Arch Virol
DOI 10.1007/s00705-016-3206-z
ANNOTATED SEQUENCE RECORD
Sequence errors in foamy virus sequences in the GenBank
database: resequencing of the prototypic foamy virus proviral
plasmids
Tobias C. Wagner1 Jochen Bodem1
Received: 10 November 2016 / Accepted: 18 December 2016
Springer-Verlag Wien 2016
Abstract Nucleotide sequences are the fundamental basis
for work on molecular mechanisms and for phylogenetic
analysis. Recently, we identified sequence errors in all of
the LTR sequences of the prototypic foamy virus stored in
the GenBank database. Here, we report the resequencing of
the proviral plasmids pHSRV13 and pHSRV2. Sequence
comparisons revealed an error rate for the foamy virus
sequences stored in the database of up to 10 errors per 1000
bp. Even the newest sequences of the codon-optimized
foamy virus synthetic Gag, Pol, and Env amino acid
sequences showed exchanges compared to the new proviral
pHSRV13n sequence. Our results provide evidence that
some prototypic foamy virus sequences contain errors and
should be revised.
Introduction
Foamy viruses were already described as a cytopathic
agent in cell cultures in the year 1954 [1], and they were
later isolated from a nasopharyngeal carcinoma of an
African patient [2]. This isolate previously called human
foamy virus is the best-analysed foamy virus so far and
represents the prototypic foamy virus (PFV). The genome
of PFV was cloned, and the first PFV sequences were
Electronic supplementary material The online version of this
article (doi:10.1007/s00705-016-3206-z) contains supplementary
material, which is available to authorized users.
& Jochen Bodem
Jochen.Bodem@vim.uni-wuerzbuerg.de
1
Institut fur Virologie und Immubiologie, Julius-Maximilians-
Universitat, Wurzburg, Germany
published in 1987 [35]. Finally, two infectious proviral
plasmids, pHSRV2 [6] and pHSRV13 [7], were generated
from the same parental plasmids. These proviral sequences
have been the basis for molecular and phylogenetic studies
up to now. The first foamy virus sequences were obtained
by the Maxam-Gilbert sequencing method, while in the
eighties and early nineties, newer PFV nucleotide sequence
data were obtained by radioactive Sanger sequencing and
recording the sequence by hand. Especially long GC-rich
regions and repetitive sequences due to DNA/RNA struc-
tures might have led to sequence readout errors from urea
polyacrylamide sequencing gels.
Recently, we analysed the regulation of foamy virus
polyadenylation [8]. During these studies, we discovered
that none of the PFV sequences deposited in the GenBank
database matched the true nucleotide sequences of the two
proviral plasmids. In fact, two nucleotides were missing in
the R region (193 nt) of all HFV/PFV sequences in the
GenBank database. Furthermore, the sequences of the 5
and 3 LTRs stored in the database were not identical, and
sequence data directly obtained from the Rethwilm
(pHSRV2) and Flugel/Lochelt (pHSRV13) laboratories
differed from the database sequences as well. Based on
these inconsistencies, we decided to resequence both
proviral plasmids.
Sequence
To characterize defined plasmid stocks, we used a
pHRSV13 plasmid preparation received from Martin
Lochelt and a pHSRV2 plasmid preparation used in a
published study by the Rethwilm laboratory in 2008 [9]. In
order to ensure overlapping sequence data, sequencing
primers were located every 500 bp. Furthermore, all
123
 T. C. Wagner, J. Bodem

deviations from the original sequences were verified by
resequencing. Additional LTR sequences were determined
separately [8]. The obtained sequence data were aligned,
and consensus sequences for pHSRV2 and pHSRV13 were
generated using the SeqMan software package version
9.0.4 (DNASTAR, Madison, Wisconsin USA). The auto-
mated sequence readout, in addition to large sequence
overlaps and computerized sequence assembly, led to high-
quality nucleotide sequences. Furthermore, to insure
sequence quality, trace files containing any sequence
deviation were inspected in detail.
The resulting new sequences were named pHSRV2n und
pHSRV13n, respectively, and submitted to GenBank (ac-
cession number KX087159). These sequences were com-
pared to the human foamy virus and PFV sequences stored
in GenBank and to the pHSRV2 sequences obtained
directly from the Rethwilm laboratory, using the matcher
program of the EMBOSS software suite [10] (Table 1). For
clarity and consistency with older reports, we maintained
the abbreviation pHSRV2 for the sequence comparisons
and used the GenBank accession numbers for GenBank-
derived sequence data (Table 1 and Table S1).
These sequence comparisons revealed that the previous
Gag amino acid sequences contained two or three sequence
errors (Table 1 and Table S1). Recently, codon-optimized
expression plasmids containing HFV/PFV gag, pol, and
env were generated and patented [11]. Unfortunately, even
this synthetic codon-optimized Gag encoded one amino
acid exchange in Gag (D60N) compared to the pHSRV13n
sequence.
Furthermore, sequence comparisons showed severe
errors in some of the previously published pol sequen-
ces, which resulted in two frameshifts, leading to a
stretch of 14 amino acid residues in a different reading
frame (Tables 1 and S1). Surprisingly, these errors are
present in the earliest sequence records (accession
number HSU21247) and in sequences submitted in 2005
(accession numbers Y07723 and Y07725). However, this
part of the sequence was corrected in the synthetic,
codon-optimized pol expression plasmid [11, 12]. Nev-
ertheless, one error in the synthetic Pol amino acid
sequence was detected in the coding region of the inte-
grase (Q1141K).
A comparison of the coding sequence of pHSRV2n and
pHSRV13n showed an amino acid polymorphism located
in the Env protein (I72M) and one in the Bel2 protein
(Table 1). Unfortunately, the synthetic env expression
plasmid encodes one amino acid exchange in the Env

Table 1 Comparison of foamy virus sequences: nucleotide and amino acid exchanges in comparison to pHSRV13n
GenBank acc. no. Region (length in nucleotides/amino acid residues) Ref.
P
LTR 50 /30 Gag Pol Env Tas Bel2 exchanges
(1123/-) (1947/649) (3432/1144) (2967/989) (903/301) (1071/357) (nt/[nt/1kbp])

pHSRV2n* 5: 0; 3:1 0/0 0/0 3/1 0/0 1/1 5/0.40

KX087159
pHSRV2* 5: 72 3:74 3/2 72/14 8/3 2/1 6/3 40/3.18
HSU21247 5: 72; 3:73 4/3 72/14 111/41 1/0 8/5 45/3.58 [3, 4, 7, 15]
Y07723* 5:122,6; 3:82,6 3/2 72/14 8/3 2/1 6/3 46/3.66 [16]
Y07724 5: 72; 3:74 3/2 72/14 8/3 2/1 6/3 40/3.18 [16]
2,7 2,7 2
Y07725* 5: 12 ; 3:8 3/2 7 /14 8/3 2/1 6/3 46/3.66 [16]
M19427 5:7; - 73/3 -/- -/- -/- -/- 14/4.56 [15]
EU381420 - -/- -/- -/- -/- 4/2 4/3.73 [17]
X05591 - -/- -/- 111/41 -/- -/- 11/3.71 [4]
X05592 i1366 -/- -/- -/- -/- 12/8 22/10.03 [3]
Codon opt. - -/1 -/1 -/2 (15) -/- -/- [11, 12]
1
Insertion of three nucleotides/one amino acid
2
Deletion of two nucleotides
3
One missing nucleotide
4
Three missing nucleotides
5
Differences in comparison to the amino acid sequence of pHSRV2n Env
6
Insertion of 136 nucleotides
7
Insertion of 645 nucleotides
* Dissimilarity in 5- and 3-LTR

123
Resequencing of foamy virus clones
leader peptide (N14K) compared to the new sequence of
the proviral plasmids.
The error frequency of GenBank sequences has been
determined to be 2.887 errors per 1000 bases [13], and it
was later found to be one error per 1000 bp in the mouse
genome [14]. The frequency of errors in the foamy virus
sequences with up to 10.03 errors per 1000 bp, even
without taking any insertions into account, was found to be
more than three times higher than the described error rates.
The described sequence deviations, especially in the ear-
liest human foamy virus sequences, might have resulted
from naturally occurring polymorphisms, cell culture
adaption, errors during reverse transcription, or technical
problems during sequence determination. Errors in previ-
ous sequences of pHSRV2 and pHSRV13 can be attributed
to the latter, since the plasmid stocks were derived from the
original sources, and thus the sequences that were obtained
should have matched those in GenBank. Some errors, like
the missing two nucleotides in the R region of the LTR can
be explained by the history of both proviral plasmids. The
prototypic foamy virus was first cloned in the Flugel lab-
oratory [3, 4, 15], and the resulting plasmids were used to
generate the infectious pHSRV13 plasmid. The missing
dinucleotide within the 5LTR was present in the first
foamy viral sequences of the 3 LTR [3, 4], but it was not
recorded in the first 5LTR sequence [15]. It is likely that
the LTRs were not resequenced and verified in pHSRV13,
since it was assumed that the sequences of the parental
plasmids had been determined correctly. Therefore, the
missing dinucleotide remained absent in the pHSRV13
5LTR nucleotide sequence. The pHSRV2 plasmid was
constructed by the Rethwilm laboratory using the very
same parental plasmids produced in the Flugel laboratory
[6], and again, the sequences were not verified after the
construction of the plasmid.
Differences in the cloning strategies might explain
deviations between the proviral plasmids, since Lochelt
et al. mainly used the parental plasmid C11, whereas the
env sequence of pHSRV2 is based on the C55 plasmid
[37]. However, the cloning strategy of pHSRV2 docu-
ments the exchange of an XbaI site for an Asp718 site in
env [6]. Unfortunately, this Asp718 site is not present in
any of the sequences analysed, and our sequence data show
that the parental XbaI site is still present in the pHSRV2n
sequence.
In summary, we provide evidence that the PFV
sequences available in the GenBank database contain
several sequencing errors and should not be used in the
future. Furthermore, we would like to suggest that labo-
ratories use the pHSRV13n sequence and plasmid to ensure
that experiments and data will be comparable within the
foamy virus field.
Acknowledgements We would like to thank M. Lochelt for the
pHSRV13 plasmid.
Compliance with ethical standards
Ethical approval This article does not contain any studies with
human participants or animals performed by any of the authors.
Conflict of interest The authors declare that they have no conflict of
interest.
References
1. Enders JF, Peebles TC (1954) Propagation in tissue cultures of
cytopathogenic agents from patients with measles. Proc Soc Exp
Biol Med 86(2):277286
2. Achong BG, Mansell PW, Epstein MA (1971) A new human
virus in cultures from a nasopharyngeal carcinoma. J Pathol
103(2):P18
3. Flugel RM, Maurer B, Bannert H, Rethwilm A, Schnitzler P,
Darai G (1987) Nucleotide sequence analysis of a cloned DNA
fragment from human cells reveals homology to retrotransposons.
Mol Cell Biol 7(1):231236
4. Flugel RM, Rethwilm A, Maurer B, Darai G (1987) Nucleotide
sequence analysis of the env gene and its flanking regions of the
human spumaretrovirus reveals two novel genes. Embo J
6(7):20772084
5. Rethwilm A, Darai G, Rosen A, Maurer B, Flugel RM (1987)
Molecular cloning of the genome of human spumaretrovirus.
Gene 59(1):1928
6. Rethwilm A, Baunach G, Netzer KO, Maurer B, Borisch B, ter
Meulen V (1990) Infectious DNA of the human spumaretrovirus.
Nucleic Acids Res 18(4):733738
7. Lochelt M, Zentgraf H, Flugel RM (1991) Construction of an
infectious DNA clone of the full-length human spumaretrovirus
genome and mutagenesis of the bel 1 gene. Virology
184(1):4354
8. Schrom E-M, Moschall R, Weitner H, Fecher D, Langemeier J,
Bohne J, Wohrl BM, Bodem J (2013) U1snRNP-mediated sup-
pression of polyadenylation in conjunction with the RNA struc-
ture controls poly(A) site selection in Foamy Viruses.
Retrovirology 10:55
9. Peters K, Barg N, Gartner K, Rethwilm A (2008) Complex
effects of foamy virus central purine-rich regions on viral repli-
cation. Virology 373(1):5160. doi:10.1016/j.virol.2007.10.037
10. McWilliam H, Li W, Uludag M, Squizzato S, Park YM, Buso N,
Cowley AP, Lopez R (2013) Analysis Tool Web Services from
the EMBL-EBI. Nucleic Acids Res 41 (Web Server
issue):W597W600. doi:10.1093/nar/gkt376
11. Lindemann D, Rethwilm A (2011) Foamy virus biology and its
application for vector development. Viruses 3(5):561585.
doi:10.3390/v3050561
12. Hartl MJ, Bodem J, Jochheim F, Rethwilm A, Rosch P, Wohrl
BM (2011) Regulation of foamy virus protease activity by viral
RNAa novel and unique mechanism among retroviruses.
J Virol 85(9):44624469
13. Krawetz SA (1989) Sequence errors described in GenBank: a
means to determine the accuracy of DNA sequence interpretation.
Nucleic Acids Res 17(10):39513957
14. Wesche PL, Gaffney DJ, Keightley PD (2004) DNA sequence
error rates in Genbank records estimated using the mouse genome
as a reference. DNA Seq 15(56):362364. doi:10.1080/
10425170400008972
123

15. Maurer B, Bannert H, Darai G, Flugel RM (1988) Analysis of the
primary structure of the long terminal repeat and the gag and pol
genes of the human spumaretrovirus. J Virol 62(5):15901597
16. Schmidt M, Herchenroder O, Heeney J, Rethwilm A (1997) Long
terminal repeat U3 length polymorphism of human foamy virus.
Virology 230(2):167178. doi:10.1006/viro.1997.8463
123
T. C. Wagner, J. Bodem
17. Perkovic M, Schmidt S, Marino D, Russell RA, Stauch B, Hof-
mann H, Kopietz F, Kloke BP, Zielonka J, Strover H, Hermle J,
Lindemann D, Pathak VK, Schneider G, Lochelt M, Cichutek K,
Munk C (2009) Species-specific inhibition of APOBEC3C by the
prototype foamy virus protein bet. J Biol Chem 284(9):
58195826. doi:10.1074/jbc.M808853200