Plant Cell Rep (2003) 21:511516

DOI 10.1007/s00299-002-0557-6

CELL BIOLOGY AND MORPHOGENESIS

A. Assani F. Bakry F. Kerbellec R. Hacour

G. Wenzel B. Foroughi-Wehr

Production of haploids from anther culture

of banana [Musa balbisiana (BB)]
Received: 17 May 2002 / Revised: 15 October 2002 / Accepted: 19 October 2002 / Published online: 17 December 2002
Springer-Verlag 2002

Abstract We report here, for the first time, the production of haploid plants of banana Musa balbisiana (BB).
Callus was induced from anthers in which the majority
of the microspores were at the uninucleate stage. The
frequency of callus induction was 77%. Callus proliferation usually preceded embryo formation. About 8% of
the anthers developed androgenic embryos. Of the 147
plantlets obtained, 41 were haploids (n=x=11). The frequency of haploid production depended on genotypes
used: 18 haploid plants were produced from genotype
Pisang klutuk, 12 from Pisang batu, seven from Pisang
klutuk wulung and four from Tani. The frequency of regeneration was 1.1%, which was based on the total number of anthers cultured. Diploid plants (2n=2x=22) were
also observed in the regenerated plants. The haploid
banana plants that were developed will be important
Communicated by H. Lrz
A. Assani () R. Hacour
Universit de Paris Sud XI,
Laboratoire de Morphogense Vgtale Exprimentale,
Btiment 360, 91405 Orsay Cedex, France
e-mail: akym.assani@u-bourgogne.fr
Tel.: +33-03-80396654, Fax: +33-03-80396611
F. Bakry F. Kerbellec
CIRAD-FLHOR,
Avenue du Val Montferrand,
BP 5035, 34032 Montpellier Cedex, France
G. Wenzel
Lehrstuhl fr Pflanzenbau und Pflanzenzchtung,
Technische Universitt Mnchen,
85350 Freising-Weihenstephan, Germany
B. Foroughi-Wehr
Bundesanstalt fr Zchtungsforschung an Kulturpflanzen,
Institut fr Resistenzgenetik,
Graf-Seinsheim-Strasse 23, 85461 Grnbach, Germany
A. Assani
Ecole Nationale Suprieure de Biologie Applique la Nutrition
et lAlimentation,
Laboratoire de Gnie des Procds Alimentaires
et Biotechnologiques 1,
Esplanade Erasme, 21000 Dijon, France

material for the improvement of banana through breeding programmes.

Keywords Musa balbisiana Anther culture
Androgenesis Haploids Plant regeneration
Abbreviations BAP: Benzylaminopurine
CIRAD: Centre de coopration internationale en
recherche agronomique pour le dveloppement
DAPI: 4,6-Diamidino-2-phenylindole
IAA: Indole-3-acetic acid
MOPS: 3-(Nmorpholino)propanesulfonic acid

Introduction
Haploids can result through natural parthenogenesis
(development of haploid plants from unfertilised eggs)
or androgenesis (development of haploid sporophytes
from pollen). They can also be artificially induced
through the culture of ovaries (Muren 1989), ovules
(Hansen et al.1994), anthers (Foroughi-Wehr et al.
1982), microspores (Khler and Wenzel 1985) and sexual hybridisation [e.g. cross between cultivated barley and
the wild species (Kasha and Kao 1970)]. However, anther or microspore cultures have been found to be the
most efficient techniques for obtaining a large number of
haploid plants (De Buyser and Henry 1980). The regenerated haploid plants are generally sterile, requiring
chromosome doubling for use in breeding programmes.
Chromosomes can be doubled either spontaneously or
artificially, and haploid plantlets are usually treated with
colchicine as a means of inducing chromosome doubling
(Foroughi-Wehr and Friedt 1984).
Whereas many years are needed to obtain inbreds
by conventional breeding methods, homozygous lines
can be produced in only 1 year by chromosome doubling of haploid plantlets obtained by anther culture
(Foroughi-Wehr et al. 1982; Foroughi-Wehr and Wenzel
1989). The inbred line produced by anther culture has
fixed genotypes so that continued selection for an inter-

512

esting gene combination is not required (Wan and

Widholm 1993).
Since a reproducible method for producing haploid
plants is available in banana, it may be used to produce
haploid plants which could be used in conventional
breeding programmes for banana improvement. Further,
the establishment of haploid cell cultures may be useful
in somatic fusion procedures (unconventional breeding)
to obtain directly triploid cells starting from diploid and
haploid cells. Kerbellec (1996) was first to report
successful haploid plant regeneration in banana Musa
acuminata (AA). However, to date, no report is available
on haploid plant production in M. balbisiana (BB).
In the present investigation, we report, for the first
time, the successful regeneration of haploid plants
(x=n=11) of four genotypes of the species M. balbisiana
(BB), which is known to carry resistant genes against
economically important banana diseases.

Materials and methods

Plant materials
The plants, diploid (2n=2x=22) Pisang klutuk (BB), Pisang klutuk
wulung (BB), Pisang batu (BB) and Tani (BB), all derived from
the diploid wild species Musa balbisiana (BB), which contributes
the B genome, were provided by CIRAD-Guadeloupe.

Fig 1 A Male flower bud of banana (adapted from Champion

1963). Bar 6 cm. B Male flower of banana (adapted from Champion
1963). Bar 2 cm

Methods

Plant regeneration

Anther cultures

Androgenic embryos were transferred onto regeneration medium

containing MS macro- and micro-nutrients, vitamins of Morel,
88 mM sucrose, 2.2 M BAP, 2.3 M IAA, 7.5 g l-1 agarose sea
plaque (Sigma) (pH 5.7). The cultures were kept at 27C in the
dark. Plant regeneration occurred 8 months after the anthers had
been plated. Regenerated plantlets were transferred to solidified
growth regulator-free MS medium with 1.2 mM NH4NO3 (pH 5.7)
and cultured at 27C under a 16/8-h (day/night) photoperiod (light
intensity: 65 mol m-2 s-1).

The anthers were isolated according to procedure described by

Kerbellec (1996), with some modifications. The male flower bud
(Fig. 1A), which contains male flowers in all developmental stages (immature and mature flowers), was used as donor material for
anther isolation. For sterilisation, male flower buds 1015 cm in
length were immersed in 95% (v/v) ethanol, followed by flaming.
This sterilisation procedure was repeated three times. The bract
was separated from the immature male flowers (Fig. 1B) and eliminated. An immature male flower (56 cm), which contained five
stamens with fully developed anthers was isolated and transferred
into a petri dish (diameter: 9.5 cm) containing filter paper that had
been moistened with sterile water. The surrounding tepals (compound and free tepals) and the ovary on which the anthers were
fixed were eliminated. One anther from each flower was squashed
in 2.8% NaH2PO4 buffer or 1% aceto-carmine in order to determine the appropriate microspore developmental stage, i.e. uninucleate stage (Fig 2A). Banana anthers are relatively large
1.01.5 cm so that their isolation and transfer onto culture medium can be done with the naked eye. Ten anthers (each approx.
1 cm in length) containing microspores at the uninucleate stage
were placed in a petri dish (diameter: 9.5 cm) containing 30 ml
medium. The solid culture medium contained MS macro- and
micro-nutrients (Murashige and Skoog1962), vitamins of Morel
(Morel and Wetmore1951), 500 mg l-1 casein hydrolysate, 73 mM
sucrose, 4.4 M BAP, 2.3 M IAA and 6 g l-1 agarose (type II;
Sigma, St. Louis, Mo.). The pH of the medium was adjusted to 5.7
before autoclaving. Cultures were kept at 27C in darkness.
Anthers were maintained on the same culture medium without
subculture until callus formation, which occurred after 4 months
on the culture medium. Androgenic embryos were formed
6 months after the anthers had been plated.

Determination of ploidy level

The flow cytometric technique was used to determine the ploidy
level of the anther-derived plants. Leaf pieces (11 cm) from in
vitro plants derived from anther cultures were chopped with a
razor blade in 600 l buffer solution consisting of 45 mM MgCl2,
30 mM Tri-sodium citrate (Na3C6H5O7, 2H2O), 20 mM MOPS,
1% triton X-100, 10 mM sodium-metabisulfite (Na2S2O5) (pH 7)
to obtain a nuclei suspension. The whole sample was sieved
through a 40-m mesh nylon filter and stained with 6 g ml-1
DAPI solution. A diploid parental plant was used as an internal
standard. Nuclei analysis was done using a Partec CA II flow
cytometer.

Results
Androgenic callus
About 4 months after being transferred onto culture medium, banana anthers formed the first androgenic calli.
Callus was only initiated from cultured anthers between

513

Fig. 2 A Microspores isolated from banana anthers. Bar 14 m. B Development of callus from anthers after 5 months on induction
medium. Bar 2 mm. C Androgenic embryos formed on calli. Bar 4 mm. D Haploid plantlets regenerated from anthers. Bar 3 cm

514
Table 1 Frequency of embryo
formation in anther culture of
four genotypes of banana Musa
balbisiana (BB)
a Number

of anthers forming
callus/total number of anthers
cultured
b Number of embryos formed/
total number of anthers cultured

Table 2 Frequency of plant

regeneration in anther culture
of four genotypes of banana
Musa balbisiana (BB)

a Number of plantlets/total
number of anthers cultured

Genotype

Number of
cultured anthers

Pisang klutuk wulung

Pisang klutuk
Pisang batu
Tani

530
1,650
1,420
290
3,890

Genotype

Total number
of plantlets

Pisang klutuk wulung

Pisang klutuk
Pisang batu
Tani

31
63
41
12
147

Anthers forming calli

Embryos formed:

Number

Percentagea

Number

Percentageb

469
1,345
957
215
2,986

88.5
81.5
67.4
74.1
76.8

49
120
124
14
307

9.2
7.3
8.7
4.8
7.9

Diploid plantlets:

Haploid plantlets:

Number

Percentagea

Number

Percentagea

24
45
29
8
106

4.5
2.7
2.0
2.8
2.7

7
18
12
4
41

1.3
1.1
0.8
1.4
1.1

scales 17 and 28 relative to the floral apex (floral apex

has a scale of 0). Squashes of anthers (1.01.5 cm) from
these scales showed microspores at the uninucleate stage
(Fig. 2A). A subculture of the anthers before the induction of first calli led to necrosis; therefore, a subculture
of anthers before callus formation should be avoided.
The calli formed were either friable or compact
(Fig. 2B). There was no correlation between the type of
calli and the genotype; friable and compact calli were
observed in all genotypes. Anther-derived calli were produced from all four of the genotypes tested. The number
of anthers producing calli was generally very high
(Table 1) and appeared to be genotype-dependent.
Among the genotypes studied, the best results with respect to callus formation were obtained with Pisang
klutuk wulung (88.5%) and Pisang klutuk (81.5%), with
the average rate being 76.8%.
Androgenic embryos
Androgenic embryos (Fig. 2C) were formed on calli
after 6 months of anther culture. The frequency of embryo formation was found to be genotype-dependent
(Table 1). The highest relative embryo number per anther was obtained with Pisang klutuk wulung (9.2%)
followed by Pisang batu (8.7%), Pisang klutuk (7.3%)
and Tani (4.8%). The average rate of embryos formed
was 7.9% based on the total number of anthers cultured.
Haploid plantlet regeneration
Plant regeneration occurred after 8 months of anther culture. Of the 147 plantlets obtained, 41 were haploid
plantlets (Table 2): four (1.4%, based on the total num-

Fig. 3 Flow scheme for anther culture of banana

ber of anthers cultured) were from Tani, seven (1.3%)

from Pisang klutuk wulung, 18 (1.1%) from Pisang
klutuk and 12 (0.8%) from Pisang batu. The average
relative number of haploid plantlets that developed from
the cultured anthers was 1.1%. Flow cytometry analysis
revealed that diploid plants were also present among the
regenerated plants. The relative number of diploid plants

515

formed was 2.7%, based on the total number of anthers

cultured. The haploid plants (Fig. 2D) obtained have
been kept as an in vitro collection, and some were transplanted to the field for further testing. A flow scheme for
anther culture of banana is presented in Fig. 3.

Discussion
The results of the investigation reported here show, for
the first time, that haploid plants can be produced efficiently in Musa balbisiana (BB). This species plays an
important role in banana breeding because it contains
resistant genes against banana diseases (Four 1993). To
the best of our knowledge, there is only one report on
haploid plant regeneration in banana M. acuminata (AA)
(Kerbellec 1996).
We showed that callus formation was obtained if the
cultured anthers contained microspores at the uninucleate stage. This has also been observed in banana
M. acuminata (Kerbellec 1996) and in many other
monocots like barley (Foroughi-Wehr et al. 1982), wheat
(Hu and Kasha 1997), maize (Wan et al. 1991) and rice
(Alemanno and Guiderdoni 1994). The average frequency of calli formed was 76.8% under our culture conditions, which is higher than that obtained for rice (20%)
(Sathish et al. 1995).
Some of the androgenic calli were also embryogenic.
The embryos produced were similar to those obtained in
previous studies on embryogenesis in banana (Assani et
al. 2001, 2002). The percentage of anthers developing
into embryos was 7.9%, which is similar to the value obtained for wheat (Stober and Hess 1997). The frequency
of haploid plant regeneration was 1.1%. The differential
response among the genotypes with respect to plant regeneration suggests that genetic factors affect anther cultures. According to Powell (1988), the genotype dependency of anther culture response is the major limitation
to a wider exploitation of anther culture in breeding.
However, genotype difference can be overcome by
crossing a highly responsive genotype to a non-responsive genotype (Hou et al. 1994). However, this is only
relevant for conventional cross breeding.
The presence of diploid plants was also observed
among the regenerants. While this could be a consequence of the regeneration of diploid anther tissues such
as anther wall or connective tissue, in monocots the regeneration of somatic anther tissue has been very rarely
reported. The other possibility is spontaneous chromosome doubling in haploid cells. The possible factors
leading to spontaneous doubled-haploid plants could be
nuclear fusion in the early divisions of the microspores,
endomitosis, endoreduplication or multipolar mitosis
during the callus phase (Chen et al.1982; De Buyser and
Henry1986). Spontaneous chromosome doubling in
anther cultures is considered to be advantageous
because artificially induced chromosome doubling
through colchicine treatment of haploid plants is then
not necessary.

Conclusion
Haploid plant production can be achieved in banana
M. balbisiana (BB). Our results show that: (1) the callus
phase usually preceded embryo formation, (2) the frequency of haploid plant regeneration was genotypedependent, (3) diploid plants were present among the
regenerants; this could be a result of spontaneous chromosome doubling occurring during androgenesis. The
haploid plants developed here may be important for the
improvement of banana through breeding programmes.
Acknowledgements We thank Dr. M.V. Rajam, University of
Delhi South Campus, New Delhi for reading the manuscript and
Mr. D. Froger for his help with the photography. This investigation was generously supported by the European Union (INCODC-Contract no. IC18-CT97-02-04).

References
Alemanno L, Guiderdoni E (1994) Increased doubled haploid
plant regeneration from rice (Oryza sativa L.) anthers cultured on colchicine-supplemented media. Plant Cell Rep 13:
432436
Assani A, Hacour R, Wenzel G, Cte FX, Bakry F, ForoughiWehr B, Ducreux G, Aguillar ME, Grapin A (2001) Plant regeneration from protoplasts of dessert banana Grande Naine
(Musa spp., Cavendish sub-group AAA) via somatic embryogenesis. Plant Cell Rep 20:482488
Assani A, Hacour R, Wenzel G, Foroughi-Wehr B, Bakry F, Cte
FX, Ducreux G, Ambroise A, Grapin A (2002) Influence of donor material and genotype on protoplast regeneration in banana
and plantain cultivars (Musa spp.). Plant Sci 162:355362
Champion J (1963) Le bananier. G.-P. Maisonneuve et Larose, Paris
Chen Z, Qian C, Qin M, Xu X, Xiao Y (1982) Recent advances in
anther culture of Hevea brasiliensis (Muell.-Arg.). Theor Appl
Genet 62:103108
De Buyser J, Henry Y (1980) Induction of haploid and diploid
plants through in vitro anther culture of haploid wheat
(n=3x=21). Theor Appl Genet 57:5758
De Buyser J, Henry Y (1986) Wheat: production of haploids, performance of doubled haploids, and yield trials. In: Bajaj YPS
(ed) Biotechnology in agriculture and forestry, vol 2. Crops I.
Springer, Berlin Heidelberg New York, pp 7388
Foroughi-Wehr B, Friedt W, Wenzel G (1982) On the genetic improvement of androgenetic haploid formation in Hordeum vulgare L. Theor Appl Genet 62:233239
Foroughi-Wehr B, Friedt W (1984) Rapid production of recombinant barley yellow mosaic virus resistant Hordeum vulgare
lines by anther culture. Theor Appl Genet 67:377382
Foroughi- Wehr B, Wenzel G (1989) Androgenic haploid production. Int Assoc Plant Tissue Cult Newsl 58:1118
Four E (1993) Characterization of the reactions of banana cultivars to Mycosphaerella fijiensis Morelet in Cameroon and genetics of resistance. In: Ganry J (ed) Breeding banana and
plantain for resistance to diseases and pests. Proc Int Symp
Genet Improve Bananas Resistance Dis Pests. Cirad, Montpellier, pp 159170
Hansen AL, Plever C, Pedersen HC, Keimer B, Andersen SB
(1994) Efficient in vitro chromosome doubling during Beta
vulgaris ovule culture. Plant Breed 112:8995
Hou L, Ullrich SE, Kleinhofs A (1994) Inheritance of anther culture traits in barley. Crop Sci 34:12431247
Hu T, Kasha KJ (1997) Improvement of isolated microspore culture of wheat (Triticum aestivum L.) through ovary co-culture.
Plant Cell Rep 16:520525
Kasha KJ, Kao KN (1970) High frequency haploid production in
barley (Hordeum vulgare L.). Nature 225:874875

516
Kerbellec F (1996) Etablissement dune technique dandrogense
pour lamlioration gntique du bananier (Musa spp.). PhD
thesis, Ecole Nationale Agronomique de Rennes, France
Khler F, Wenzel G (1985) Regeneration of isolated barley microspores in conditioned media and trials to characterise the responsible factor. J Plant Physiol 121:181191
Morel G, Wetmore RH (1951) Fern callus tissue culture. Am J Bot
38:141143
Murashige T, Skoog F (1962) A revised medium for rapid growth
and bioassays with tobacco tissue cultures. Physiol Plant
15:473497
Muren RC (1989) Haploid plant induction from unpollinated ovaries in onion. Hortic Sci 24:833834
Powell W (1988) Diallel analysis of barley anther culture response.
Genome 30:152157

Sathish P, Gamborg OL, Nabors MW (1995) Rice anther culture:

callus initiation and androclonal variation in progenies of regenerated plants. Plant Cell Rep 14:432436
Stober A, Hess D (1997) Spike pretreatments, anther culture conditions, and anther culture response of 17 German varieties of
spring wheat (Triticum aestivum L.). Plant Breed 116:443447
Wan Y, Duncan DR, Rayburn AL, Petolino JF, Widholm JM
(1991) The use of antimicrotubule herbicides for the production of doubled haploid plants from anther-derived maize callus. Theor Appl Genet 81:205211
Wan Y, Widholm JM (1993) Anther culture of maize. Plant Breed
Rev 11:199224

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.