and clams were fed Algamac (Aqua-
fauna Biomarine Inc., Hawthorne, CA) every other day.
Sentinel clams used for the outplanting studies were
collected near Oakley or Davis, CA, with a batch of 30
clams pre-tested at the time of collection.
For all experiments, clams 23 cm long were dissected to
obtain digestive gland and hemolymph samples. Hemo-
lymph was extracted first by filing a notch in the clam shell
and aspirating 0.51.5 ml of hemolymph from the adductor
muscle with a sterile syringe. Hemolymph was centrifuged
to concentrate the cell pellet that was stored frozen for later
DNA extraction and PCR analysis. Next, the adductor
muscle was cut to open the clam, and the digestive gland
was excised. Half of the digestive gland was frozen for PCR
and half was sieved through a 100 mm cell strainer and
centrifuged for 15 min at 1000!g before DFA and IMS-
DFA analysis.
676 W.A. Miller et al. / International Journal for Parasitology 35 (2005) 673684
2.4. Immunomagnetic separation
Digestive gland pellets of up to 0.5 ml were concentrated
and purified with IMS (Dynal Biotech, Oslo, Norway) per
manufacturers instructions, followed by DFA analysis for
parasite detection. Briefly, the tissue pellet was suspended in
an L10 glass tube in 2 ml buffer and 10 ml millipore water.
Next 100 ml IMS beads were added and the samples were
rotated for 1 h at setting 18. The IMS beadparasite
complexes were bound to a magnetic holder while the
supernatant and debris were discarded. Beadparasite
complexes were re-suspended in 1 ml buffer and transferred
to 1.5 ml microcentrifuge tubes that were put into a smaller
magnetic holder to again bind the beadparasite complexes
while the supernatant was discarded. For parasite dis-
sociation from the beads, two acid washes of 50 ml 0.1 N
HCl each were vortexed at the beginning and end of the
10 min incubation. Tubes were then put in a small magnetic
holder and the supernatant was transferred to a DFA slide or
PCR tube containing 5 ml 1 N NaOH for neutralisation,
while the beads were held on the magnet. Hemolymph was
not considered suitable for IMS because preliminary studies
revealed that hemocytes adhered to the glass tubes.
In the Giardia IMS clam spiking experiment, two IMS
modifications were evaluated with six replicates per
modification and spike dose. First, selected samples were
processed using 50 ml of IMS beads instead of the usual
100 ml. All other IMS parameters were the same as
previously described. Second, selected samples were
dissociated from the IMS beads in the final step using
95 8C sterile PBS instead of HCl, again without changing
other IMS parameters. Cyst recovery data were then
analysed using negative binomial regression to determine
the most sensitive and cost effective protocol for use on the
wild clam samples.
2.5. Direct immunofluorescence
Cryptosporidium oocyst DFA detection was performed
on 10 ml digestive gland suspension dried onto a 3-well
Merifluor slidewell (Meridian Bioscience Inc., Cincinnati,
OH). For IMS-DFA analysis of digestive gland, the 50 ml
IMS product containing parasites was dried onto a DFA well
with two wells per tissue sample. The DFA slides were
incubated with anti-Cryptosporidium and anti-Giardia
antibody, counterstained (Meridian Bioscience Inc.) for
30 min, and examined at 200400! magnification on a
Zeiss Axioscop epi-fluorescent microscope. All slides were
examined by the same microscopist. Cryptosporidium
parvum-like oocysts were identified as 5 mm diameter
spheres that were outlined in apple green fluorescence,
often with a midline seam, whereas Cryptosporidium
andersoni/muris-like oocysts were 5!7 mm diameter and
elliptical in shape. Giardia-like parasites were also elliptical
and outlined in apple green, but were larger in size
(1014 mm diameter). After parasite quantification,
the slide well was isolated from other wells with a grease
pencil, scraped with a scalpel blade, and washed with sterile
PBS into a microcentrifuge tube for DNA extraction, PCR
amplification, and DNA sequence analysis. Testing of
hemolymph by DFA was not done because preliminary
studies revealed that hemocytes autofluoresced, making
oocyst identification difficult.
2.6. Polymerase chain reaction
Two conventional 18S rRNA PCR protocols, designated
PCR1 (Morgan et al., 1997) and PCR2 (Xiao et al., 1999),
were evaluated for Cryptosporidium spp. detection in clam
tissues. For DNA extraction from hemolymph and digestive
gland, a 50 ml maximum pellet was mixed in 180 ml ATL
buffer in a microcentrifuge tube and suspended in liquid
nitrogen for 4 min followed by boiling water for 4 min.
Using the QIAGEN DNA Mini Kit tissue protocol
(QIAGEN Inc., Valencia, CA), mussel tissue was digested,
bound to a QIAamp column, washed, and the DNA eluted
with 50 ml of 95 8C PCR water. Extracted DNA was frozen
until PCR analysis.
A direct PCR protocol, PCR1, was used to amplify a
300 bp variable segment of the Cryptosporidium 18s rRNA
gene. The Cryptosporidium primers consisted of 18sif: AGT
GAC AAG AAA TAA CAA TAC AGG and 18sir: CCT
GCT TTA AGC ACT CTA ATT TTC. The PCR reactions
contained 5 ml 10! PCR buffer, 3 mM MgCl2, 200 mM
each dNTP, 200 nm each primer, 1.5 U Taq polymerase,
and 2 ml DNA in a 50 ml total volume. Amplification
conditions for the PCR reactions started with 2 min at 96 8C,
followed by 35 cycles of 94 8C for 30 s, 60 8C for 30 s, and
72 8C for 30 s, with a final cycle of 72 8C for 7 min.
The PCR2 18S rRNA protocol used an outer primer set to
amplify a 1325 bp DNA segment, followed by amplification
of an 850 bp DNA segment using the inner primer set. The
outer primers sequences were C1F: TTC TAG AGC TAA
TAC ATG CG and C1R: CCC TAA TCT TTC GAA ACA
GGA, and the inner primer sequences were C2F: GGA AGG
GTT GTA TTT ATT AGA TAA AG and C2R: AAG GAG
TAA GGA ACA ACC TCC A. The outer primer PCR
reactions contained 5 ml PerkinElmer (Norwalk, CN) 10!
PCR buffer, 6 mM MgCl2, 200 mM each deoxynucleoside
triphosphate (dNTP), 200 nm each primer, 1.5 U Taq
polymerase, and 2 ml DNA in a 50 ml total volume. The
inner primer PCR reactions were the same except that the
MgCl2 was reduced to 3 mM. Amplification conditions for
the outer and inner PCR reactions consisted of one cycle at
94 8C for 3 min, followed by 35 cycles of 94 8C for 45 s,
55 8C for 45 s, and 72 8C for 1 min, with a final extension at
72 8C for 7 min.
The Giardia PCR protocol was designed to amplify a
432 bp segment of the glutamate dehydrogenase gene from
multiple genotypes of Giardia duodenalis (Read et al.,
2004). Primers for the semi-nested protocol consisted of
external forward primer GDHeF: TCA ACG TYA AYC
 W.A. Miller et al. / International Journal for Parasitology 35 (2005) 673684 677

GYG GYT TCC GT, internal forward primer GDHiF: CAG regression for the IMS-DFA data (Pereira et al., 1999;
TAC AAC TCY GCT CTC GG and reverse primer GDHiR: Hardin and Hilbe, 2001; Atwill et al., 2003). Both Poisson
GTT RTC CTT GCA CAT CTC C. PCR reactions and negative binomial regression are designed to model
contained 12.5 rmol of each primer, 200 um dNTPs, count data, but negative binomial regression is more
1.5 mM MgCl2, 0.5 U Taq polymerase, and 2 ml DNA in a appropriate when the variance exceeds the mean (Hardin
50 ml total reaction volume. Amplification conditions and Hilbe, 2001).
started with two cycles of 94 8C for 2 min, 56 8C for In addition to the percent recovery estimate, the assay
1 min, and 72 8C for 2 min, followed by 55 cycles of 94 8C sensitivity S(c), defined as the probability of detecting at
for 30 s, 56 8C for 20 s, and 72 8C for 45 s, with a final least one oocyst per sample, was calculated as shown in
extension of 72 8C for 7 min. equations (1) and (2), respectively, where ebxi is the percent
Amplified PCR products were separated by gel electro- recovery of the assay, a is an ancillary parameter for
phoresis on a 2% agarose gel containing GelStar (Cambrex modelling dispersion, ci is the number of oocysts spiked per
Co., East Rutherford, NJ). Product size and band intensity digestive gland, and Wi is the proportion of the digestive
were compared to a 100 bp ladder included with each gel. gland tested in the assay.
For selected isolates, the PCR product was purified
according to QIAGEN Qiaquick protocol, sequenced on Poisson Sci Z 1 K ebxi (1)
an automated sequencer, and analysed with Chromas
(Technelysium Pty Ltd, Tewantin, Qld, Australia) Negative binomial Sci Z 1 K 1=1 C aci Wi ebxi 1=a (2)
and ClustalX (Thompson et al., 1997) software for The clam tank experiment IMS-DFA oocyst count data
Cryptosporidium genotype identification. were also analysed using the negative binomial model to
assess the effect of water temperature, oocyst dose, and clam
2.7. Risk factor data collection timepoint on the number of oocysts detected per
digestive gland. For the wild sentinel clam study, each pool
Covariate data on potential risk factors were collected for was first classified as positive or negative for each pathogen.
the three study regions along the San Lorenzo River, Salinas Exact logistic regression was used to evaluate the
River, and Putah Creek, CA. Clam collection site data were association between the putative risk factors and the
categorised by study region and by upstream or downstream probability of predicting a pathogen-positive clam pool.
status. Information on sewage management was obtained Statistical significance was defined as a P!0.05.
from the Central Coast Regional Water Quality Control
Board (http://www.waterboards.ca.gov/centralcoast/). The
remaining risk factors were all coded as continuous
3. Results
variables. Precipitation data for the previous day, week,
and month before each clam collection were obtained from
the California Department of Water Resources (http://cdec. 3.1. Clam tissue spiking
water.ca.gov). Livestock density data was collected from
the 2002 agricultural census (http://www.nass.usda.gov/ Table 1 shows the proportions of Cryptosporidium-
census/census02). Human population density data was positive samples from six clam digestive gland and six
collected from the Census 2000 (http://factfinder.census. hemolymph samples spiked with 10-fold dilutions from 1 to
gov). The risk factor data could then be evaluated as 1000 oocysts and processed by DFA, IMS-DFA, and PCR
explanatory variables for detecting Cryptosporidium methods. All negative control clams (not spiked with
oocysts and Giardia cysts in clam tissues. C. parvum oocysts) tested negative by all methods. The
minimum oocyst detection limit for digestive gland tested
2.8. Data analysis Table 1
Percentage of Cryptosporidium-positive clams detected by direct fluor-
In the clam tissue spiking experiments, the proportion of escent antibody (DFA), immunomagnetic separation (IMS) with DFA, and
tissues positive per technique was calculated as the number PCR methods for spiked hemolymph and digestive gland tissues
of tissues testing positive divided by the total number of Oocysts Digestive gland % positive Hemolymph %
tissue samples tested for each technique. Percent recovery positive
values for DFA and IMS-DFA testing were calculated in DFA IMS-DFA PCR1 PCR2 PCR1 PCR2
two ways. First, by dividing the number of oocysts counted 0 0 0 0 0 0 0
on the DFA slide by the number of oocysts expected per test 1 0 17 17 0 100 67
aliquot. Expected oocysts per test aliquot were calculated by 10 0 83 17 50 100 67
multiplying the number of oocysts spiked into a digestive 100 83 100 83 67 100 50
gland sample by the proportion of spiked tissue analysed. 1000 nda nd 100 100 100 100
Second, the oocyst count data were modelled using Poisson nZsix replicates per technique and oocyst dose.
a
regression for the DFA data and negative binomial nd, not done.
678 W.A. Miller et al. / International Journal for Parasitology 35 (2005) 673684

Table 2
Recovery efficiency of direct fluorescent antibody (DFA) and immunomagnetic separation (IMS) with DFA methods for Cryptosporidium oocysts spiked into
clam digestive glands

C. parvum oocysts DFA IMS-DFA

spiked per clam
Mean no. oocysts Mean % recovery per Mean % recovery per Mean no. oocysts Mean % recovery
detected (range) test (range) clam (range) detected (range) (range)a
0 0 0 0 0 0
1 0 0 0 0.17 (01) 17 (0100)
10 0 0 0 5 (09) 50 (090)
100 2 (06) 46 (071) 2 (06) 69 (4375) 69 (4375)

nZsix replicates per technique and oocyst dose.

a
IMS-DFA mean % recovery represents the recovery per test and per clam because the whole clam digestive gland can be processed in one test.

by DFA was 100 oocysts. Concentrating the digestive gland
samples with IMS before DFA analysis increased the
minimum oocyst detection limit by 12 log10 units. All
digestive glands spiked with 100 oocysts were detected by
IMS-DFA, as well as 83% of samples spiked with 10
oocysts and 17% of samples spiked with one oocyst. For
digestive gland samples tested by PCR1, 83% or more of all
samples spiked with at least 100 oocysts, and 17% of
digestive samples spiked with one or 10 oocysts were
detected. In contrast, when tested by PCR2, 50% or more of
samples spiked with 10 or more oocysts were detected, but
no samples spiked with a single oocyst were detected. For
spiked hemolymph samples, PCR1 detected 100% and
PCR2 detected at least 50% of all samples spiked with one
or more oocysts. In conclusion, IMS-DFA and PCR1 were
the only methods able to detect a single oocyst spiked into
digestive gland samples, while a single oocyst spiked into
clam hemolymph samples was detectable by both conven-
tional PCR protocols.
In order to estimate the true number of oocysts present in
a digestive gland sample processed by our quantitative DFA
methods, the recovery efficiency was determined from the
spiking experiment data. Table 2 shows the percentage of
oocysts that were detected in C. parvum spiked digestive
gland samples tested by DFA and IMS-DFA. For digestive
gland samples spiked with 100 oocysts, a mean of two
oocysts were counted per test using DFA alone, which
regress the observed oocyst count data based on the known
spike doses, leading to a percent recovery estimate of 58%
per test. Fig. 1 shows the sensitivity of DFA and IMS-DFA;
the detection threshold where there was a 50% probability
of detecting one or more oocysts (DT50) in a digestive gland
sample processed by DFA was 80 oocysts and DT90 was 200
oocysts. For IMS-DFA, the DT50 was two oocysts and DT90
was eight oocysts per digestive gland sample.
3.2. Clam tank experiment
The DFA, IMS-DFA, and PCR Cryptosporidium detec-
tion methods were next applied to clams that had filtered
oocysts from experimentally contaminated water. Clam
mortality during the 3-week experiment was 12%. The
oocyst concentration in clam tanks decreased by 55% over
the first 6 h after oocyst inoculation compared to positive
control tanks that contained oocysts but no clams. All water
and clam samples from the negative control clam tanks that
contained water and clams but were not inoculated with
oocysts tested negative at all timepoints.
First a subset of tissue samples was analysed by DFA,
IMS-DFA, and PCR methods to determine the best
method for use on the rest of the tank experiment samples.
1
0.9
Probability of detecting

IMS-DFA
represented 46% of the oocysts expected in a 10 ml test
C. parvum oocysts

0.8
DFA
aliquot, but only 2% of the total oocyst dose spiked into the 0.7
entire digestive gland sample. In contrast, for IMS-DFA 0.6
analysis of clam digestive gland samples spiked with 100 0.5
oocysts, a mean of 69 oocysts were detected per test, which 0.4
represented 69% recovery for both the test and total spiked 0.3
digestive gland, because the whole digestive gland sample 0.2
could be processed in the IMS-DFA test. The DFA method 0.1
did not detect any of the clam tissues spiked with one or 10 0
oocysts. In contrast, the IMS-DFA method detected 50% of 0.01 0.1 1 10 100 1000
digestive gland samples spiked with 10 oocysts and 17% of No. oocysts in digestive gland
samples spiked with one oocyst. By fitting a Poisson model
Fig. 1. Sensitivity of direct fluorescent antibody (DFA) and immunomag-
to the DFA oocyst count data, the percent recovery per test netic separation (IMS) with DFA for detection of Cryptosporidium oocysts
was approximately 54%. The IMS-DFA count data was spiked into clam digestive glands, using Poisson and negative binomial
more variable, so a negative binomial model was used to regression, respectively.
 W.A. Miller et al. / International Journal for Parasitology 35 (2005) 673684 679

Table 3
Proportion of Cryptosporidium-positive clams detected after a 6 h tank exposure to 250 oocysts/clam

Clam Cryptosporidium Digestive gland % positive Hemolymph % positive

oocyst exposure status
DFA IMS-DFA PCR1 PCR2 PCR1 PCR2
Exposed (nZ21) 24 52 24 0 24 0
Unexposed (nZ3) 0 0 33a 0 66a 0
a
DNA sequence analysis of PCR products showed that dinoflagellate and clam DNA were amplified, not Cryptosporidium DNA.

Table 3 shows the comparison of Cryptosporidium detection
techniques on clam digestive gland and hemolymph
samples from oocyst exposed and unexposed clams. The
IMS-DFA method was the most sensitive test evaluated,
detecting Cryptosporidium in 52% of exposed clam
digestive gland samples and none of the unexposed clams.
For exposed clam digestive glands tested by other methods,
24% were positive by DFA and PCR1, while all samples
were negative by PCR2. Hemolymph samples tested by
PCR2 were all negative but 24% were again positive by
PCR1. Several samples tested by PCR1 from clams in the
oocyst-unexposed tanks also produced gel bands of the
appropriate size. Sequence analysis of the PCR1 products
confirmed that these were false-positive Cryptosporidium
results and identified the sequences as clam and dino-
flagellate DNA.
The most sensitive method, IMS-DFA of digestive gland
samples, was then used to analyse all the tank experiment
samples. The number of oocysts detected per digestive
gland varied widely, ranging from 0 to 242 oocysts. The
majority of oocysts were detected during the first 9 h, but
low concentrations of oocysts continued to be detected in
clams exposed to low or high doses for up to 1 and 3 weeks,
respectively. Dose and time since exposure of oocysts were
highly associated with the ability of IMS-DFA to detect
oocysts in clam digestive glands (Table 4). Coefficients
from the negative binomial model can be interpreted as
the natural logarithm of the mean number of oocysts
detected per assay for each treatment effect. For example,
relative to the referent condition (low dose, time point at 3 h
post-exposure), the mean number of oocysts detected per
assay for the middle dose and high dose at 3 h post-exposure
would be 5.1 and 36.2 additional oocysts (e1.62, e3.59),
respectively. The number of oocysts detected per sample
was not significantly different (PZ0.5) at 6 h compared to
3 h post-exposure, but samples tested at O24 h post-
exposure had significantly fewer oocysts compared to 3 h
post-exposure. Testing clam tissues at these latter times
increased the likelihood of a false negative with this assay.
Incubating clams at 10 compared to 20 8C water tempera-
ture did not affect the number of oocysts detected with this
assay (PZ0.89).
Fig. 2 illustrates the relationship between the mean
observed oocysts per clam sample in relation to the
predicted oocysts and 95% confidence interval for the
three oocyst exposure doses over the 3 week experiment.
The significant effect of oocyst dose is shown in the different
y-axis values for the three graphs: the mean oocysts detected
were 40 per clam in the high dose (250 oocysts/clam) tank at
3 h post-exposure, whereas the mean observed oocysts were
only eight per clam at the middle dose (25 oocysts/clam),
and one per clam at the low dose exposure (2.5 oocysts/
clam). The confidence limits are widest at the early
collection timepoints because some clams had over 100

Table 4
Estimated maximum likelihood coefficients of the negative binomial regression model fitted to oocyst recovery data from the clam Cryptosporidium parvum
oocyst tank exposure experiment

Parameter Coefficient 95% Confidence interval P-value

Oocyst dose/clam
2.5 oocystsa 0.0
25 oocysts 1.62 0.73 2.52 !0.001
250 oocysts 3.59 2.61 4.57 !0.001
Clam collection time post-oocyst exposure
3 ha 0.0
6h K0.57 K2.24 1.1 0.5
9h K1.69 K2.76 K0.62 0.002
1 day K1.99 K3.31 K0.66 0.003
3 days K6.28 K8.59 K3.98 !0.001
7 days K3.37 K5.61 K1.12 0.003
14 days K18.9 K20.08 K17.74 !0.001
21 days K6.28 K8.56 K4.01 !0.001
Intercept 0.012 K0.88 0.91 0.98

Oocyst dose and timepoint coefficients are in relation to the low dose and 3 h timepoint as reference values in the negative binomial model.
a
Referent condition for the negative binomial regression model.
680 W.A. Miller et al. / International Journal for Parasitology 35 (2005) 673684

C. parvum 6 hr tank exposure to 250 oocysts/clam

160
140 Predicted Oocysts from
Negative Binomial Model
Oocysts per clam

120 Predicted Oocyst 95% CI

Lower Limit
100 Predicted Oocyst 95% CI
80 Upper Limit
Mean Observed Oocysts
60
40
20
0
3 6 9 24 72 168 336 504
Time post-exposure (hrs)

C. parvum 6 hr tank exposure to 25 oocysts/clam

18 Fig. 3. Cryptosporidium spp. prevalence detected by immunomagnetic
16 separation with direct fluorescent antibody testing of clam batches collected
Oocysts per clam

14 from three California rivers in 2003.

12
10 the dry and wet seasons. Both C. parvum-like and
8 C. andersoni/muris-like oocysts were detected significantly
6
more often in the wet season than in the dry season
4
2
(P!0.005). Using PCR and DNA sequence analysis,
0 C. parvum sequences were identified from DFA-positive
3 6 9 24 72 168 336 504 slide scrapings from a San Lorenzo River clam and from a
Time post-exposure (hrs) Salinas River clam (Genbank Accession Nos. AY864316
C. parvum 6 hr tank exposure to 2.5 oocysts/clam
and AY864317). Individual hemolymph testing by PCR and
3 DNA sequence analysis detected the same C. parvum
sequences in a Salinas River clam. No other Cryptospor-
Oocysts per clam

2.5
idium genotypes were identified by PCR amplification of
2
slide scrapings or hemolymph samples.
1.5 Risk factor data were then analysed to look for variables
1
that might affect the likelihood of detecting Cryptospor-
idium in sentinel clam batches. The covariates for
0.5 precipitation accumulation, human density, and animal
0 density across the three study sites did not vary and so were
3 6 9 24 72 168 336 504 excluded from further analysis. Using Exact logistic
Tim e post-exposure (hrs) regression, the effect of season (P!0.001) and year
Fig. 2. Cryptosporidium oocysts detected in clam digestive gland by
(P!0.005) were significantly associated with the prob-
immunomagnetic separation with direct fluorescent antibody testing for ability of detecting Cryptosporidium-positive clam pools.
three oocyst exposure doses used in a 6 h clam tank exposure experiment. The clam collection region, upstream versus downstream
Clams were tested at 3, 6, 9 h, 1, 3, 7, 14, and 21 days post-exposure. site location, and sewage management practices were not
significantly associated with Cryptosporidium detection in
oocysts detected while others had very low numbers, this study (PO0.2).
whereas at later timepoints oocysts were only detected at Giardia testing was initiated in the second year of the
low numbers in exposed clam samples. study. Table 5 shows the recovery efficiency of Giardia
cysts spiked into clam digestive gland when using three
3.3. Wild clam testing
Table 5
In 2002, all sentinel clam samples from the three riverine Recovery efficiency of Giardia cysts spiked into clam digestive gland and
processed by three immunomagnetic separation (IMS) protocols
regions were negative by IMS-DFA of pooled digestive
gland samples and PCR of individual hemolymph samples. Giardia cysts Full dose IMS Half-dose IMS Half-dose IMS
However, in 2003, Cryptosporidium oocysts were detected spiked per clam beadsCacid beads C acid beadsCheat
dissociationa dissociationa dissociationa
in clams from all three riverine study regions. The mean
number of Cryptosporidium detected per positive clam pool 0 0 0 0
10 60 (5075) 66 (3888) 2 (06)
in 2003 was two oocysts (range, 17). Fig. 3 shows the
100 55 (4464) 65 (5771) 41 (2663)
prevalence of C. parvum and C. andersoni/muris-like
a
oocysts detected in clam batches by IMS-DFA during Mean % recovery for six replicates (% recovery range).
 W.A. Miller et al. / International Journal for Parasitology 35 (2005) 673684
Fig. 4. Giardia spp. prevalence detected by immunomagnetic separation
with direct fluorescent antibody testing of clam batches collected from three
California rivers in 2003.
variations of the IMS-DFA protocol. There was no
significant difference between using a full and half dose of
IMS-beads with acid dissociation (PZ0.126), while using
acid dissociation was significantly better than heat dis-
sociation (P!0.001). The most sensitive and cost efficient
method, a half dose of IMS beads with acid dissociation,
detected a mean of 65% of 100 oocysts and 66% of 10 cysts
spiked into digestive gland samples.
This IMS-DFA method was then applied to the 2003
sentinel clams. Giardia was detected in clams from all three
riverine study regions. The mean number of cysts detected
per clam pool was three (range, 126). Using the same risk
factor analyses that were applied for Cryptosporidium,
Giardia-positive clam pools were significantly associated
with wet season sampling (PZ0.015) and downstream
location (P!0.001). Clam region and sewage management
practices were not significantly associated with Giardia-
positive clams. Fig. 4 shows the prevalence of Giardia
detected in clam digestive gland pools during the wet and
dry seasons of 2003. No positive PCR results were obtained
from amplifying DNA from hemolymph samples and
Giardia-positive IMS-DFA slide scrapings.
4. Discussion
This study showed that IMS, DFA and PCR methods can
be used to detect Cryptosporidium in spiked clam tissues,
clams that have filtered oocysts from water inoculated with
environmentally plausible oocyst doses, and clams out-
planted in naturally contaminated freshwater ecosystems in
California, USA. The most analytically sensitive technique
for oocyst detection in digestive gland was IMS concen-
tration followed by DFA analysis, which detected oocysts in
17 and 83% of the samples exposed to 1 and 10 oocysts,
respectively. For oocyst detection in hemolymph, both PCR
techniques were able to detect a single oocyst in a 1 ml
hemolymph aliquot at least 67% of the time. In the tank
experiment, the most sensitive Cryptosporidium detection
technique was IMS concentration of digestive gland tissues
681
followed by DFA analysis for parasite quantification, while
PCR did not perform as well. The greatest number of
oocysts were detected in clams exposed to high oocyst doses
and collected during the first day after oocyst exposure,
though low numbers of oocysts were detected as long as 3
weeks post-exposure. The wild clam study used IMS-DFA
and PCR to detect Cryptosporidium and Giardia spp. in
sentinel clams from three riverine study regions in
California. Risk factor analysis showed that clams sampled
during the wet season were more likely to contain
Cryptosporidium oocysts and Giardia cysts. Cryptospor-
idium parvum was confirmed in outplanted riverine clams
using PCR and DNA sequence analysis.
This study is the first to evaluate the utility of IMS
concentration for improving the analytical sensitivity of
DFA techniques on clam digestive gland samples. Pathogen
detection in fecal and environmental samples is challenging
due to the small sample volume that can be analysed per
test, and the presence of substances inhibitory to PCR
assays. The IMS concentration method overcomes these
problems in two ways. First, the volume of processed
sample is increased from 10 ml per DFA test or 50 ml per
PCR test to 500 ml per IMS test. Second, the IMS procedure
removes inhibitors by using several wash steps while the
parasites are bound to a magnetic holder via the immuno-
magnetic beads. Our finding that IMS concentration
improved the minimum oocyst detection limit in digestive
glands by 12 log10 units is similar to the findings in another
study that applied IMS to bovine fecal samples (Pereira
et al., 1999). The log10 increase in IMS-DFA sensitivity
from 100 to 10 oocysts per clam digestive sample is
important for accurate detection of the low numbers of
oocysts that may be expected in environmental samples.
One other study (Gomez-Bautista et al., 2000) used IMS to
improve assay sensitivity by concentrating bivalve hom-
ogenates and water samples by IMS when samples had
tested negative by fluorescent antibody testing alone. This
approach has merit because the investigators were able to
maximise their assay sensitivity using IMS concentration,
while minimising the number of assays run by testing
pooled bivalve homogenates.
Two modifications of the IMS technique were evaluated
in this study. First, to reduce the cost of reagents of IMS
testing (approximately $US 50/test), a half dose of IMS
beads was compared to using the full dose of beads. In the
Giardia spiking trial, we found that there was no significant
difference in the cyst recovery efficiency when using a half
or full dose of IMS beads, suggesting that the cost per IMS
test could be halved by using a half dose of beads per sample
in future studies. The second IMS modification assessed in
the Giardia spiking trial was acid dissociation compared to
heat dissociation of oocystbead complexes, based on a
recent publication that found increased parasite recovery
efficiencies in water samples using heat dissociation
(Ware et al., 2003). In our spiking experiment, cyst
recovery efficiencies for samples with acid dissociation of
682 W.A. Miller et al. / International Journal for Parasitology 35 (2005) 673684
oocystbead complexes during the IMS protocol were
significantly better than recovery efficiencies for samples
with heat dissociation. Possible explanations for this
discrepancy include differences in the study samples
analysed (clam tissues vs water samples), parasite charac-
teristics (Cryptosporidium vs Giardia), or study design.
Most IMS studies have used acid dissociation, including
studies based on the Environmental Protection Agency
Method 1623 for detection of Cryptosporidium and Giardia
in water samples (United States Environmental Protection
Agency, 2001; Sturbaum et al., 2002), and in our hands the
heat dissociation did not improve assay sensitivity for
oocyst detection in bivalve digestive glands.
Whether or not IMS is used for sample concentration, all
samples must still be analysed using a parasite detection
method such as DFA or PCR. The most widespread detection
methods used in bivalve studies to date have been fluorescent
antibody techniques (Fayer et al., 1998; Graczyk et al., 1998;
Gomez-Bautista et al., 2000). The advantages of these
methods are that results are quantitative and that fresh or
formalin-preserved samples can be analysed. However,
frozen tissues are not suitable for DFA analysis because
oocysts and cysts often rupture during the freezethaw
process. Another disadvantage of DFA is that it cannot
distinguish between the many Cryptosporidium genotypes
that look morphologically identical. In contrast, PCR
methods can be used for genotype identification and can be
used on fresh or frozen tissues, but formalin-preserved
samples are problematic. A limitation of conventional PCR is
that results are not quantitative, though the increasing
accuracy of quantitative PCR techniques such as TaqMan
PCR may solve this problem. As with our study, Sturbaum
et al. (2002) observed that false positives can occur with PCR
detection if DNA sequence analysis is not performed. The
ultimate choice between DFA and PCR detection methods
may then depend on whether the goal is parasite quantifi-
cation or genotype identification, or both.
PCR has been used in wild bivalve studies for parasite
genotype identification, and in some cases the Cryptospor-
idium genotypes detected in wild bivalves have provided
clues about the fecal loading sources (Fayer et al., 1998;
Gomez-Bautista et. al., 2000). For example, a study in the
Chesapeake Bay detected Cryptosporidium hominis, Cryp-
tosporidium baileyi, and C. parvum in oysters (Fayer et al.,
1998). Cryptosporidium hominis and C. baileyi oocysts are
primarily shed by humans and chickens, respectively, while
C. parvum oocysts are shed by both human and animal
hosts. After detecting the host-specific genotypes C. hominis
and C. baileyi in bivalves, nearby sources of human and
poultry feces could then be identified for further investi-
gation into the sources of fecal pollution and possible
remediation strategies to reduce fecal pollution. For non-
host-specific genotypes such as C. parvum, only the
geographic location of contaminated bivalves could provide
clues as to sources of fecal pollution. In the case of riverine
ecosystems, water flow is unidirectional and so fecal
pollution sources will usually be found upstream. However,
in estuarine and marine waters there is continuous mixing,
making identification of pollution source locations more
difficult. Another advantage of testing bivalves from
riverine systems is that the pathogen load may be more
concentrated than in the larger estuarine or marine bodies of
water that tend to dilute the incoming freshwater pollutants.
In an effort to simulate a riverine clam environment, our
tank exposure experiment was designed to evaluate a range
of environmentally plausible water temperatures and oocyst
doses that may affect oocyst detection in individual clams
over time (Rose, 1997). Several important findings became
apparent from the tank experiment data. First, there was
substantial variability in the number of oocysts detected in
individual clams (range, 1242 oocysts/digestive gland
sample). Each of these samples represented the total number
of oocysts detected by IMS-DFA in half a clam digestive
gland, unadjusted for the IMS 58% oocyst recovery
efficiency. Because the whole tissue pellet could be
concentrated by IMS for each clam sample, the variability
is more likely attributable to biologic variability in clams
filtering oocysts than assay limitations that occur when only
a sub-sample of the tissue pellet can be analysed. This
suggests that sampling a larger number of clams by testing
pooled samples will be more likely to represent the true
Cryptosporidium status within a clam batch than testing a
smaller number of clams individually. Second, there was no
significant difference in the number of oocysts detected
between clams held at 10 and 20 8C, so water temperatures
in this range should not limit the utility of clams as
bioindicators of fecal protozoal contamination in freshwater
ecosytems. Third, the greatest number of oocysts was
detected within the first day post-exposure but at very low
levels for as long as 3 weeks post-exposure, suggesting that
clams collected within a day of oocyst exposure events
such as storm runoff may be the best bioindicators of
Cryptosporidium load in freshwater ecosystems. Finally,
oocysts were detected at all three exposure doses ranging
from 20 to 2000 oocysts/L (3300 oocysts/clam), and more
oocysts were detected in clams exposed to the high dose
tanks than in clams from the medium or low dose tanks. This
suggests that clams may accumulate oocysts in proportion to
the exposure dose, a finding that is consistent with a recent
study (Graczyk et al., 2003) in which the authors exposed
clams to daily low oocyst doses for a month. They found
that the number of oocysts detected in clams increased over
the exposure month and decreased over the following three
weeks post-exposure.
Our wild clam study utilised the best methods from the
spiking and tank exposure experiments and applied them to
clams that had filtered fecal pathogens from naturally
contaminated waters. The three riverine study regions were
carefully selected sites where C. fluminea clams were
already in existence, because these clams are an aggressive
non-native species that we did not want to introduce into
new watersheds. During the 2002 dry and wet season
 W.A. Miller et al. / International Journal for Parasitology 35 (2005) 673684
sentinel clam testing, all three study regions were negative
for Cryptosporidium spp. by IMS-DFA and PCR testing.
However, during the 2003 dry and wet season testing, both
Cryptosporidium and Giardia spp. were detected in clams
from the three study regions by IMS-DFA. The year-to-year
variability in results is not surprising considering the
dynamic ecosystems in which clams are exposed to a
variety of water quality events over time. Because clams
slowly depurate the pathogens they filter, any one clam
batch represents a snapshot of water quality over the past
days to weeks, and thus repeated sampling will be needed to
better understand how fecal pathogens flow through riverine
ecosystems.
Risk factor analysis evaluated the association between
environmental variables and pathogen prevalence in clam
batches. Cryptosporidium and Giardia were detected more
often in the wet than the dry seasons in this study. An
explanation for the observed seasonality is that Cryptospor-
idium oocysts and Giardia cysts enter the riverine systems
primarily via overland runoff during storm events. Another
explanation is that oocyst and cyst shedding may occur
more often or at a greater magnitude during the wet season
(NovemberApril) than the dry season (MayOctober)
in California. Some studies have found seasonality in
Cryptosporidium detection in environmental bivalve and
water samples (Fayer et al., 2002; Tsushima et al., 2003),
while others have not (Robertson and Gjerde, 2001;
Gomez-Couso et al., 2003). This variability in outcomes
may be attributable to the different environmental factors
and study designs used.
This sentinel clam study was the first to utilise clams as
bioindicators of fecal protozoal pollution with Cryptospor-
idium and Giardia in natural riverine ecosystems. In future
studies, bivalves could be used as part of ambient water
quality monitoring programs focused on understanding the
dynamics of fecal pathogen loading in recreational waters
and evaluating whether mitigation of terrestrial fecal
loading sources will decrease the pathogen load in down-
stream waterways. Bivalves could also be used to monitor
for other fecal pathogens such as Toxoplasma gondii, a
parasite that causes neurologic disease in humans and
animals (Miller et al., 2004; Montoya and Liesenfeld, 2004).
A recent study of risk factors for Toxoplasma infection in
southern sea otters (Enhydra lutris nereis) found a
significant correlation with freshwater outflow (Miller
et al., 2002), and this finding is consistent with our clam
study that documented the presence of the fecal protozoal
pathogens Cryptosporidium and Giardia in two of the major
freshwater tributaries that feed into the southern sea otter
home range.
Acknowledgements
The authors would like to thank Gary Ichikawa from the
California State Mussel Watch Program, Kristen Arkush
683
from the Bodega Marine Laboratory, and Karen Worcester
from the Central Coast Regional Water Quality & Control
Board for their expertise on bivalve and water quality issues.
The assistance from student volunteers to collect and dissect
bivalves is also much appreciated. This study was supported
by grants from the Coastal Environmental Quality Initiative
and the University of California Center for Water Resources
and Center for Food Animal Health.
References
Arrowood, M.J., Sterling, C.R., 1987. Isolation of Cryptosporidium oocysts
and sporozoites using discontinuous sucrose and isopycnic Percoll
gradients. J. Parasitol. 73, 314319.
Atwill, E.R., Rose, T.L., 1998. Corbicula fluminea as an environmental
assay for detection of Cryptosporidium parvum in surface waters.
Technical Completion Report, Water Resources Center, University of
California, pp. 146.
Atwill, E.R., Hoar, B.R., Pereira, M.D.G., Tate, K.W., Rulofson, F.,
Nader, R., 2003. Improved quantitative estimates of low environmental
loading and sporadic periparturient shedding of Cryptosporidium
parvum in beef cattle. Appl. Environ. Microbiol. 89, 46044610.
Fayer, R., Graczyk, T.K., Lewis, E.J., Trout, J.M., Farley, C.A., 1998.
Survival of infectious Cryptosporidium parvum oocysts in seawater and
eastern oysters (Crassotrea virginica) in the Chesapeake Bay. Appl.
Environ. Microbiol. 64, 10701074.
Fayer, R., Lewis, E.J., Trout, J.M., Graczyk, T.K., Jenkins, M.C.,
Higgins, J., Xiao, L., Lal, A., 1999. Cryptosporidium parvum in oysters
from commercial harvesting sites in the Chesapeake Bay. Emerg.
Infect. Dis. 5, 706710.
Fayer, R., Trout, J.M., Lewis, E.J., Xiao, L., Lal, A., Jenkins, M.C.,
Graczyk, T.K., 2002. Temporal variability of Cryptosporidium in the
Chesapeake Bay. Parasitol. Res. 88, 9981003.
Freire-Santos, F., Oteiza-Lopez, A.M., Vergara-Castiblanco, C.A., Arees-
Mazas, M.E., Alvarez-Suarez, E., Garcia-Martin, O., 2000. Detection of
Cryptosporidium oocysts in bivalve molluscs destined for human
consumption. J. Parasitol. 86, 853854.
Gomez-Bautista, M., Ortega-Mora, L.M., Tabares, E., Lopez-Roda, V.,
Costas, E., 2000. Detection of infectious Cryptosporidium parvum
oocysts in mussels (Mytilus galloprovincialis) and cockles (Cerasto-
derma edule). Appl. Environ. Microbiol. 66, 18661870.
Gomez-Couso, H., Freire-Santos, F., Martinez-Urtaza, J., Garcia-
Martin, O., Mazas, M.E., 2003. Contamination of bivalve molluscs
by Cryptosporidium oocysts: need for new quality control standards.
Int. J. Food Microbiol. 87, 97105.
Graczyk, T.K., Fayer, R., Cranfield, M.R., Conn, D.B., 1998. Recovery of
waterborne Cryptosporidium parvum oocysts by freshwater benthic
clams (Corbicula fluminea). Appl. Environ. Microbiol. 64, 427430.
Graczyk, T.K., Fayer, R., Lewis, E.J., Trout, J.M., Farley, C.A., 1999.
Cryptosporidium parvum in bent mussels (Ischadium recurvum) in the
Chesapeake Bay. Parasitol. Res. 85, 518521.
Graczyk, T.K., Conn, D.B., Marcogliese, D.J., Graczyk, H., de
Lafontaine, Y., 2003. Accumulation of human waterborne parasites
by zebra mussels (Dreissena polymorpha) and Asian freshwater clams
(Corbicula fluminea). Parasitol. Res. 89, 107112.
Hardin, J., Hilbe, J., 2001. Negative binomial family. In: Generalized
Linear Models and Extensions. Stata Press, College Station, TX, pp.
141158.
Heitman, T.L., Frederick, L.M., Viste, J.R., Guselle, N.J., Morgan, U.M.,
Thompson, R.C.A., Olson, M.E., 2002. Prevalence of Giardia and
Cryptosporidium and characterization of Cryptosporidium spp. isolated
from wildlife, human, and agricultural sources in the North Saskatch-
ewan River Basin in Alberta, Canada. Can. J. Microbiol. 48, 530541.
684 W.A. Miller et al. / International Journal for Parasitology 35 (2005) 673684
Izumi, T., Itoh, Y., Yagita, K., Endod, T., Ohyama, T., 2004. Brackish
water benthic shellfish (Corbicula japonica) as a biological indicator
for Cryptosporidium parvum oocysts in river water. Bull. Environ.
Contam. Toxicol. 72, 2937.
McGlade, T.R., Robertson, I.D., Elliot, A.D., Read, C., Thompson, R.C.A.,
2003. Gastrointestinal parasites of domestic cats in Perth, Western
Australia. Vet. Parasitol. 117, 251262.
McMahon, R.B., Bogan, A.E., 2001. Mollusca:Bivalvia. In: Thorp, J.H.,
Covich, A.P. (Eds.), Ecology and Classification of North American
Freshwater Invertebrates. Academic Press, San Diego, CA, pp. 331397.
Miller, M.A., Gardner, I.A., Kreuder, C., Paradies, D.M., Worcester, K.R.,
Jessup, D.A., Dodd, E., Harris, M.D., Ames, J.A., Packham, A.E.,
Conrad, P.A., 2002. Coastal freshwater runoff is a risk factor for
Toxoplasma gondii infection in southern sea otters (Enhydra lutris
nereis). Int. J. Parasitol. 32, 9971006.
Miller, M.A., Grigg, M.E., Kreuder, C., James, E.R., Melli, A.C.,
Crosbie, P.R., Jessup, D.A., Boothroyd, J.C., Brownstein, D.,
Conrad, P.A., 2004. An unusual genotype of Toxoplasma gondii is
common in California sea otters (Enhydra lutris nereis) and is a cause
of mortality. Int. J. Parasitol. 34, 275284.
Monis, P.T., Thompson, R.C.A., 2003. Cryptosporidium and Giardia-
zoonoses: fact or fiction? Infect. Genet. Evol. 3, 233244.
Montoya, J.G., Liesenfeld, O., 2004. Toxoplasmosis. Lancet 363, 1965
1976.
Morgan, U.M., Constantine, C.C., Forbes, D.A., Thompson, R.C.A., 1997.
Differentiation between human and animal isolates of Cryptosporidium
parvum using rDNA sequencing and direct PCR analysis. J. Parasitol.
83, 825830.
OConnor, T.P., 2002. National distribution of chemical concentrations in
mussels and oysters in the USA. Mar. Environ. Res. 53, 117143.
Okhuysen, P.C., Chappell, C.L., Crabb, J.H., Sterling, C.R., DuPont, H.L.,
1999. Virulence of three distinct Cryptosporidium parvum isolates from
healthy adults. J. Infect. Dis. 180, 12751281.
Ong, C.S., Eisler, D.L., Goh, S.H., Tomblin, J., Awad-El-Kariem, F.M.,
Beard, C.B., Xiao, L., Sulaiman, I., Lal, A., Fyfe, M., King, A.,
Bowie, W.R., Isaac-Renton, J.L., 1999. Molecular epidemiology of
cryptosporidiosis outbreaks and transmission in British Columbia,
Canada. Am. J. Trop. Med. Hyg. 61, 6369.
Pereira, M.D.G., Atwill, E.R., Jones, T., 1999. Comparison of sensitivity of
immunofluorescent microscopy to that of a combination of immuno-
fluorescent microscopy and immunomagnetic separation for detection
of Cryptosporidium parvum oocysts in adult bovine feces. Appl.
Environ. Microbiol. 65, 32263239.
Pommepuy, M., Dumans, F., Caprais, M.P., Camus, P., LeMennec, C.,
Parnaudeau, S., Haugarreau, L., Sarrette, B., Vilagines, P., Pothier, P.,
Kholi, E., LeGuyader, F., 2004. Sewage impact on shellfish microbial
contamination. Water Sci. Technol. 50, 117124.
Read, C.M., Monis, P.T., Thompson, R.C.A., 2004. Discrimination of all
genotypes of Giardia duodenalis at the glutamate dehydrogenase locus
using PCR-RFLP. Infect. Genet. Evol. 4, 125130.
Robertson, L.J., Gjerde, B., 2001. Occurrence of Cryptosporidium oocysts
and Giardia cysts in raw waters in Norway. Scand. J. Public Health 29,
200207.
Robertson, L.J., Campbell, A.T., Smith, H.V., 1992. Survival of
Cryptosporidium parvum oocysts under various environmental press-
ures. Appl. Environ. Microbiol. 58, 34943500.
Rose, J.B., 1997. Environmental ecology of Cryptosporidium and public
health implications. Ann. Rev. Public Health 18, 135161.
Sturbaum, G.D., Klonicki, P.T., Marshall, M.M., Jost, B.H., Clay, B.L.,
Sterling, C.R., 2002. Immunomagnetic separation (IMS)-fluorescent
antibody detection and IMS-PCR detection of seeded Cryptosporidium
parvum oocysts in natural waters and their limitations. Appl. Environ.
Microbiol. 68, 29912996.
Tamburrini, A., Pozio, E., 1999. Long-term survival of Cryptosporidium
parvum oocysts in seawater and in experimentally infected mussels
(Mytilus galloprovincialis). Int. J. Parasitol. 29, 711715.
Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, D.G.,
1997. The ClustalX windows interface: flexible strategies for multiple
sequence alignment aided by quality analysis tools. Nucl. Acids Res.
25, 48764882.
Tsushima, Y., Karanis, P., Kamada, T., Makala, l., Xuan, X., Tohya, Y.,
Akashi, A., Nagasawa, H., 2003. Seasonal change in the number of
Cryptosporidium parvum oocysts in water samples from the rivers in
Hokkaido, Japan, detected by ferric sulfate flocculation method. J. Vet.
Med. Sci. 65, 121123.
United States Environmental Protection Agency, 2001. Method 1623:
Giardia and Cryptosporidium in water by filtration/IMS/FA. Office of
Water, Washington, DC, EPA 821-R01-025.
Ware, M.W., Wymer, L., Lindquist, H.D., Schaefer, F.W., 2003.
Evaluation of an alternative IMS dissociation procedure for use with
Method 1622: detection of Cryptosporidium in water. J. Microbiol.
Methods 55, 575583.
Xiao, L., Morgan, U.M., Limor, J., Escalante, A., Arrowood, M.,
Shulaw, W., Thompson, R.C.A., Fayer, R., Lal, A.A., 1999. Genetic
diversity within Cryptosporidium parvum and related Cryptosporidium
species. Appl. Environ. Microbiol. 65, 33863391.
Zhou, L., Fayer, R., Trout, J.M., Ryan, U.M., Schaefer, F.W., Xiao, L.,
2004. Genotypes of Cryptosporidium species infecting fur-bearing
mammals differ from those of species infecting humans. Appl. Environ.
Microbiol. 70, 75747577.
 International Journal for Parasitology 35 (2005) 685692
www.parasitology-online.com

Developmental expression and molecular analysis of two Meloidogyne

incognita pectate lyase genes*
Guozhong Huanga, Ruihua Donga, Rex Allena, Eric L. Davisc,
Thomas J. Baumb, Richard S. Husseya,*
a
Department of Plant Pathology, University of Georgia, Athens, GA 30602-7274, USA
b
Department of Plant Pathology, Iowa State University, 351 Bessey Hall, Ames, IA 50011, USA
c
Department of Plant Pathology, North Carolina State University, Box 7616, Raleigh, NC 27695-7616, USA

Received 8 December 2004; received in revised form 18 January 2005; accepted 19 January 2005

Abstract
Proteinaceous secretions from the oesophageal glands of plant-parasitic nematodes have crucial roles in nematode parasitism of plants.
Two cDNAs (designated Mi-pel-1 and Mi-pel-2) encoding pectate lyases were isolated from the root-knot nematode, Meloidogyne incognita,
oesophageal gland-cell subtractive cDNA libraries, and the corresponding genomic DNAs were subsequently cloned. Southern blot analyses
revealed that homologues to these pectate lyase genes were broadly distributed in Meloidogyne species, and present as members of a small
multigene family. Mi-pel-1 and Mi-pel-2 encoded, respectively, predicted proteins of 271 and 280 amino acids, each of which was preceded
by a signal peptide for secretion. Interestingly, these pectate lyases showed diversity at the amino acid level, with only 31% identity and 49%
similarity. These pectate lyases were classified into the same family of pectate lyases with those of other phytoparasitic nematodes that
contain four conserved regions characteristic of the class III pectate lyases of microbes. In situ mRNA hybridisation analyses showed the
transcripts of Mi-pel-1 and Mi-pel-2 accumulated exclusively within the subventral oesophageal gland cells of M. incognita. RT-PCR
analysis confirmed that their transcriptions were strong at the pre-parasitic and early parasitic second-stage juveniles, and not detectable at
the late parasitic stages of the nematodes. These results indicated that these pectate lyases, like cellulases, could be secreted into plant tissues
to facilitate the penetration and intercellular migration of M. incognita during the early stages of plant parasitism.
q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: Root-knot nematode; Pectate lyase; Oesophageal gland cells; Plant parasitic nematode

1. Introduction bonds (Yoder et al., 1993). Pectate lyases, therefore,

Pectate lyases are a class of polysaccharidases produced
by a diverse group of organisms including bacteria and
fungi (Herron et al., 2000). Plant pathogens secret an array
of pectate lyases to cleave a-1,4 linked galacturonosyl
residues of the pectate component of the primary plant cell
wall and middle lamella by b-elimination of the glycosidic
*
Note: Nucleotide and/or amino acid sequence data reported in this paper
are available in the GenBanke, DDBJ, and EMBL databases under the
accession numbers AF527788 for the Mi-pel-1 cDNA and AY515702 for
the genomic clone, AY327873 for the Mi-pel-2 cDNA and AY515703 for
the genomic clone.
* Corresponding author. Tel.: C1 706 542 1254; fax: C1 706 542 1262.
E-mail address: hussey@uga.edu (R.S. Hussey).
0020-7519/$30.00 q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijpara.2005.01.006
function as pathogenicity factors for many phytopathogenic
microbes by degrading host cell wall to allow penetration
and colonisation (Alghisi and Favaron, 1995).
Secretory proteins synthesised in the oesophageal glands
of plant-parasitic nematodes play essential roles in nema-
tode infection and parasitism of plants (Hussey, 1989; Davis
et al., 2000, 2004). Recently, three pectate lyase genes
expressed in the oesophageal subventral gland cells of cyst
and root-knot nematodes have been identified, and all show
sequence similarity on the amino acid level to class III
pectate lyases of bacteria and fungi (Popeijus et al., 2000;
De Boer et al., 2002; Doyle and Lambert, 2002). Pectate
lyases are divided into five classes, of which only members
of class I and class II have received extensive structural and
functional analyses (Shevchik et al., 1997). The biochemical
686 G. Huang et al. / International Journal for Parasitology 35 (2005) 685692
analyses of class III pectate lyases are now emerging
(Shevchik et al., 1997; Popeijus et al., 2000; Doyle and
Lambert, 2002).
Root-knot nematodes, Meloidogyne species, are obligate
sedentary endoparasites that infect over 2000 plant species
(Sasser, 1980). The second-stage juvenile (J2) penetrates
susceptible plant roots in the zone of elongation and
migrates intercellularly in the cortex to the differentiating
vascular cylinder where it releases proteinaceous secretions
through its stylet into five to seven parenchyma cells. The
affected root cells are transformed into multinucleate giant-
cells, which serve as permanent feeding sites for the
subsequent sedentary parasitic stages (Hussey, 1989; Davis
et al., 2004). Pectin is a major structural component of
primary cell walls and highly concentrated in the middle
lamella between plant cells (Barras et al., 1994). Similar to
the secretion of endoglucanases by cyst nematode juveniles
migrating through roots (Wang et al., 1999) root-knot
nematodes likely secrete pectin-degrading enzymes includ-
ing pectate lyases that are needed for weakening the cell
walls of root tissue during penetration and intercellular
migration of the parasite.
In attempting to obtain a comprehensive profile of
parasitism genes expressed in the oesophageal gland cells,
we constructed two gland-cell specific cDNA libraries from
mRNA in the cytoplasm microaspirated from the oesopha-
geal gland cells of parasitic stages of M. incognita (Huang
et al., 2003, 2004). Expressed sequence tag (EST) analyses
of these M. incognita gland-cell cDNA libraries identified
two cDNA clones that had high similarities to class III
pectate lyases. We report the isolation and molecular
analysis of full-length cDNA and genomic clones of both
M. incognita pectate lyases and developmental expression
profiles of both genes in the life cycle of M. incognita.
2. Materials and methods
2.1. Biological materials
M. incognita, Meloidogyne javanica and Meloidogyne
arenaria were cultured on greenhouse-grown tomato Lyco-
persicon esculentum cv. Marion. Meloidogyne eggs were
collected from galled roots using 0.5% NaOCl (Hussey and
Barker, 1973). Second-stage juveniles (J2) were hatched from
eggs placed on 25-mm opening sieves. Tomato roots infected
with M. incognita J2 (w25,000) were harvested at 2, 7 or 13
days after inoculation, and the different parasitic stages of
M. incognita were extracted by root maceration and sieving
(Ding et al., 1998). Adult females of M. incognita were hand-
picked under dissecting microscope from galled roots at 30 to
35 days p.i. (Ray et al., 1994). Mixed parasitic stages of
M. incognita for in situ hybridisations were collected at 13 to 15
days p.i. as described (De Boer et al., 1998). Caenorhabditis
elegans was grown on agar plates containing a lawn of
Escherichia coli OP50 (Brenner, 1974). Young tomato leaves
(1-month-old) were collected from greenhouse-grown
L. esculentum cv. Marion.
2.2. Nucleic acid extraction and synthesis
Genomic DNA was isolated from tomato leaves as
described by Dellaporta (1993). Genomic DNA was isolated
from J2 of Melodogyne species and mixed stages of
C. elegans as described by Ray et al. (1994).
Poly(A)C RNA (mRNA) was isolated from different
stages of M. incognita (Huang et al., 2003) with Dynabeads
Oligo(dT)25 magnetic beads (Dynal, Lake Success, NY,
USA) following the manufacturers instructions, eluted with
10 ml diethylpyrocarbonate (DEPC)-treated water, and then
converted to first-strand cDNA by reverse transcription
(RT)-PCR using a cDNA synthesis kit (Clontech Labora-
tories, Palo Alto, CA, USA) according to the manufacturers
instructions.
2.3. Isolation of Mi-pel-1 and Mi-pel-2 cDNA clones
The C-terminal cDNA fragment (448 bp) of Mi-pel-1
was identified from clone 34C04 during random sequencing
of clones from a M. incognita oesophageal gland cell long-
distance PCR cDNA library (Huang et al., 2003). The full-
length cDNA sequence of Mi-pel-1 was subsequently
obtained by 5 0 -rapid amplification of cDNA ends
(5 0 -RACE) by using first-strand cDNA generated from
mixed parasitic stages of M. incognita and 5 0 end gene-
specific primer 34C04RC with the SMART RACE cDNA
amplification kit (Clontech) according to the manufacturers
instructions. The oligonucleotides used in this study
(Table 1) were synthesised by Integrated DNA Technolo-
gies (Coraville, IA, USA). The amplified PCR products
were cloned into pGEM-T Easy vector (Promega, Madison,
WI, USA) for sequencing and delineation of the full-length
Mi-pel-1 cDNA.
Mi-pel-2 was identified from clone 2B02B during
random sequencing of clones from a subtractive gland cell
cDNA library (Huang et al., 2004). The full-length double-
strand cDNA sequence of Mi-pel-2 in pGEM-T Easy vector
Table 1
Primers used in this study
Name Sequences (5 0 /3 0 )
T7 TAATACGACTCACTATAGGG
SP6 ATTTAGGTGACACTATAG
PL1GF ATGTTGACAAGCAAAACTAGCTTC
34D04RC GGTTCGGTTCAGAACCGACAACCCTAC
PL2GF TTTAAAATTATCAAAAAATGCCAC
PL2GR ATTACTATTTTCTAAATATCAAGC
34C04F GCCAACTTCGAATAACGAC
34C04R TCTTTGGATATCCGGAAC
2B02BF AATTTCCAACGGCTAAGGG
2B02BR GCATTAAAAGCTGCGGCGTG
MIAF TGACTCTGGAGATGGTGTTACG
MIAR GTGATGACTTGACCGTCAGGC
 G. Huang et al. / International Journal for Parasitology 35 (2005) 685692
was obtained using T7, SP6 and specific primers in
sequencing reactions.
2.4. Genomic clones
Two pairs of the gene-specific primers PL1GF and
34C04RC, and PL2GF and PL2GR (Table 1) designed from
the 5 0 - and 3 0 -terminal sequences of the Mi-pel-1 and
Mi-pel-2 cDNAs, respectively, were used to amplify the
corresponding genomic sequences from 200 ng of
M. incognita genomic DNA. The PCR products separated
in a 1.2% agarose gel were cut and purified with a QIAquick
gel extraction kit (Qiagen, Valencia, CA, USA). The gel-
purified products were cloned into pGEM-T easy vector for
sequencing.
2.5. Southern blot analysis
Ten micrograms of genomic DNA from each sample was
digested with Eco RI and Hind III, fractionated in a 0.7%
agarose gel and transferred onto Hybond-N Nylon
membranes (Amersham Pharmacia Biotech, Piscataway,
NJ, USA), following standard procedures (Sambrook et al.,
1989). Mi-pel-1 and Mi-pel-2 probes were prepared by
amplification of full-length cDNAs from recombinant
pGEM-T Easy vectors using T7 and SP6 primers. The
gel-purified PCR products were labeled by PCR using a
PCR-DIG probe synthesis system (Roche Applied Science,
Indianapolis, IN, USA). High-stringency hybridisations
were performed at 40 8C in DIG Easy Hyb solution
(Roche Applied Science) overnight, and washed twice
with 0.5!SSC/0.1% SDS solution at 68 8C. Probes that
hybridised to nematode or tomato genomic DNAs on
Southern blots were detected by alkaline phosphatase-
conjugated anti-DIG antibody and disodium 3-(4-methox-
yspiro{1,2-dioxetane-3,2 0 -(5 0 -chloro)tricyclo[3.3.1.13,7]de-
can}-4-yl)pheryl phosphate (CSPD) (Roche Applied
Science) chemiluminescent substrate reaction. The mem-
branes were exposed to high performance chemilumines-
cence films (Amersham Pharmacia Biotech) at RT for 1.5 h.
2.6. Sequence analyses, alignment, and phylogenetics
Sequence comparisons of M. incognita pectate lyases
with others in the databases were carried out using NCBI
BLASTX server www.ncbi.nlm.nih.gov/blastx/index.html
(Benson et al., 1998). Signal peptide predictions were
made using the SignalP program http://www.cbs.dtu.dk/
services/SignalP/ (Nielsen et al., 1997). Sequence align-
ments were made with a multiple alignment program
(MAP) http://searchlauncher.bcm.tmc.edu/multi-align/
multi-align.html and the publishable output was produced
using BOXSHADE 3.21 http://www.ch.embnet.org/soft-
ware/BOX_form.html. Protein theoretical pI and molecular
weight were predicted using the ProtParam tool (Wilkins
et al., 1998). http://ca.expasy.org/cgi-bin/protparam.
687
ClustalW1.8 multiple alignment was performed at
European Bioinformatics Institute (http://www.ebi.ac.uk.
clustalw/) with default parameters. The unrooted phyloge-
netic tree was generated based on the neighbor-joining
method using PAUP8 4.0b. Bootstrap values were calcu-
lated from 1000 replicates.
2.7. In situ hybridisation
The specific primers 34C04F and 34C04R, 2B02BF and
2B02BR were used to amplify the cDNA fragments (250 bp
in Mi-pel-1 and 283 bp in Mi-pel-2) from the full-length
cDNA clones of Mi-pel-1 and Mi-pel-2, respectively. The
PCR products were used to synthesise digoxigenin-labeled
sense and antisense cDNA probes (Boehringer Mannheim,
Mannheim, Germany) by asymmetric PCR amplification
with the four primers. All PCR reactions were performed in
20 ml reaction mixture with PCR digoxigenin labeling mix
(Boehringer Mannheim) in the asymmetric PCR instead of
dNTPs in the normal PCR (Huang et al., 2003). In situ
hybridisation was performed with pre-parasitic J2 and
mixed parasitic stages of M. incognita as described by
De Boer et al. (1998) except that nematodes were pre-
incubated at 37 8C for 1 h in proteinase K (1 mg/ml) (Roche
Applied Science) and the hybridisations were performed at
42 8C.
2.8. RT-PCR analysis
Transcripts of Mi-pel-1 and Mi-pel-2 were analysed at
different developmental stages of M. incognita by reverse
transcription (RT)-PCR. First-strand cDNAs from each
sample were used in PCR reactions with the same set of
gene-specific primers as in the above in situ hybridisations.
Control RT-PCR reactions were conducted on each sample
with the primers MIAF and MIAR designed from the
constitutively expressed M. incognita beta-actin gene
(GenBank accession number BE225475). PCR products
were run on 1.2% agarose gel with the PCR products of
Mi-pel-1 and Mi-pel-2 full-length cDNA as another positive
control and stained in ethidium bromide.
3. Results
3.1. Isolation of the M. incognita pectate lyase cDNAs
A cDNA fragment (448 bp) encoding 56 amino acids
with similarity to the C-termini of phytonematode and
microbial pectate lyases was identified from clone 34C04
during analysis of 2,452 cDNA sequences of a M. incognita
long-distance (LD) PCR cDNA library constructed from
cytoplasm microaspirated from oesophageal gland cells
(Huang et al., 2003). The full-length cDNA sequence
containing 1124 bp was obtained by 5 0 -rapid amplification
of cDNA ends (5 0 -RACE) by using mRNA of mixed
688 G. Huang et al. / International Journal for Parasitology 35 (2005) 685692

Mi-pel-1
31 99 51 261 190 109 420 344 277
(1782 bp)

Mi-pel-2
34 325 577 295 572 220 85
(2108 bp)

Fig. 1. Organisation of Meloidogyne incognita extracellular pectate lyase genes, Mi-pel-1 and Mi-pel-2.UTRs (grey boxes), exons (dark boxes) and introns
(open boxes) are drawn schematically to indicate their relative positions and sizes.

parasitic stages of M. incognita as starting material and each M. incognita cDNA and genomic DNA as template.
designated Mi-pel-1. The Mi-pel-1 cDNA contained an open The Mi-pel-1 and Mi-pel-2 genomic DNA products were
reading frame (ORF) of 813 bp encoding a deduced protein 1782 and 2108 bp, respectively (Fig. 1). The exon/intron
of 271 amino acids. The 5 0 UTR and 3 0 UTR sequences of boundaries of the genomic sequences were determined by
this ORF were 31 and 277 bp, respectively (Fig. 1). The 3 0 aligning the genomic sequences with the corresponding
UTR contained a typical polyadenylation signal cDNA sequences. Mi-pel-1 genomic DNA contained one
(AATAAA) at 12 nucleotides (nt) upstream from the short intron (51 bp) and two longer introns (190 and
poly(dA) tail. The deduced amino acid sequence of 420 bp). Surprisingly, Mi-pel-2 genomic DNA contained
Mi-pel-1 contained a N-terminal hydrophobic signal peptide two very large introns (577 and 572 bp). All sequences of
(22 amino acids) as predicted by Signal P (Nielsen et al., the introns followed the canonical GU-AG rule for cis-
1997). The signal peptide terminated immediately upstream splicing (Blumenthal and Steward, 1997). The four exon
of a potential eukaryotic signal peptide cleavage site sizes for Mi-pel-1 ranged from 99 to 344 bp with an average
between amino acids Ala22 and Asp23, giving rise to a exon size of 203 bp, and the three exon sizes for Mi-pel-2
mature protein with a Mr of 27,234 and a pI of 5.59, as ranged from 220 to 325 bp with an average exon size of
predicted by ProtParam tool (Wilkins et al., 1998). A motif 280 bp (Fig. 1).
search in PROSITE database (Hofmann et al., 1999) Southern blots were used to investigate the copy number
identified two possible N-glycosylation sites at amino of Mi-pel-1 and Mi-pel-2 in M. incognita and to assess the
acids Asn31 and Asn183. The Mi-pel-1 cDNA contained a presence of the homologues in Meloidogyne species. The
nematode trans-spiced leader sequence (SL1M) at 9 nt Mi-pel-1 and Mi-pel-2 cDNAs were DIG-labeled and used
upstream of the initiating methionine codon (Ray et al.,
as probes for a DNA gel blot containing Eco RI- and Hind
1994; Blumenthal and Steward, 1997).
III-digested DNA from the nematode species M. incognita,
Another cDNA (clone 2B02B) with similarity to phytone-
M. javanica, M. arenaria and C. elegans. As a control, the
matode and microbial pectate lyases was identified during the
blot also contained genomic DNA from tomato. Both
sequencing of 711 cDNA clones of a second M. incognita
Mi-pel-1 and Mi-pel-2 probes hybridised specifically with
subtractive gland cell cDNA library and designated Mi-pel-2
DNA of all three Meloidogyne species (Fig. 2). The Mi-pel-1
(Huang et al., 2004). The putative full-length Mi-pel-2 cDNA
or Mi-pel-2 cDNA hybridised to multiple fragments and
contained 963 bp containing an ORF of 840 bp that ended with
neither the cDNAs nor the genomic clone contains Eco RI or
a TAA stop codon. The cDNA contained a 34 bp 5 0 UTR and a
85 bp 3 0 UTR (Fig. 1), which contained a polyadenylation
signal (AATAAA) 11 nt upstream from the poly(dA) tail. The
ORF encoded a deduced protein of 280 amino acids with a
putative N-terminal signal sequence of highly hydrophobic
amino acids (24 amino acids), as predicted by SignalP
(Nielsen et al., 1997). The signal peptide terminated
immediately upstream of a potential eukaryotic signal peptide
cleavage site between amino acids Ser24 and Asp25, resulting
in a mature protein with a Mr of 27,793 and a pI of 8.86, as
predicted by ProtParam (Wilkins et al., 1998). A motif search
(Hofmann et al., 1999) identified four possible N-glycosyla-
tion sites at amino acids Asn35, Asn121, Asn198 and Asn229.
The Mi-pel-2 cDNA did not contain a nematode trans-spiced
leader sequence at the 5 0 end (Ray et al., 1994; Blumenthal and
Steward, 1997).

Fig. 2. Southern blot analyses of Mi-pel-1 and Mi-pel-2. A blot containing

3.2. Genomic clones
Genomic clones of Mi-pel-1 and Mi-pel-2 were obtained
by PCR amplification using the gene-specific primers from
genomic DNA from Meloidogyne incognita (Mi), Meloidogyne javanica
(Mj), Meloidogyne arenaria (Ma) digested with Eco RI (E) or Hind III (H)
was hybridised with a Mi-pel-1 cDNA probe (A) or a Mi-pel-2 cDNA probe
(B), respectively. M: 80 ng DIG-labeled molecular weight marker III in kb.
 G. Huang et al. / International Journal for Parasitology 35 (2005) 685692 689

Hind III sites; therefore, Mi-pel-1 and Mi-pel-2 are probably
members of a small multigene family. No hybridisation of
Mi-pel-1 or Mi-pel-2 was detected with genomic DNA of
C. elegans and tomato (data not shown).
3.3. Sequence comparison
Protein homology search of mature MI-PEL-1 and
MI-PEL-2 minus the putative signal peptide sequences
using PSI-BLASTP indicated high similarity to class III
pectate lyases from plant parasitic nematodes, bacteria and
fungi. MI-PEL-1 showed the highest similarity (96% identity
and 98% similarity) to the pectate lyase (MJ-PEL-1) from
M. javanica (Doyle and Lambert, 2002). MI-PEL-2 showed
the highest similarity (36% identity and 55% similarity) to
the pectate lyase (GR-PEL-1) from Globodera rostochiensis
(Popeijus et al., 2000). MI-PEL-1 had 31% identity and 49%
similarity to MI-PEL-2. The pectate lyases from M. incognita
shared 29 to 38% identity and 45 to 57% similarity with
pectate lyases from other plant-parasitic nematode species
and microbes.
Multiple sequence alignment of MI-PEL-1 and MI-PEL-2
with three pectate lyases from other plant-parasitic nematodes
and two from microbes is shown in Fig. 3. MI-PEL-1 and
MI-PEL-2 contained four highly conserved regions charac-
teristic of class III pectate lyases in bacteria and fungi, 1013

MI-PEL-1 1 ----MFSSKTSFNFLLLISSFALCKADFWPKARNNITVSETIQITNRDCNFDRYIPDPSK
MJ-PEL-1 1 ----MLTSKTSFNFLFLISSLALCKADFWPKARNNITVSETIQITNFDCHFDRYIPDPSK
GR-PEL-1 1 ----MLFVIISIVFAQIFQVHA---LCTFPSSTKTITVQATMNVASNTDYKYTTFVGGSG
HG-PEL-1 1 ----MLFILLVITFVQIGQLNAG--ICTFPNPSKSVTVQSMMTVSSSTDYKNTLFVGGSG
MI-PEL-2 1 MPHFYLKFLINLILLNLFPLLIKSDLCKFPTAKGNQTVDETIPLNKDKDFGFIRLIASPK
FS-PEL-A 1 ------MKFTAAFVAALVGTSSAAVTKTLPKSAGATSFPTAVPVKGSYDGGMKRFEREPK
PC-PEL-1 77 DTNSGLNYWYWVDVVSENQAQVVSNAVTTAPNAGPLRAAKASSECKPGATFENRTVDCGG

MI-PEL-1 57 LGNGGQNEHQG---YVFEIKNGGSLSNCIIGARPGTKGSAHGVLCDG-DCDINNVWFEDV
MJ-PEL-1 57 LGNGGQNEHQG---YVFEIKDGGSLSNCIIGARPGTKGSAHGVLCDG-DCDINNVWFEDV
GR-PEL-1 54 ILNGACDVKNGKMKYLMVLKHGVTIKNAIINT-P---G--LGIYCEG-SCVLENIYYKKL
HG-PEL-1 55 ILNGACDVNNDKLKYLMTLKNGVTIKNAILDT-P---G--LGIYCEG-NCVLENIYYKRL
MI-PEL-2 61 LGSCTIDFSKKMS-PILWLSDGVTVSNLIIGTES-----SSGIWCSG-SCTLKNVYFERV
FS-PEL-A 55 VCKGQDETGEKD--AMFILENGATLSNVIIGASQ-----AEGVHCKG-TCTLNNVWWADV
PC-PEL-1 135 VTIGTSCPNDSDKQKPLIILKNATVKNLRISASG----RADGIHCDSGNCTIENVIWEDI
* *
I II
MI-PEL-1 113 GEDAINFNG-----DSDGCVYN--VNGGGAKNGEDKVMQFDGKG--TLNVNNYYVD-NYV
MJ-PEL-1 113 GEDALNFNG-----DNDNCVYN--VNGGGAKNGEDKVMQFDGKG--TLYVNNYYVD-NYV
GR-PEL-1 107 CYHATGFGY-----KSTGTSYTYQVIGGAGQGSPDKYFTQSGRG--TTIIKNFCAEGKYG
HG-PEL-1 108 CYHATGFGY-----KSTSTSYTYQVIGGAGQGSPDKYFTQSGKG--TTIIKNFCAEGKYG
MI-PEL-2 114 CTHAAAFNATTDFTKTDRRSFTYTVEGGAGLHALDKMFVQSGPG--KTIINNFCGD-GFQ
FS-PEL-A 107 CEDAVTLKQ---------TSGTSYINGGGAFHASDKIIQFNGRG--TVHVKDFYAE-DYG
PC-PEL-1 191 CEDAATNNGK----TMTIVGGIAHNAKDGYGGKPDKVLQHNSKNSTTVVKGNFTLTGEHG
##
III IV
MI-PEL-1 163 RFCRSCGDCGDQH-----QRHIVITNLTAVHGQAGQFVCGVNSNYQDTCTLHDIKMEKGI
MJ-PEL-1 163 RFCRSCGNCGDQH-----QRHIVITNLTAVHGQAGQFVCGVNSNYQDTCTLHDIKMEKGI
GR-PEL-1 160 KVWCSCGNCIDQM-----PRSVQISNTKIQG--PGLAIISANSNYGDKISISGLTLYGQG
HG-PEL-1 161 KLWCSCGNCPFQT-----ARTVQISNTVLKG--PGLSVVSLNSNYGDKMSLSGLTLHGQK
MI-PEL-2 171 KVWRSCGTCNDEVSQNSKQRTVTITNSNFTG--KGHVIASGNAPYKDKVSFNNVKIFGYK
FS-PEL-A 155 KLSRSCGNCKDNGG----PRNVIVENSVAVD---GGVLCGINTNYGDTCKVINSCQDKGK
PC-PEL-1 247 KLWRSCGDCSNNGG----PRFLTVTSATVNG--TIDSIAGVNRNYGDVATISGLKIKNYK
# * * #

MI-PEL-1 218 HPCKVFDGNSDGSEPTSNNDEEDHGDGKFCIYKKGDIKYIGSKPKPKSKKSAKN

MJ-PEL-1 218 HPCKVFDGNSDGSEPTSNNDEEDHGDGKFCIYKKGDIKYIGSKPKPKSKKSAKN
GR-PEL-1 213 SPNTLTKYICQSYNGLTTMATMQP-NAKFRPTQSG----TGTCSYSTSAIKIVN
HG-PEL-1 214 SASTKTSYICQEYKGLTYMAAMSP-QANYEPTKSG----SGTCAYSASAVKIAS
MI-PEL-2 229 NRSTRVVYACGEVKPEISEDHLDTGASNWYIPGRAG--TGTVCNYPASAVKIVN
FS-PEL-A 208 -YCDRYEGNSSGKEPTKIGSGPDG----------------KYCTVTGSTTSC--
PC-PEL-1 301 ---EGKPPVCEEFKGVVKGQGSTEKYG--EKWDT------TNCKVSRSGVSKL-

Fig. 3. Multiple sequence alignment of Meloidogyne incognita pectate lyase 1 (MI-PEL-1, AF52778) and 2 (MI-PEL-2, AY327873) protein sequences with the
sequences of pectate lyases from other plant-parasitic nematodes, bacterium, and a fungus. MJ-PEL-1 (AF455757) from Meloidogyne javanica; HG-PEL-1
(AY026357) from Heterodera glycines; and GR-PEL-1 (AF127915) from Globodera rostochiensis, PC-PEL-1 (Y13340) from Pectobacterium chrysanthemi;
and FS-PEL-A (M94652) from Fusarium solani f. sp. pisi. The putative signal sequences of M. incognita MI-PEL-1 and MI-PEL-2 proteins are boxed. Black
bars (I to IV) indicate the conserved regions characteristic of Class III pectate lyases. Identical residues are highlighted in black, highly conserved are in gray.
Very highly conserved charged residues are indicated by a number symbol (#), while an asterisk (*) indicates conserved cysteine residues.
690 G. Huang et al. / International Journal for Parasitology 35 (2005) 685692

100 MI-PEL-1 [AF527788]

MJ-PEL-1 [AF455757]
Phyto-
MI-PEL-2 [AY327873] nematodes
100
100 GR-PEL-1 [AF127915]
89
HG-PEL-1 [AY026357]
100 PC-PEL-1 [Y13340]
Bacteria
EC-PEL-B [X79232]
47

100 FS-PEL-A [M94692]

95 FS-PEL-B [U13051] Fungi
100 FS-PEL-C [U13049]
FS-PEL-D [U13050]

BS-PEL-A [AJ237980] Bacteria

Fig. 4. Unrooted phylogenetic tree of pectate lyases is generated based on
the protein sequences without signal peptides using the neighbor-joining
method. GenBank accession numbers of pectate lyases from Meloidogyne
incognita (MI-PEL-1 and MI-PEL-2), Meloidogyne javanica (MJ-PEL-1),
Globodera rostochiensis (GR-PEL-1), Heterodera glycines (HG-PEL-1),
Pectobacterium chrysanthemi (PC-PEL-1), Erwinia carotovora (EC-PEL-
B), Fusarium solani f. sp. pisi, (FS-PEL-A, B, C and D) and Bacillus sp.
BP-23 (BS-PEL-A) are indicated in brackets. The bootstrap values are
calculated from 1000 replicates.
cysteine residues and four charged residues (Fig. 3) that are
potentially involved in catalysis (Shevchik et al., 1997;
Popeijus et al., 2000).
A phylogenetic tree was generated from an alignment of
the distinct MI-PEL-1 and MI-PEL-2 protein sequences
with their homologous class III pectate lyases from
phytonematodes, bacteria and fungi (Fig. 4). MI-PEL-1
fell into a subgroup with MJ-PEL-1 from M. javanica, and
MI-PEL-2 fell into another subgroup with GR-PEL-1 from
G. rostochiensis and HG-PEL-1 from H. glycines.
3.4. Developmental expression
Fig. 5. Hybridisation of the digoxygenin-labeled anti-sense cDNA probes
(dark staining) of Mi-pel-1 to transcripts accumulating exclusively within
the subventral oesophageal gland cells of Meloidogyne incognita. Anterior
end of a pre-parasitic second-stage juvenile. SvG, subventral gland cells;
M, metacorpus.
(eggs and 13-days p.i. juveniles), but not detectable in
adult females (Fig. 6). The transcription of Mi-pel-2 was
strong in two nematode stages (pre-J2s, 2 days p.i. J2s,
presence of a 283-bp size band) and weak in two nematode
stages (eggs and 7-days p.i. J2s), but not detectable in other
parasitic stages (13-days p.i. juveniles and adult females)
(Fig. 6). Since the primers 2B02BF and 2B02BR spanned
the 325-bp intron of Mi-pel-2 genomic DNA and allowed
differentiation of PCR products from genomic DNA and
cDNA, the amplification of only the 283-bp PCR product in
all the reactions indicated genomic DNA was not a
contaminate. This result revealed that both Mi-pel-1 and
Mi-pel-2 were strongly transcribed in pre-parasitic and early
parasitic J2s, while they had some different developmental
expression patterns in the life cycle stages (Fig. 6).

The tissue localisation and the developmental expression 4. Discussion

of Mi-pel-1 and Mi-pel-2 were analysed in different life
stages of M. incognita by in situ mRNA hybridisation. The Plant cell wall polysaccharides are primarily composed
digoxigenin-(DIG)-labeled antisense cDNA probes of of cellulose, hemicellulase and pectin (Herron et al., 2000).
Mi-pel-1 (Fig. 5) and Mi-pel-2 (data not shown) specifically In order to parasitise plants, plant-parasitic nematodes secret
hybridised with transcripts accumulated within the oeso-
phageal subventral gland cells of M. incognita. No
hybridisation was observed within sections of M. incognita
with the control sense cDNA probes of Mi-pel-1 or Mi-pel-2.
The hybridisation signals for Mi-pel-1 and Mi-pel-2 were
weak within the subventral gland cells in the pre-parasitic J2
and moderately strong within these gland cells in early
parasitic second-stage juveniles (data not shown).
The developmental expression pattern of Mi-pel-1 and
Mi-pel-2 were also determined by PCR amplification of
stage specific cDNAs using the gene specific primers Fig. 6. Determination of Mi-pel-1 and Mi-pel-2 developmental expression
34C04F and 34C04R of Mi-pel-1 and the gene specific by PCR amplification of cDNAs from different stages of Meloidogyne
incognita. Lanes 1 to 7 indicate cDNA templates from eggs, pre-parasitic
primers 2B02BF and 2B02BR of Mi-pel-2, respectively.
second-stage juveniles, 2-days p.i. juveniles, 7-days p.i. J2, 13-days p.i.
The transcription of Mi-pel-1 was strong in three nematode juveniles, adult females, and Mi-cm-1 or Mi-cm-2 full-length cDNA clone
stages (pre-J2s, 2 days p.i. and 7-days p.i. J2s, presence of a (positive control), respectively. Mi-act, M. incognita b-actin gene is used as
250-bp size band) and weak in two nematode stages positive control for each cDNA template.
 G. Huang et al. / International Journal for Parasitology 35 (2005) 685692
a battery of cell-wall-degrading enzymes including cellu-
lases and pectinases to break down the complex matrices of
primary cell walls (Smant et al., 1998; Rosso et al., 1999;
Popeijus et al., 2000; De Boer et al., 2002; Doyle and
Lambert, 2002). Root-knot nematode J2s move within plant
roots by migrating between plant cells along the middle
lamella that contains the most concentrated pectin; there-
fore, a secreted pectinase would facilitate root-knot
nematode infection of roots. In this study, we cloned and
characterised two pectate lyase genes (Mi-pel-1 and Mi-pel-2)
from the root-knot nematode, M. incognita.
Pectate lyases are diverse and can be divided into five
classes, but only some classes have been structurally and
functionally characterised (Shevchik et al., 1997). Based on
sequence similarities, class III pectate lyases have four
highly-conserved regions and are characterised by the
presence of several cysteine and charged residues, which
are potentially involved in catalysis (Shevchik et al., 1997;
Popeijus et al., 2000). The M. incognita pectate lyases are
distinctly different at the amino acid level, but they contain
all the features of enzymes that are characteristic of
microbial class III pectate lyases (Shevchik et al., 1997;
Popeijus et al., 2000).
Three class III pectate lyases have been identified from
other plant-parasitic nematodes including G. rostochiensis
(Popeijus et al., 2000), H. glycines (De Boer et al., 2002)
and M. javanica (Doyle and Lambert, 2002). Here we add
two examples of a Class III pectate lyase, indicating that the
pectin-degrading enzymes have a wide distribution among
endo-parasitic nematodes. Southern blot analysis of Mi-pel-1
and Mi-pel-2 supports this hypothesis, and reveals that both
pectate lyase genes are present in three Meloidogyne species
and exist as members of a small multigene family. In most
plant pathogens, pectate lyases usually exist as multiple,
independently regulated isozymes that are 27 to 80%
identical at the amino acid level. The pectate lyase isozymes
each use a similar catalytic mechanism, but recognise
different methylated and non-methylated oligogalacturonate
units (Herron et al., 2000). Because root-knot nematodes
infect plant species from diverse plant families, they may
have evolved an array of pectate lyases to infect such a broad
range of host plants.
Like other cell-wall degrading enzymes such as cellulase
(Rosso et al., 1999) and polygalacturonase (Jaubert et al.,
2002) in M. incognita, MI-PEL-1 and MI-PEL-2 are more
similar to microbial than eukaryotic enzymes. However,
several lines of evidence including the presence of poly(A)
at their 3 0 -ends and introns in their genomic DNAs, gene
expression in the oesophageal gland cells, and hybridisation
to M. incognita genomic DNA indicate the nematode origin
of these enzymes. The microbial similarity has led to
speculation that some nematode parasitism genes may have
been acquired by horizontal gene transfer from microbes
(Davis et al., 2000).
Most nematode introns are small, e.g. the majority of
C. elegans introns are less than 60 bp (Blumenthal
691
and Steward, 1997). One intron in Mi-pel-1 has 420 bp and
both introns in Mi-pel-2 are over 570 bp. Intriguingly, some of
the introns in other plant-parasitic nematode genes acquired by
putatively horizontal gene transfer are also considerably larger
than standard nematode introns. For example, the introns of
the G. rostochiensis Gp-cm-1 (chorismate mutase) are 547 and
335 bp (Jones et al., 2003) and the first intron of the
G. rostochiensis eng-1 and eng-2 (cellulase) sequences are
309 and 734 bp, respectively (Yan et al., 1998).
MI-PEL-1 without the signal peptide shares 96% identity
and 98% similarity to MJ-PEL-1 from M. javanica. MJ-
PEL-1, a secreted enzyme, displays biochemical features
typical of class III pectate lyases including an absolute
requirement for Ca2C and optimal activity at basic pH
(Doyle and Lambert, 2002). The high similarity of MI-PEL-1
to MJ-PEL-1 suggests that MI-PEL-1 in M. incognita may
have a similar function as MJ-PEL-1. MI-PEL-2 has the
highest similarity (E-value of 1eK34) to GR-PEL-1 from
G. rostochiensis, which produces an active enzyme in the
heterologous host Pichia pastoris (Popeijus et al., 2000). The
potential for distinct biochemical characteristics of MI-PEL-
2 from MI-PEL-1 are under investigation.
The gland-cell specific expression and presence of
predicted secretion signal peptides in both MI-PEL-1 and
MJ-PEL-2 suggest these proteins are secreted from the
nematode and, therefore may play an essential role in the
infection of host plants by this parasite. That these pectate
lyase genes are expressed in the early nematode develop-
mental stages suggests that their pectate lyase activities
would help to degrade the pectin matrix and assist nematode
migration between the cells in host roots. The Mi-pel-1 gene
is also expressed in the late second-stage juveniles,
indicating that it could also have a role in the maintenance
of giant cell plasticity.
Acknowledgements
Support for this research was provided by the National
Research Initiative Competitive Grants Program of the
Cooperative State Research, Education, and Extension
Service of the United States Department of Agriculture
under Agreement No. 99-35302-8080, the Iowa Soybean
Promotion Board, the Iowa Agriculture and Home Economics
Experiment Station by Hatch Act and State of Iowa funds, the
N. C. Agricultural Research Service, and by State and Hatch
Funds allocated to the Georgia Agricultural Experiments.
References
Alghisi, P., Favaron, F., 1995. Pectin-degrading enzyme and plant-parasite
interactions. Eur. J. Plant Path. 101, 365375.
Barras, F., Gijsegem, F.V., Chatterjee, A.K., 1994. Extracellular enzymes
and pathogenesis of soft-rot Erwinia. Annu. Rev. Phytopathol. 32,
201234.
692 G. Huang et al. / International Journal for Parasitology 35 (2005) 685692
Benson, D.A., Boguski, M.S., Lipman, D.J., Ostell, J., Ouellette, B.F.F.,
1998. GenBank. Nucleic Acids Res. 26, 17.
Blumenthal, T., Steward, K., 1997. RNA processing and gene structure. In:
Riddle, D.L., Blumenthal, T., Meyer, B.J., Priess, J.R. (Eds.), C. elegans
II. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY,
pp. 117146.
Brenner, S., 1974. The genetics of the nematode Caenorhabditis elegans.
Genetics 77, 7194.
Davis, E.L., Hussey, R.S., Baum, T.J., Bakker, J., Schots, A., Rosso, M.N.,
Abad, P., 2000. Nematode parasitism genes. Annu. Rev. Phytopathol.
38, 365396.
Davis, E.L., Hussey, R.S., Baum, T.J., 2004. Getting to the roots of
parasitism by nematodes. Trends Parasitol. 20, 134141.
De Boer, J.M., Yan, Y., Bakker, J., Davis, E.L., Baum, T.J., 1998. In situ
hybridization to messenger RNA of Heterodera glycines. J. Nematol.
30, 309312.
De Boer, J.M., Mcdermott, J.P., Davis, E.L., Hussey, R.S., Popeijus, H.,
Smant, G., Baum, T.J., 2002. Cloning of a putative pectate lyase gene
expressed in the subventral esophageal glands of Heterodera glycines.
J. Nematol. 34, 911.
Dellaporta, S.L., 1993. Plant DNA miniprep and microprep version 2.1-2.3.
In: Freeling, M., Walbot, V. (Eds.), The Maize Handbook. Springer,
New York, pp. 522525.
Ding, X., Shields, J., Allen, R., Hussey, R.S., 1998. A secretory cellulose-
binding protein cDNA cloned from the root-knot nematode (Meloidogyne
incognita). Mol. Plant-Microbe. Interact. 11, 952959.
Doyle, E.A., Lambert, K.N., 2002. Cloning and characterization of an
esophageal-gland-specific pectate lyase from the root-knot nematode
Meloidogyne javanica. Mol. Plant-Microbe. Interact. 15, 549556.
Herron, S.R., Benen, J.A.E., Scavetta, R.D., Visser, J., Jurnak, F., 2000.
Structure and function of pectic enzymes: virulence factors of plant
pathogens. Proc. Natl Acad. Sci. USA 97, 87628769.
Hofmann, K., Bucher, P., Falquet, L., Bairoch, A., 1999. The PROSITE
databases, its status in 1999. Nucleic Acids Res. 27, 215219.
Huang, G.Z., Gao, B., Maier, T., Allen, R., Davis, E.L., Baum, T.J.,
Hussey, R.S., 2003. A profile of putative parasitism genes expressed in
the esophageal gland cells of the root-knot nematode Meloidogyne
incognita. Mol. Plant-Microbe. Interact. 16, 376381.
Huang, G.Z., Dong, R., Maier, T., Allen, R., Davis, E.L., Baum, T.J.,
Hussey, R.S., 2004. Use of solid-phase subtractive hybridization for the
identification of parasitism gene candidates from the root-knot
nematode Meloidogyne incognita. Mol. Plant Path. 5, 217222.
Hussey, R.S., 1989. Desease-inducing secretions of plant-parasitic
nematodes. Annu. Rev. Phytopathol. 27, 123141.
Hussey, R.S., Barker, K.R., 1973. A comparison of methods of collecting
inocula of Meloidogyne spp., including a new technique. Plant Dis. Rep.
57, 10251028.
Jaubert, S., Laffaire, J.-B., Abad, P., Rosso, M.-N., 2002. A polygalactur-
onase of animal origin isolated from the root-knot nematode
Meloidogyne incognita. Fed. Eur. Biochem. Soc. Lett. 522, 109112.
Jones, J.T., Furlanetto, C., Bakker, E., Banks, B., Blok, V., Chen, Q.,
Phillips, M., Prior, A., 2003. Characterization of a chorismate mutase
from the potato cyst nematode Globodera pallida. Mol. Plant Pathol. 4,
4350.
Nielsen, H., Engelbrecht, J., Brunak, S., von Heijne, G., 1997. Identification
of prokaryotic and eukaryotic signal peptides and prediction of their
cleavage sites. Protein Eng. 10, 16.
Popeijus, H., Overmars, H., Jones, J., Blok, V., Goverse, A., Helder, J.,
Schots, A., Bakker, J., Smant, G., 2000. Degradation of plant cell walls
by a nematode. Nature 406, 3637.
Ray, C., Abbott, A.G., Hussey, R.S., 1994. Trans-splicing of a Meloidogyne
incognita mRNA encoding a putative esophageal gland protein. Mol.
Biochem. Parasitol. 68, 93101.
Rosso, M.-N., Favery, B., Piotte, C., Arthaud, L., De Boer, J.M.,
Hussey, R.S., Bakker, J., Baum, T.J., Abad, P., 1999. Isolation of a
cDNA encoding a b-1,4-endoglucanase in the root-knot nematode
Meloidogyne incognita and expression analysis during plant parasitism.
Mol. Plant-Microbe. Interact. 12, 585591.
Sambrook, J., Fristch, E.F., Maniatis, T., 1989. Molecular Cloning: A
Laboratory Manual, second ed. Cold Spring Harbor Laboratory Press,
Cold Spring Harbor, NY, USA.
Sasser, J.N., 1980. Root-knot nematode: a global menace to crop
production. Plant Dis. 64, 3641.
Shevchik, V., Robert-Baudouy, J., Hugouviux-Cotte-Pattat, N., 1997.
Pectate lyase PelI of Ervinia chrysanthemi 3937 belongs to a new
family. J. Bacteriol. 179, 73217330.
Smant, G., Stokkermans, J.P.W.G., Yan, Y., de Boer, J.M., Baum, T.J.,
Wang, X., Hussey, R.S., Gommers, F.J., Henrissat, B., Davis, E.L.,
Helder, J., Schots, A., Bakker, J., 1998. Endogenous cellulases in animals:
isolation of b-1,4-endoglucanase genes from two species of plant-parasitic
cyst nematodes. Proc. Natl Acad. Sci. USA 95, 49064911.
Wang, X., Meyers, D., Yan, Y., Baum, T., Smant, G., Hussey, R., Davis, E.,
1999. In planta localization of a b-1,4-endoglucanase secreted by
Heterodera glycines. Mol. Plant-Microbe. Interact. 12, 6467.
Wilkins, M.R., Gasteiger, E., Bairoch, A., Sanchez, J.-C., Williams, K.I.,
Hochstrasser, D.F., 1998. Protein identification and analysis tools in
ExPASy server. In: Link, A.J. (Ed.), 2-D Proteome Analysis Protocols,
2. Humana Press, New Jersey, NY, pp. 531552.
Yan, Y., Smant, G., Stokkermans, J.P.W.G., Qin, L., Helder, J., Baum, T.,
Schots, A., Davis, E.L., 1998. Genomic organization of four b-1,4-
endoglucanase genes in plant parasitic cyst nematodes and its
evolutionary implications. Gene 220, 6170.
Yoder, M.D., Keen, N.T., Jurnak, F., 1993. New domain motif: the structure
of pectate lyase C, a secreted plant virulence factor. Science 260,
15031507.
 International Journal for Parasitology 35 (2005) 693701
www.parasitology-online.com

Echinococcus shiquicus n. sp., a taeniid cestode from Tibetan

fox and plateau pika in China*
Ning Xiaoa,b, Jiamin Qiub, Minoru Nakaoa, Tiaoying Lib, Wen Yangb,
Xingwang Chenb, Peter M. Schantzc, Philip S. Craigd, Akira Itoa,*
a
Department of Parasitology, Asahikawa Medical College, Asahikawa 078-8510, Japan
b
Institute of Parasitic Diseases, Sichuan Center for Disease Control and Prevention, Chengdu, Sichuan 610041, China
c
Division of Parasitic Diseases, National Centers for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, GA 30341, USA
d
Cestode Zoonoses Research Group, Bioscience Research Institute and School of Environment and Life Sciences,
University of Salford, Great Manchester M5 4WT, UK
Received 26 October 2004; received in revised form 4 January 2005; accepted 4 January 2005

Abstract

The taeniid cestode Echinococcus shiquicus n. sp. was found from the Tibetan fox Vulpes ferrilata and the plateau pika Ochotona
curzoniae in the Qinghai-Tibet plateau region of China. In the adult stage, E. shiquicus from the foxes is morphologically similar to
Echinococcus multilocularis. However, the new species is differentiated by its smaller rostellar hooks, fewer segments, distinct position of
genital pore in the mature segment and fewer eggs in the gravid segment. Hydatid cysts of E. shiquicus found in the livers from the pikas were
essentially unilocular but an oligovesicular cyst was also found. The data of mitochondrial and nuclear DNA sequences proved E. shiquicus
to be a valid taxon.
q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: Qinghai-Tibet plateau; Tibetan fox; Plateau pika; Echinococcus shiquicus n. sp.

1. Introduction distributed, whereas the latter two species are restricted to

Species of Echinococcus Rudolphi, 1801 (Cestoda:
Taeniidae) are minute tapeworms of carnivores. Their
larvae are known as hydatids, which proliferate asexually
in various mammals including humans. The taxonomy of
this genus has been controversial owing to inadequate
descriptions and sympatric occurrences of subspecies.
A total of 16 species and 13 subspecies have been described
but only four species (i.e. Echinococcus granulosus,
Echinococcus multilocularis, Echinococcus oligarthrus
and Echinococcus vogeli) are generally accepted as valid
taxa (Rausch and Bernstein, 1972; Kumaratilake and
Thompson, 1982). The former two species are widely
*
Note: Nucleotide sequence data reported in this paper are available in
DDBJ/EMBL/GenBank databases under the accession numbers
AB159136AB159143.
* Corresponding author. Tel.: C81 166 68 2420; fax: C81 166 68 2429.
E-mail address: akiraito@asahikawa-med.ac.jp (A. Ito).
Central and South America. These species are distinguish-
able by a number of morphological characteristics of both
adult and larval stages. However, several strains of
E. granulosus, which show substantial genetic diversity,
have been classified into 10 genotypes (G110) (Bowles
et al., 1992, 1995; Bowles and McManus, 1993; Scott et al.,
1997; Lavikainen et al., 2003). Recently, Thompson and
McManus (2003) proposed the following taxonomic revi-
sion; the G1 (sheep strain) genotype is the prototypical
species of E. granulosus but the G4 (horse strain) and G5
(cattle strain) genotypes are distinct species of Echinococ-
cus equinus and Echinococcus ortleppi, respectively. Thus,
the biological entity of sibling or cryptic species should be
considered in the taxonomy of Echinococcus.
Our research group is currently collecting specimens of
E. multilocularis throughout the Holarctic region for a
large-scale genetic study of the species. During the course of
mitochondrial DNA (mtDNA) sequencing, we noticed that

0020-7519/$30.00 q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijpara.2005.01.003
694 N. Xiao et al. / International Journal for Parasitology 35 (2005) 693701

a larval specimen from the plateau pika, Ochotona
curzoniae, in the Qinghai-Tibet plateau region of China
showed a characteristic sequence, which was dissimilar to
any published sequences of Echinococcus spp. The same
sequence was subsequently found in adult specimens from
the Tibetan fox, Vulpes ferrilata. This unknown Echino-
coccus species is distributed sympatrically with E. multi-
locularis and the E. granulosus G1 genotype. Qiu et al.
(1995) have already observed its morphological character-
istics but considered it to be a variant of E. multilocularis.
Taxonomic criteria including morphology, host preference,
molecular genetics and geographical distribution have led
us to describe a new species. In this article, we present the
morphological features of both adult and larval stages and
provide molecular evidence to support the validity of the
new species.
2. Materials and methods
2.1. Parasite samples and morphological observations
From July 2001 to November 2003, larval and adult
specimens of Echinococcus spp. were collected from foxes,
dogs, pikas, voles and sheep in Shiqu County, the Qinghai-
Tibet plateau region of western Sichuan, China (Table 1).
All samples were collected following the local laws for the
preservation of domestic animals and wildlife. Since the
Tibetan foxes were strictly protected from hunting, parasites
were taken from the carcasses killed by attacks of stray
dogs. Tapeworms from canine intestines were relaxed in tap
water and then fixed in 4% formalin. Hydatid tissues from
intermediate hosts were also fixed in 4% formalin. Parts of
both larval and adult samples were stored in 7099%
ethanol for DNA preservation. The formalin-fixed samples
were subjected to morphological observations. The tape-
worms were stained overnight with Delafields haematoxy-
lin, destained with 70% ethanol containing 1% hydrochloric
acid, dehydrated in ethanol, cleared with xylene and
mounted in Canada balsam. Eggs were obtained from
broken gravid segments. To examine rostellar hooks,
Table 1
Origins of Echinococcus samples collected in Shiqu County, the Qinghai-Tibet plateau region of China
tapeworms placed on a glass slide were crushed with
pressure on a coverslip. The hydatid tissues were embedded
in paraffin-wax. Sections (35 mm thick) were stained with
H&E.
2.2. Sequence analysis
DNA was purified from hydatid tissues by using a spin
column kit (DNeasy tissue kit; Qiagen, Germany).
As reported previously (Nakao et al., 2003a), tapeworms
were individually lysed in 10 ml of 0.02 N NaOH at 95 8C
for 10 min. The larval DNA or the adult lysate was used as a
template for polymerase chain reaction (PCR). A DNA
polymerase with 3 0 5 0 exonuclease proofreading activity
(Ex-Taq; Takara Biomedicals, Japan) was used for PCR
amplification. PCR was carried out in a 50 ml reaction
mixture containing 1 ml template, 200 mM of each dNTP,
0.2 mM of each primer, 1 U of Ex-Taq polymerase and the
manufacturer-supplied reaction buffer. Thermal reactions
were performed for 35 cycles of denaturation (94 8C for
30 s), annealing (5456 8C for 30 s) and extension (72 8C
for 6090 s). Primer pairs used for the amplification of
mitochondrial or nuclear DNA regions are shown in Table
2. The PCR products were directly sequenced by using a dye
terminator cycle sequencing kit (DYEnamic ET terminator;
Amersham Biosciences, UK) and an automated sequencer
(ABI PRISM 377; Applied Biosystems, USA).
The mitochondrial genomes of E. multilocularis (database
accession no. AB018440), E. granulosus (AF297617 and
AF346403) and Taenia solium (AB086256) served as
reference sequences (Nakao et al., 2002, 2003b; Le et al.,
2002). Published mtDNA sequences of cox1 (M84661
M84671 and AF525457), nad1 (AJ237632AJ237643) and
atp6 (AY056611AY056615) genes were used for compari-
son (Bowles et al., 1992, 1994; Bowles and McManus, 1993;
McManus et al., 2002; Lavikainen et al., 2003). The elp locus
of an ezrinradixinmoesin (ERM)-like protein (AJ012663)
was used to compare nuclear DNA (Brehm et al., 1999).
Multiple alignments of sequences were achieved by
the Clustal W program (http://www.ddbj.nig.ac.jp). Gaps
and missing data were deleted from the alignments.

Species (developmental stage) Final or intermediate hosts No. samples used for DNA sequencing (no. hosts)a
(no. infected) cob elp
E. shiquicus n. sp. (adult) Fox, Vulpes ferrilata (6) 18 (6) 12 (4)
E. shiquicus n. sp. (larva) Pika, Ochotona curzoniae (5) 5 (5) 4 (4)
E. multilocularis (adult) Fox, Vulpes ferrilata (1) 4 (1) 4 (1)
E. multilocularis (adult) Fox, Vulpes vulpes (1) 4 (1) 4 (1)
E. multilocularis (adult) Dog, Canis familiaris (3) 6 (3) 6 (3)
E. multilocularis (larva) Vole, Microtus fuscus (4) 4 (4) 4 (4)
E. multilocularis (larva) Vole, Pitymys irene (1) 1 (1) 0 (0)
E. granulosus G1 (adult) Dog, Canis familiaris (5) 8 (5) 6 (3)
E. granulosus G1 (larva) Sheep, Ovis aries (1) 1 (1) 1 (1)
a
cob, cytochrome b; elp, ezrinradixinmoesin (ERM)-like protein. The partial nucleotide sequences of mitochondrial cob gene and the intron VII
sequences of nuclear elp locus were determined to confirm the identification of species. In the adult stage, one to four worms per host were used for sequencing.
 N. Xiao et al. / International Journal for Parasitology 35 (2005) 693701
Table 2
Primer pairs used for PCR amplification
Target genesa Sequences (5 0 3 0 ) of primer pairsb
cox1 (mtDNA) F: AGAGAAAATTGTGGAGTTACTGCT
R: ATTACTAATCAACTTAGACTTACA
nad1 (mtDNA) F: TAGTTTAATTAGAATGTCGGTTTG
R: TCTTGAAGTTAACAGCATCACGA
atp6 (mtDNA) F: GCATCAATTTGAAGAGTTGGGGATAAC
R: CCAAATAATCTATCAACTACACAACAC
cob (mtDNA) F1: GTTTAAACTGGTAGATTGTGGTTC
R1: CTCCACAGTAGAAATCACCATCA
F2: GTCAGATGTCTTATTGGGCTGC
R2: TCTGGGTGACACCCACCTAAATA
rrnL (mtDNA) F: ATGCGTTGGATTGATGATTGTAAT
R: AAACAAACTTCATGCAGCCAATG
elp (nuclear DNA) F: ATGCGCGTGAGAGTCTTCAGAAGA
R: ATTCTGCGAAGCTCAGCTTCA
a
cox1, cytochrome c oxidase subunit 1; nad1, NADH dehydrogenase
subunit 1; atp6, ATPase subunit 6; cob, cytochrome b; rrnL, large-subunit
rRNA; elp, ezrinradixinmoesin (ERM)-like protein. Primers were
designed from the mtDNA genome (Nakao et al., 2002) and the elp
exons (Brehm et al., 1999) of Echinococcus multilocularis.
b
Forward (F) and reverse (R) primers. Partial fragments of cob were
amplified and sequenced by using primers F2 and R2 to confirm the
identification of species.
Percentage divergences of nucleotide sequences were cor-
rected by Kimuras two parameter model (Kimura, 1980).
Phylogenetic trees were constructed from the alignments by
using the neighbour-joining method in the MEGA2 software
(Saitou and Nei, 1987; Kumar et al., 2001). All three codon
positions were used to analyse nucleotide sequences.
Confidence values for each branch of the trees were
determined by 1000 bootstrap replications.
3. Results
3.1. Description of adult worm
Adults of Echinococcus shiquicus n. sp. were found only
in Tibetan foxes. During the survey period, six (37.5%) of
16 Tibetan foxes were confirmed to be infected with
E. shiquicus by DNA sequencing. Adult specimens from
two foxes, whose morphological conditions remained
better, were used for observation. As shown in Fig. 1, the
adults containing a gravid segment were divided into two
types. The first type consisted of only pre-mature and gravid
segments (Fig. 1A). Early ovary and testes were formed in
the pre-mature segment but its genital pore was closed. This
unique type constituted the majority of the specimens. The
second type consisted of immature, mature and gravid
segments (Fig. 1B). The number of segments in fully
developed adults did not exceed three. The adults of the
second type (nZ20) were used for the following
description.
695
A B
sc
sc
is
pms
ms
gs
gs
1 mm
Fig. 1. Adults of Echinococcus shiquicus n. sp. in a naturally infected
Tibetan fox. The adults containing a gravid segment were classified into
two types (A and B). gs, gravid segment; is, immature segment; ms, mature
segment; pms, pre-mature segment; sc, scolex.
3.1.1. Description
All measurements are in micrometers, except where
indicated. Length of whole body 1.31.7 (mean 1.5) mm.
Strobila extremely small, with only three segments
(Fig. 2A). Genital pores irregularly alternating. Lateral
osmoregulatory canals running through scolex to gravid
segment. Scolex with four suckers. Suckers oval, 6373
(mean 69) in maximum diameter. Rostellum armed with
tiny hooks. Large hooks 2023 (mean 21, nZ19) long,
small hooks 1617 (mean 17, nZ6) long (Fig. 2B).
Neck absent. Immature segment 80150 (mean 115) long by
160230 (mean 192) wide, genital primordium present.
Mature segment 300475 (mean 386) long by 250350
(mean 285) wide. Genital pore lateral, opened at 1/4 anterior
portion of mature segment. Cirrus pouch pyriform, enclosing
minute cirrus and coiled vas deferens, 120138 (mean 131)
long by 4563 (mean 56) wide, located in anterior portion of
segment extending beyond osmoregulatory canals to midline.
Ovary bilobed, lobes subcircular in dorso-ventral view, 5375
(mean 63) in maximum diameter. Vitelline gland subsphe-
rical, 6883 (mean 77) in maximum diameter, postovarian.
Ovary and vitelline gland located in centre of mature segment.
Testes spherical, 2545 (mean 36) in diameter, 1220 (mean
16) in number, mainly distributed posterior to vitelline gland.
Few testes anterior to genital pore, 02 in number. Gravid
segment 625800 (mean 708) long by 275350 (mean 325)
696 N. Xiao et al. / International Journal for Parasitology 35 (2005) 693701
r
A B 20 m
s
oc
gpr sh
cp
u lh
gpo
o
t vg 30 m
C
e pp
gu
1 mm
Fig. 2. Morphological features of adult Echinococcus shiquicus n. sp. (A)
Fully developed adult. cp, cirrus pouch; e, eggs; gpo, genital pore; gpr,
genital primordium; gu, gravid uterus; o, ovary; oc, osmoregulatory canals;
r, rostellum; s, sucker; t, testes; u, uterus; vg, vitelline gland. (B) Adult
hooks. lh, large hook; sh, small hook. (C) Eggs in gravid uterus.
wide. Genital pore located at 1/3 anterior portion of gravid
segment. Cirrus pouch, vagina and seminal receptacle still
remaining in gravid segment. Gravid uterus branchless, sac-
like, extending to posterior 1/3 of segment. Number of eggs in
gravid uterus 3794 (mean 76). Mature eggs 3440 (mean 38)
in diameter (nZ24, from three worms), containing hexacanth
embryo (Fig. 2C).
We could not describe the exact number of rostellar
hooks because most hooks were lost during relaxation and
fixation. According to the original description of Qiu et al.
(1995), the number is 1834. To measure the length of the
hooks, six worms retaining both large and small hooks were
selected from several hundred worms (see Table 5).
3.2. Morphological features of larva
The metacestode of E. shiquicus was found only in
plateau pikas. DNA sequencing revealed that five (5%) of
101 pikas harboured hydatid cysts of E. shiquicus in their
livers. A pulmonary hydatid cyst was also found in one pika.
Most of the larval forms were unilocular cysts approxi-
mately 10 mm in diameter but one showed an oligovesicular
form (Fig. 3A). The cysts included no daughter cysts. Fully
developed brood capsules containing many protoscoleces
were attached firmly to germinal layers. The protoscoleces
were 125140 mm (mean 128 mm) long by 105125 mm
A
bc
1 cm
B
ps
0.5 mm
Fig. 3. Larval Echinococcus shiquicus n. sp. developed in a plateau pika.
(A) Hepatic hydatid. bc, brood capsule. (B) Cross-section of the hydatid.
p, protrusion; ps, protoscolex.
(mean 117 mm) wide (nZ20, from one cyst). Numbers of
hooks in the protoscolex were 1924 (mean 21, nZ12), and
their length ranged from 16 to 21 mm (mean 18 mm, nZ30,
from five protoscoleces). It was difficult to discriminate
between large and small hooks. Host inflammatory reactions
to cysts appeared minimal. The adventitial layer around
cysts was thin, but the laminated layer of cysts was
relatively broad, being 538 mm in thickness. A protrusion
of the cyst was found, suggesting that exogenous budding
may occur (Fig. 3B).
3.3. Molecular analyses
Since three species of Echinococcus are distributed
sympatrically in the survey area, partial fragments of
mitochondrial cob and nuclear elp were amplified and
sequenced to confirm the identification of species (Table 1).
The length of cob sequenced was 549 bp. Intraspecific
variation of the cob sequence was observed in E. shiquicus
and E. granulosus but not in E. multilocularis (Table 3).
Numbers of variable nucleotide sites were 15 (2.7% of total
length) in E. shiquicus and one (0.2%) in E. granulosus.
Transitional substitutions occurred at all variable sites. The
maximum percentage of divergence reached 1.3 when 23
sequences of E. shiquicus were compared with each other.
 N. Xiao et al. / International Journal for Parasitology 35 (2005) 693701 697

Table 3 to those of E. multilocularis (Nakao et al., 2002) and

Intraspecific variation of mitochondrial cob sequences and pairwise E. granulosus G1 and G4 (Le et al., 2002). Table 4 shows
comparison of the intron VII sequences of nuclear elp locus among three
species of Echinococcus collected in Shiqu County
the pairwise divergence values of nucleotide sequences
between E. shiquicus and other Echinococcus species. The
Species Maximum per- Pairwise divergence (%) values indicated that E. shiquicus is almost equidistant from
cent divergence of elp intronb
of cob within other Echinococcus species regardless of the genes
E. shiquicus E. multilocularis
speciesa examined. Moreover, the values were at interspecific level
E. shiquicus n. sp. 1.3 (23)
when compared with those between E. multilocularis and
E. multilocularis 0 (19) 5.2 E. granulosus G1 genotype. Among the genes examined,
E. granulosus G1 0.2 (9) 4.7 5.3 cox1 showed the minimum divergence values (7.810.6%).
a
The partial nucleotide sequences (549 bp) were determined and the In contrast, the maximum values (18.422.1%) were
maximum values of percentage divergence were compared within the observed in atp6. Of the Echinococcus spp. and genotypes
species. The number of samples examined was shown in parentheses. examined, only 79 (21.6%) of 366 nucleotide sites were
b
Lengths of the intron sequences were 866 bp in Echinococcus variable in cox1, whereas 196 (38.2%) of 513 sites were
shiquicus, 864 bp in Echinococcus multilocularis and 872 bp in
variable in atp6. The extreme bias toward thymine base was
Echinococcus granulosus G1. There were no intraspecific variations in
18 samples of E. multilocularis and seven samples of the E. granulosus G1. observed in the coding strand of all protein genes examined.
In E. shiquicus, one out of 16 samples showed a variation (1 base The thymine contents were 46.248.7% in cox1, 46.1
substitution). 49.9% in nad1, 51.953.4% in atp6 and 47.748.8% in cob.
The phylogenetic trees of Echinococcus were obtained
from the neighbour-joining analysis using nucleotide
The intron VII sequences of nuclear elp locus were
sequences of partial cox1, partial nad1 and complete atp6.
determined in 16 samples of E. shiquicus, 18 samples of
As shown in Fig. 4, the resultant trees depicted that
E. multilocularis and seven samples of E. granulosus.
E. shiquicus, E. multilocularis, E. vogeli, E. oligarthrus,
As shown in Table 3, the intron sequences of E. shiquicus
E. granulosus G1 (ZE. granulosus), E. granulosus G4
were different from those of E. multilocularis and
(ZE. equinus) and E. granulosus G5 (ZE. ortleppi) are
E. granulosus. Similarly, both E. multilocularis and
distantly related to each other. However, the branching
E. granulosus had unique sequences.
patterns of the trees were different from each other. The
The DNA fragments containing complete mitochondrial
phylogenetic positions of these seven species were unclear
genes were amplified from the hydatid tissue of E. shiquicus
because of low bootstrap values in each tree. On the other
and sequenced for phylogenetic analyses. The full lengths of
hand, the genotypes G610 of E. granulosus (camel, pig and
E. shiquicus mitochondrial genes determined in this study
cervid strains) formed a single cluster in the nad1-tree,
were 1608 bp in cox1, 897 bp in nad1, 513 bp in atp6,
suggesting that these genotypes may belong to a single
1068 bp in cob and 985 bp in rrnL. The lengths were similar
species. Phylogenies were reconstructed using deduced

Table 4
Percentage divergences of mitochondrial nucleotide sequences between Echinococcus shiquicus n. sp. and other Echinococcus species

E. shiquicus compared Mitochondrial genesa

with: cox1 nad1 atp6 cob rrnL
E. multilocularis 9.0 16.9 21.5 13.1 14.2
E. oligarthrus 9.0 20.0 21.1 b
E. vogeli 7.8 18.5 19.4
E. granulosus G1 9.4 21.0 22.1 14.2 14.2
E. granulosus G2 8.8 20.1
E. granulosus G3 9.1 20.1
E. granulosus G4 8.1 17.6 18.4 11.1 13.4
E. granulosus G5 9.0 18.1
E. granulosus G6 10.2 18.1 21.1
E. granulosus G7 10.6 17.5 20.8
E. granulosus G8 18.2 20.5
E. granulosus G9 20.3c
E. granulosus G10 9.7 17.5
(9.9)d (19.3) (22.8) (13.6) (11.3)
a
The alignments of cox1 and nad1 were made by using partial sequences, whereas complete sequences were aligned in atp6, cob and rrnL. The numbers of
nucleotide sites examined were 366 in cox1, 442 in nad1, 513 in atp6, 1068 in cob and 970 in rrnL.
b
Sequence data were unavailable in databases.
c
The sequence of Echinococcus granulosus G9 (human isolate GS) was taken from published data (Scott et al., 1997).
d
Percentage divergences between Echinococcus multilocularis and E. granulosus G1 were shown in parentheses.
698 N. Xiao et al. / International Journal for Parasitology 35 (2005) 693701

81 Egra G1 4. Discussion
100 Egra G3
A
37 Egra G2 Apart from the neotropical species E. oligarthrus and
Egra G5 E. vogeli, E. shiquicus n. sp. must be differentiated from
38 96 Egra G6 E. granulosus and E. multilocularis. In the adult stage,
99 Egra G7
100
E. shiquicus is easily distinguishable from E. granulosus by
35 Egra G10
its shorter length, branchless gravid uterus and anterior
0.02 Emul
51 Egra G4 position of genital pore in the gravid segment. As shown in
45 Evog Table 5, E. shiquicus overlapped in most morphological
Eshi features with E. multilocularis reported in China (Zhu et al.,
Eoli 1983; Li et al., 1985; Tang et al., 1988; Wang et al., 1989).
However, smaller rostellar hooks and the upper position of
Tsol
genital pore in the mature segment are characteristic
Egra G1
of E. shiquicus. The strobila of Echinococcus consists of
100 several segments, whose reproductive organs gradually
B Egra G2
82 Egra G3
develop toward the posterior end. In most species of
Egra G4 Echinococcus, the gravid segment is connected to the
45
63 Egra G5 mature segment. However, a strobila consisting of only two
99 Egra G6 segments (a gravid segment directly attaches to a pre-
0.03 98 Egra G7 mature segment) is unique to E. shiquicus (Fig. 1A). Fewer
39 94 Egra G9 eggs in the gravid segment of E. shiquicus (less than 100) is
99 Egra G8
39
also useful for differentiation because E. multilocularis in
85 Egra G10
72 Emul China shows higher fecundity (200400 eggs per gravid
Eshi segment) as reported by Zhu et al. (1983). In the larval
Evog stage, E. shiquicus is quite different from E. multilocularis.
Eoli A unilocular minicyst containing fully developed brood
capsules is typical of E. shiquicus. Unlike E. granulosus, no
Tsol daughter cysts appear within the fertile cyst of E. shiquicus.
The larval development of E. shiquicus in hosts other than
54 Egra G1
plateau pika is unknown. In morphologically questionable
C Egra G4
cases of both adult and larva, the sequencing of mitochon-
50 100 Egra G6
100
drial DNA is recommended for the identification of species.
Egra G7
0.05 The mammalian fauna of the Qinghai-Tibet plateau
Egra G8
consists of elements of the Palaearctic and Oriental realms
41 61 Emul
(Feng et al., 1980). In this region approximately 4000 m
Evog
42 Eoli
above sea level, many wild and domestic mammals
Eshi
including foxes, dogs, voles, pikas, hares, sheep and yaks
are involved in the transmission cycles of Echinococcus
Tsol (Qiu et al., 1995; Xiao et al., 2003, 2004). In this study, we
found that the Tibetan fox V. ferrilata and the plateau pika
Fig. 4. The neighbour-joining phylogenetic trees of Echinococcus. The
trees were constructed from mitochondrial nucleotide sequences of partial O. curzoniae, which are endemic to the plateau, serve as
cox1 (A), partial nad1 (B) and complete atp6 (C). EgraG110, natural hosts for E. shiquicus. A high density of the pika
Echinococcus granulosus genotypes; Emul, Echinococcus multilocularis; (Lai and Smith, 1996) is probably important to maintain the
Eoli, Echinococcus oligarthrus; Eshi, Echinococcus shiquicus n. sp.; Evog, life cycle of E. shiquicus. Both the pika and the fox are
Echinococcus vogeli; Tsol, Taenia solium (an outgroup). Numbers at
adapted to the high altitude steppe but do not survive in
individual nodes are the bootstrap confidence values (%). The scale bars
represent the estimated number of nucleotide substitutions per nucleotide lowlands. Accordingly, we predict that the distribution of
site. E. shiquicus is restricted within the plateau and adjacent
highlands. In contrast, it seems likely that E. granulosus was
amino acid sequences; however, the interspecific relation- recently introduced into the plateau by human activities
ships were also ambiguous (data not shown). Although associated with livestock farming. We also speculate that
mitochondrial rRNA gene is regarded as a good candidate E. multilocularis recently invaded the plateau together with
for the study of deep phylogeny (von Nickisch-Rosenegk et the red fox V. vulpes which has expanded its own niche into
al., 1999), we could not examine the usefulness of rrnL gene the high altitude steppe. The high level of intraspecific
because the sequences of rrnL have been determined only in variation in cob sequences of E. shiquicus supports its
E. shiquicus, E. multilocularis and the E. granulosus G1 and ancient endemism; however, further phylogenetical and
G4 (Table 4). ecological studies are required to verify our speculation.
 N. Xiao et al. / International Journal for Parasitology 35 (2005) 693701
Table 5
Morphological comparison between adult worms of Echinococcus shiquicus n. sp. and Echinococcus multilocularis in China
E. shiquicus E. multilocularis in four localities of China (host)
Sichuan
(Dog)
Body length (mm) 1.31.7 1.33.0
No. of segments 23 45
No. of hooks 1834a 2940
Length of hooks (mm)b
L 2023 2931
S 1617 1626
No. of testes 1220 1529
Position of testes Majority posterior to genital pore. In Nei Mongolia samples, none were located anterior to genital pore.
Position of genital Anterior to the middle of lateral margin. The pore of E. shiquicus was located more anterior than that of E. multilocularis, particularly
pore in the mature segment.
Gravid uterus Branchless and sac-like shape in all samples.
Data cited This study Zhu et al. (1983)
a
Data from Qiu et al. (1995).
b
L, large hook; S, small hook.
Little is known about interspecific mating in parasitic
flatworms under natural conditions. Molecular genetic
evidence for interspecific hybridization has been reported
in the members of the Schistosomatidae (Morgan et al.,
2003), but similar cases have not been found in taeniid
cestodes. Nuclear DNA sequence can serve as a genetic
marker to evaluate the consequence of interspecific
hybridization. The nuclear elp gene in E. multilocularis
represents a single locus, and various species of Echino-
coccus contain its homologues in their genomes (Hemmings
and McManus, 1991; Brehm et al., 1999). Therefore, its
intron VII sequences were compared among the sympatric
species of E. shiquicus, E. multilocularis and E. granulosus.
In examining specimens available to us, there was no
evidence of interspecific hybridization, suggesting that the
three species are reproductively isolated.
The segregating mechanism, which maintains the
genetic identity of these parasites, is unclear. To explain
this mechanism, we present the following two hypoth-
eses. The first is an ecological isolation, which is
associated with the predatorprey relationship of host
mammals and their susceptibility to the parasites. In the
Qinghai-Tibet plateau, domestic dogs and sheep are
involved in the life cycle of E. granulosus. On the other
hand, wild animals are natural hosts for E. multilocularis
and E. shiquicus. Rodents of the Arvicolidae most
commonly serve as intermediate hosts for E. multi-
locularis. If red foxes mainly hunt rodents whereas
Tibetan foxes show a particular preference for pikas,
E. multilocularis and E. shiquicus might acquire their
own niches. A dietary analysis of canines in the plateau
is necessary to understand the transmission dynamics of
Echinococcus spp. However, the segregating mechanism
cannot be explained from the ecological aspect alone.
The second hypothesis is a physiological isolation
concerning the reproduction of parasites. On the plateau,
699
Ningxia Xinjiang Nei Mongolia
(Red fox) (Wolf) (Corsac fox)
1.12.4 1.31.7 1.83.3
25 37 34
3032 2430 2830
2832 28 2627
2026 23 2022
1622 1216 1420
Li et al. (1985) Wang et al. (1989) Tang et al. (1988)
there are no documented records of canines concurrently
infected with different Echinococcus species. However,
we assume that mixed infections might sometimes occur.
Both male and female reproductive organs share a
common genital pore in the mature segment of
Echinococcus. Therefore, the parasite has the potential
for both cross- and self-insemination. Based on morpho-
logical observations, Kumaratilake et al. (1986) suggested
that the self-insemination by inserting cirrus into the
adjacent vagina is common in E. granulosus but is rare
in E. multilocularis. The shorter cirrus and lack of
vaginal sphincter in E. multilocularis are probable causes
of the rarity. A recent population genetic study supported
the hypothesis that cross-insemination occurs in
E. multilocularis (Nakao et al., 2003a). The frequency
of self-insemination in E. shiquicus is unknown. If self-
insemination predominates, E. shiquicus could retain its
genetic identity even though mixed infections occur in a
fox. We also assume that gamete incompatibility and
hybrid inviability may be responsible for preventing the
crossing between different Echinococcus species.
Shiqu County, located in the Qinghai-Tibet plateau
region, is a highly endemic area of human echinococcosis.
An epidemiological survey using ultrasonography, X-ray
and serological tests estimated that 97 (7.8%) of 1249
residents in three townships were infected with Echinococ-
cus (Qiu et al., 2000). Among them, 60 were diagnosed as
cystic echinococcosis and 37 as alveolar echinococcosis.
However, the diagnoses were not confirmed by inspecting
surgically removed lesions or biopsy samples (Li et al.,
2004). The Tibetan people of Shiqu who live in high altitude
steppe are in close contact with canines. Further studies are
required to examine the possibility of human infections with
E. shiquicus.
In this study, the sequence data of mitochondrial DNA
were especially useful in demonstrating the validity of
700 N. Xiao et al. / International Journal for Parasitology 35 (2005) 693701
E. shiquicus. However, the phylogenetic trees deduced from
sequences of cox1, nad1 and atp6 were insufficient to
resolve comprehensive relationships among various species
of Echinococcus. The ambiguity of the trees is probably due
to several factors, such as the short length of sequences
examined (366 bp for cox1, 442 bp for nad1 and 513 bp for
atp6), the strong mutational bias toward thymine and the
saturation of nucleotide substitutions. To infer an exact
phylogeny of Echinococcus, the DNA sequencing of
mitochondrial genomes and nuclear rRNA genes is required
in various species.
Recently, Tang et al. (2004) reported that a variant of
E. multilocularis in Inner Mongolia of China should be
regarded as a new species. In their report, the subspecies
name of Echinococcus multilocularis sibiricensis was used
for the variant. The lengths of its rostellar hooks were 2627
mm (large) and 2022 mm (small) and its hydatid cysts in
voles and mice showed an alveolar form. Considering these
morphological features, the variant may be unrelated to
E. shiquicus.
5. Taxonomic summary
5.1. Echinococcus shiquicus n. sp.
Type host: Tibetan fox, Vulpes ferrilata.
Site of infection: The lower part of small intestine
(mainly in the ileum). The number of worms ranges from
hundreds to ten thousands.
Type locality: Shiqu County, the Qinghai-Tibet plateau
region of western Sichuan, China.
Type specimens: The type series consists of fully
developed adult specimens. Holotype (slide no.
ScCDCPTE001) and nine paratypes (ScCDCPTE002-
010) are kept in Institute of Parasitic Diseases, Sichuan
Center for Disease Control and Prevention, Chengdu,
Sichuan, China.
Intermediate host: Plateau pika, Ochotona curzoniae.
The metacestode develops into unilocular cyst mainly in
the liver.
Etymology: The new species is named after its locality of
occurrence.
Acknowledgements
The authors thank Patrick Giraudoux, Francis Raoul
(Universite de Franche-Comte, France), Xiaoming Wang
(East China Normal University, China) and Kenichi
Takahashi (Hokkaido Institute of Public Health, Japan) for
their help in sample collection. This investigation was
supported by the National Institute of Health (1R01
TW01565-01; Principal Investigator, Philip S. Craig), the
Thrasher Fund to Peter M. Schantz and a Grant-in-Aid from
the Japan Society of Promotion of Science to Akira Ito
(14256001).
References
Bowles, J., McManus, D.P., 1993. NADH dehydrogenase 1 gene sequences
compared for species and strains of the genus Echinococcus. Int.
J. Parasitol. 23, 969972.
Bowles, J., Blair, D., McManus, D.P., 1992. Genetic variants within the
genus Echinococcus identified mitochondrial DNA sequencing. Mol.
Biochem. Parasitol. 54, 165174.
Bowles, J., Blair, D., McManus, D.P., 1994. Molecular genetic character-
ization of the cervid strain (northern form) of Echinococcus
granulosus. Parasitology 109, 215221.
Bowles, J., Blair, D., McManus, D.P., 1995. A molecular phylogeny of the
genus Echinococcus. Parasitology 110, 317328.
Brehm, K., Jensen, K., Frosch, P., Frosch, M., 1999. Characterization of the
genomic locus expressing the ERM-like protein of Echinococcus
multilocularis. Mol. Biochem. Parasitol. 100, 147152.
Feng, Z., Zheng, C., Cai, G., 1980. Mammals from southeastern Xizang (in
Chinese). Acta Zool. Sinica 26, 9197.
Hemmings, L., McManus, D.P., 1991. The diagnostic value and molecular
characterisation of an Echinococcus multilocularis antigen gene clone.
Mol. Biochem. Parasitol. 44, 5362.
Kimura, M., 1980. A simple method for estimating evolutionary rates of
base substitutions through comparative studies of nucleotide sequences.
J. Mol. Evol. 16, 111120.
Kumar, S., Tamura, K., Jakobsen, I.B., Nei, M., 2001. MEGA2: molecular
evolutionary genetics analysis software. Bioinformatics 17, 12441245.
Kumaratilake, L.M., Thompson, R.C.A., 1982. A review of the taxonomy
and speciation of the genus Echinococcus Rudolphi 1801. Z.
Parasitenkd. 68, 121146.
Kumaratilake, L.M., Thompson, R.C.A., Eckert, J., DAlessandro, A.,
1986. Sperm transfer in Echinococcus (Cestoda: Taeniidae). Z.
Parasitenkd. 72, 265269.
Lai, C.H., Smith, A.T., 1996. Keystone status of plateau pikas (Ochotona
curzoniae): effect of control on biodiversity of native birds, in:
Peter, J.S., Wang, S., Xie, Y. (Eds.), Conserving Chinas Biodiversity
(II). China Environmental Science Press, Beijing, pp. 222230.
Lavikainen, A., Lehtinen, M.J., Meri, T., Hirvela-Koski, V., Meri, S., 2003.
Molecular genetic characterization of the Fennoscandian cervid strain, a
new genotypic group (G10) of Echinococcus granulosus. Parasitology
127, 207215.
Le, T.H., Pearson, M.S., Blair, D., Dai, N., Zhang, L.H., McManus, D.P.,
2002. Complete mitochondrial genomes confirm the distinctiveness of
the horse-dog and sheep-dog strains of Echinococcus granulosus.
Parasitology 124, 97112.
Li, W.X., Zhang, G.C., Lin, Y.G., Hong, L.X., 1985. The occurrence of
Echinococcus multilocularis Leuckart, 1863 the natural animal host in
China and its morphological study (in Chinese). Acta Zool. Sinica 31,
365371.
Li, T.Y., Qiu, J.M., Craig, P.S., Ito, A., Yang, W., Vuitton, D.A., Xiao, N.,
Chen, X.W., Yu, W., Schantz, P.M., 2004. Review of 311 cases of
alveolar echinococcosis and criteria for classification of hepatic
ultrasound images. Southeast Asian J. Trop. Med. Public Health 35,
213217.
McManus, D.P., Zhang, L., Castrodale, L.J., Le, T.H., Pearson, M.,
Blair, D., 2002. Molecular genetic characterization of an unusually
severe case of hydatid disease in Alaska caused by the cervid
strain of Echinococcus granulosus. Am. J. Trop. Med. Hyg. 67,
296298.
Morgan, J.A., DeJong, R.J., Lwambo, N.J., Mungai, B.N., Mkoji, G.M.,
Loker, E.S., 2003. First report of a natural hybrid between Schistosoma
mansoni and S. rodhaini. J. Parasitol. 89, 416418.
 N. Xiao et al. / International Journal for Parasitology 35 (2005) 693701
Nakao, M., Yokoyama, N., Sako, Y., Fukunaga, M., Ito, A., 2002. The
complete mitochondrial DNA sequence of the cestode Echinococcus
multilocularis (Cyclophyllidea: Taeniidae). Mitochondrion 1,
497509.
Nakao, M., Sako, Y., Ito, A., 2003a. Isolation of polymorphic microsatellite
loci from the tapeworm Echinococcus multilocularis. Infect. Genet.
Evol. 3, 159163.
Nakao, M., Sako, Y., Ito, A., 2003b. The mitochondrial genome of the
tapeworm Taenia solium: a finding of the abbreviated stop codon U.
J. Parasitol. 89, 633635.
Qiu, J.M., Chen, X.W., Ren, M., Luo, C.X., Liu, D.L., Liu, X.T., He, D.L.,
1995. Epidemiological study on alveolar hydatid disease in Qinghai-
Xizang plateau (in Chinese). J. Pract. Parasit. Dis. 3, 106109.
Qiu, J.M., Liu, F.J., Schantz, M.P., Ito, A., Delker, C., He, J.G., Zhang, Y.,
Chen, X.W., 2000. Epidemiological study on human hydatidosis in
Tibetan region of western Sichuan (in Chinese). Chin. J. Zoonoses 16,
7780.
Rausch, R.L., Bernstein, J.J., 1972. Echinococcus vogeli sp. n (Cestoda:
Taeniidae) from the bush dog, Speothos venaticus (Lund). Z.
Tropenmed. Parasitol. 23, 2534.
Saitou, N., Nei, M., 1987. The neighbor-joining method: a new method for
reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406425.
Scott, J.C., Stefaniak, J., Pawlowski, Z.S., McManus, D.P., 1997.
Molecular genetic analysis of human cystic hydatid cases from Poland:
identification of a new genotypic group (G9) of Echinococcus
granulosus. Parasitology 114, 3743.
701
Tang, C.T., Cui, G.W., Qian, Y.C., Lu, S.M., Lu, H.C., 1988. On the
occurrence of Echinococcus multilocularis in Hulunbeier Pasture, Nei
Mongolian Autonomous Region (in Chinese). Acta Zool. Sinica 34,
172179.
Tang, C.T., Quian, Y.C., Kang, Y.M., Cui, G.W., Lu, H.C., Shu, L.M.,
Wang, Y.H., Tang, L., 2004. Study on the ecological distribution of
alveolar Echinococcus in Hulunbeier Pasture of Inner Mongolia, China.
Parasitology 128, 187194.
Thompson, R.C.A., McManus, D.P., 2003. Towards a taxonomic revision
of the genus Echinococcus. Trends Parasitol. 18, 452457.
von Nickisch-Rosenegk, M., Lucius, R., Loos-Frank, B., 1999. Contri-
butions to the phylogeny of the Cyclophyllidea (Cestoda) inferred from
mitochondrial 12S rDNA. J. Mol. Evol. 48, 586596.
Wang, W., Ding, Z.X., Wu, Y., Wu, J.G., 1989. The occurrence of
Echinococcus multilocularis Leuckart, 1863 in fox and wolf in
Tacheng, Xinjiang (in Chinese). Endemic Dis. Bull. 4, 811.
Xiao, N., Qiu, J.M., Nakao, M., Nakaya, K., Yamasaki, H., Sako, Y.,
Mamuti, W., Schantz, P.M., Craig, P.S., Ito, A., 2003. Identification of
Echinococcus species from a yak in the Qinghai-Tibet plateau region of
China. Am. J. Trop. Med. Hyg. 69, 445446.
Xiao, N., Li, T.Y., Qiu, J.M., Nakao, M., Chen, X.W., Nakaya, K.,
Yamasaki, H., Schantz, P.M., Craig, P.S., Ito, A., 2004. The Tibetan
hare Lepus oiostolus: a novel intermediate host for Echinococcus
multilocularis. Parasitol. Res. 92, 352353.
Zhu, Y.B., Qiu, J.M., Qiu, D.C., Yi, D.Y., 1983. The discovery of
Echinococcus multilocularis in China (in Chinese). Sichuan J. Zool. 2, 44.