Plant Cell Rep (2006) 25: 1–10

DOI 10.1007/s00299-005-0028-y

CELL BIOLOGY AND MORPHOGENESIS

E. D. J. Supena · S. Suharsono · E. Jacobsen ·

J. B. M. Custers

Successful development of a shed-microspore culture protocol

for doubled haploid production in Indonesian hot pepper
(Capsicum annuum L.)
Received: 1 April 2004 / Revised: 28 October 2004 / Accepted: 19 May 2005 / Published online: 20 September 2005


C Springer-Verlag 2005

Abstract Various systems of anther and microspore cul- in other tropical Asian countries (Anonym. 2000b). Local
tures were studied to establish an efficient doubled haploid breeding programs in Indonesia are rather traditional and
production method for Indonesian hot pepper (Capsicum generally too simple to improve the hot pepper crop with
annuum L.). A shed-microspore culture protocol was de- resistances against pests and diseases. There is clearly a
veloped which outperformed all the previously reported need in Indonesia for new and more sophisticated breeding
methods of haploid production in pepper. The critical fac- techniques for hot pepper. One such technique is haploid
tors of the protocol are: selection of flower buds with more technology, i.e. the production of (doubled) haploid plants
than 50% late unicellular microspores, a 1 day 4◦ C pre- from male or female gametes.
treatment of the buds, followed by culture of the anthers Spontaneous parthenogenesis was the first system used
in double-layer medium system for 1 week at 9◦ C and to obtain haploid plants in bell pepper (Christensen and
thereafter at 28◦ C in continuous darkness. The medium Bamford 1943), but the efficiency of production was low
contained Nitsch components and 2% maltose, with 1% (Pochard and Dumas de Vaulx 1979). Anther culture using
activated charcoal in the solid under layer and 2.5 µM solid media with relatively high concentrations of growth
zeatin and 5 µM indole-3-acetic acid in the liquid upper regulators was attempted in the 1970s, but yields were low
layer. All the ten genotypes of hot pepper tested, responded and the plants had to be regenerated via a callus phase
to this protocol. The best genotypes produced four to seven (Wang et al. 1973; George and Narayanaswamy 1973).
plants per original flower bud. This protocol can be used Sibi et al. (1979) introduced a more successful two-step
as a potential tool for producing doubled haploid plants for anther culture system, which was further optimized by Du-
hot pepper breeding. mas de Vaulx et al. (1981). In this protocol, anthers are first
incubated on a medium with high concentrations of growth
Keywords Capsicum annuum . Hot pepper . regulators, then subcultured on a medium with low amounts
Shed-microspore culture . Defective shoots of growth regulators. A heat shock (35◦ C) treatment was
applied at the beginning of the culture. Using this proto-
col, the two best performing bell pepper lines gave 42 and
Introduction 51 plants per 100 anthers, respectively, while a group of
eight F1s yielded on average 18 plants per 100 anthers
Hot pepper (Capsicum annuum L.) is the most important (Dumas de Vaulx et al. 1981). This anther culture system
vegetable crop in Indonesia, both in terms of cultivated has been used to produce many doubled haploid plants of
area and economic value (Anonym. 2000a). However, the bell pepper for use in breeding programs (Abak et al. 1982;
yield of pepper in Indonesia is relatively lower than that Pochard et al. 1983; Hendy et al. 1985; Dumas de Vaulx and
Pochard 1986; Daubèze et al. 1990; Caranta et al. 1996).
Communicated by E. Heberle-Bors Since the publication of Dumas de Vaulx et al. (1981),
E. D. J. Supena · S. Suharsono many other groups have attempted to use the new anther
Research Center for Biotechnology, Bogor Agricultural culture system for their own pepper germplasm (Vagera
University (IPB), P.O. Box 1, Bogor 16610, Indonesia, and Havránek 1985; Morrison et al. 1986; Munyon et al.
E. D. J. Supena · E. Jacobsen · J. B. M. Custers (
)
1989; Kristiansen and Andersen 1993; Maheswary and
Plant Research International, Wageningen University and Mak 1993; Qin and Rotino 1993; Ltifi and Wenzel 1994;
Research Centre, Mitykó et al. 1995; Dolcet-Sanjuan et al. 1997; Gémesné
P.O. Box 16, NL-6700 AA Wageningen, The Netherlands et al. 1998; Gyulai et al. 2000). But, in general, the results
e-mail: [email protected] were disappointing, as not more than 1–5 plants per 100 an-
Tel.: +31-317-477047
Fax: +31-317-418094 thers were obtained, and many accessions did not respond
2

at all. Therefore, several researchers attempted to improve were grown with two main stems, and young fruits were
the method of Dumas de Vaulx et al. (1981). Vagera and removed and plants were severely pruned to stimulate
Havránek (1985) added activated charcoal to the medium, the development of new young side shoots for flower
with some beneficial effect. Morrison et al. (1986) subcul- bud production. Additional light (Philips SON-T lamps)
tured the anthers after the initiation medium on a double- was given from early September onward to sustain plant
layer medium, i.e. solid charcoal medium with a liquid growth.
top layer. Mitykó et al. (1995) selected the healthy anthers Flower buds of the desired size (petals equal or slightly
every month and subcultured them on a fresh medium. longer than sepals) were harvested in the morning, disin-
Dolcet-Sanjuan et al. (1997) switched back from the two- fected for 10 min in 2% NaOCl with 0.05% (v/v) Tween-20
step procedure to a one-step culture, using a double-layer added, and then rinsed three times in sterile tap water. Iso-
medium with charcoal, but without growth regulators, and lated anthers with a faint purple tip were of the proper
with maltose instead of sucrose. developmental stage, and were used for anther culture or
These revised protocols brought about moderate im- for microspore isolation. Staining with 4 , 6-diamidino-
provements for the pepper genotypes investigated, but high 2-phenylindole (DAPI) showed that these anthers con-
yields as reported by Dumas de Vaulx et al. (1981) were tained more than 50% of microspores in the late unicel-
not obtained. Out of more than 250 pepper accessions lular stage, with amounts of mid unicellular microspores
documented in literature, only one genotype outperformed and of bicellular pollen not exceeding 20% and 50%,
the genotypes of Dumas de Vaulx et al. (1981), viz. respectively.
‘Fehézörön’ with 76 plants per 100 anthers (Mitykó et al.
1995), while another genotype ‘Emerald Giant’, which
produced 40 plants per 100 anthers (Morrison et al. 1986), Culture protocols for microspore embryogenesis
equalled the best genotypes of Dumas de Vaulx et al.
(1981). In addition to the protocols described above, Four different culture procedures were evaluated:
which are all based on anther culture systems, Regner
(1994, 1996) also studied isolated microspore culture in 1. Pepper anther culture according to Dumas de Vaulx
bell pepper. He mainly used the Brassica napus protocol et al. (1981). Anthers were incubated on solid C
(Polsoni et al. 1987; Pechan and Keller 1988) together with medium with 0.01 mg/l kinetin and 0.01 mg/l 2,4-
the Dumas de Vaulx media, but was not able to develop a dichlorophenoxyacetic acid (2,4-D) for 12 days, and
successful culture protocol. then subcultured on R1 medium with only 0.01 mg/l
In this paper, we report a procedure of shed-microspore kinetin. Cultures were kept at 35◦ C in darkness during
culture for haploid plant production in Indonesian hot the first 8 days, and then transferred to 12 h light
pepper genotypes, which are of the cayenne fruit type (25–30 µmol/m2 ·s) at 25◦ C.
(Berke and Shieh 2000). We compared four different 2. Pepper anther culture according to Dolcet-Sanjuan et al.
systems of anther and microspore cultures previously (1997). A double-layer medium was used, in which the
used in pepper and tobacco (Dumas de Vaulx et al. under layer consisted of Nitsch components (Nitsch and
1981; Johansson et al. 1982; Dolcet-Sanjuan et al. 1997; Nitsch 1969) with 2% maltose and 0.5% activated char-
Touraev and Heberle-Bors 1999) for their ability to induce coal (Duchefa) and was solidified with 0.6% Plant Agar
embryogenesis from microspores of Indonesian hot pepper. (Duchefa), while the liquid top layer contained Nitsch
components plus 2% maltose. Media for both layers
were sterilized by autoclaving, prior to which the pH
Materials and methods was adjusted to 5.8. Commonly, 3 cm Petri dishes were
used, with approximately 1.5 ml solid medium and 1 ml
Plant material and microspore stage characterization liquid top layer. Cultures were kept at 9◦ C during the first
week and then at 28◦ C, always in continuous darkness.
Ten Indonesian hot pepper (Capsicum annuum L.) geno- 3. Tobacco anther culture according to Johansson et al.
types were used: large hot pepper type ‘Galaxy’, ‘Jatilaba’, (1982) and Custers et al. (1999). A double-layer medium
‘Tit-Super’, ‘Tombak’, ‘LV-2319’ and ‘LV-2323’, and system was used, in which the solid under layer con-
curly pepper type ‘Cemeti’, ‘Laris’, ‘Tornado’ and sisted of half strength MS medium (Murashige and
‘Typhoon’. LV-2319 and LV-2323 are breeding lines from Skoog 1962) with 2% sucrose, 0.5% activated charcoal,
the Indonesian Vegetable Research Institute, Bandung. and 0.6% Plant Agar, while the liquid upper layer con-
The remaining genotypes are open-pollinated varieties tained half strength MS plus 2% sucrose. Sterilization
from Indonesian seed companies. Plants were grown in of media and preparation of the Petri dishes were as de-
glasshouses in Wageningen, The Netherlands from early scribed for the Dolcet-Sanjuan system above, with the
April to November on 12 cm thick rockwool slabs covered exception that pH of the medium was adjusted to 6. Be-
with plastic. Water and nutrients were supplied by trickle fore starting the culture, flower buds were first put in a
irrigation, each plant having its own dripper. Temperature plastic jar with screw lid on a wet filter paper and stored
set points in the glasshouse were 21◦ C day/19◦ C night. A at 9◦ C for 7 days. The flower buds were then disinfected,
whitewash cover and an internal screen were used during the anthers isolated and incubated on top of the liquid
summer to keep the temperature below 30◦ C. Plants medium. Cultures were kept at 25◦ C in darkness.
 3

4. Isolated microspore culture according to the tobacco in 6 cm Petri dishes and were kept under 16 h light (25–
protocol by Touraev and Heberle-Bors (1999). Liquid B 30 µmol/m2 .s) at 25◦ C. After 3–4 weeks, seedlings show-
medium was used for starvation of the microspores, and ing cotyledons and a first true leaf were transferred into
thereafter these were subcultured into rich liquid AT3 honey jars with a 1 cm thick layer of vermiculite wet with
medium. The AT3 medium contained 3, 6 or 9% sucrose half strength liquid MS medium supplemented with 1%
or maltose as the carbon source. The pH of B medium sucrose. Seedlings that had formed four to six leaves were
and AT3 medium was adjusted to 6.8 and 6.2, respec- transferred to soil.
tively, and both liquid media were sterilised by filtration.
Cultures were performed in 3 or 6 cm Petri dishes, with
1 or 3 ml microspore suspension, respectively, at a den- Experimental design and assessment of results
sity of 40,000 microspores per ml. The microspores were
cultured for 6 days at 25◦ C or 32◦ C, either continuously In the experiments dealing with anther culture procedures,
in starvation B medium or for the first 4 or 24 h in AT3 each experiment consisted of five treatments. Per treatment,
medium containing 6% sucrose (prefeeding) followed five Petri dishes were used. Each Petri dish consisted of
by culture in starvation medium. Thereafter, swollen six anthers taken randomly from six different buds. How-
microspores were purified by Percoll gradient centrifu- ever, in the case of experiments that were carried out to
gation according to Kyo and Harada (1986) with 50% study the effect of bud pretreatment (at 4◦ C) as well as the
Percoll in 1.65× normal strength B medium and a cen- influence of genotype, anthers from one bud (mostly six
trifugation force of 100 g (4 min), and then transferred anthers) were kept together in one Petri dish for studying
to rich medium and cultured at 25◦ C. All the cultures also the bud responsiveness. Experiments were repeated
were maintained in continuous darkness. at least three times during different periods. In the case of
microspore culture experiments, microspores were isolated
from the anthers of a batch of 20–30 buds, resulting in about
20 ml microspore suspension which was used for various
Experiments performed treatments.
Experimental data were analysed by standard analysis
Especially in the Dolcet-Sanjuan procedure, we studied of variance using General Linear Model of SPSS 10.0 for
various conditions of certain factors in more detail. Windows software, and least significant differences (LSD)
– Flower bud cold (4◦ C) pretreatment was given for dif- were calculated to determine the statistical significance of
ferent periods (0, 1, 3 and 7 days) with three varieties, treatment effects. For convenience, we present the yield
‘Cemeti’, ‘Jatilaba’, and ‘Tombak’. After the cold pre- of embryos and plants produced per bud and not per 100
treatment, the anthers were isolated, and kept in culture anthers, which is commonly done.
for 1 week at 9◦ C and then at 28◦ C, always in continuous
darkness.
– Low as well as high incubation temperatures (4◦ , 9◦ , Cytological analysis
28◦ , 32◦ , and 35◦ C) were used during the first week of
culture in ‘Tombak’. After this treatment, all cultures To follow early development of microspores in culture,
were continued at 28◦ C, always in continuous darkness. cytological examination was conducted during the first
– Ten Indonesian cayenne pepper genotypes were com- 3 weeks of culture. Anthers were carefully opened with
pared for their performance in the system after flower a scalpel and needles, and samples of microspores were
bud cold (4◦ C) pretreatment for 1 day. collected for DAPI staining and observation. Upon dehis-
– The influence of different concentrations of activated cence of the anthers, progressive development in cultures
charcoal from Duchefa (up to 2%) in the under layer was followed by observation under a stereomicroscope.
medium was analysed in ‘Tombak’.
– The effect of the addition of 2.5–25 µM zeatin or 5–
50 µM indole-3-acetic acid (IAA) in liquid upper layer Analysis of ploidy levels and characterization
medium was analysed in ‘Galaxy’. Also, the combina- of regenerants
tions of zeatin (2.5 or 10 µM) and IAA (5 or 25 µM)
were investigated. Growth regulators were added after For ploidy analysis, a piece of leaf was cut from seedlings,
autoclaving. nuclei were isolated and the amount of DNA was mea-
sured using a Coulter Epics XL-MCL (Beckman-Coulter,
USA) flow cytometer according to the protocol described
Germination of microspore embryos by Lanteri et al. (2000). Genetic proof for the gametic origin
of the diploid plants derived from anther cultures was ob-
Germination medium contained half strength MS elements, tained by checking their selfed progeny plants for complete
2% sucrose, and 0.1 µM 6-benzylaminopurine (BA), so- uniformity as compared to the heterogeneous composition
lidified with 0.6% Plant Agar. Cultures were performed of the original donor plant populations.
4

Results

Comparison of various culture systems

In the initial experiments, we used the Dumas de Vaulx an-

ther culture procedure, but the results with Indonesian hot
pepper were disappointing. Out of seven genotypes tested,
only one was responsive and gave less than 0.2 plants per
bud. The tobacco anther culture procedure of Johansson
gave better results. Out of four genotypes tested, two were Fig. 2 Early development of hot pepper (Capsicum annuum)
responsive, with an average output of 0.5 plants per bud. ‘Tombak’ microspores in shed-microspore culture as observed with
DAPI staining. A and B Microspores taken from anthers incubated
The best results were obtained with the Dolcet-Sanjuan for 1 and 2 weeks, respectively. C Proembryo after 3 weeks of culture
system; all four genotypes tested earlier in Johansson pro- that developed from a microspore dispersed in the liquid medium.
cedure were responsive, yielding on average 1 plant per Bars = 10 µm for A–C
bud.
In addition to the anther culture, we also studied isolated about isolated microspore culture, we decided to continue
microspore culture (Fig. 1A). Our initial results were very further experiments on the anther culture systems with a
promising with hot pepper. After two and a half weeks double-layer medium, especially that of Dolcet-Sanjuan.
of culture, relatively high percentages (4–7%) of sporo-
phytically dividing microspores with five to eight nuclei
were found, whereas this frequency never exceeded 0.5% Anther culture resembles shed-microspore culture
in the anther culture systems. In rich medium with su-
crose, however, the sporophytic microspores, which con- Upon incubation in a double-layer culture system, the hot
tained relatively small nuclei, exhibited excessive starch pepper anthers floated on the surface of the liquid upper
accumulation and had a strong tendency to burst preco- layer medium, and after 2–3 weeks they showed normal
ciously (Fig. 1B), leading to dispersal of loose cells in dehiscence, releasing the microspores into the medium.
the medium. This could be avoided by performing the mi- Notably, most embryos were formed from released mi-
crospore isolation in rich medium with 6% sucrose and crospores rather than from microspores attached to the an-
by prefeeding in the same medium for 4 h prior to star- ther walls. As the development of these microspores re-
vation in B medium. This treatment allowed sporophytic sembles the development in shed-microspore culture, as
divisions to start during the starvation period. High per- reported for instance in barley (Ziauddin et al. 1990), we
centages of up to 15% of sporophytic microspores were decided to use this term for our cultures.
obtained, which could be increased to 40% by Percoll gra- Figure 2 shows early cytological observations on hot pep-
dient centrifugation. Subculture in rich medium with 3 and per ‘Tombak’ shed-microspore cultures. After 1 week of
6% maltose stimulated the development of high quality culture at 9◦ C, some of the microspores have started di-
multicellular microspores that did not show early bursting viding sporophytically, resulting in two vegetative-like nu-
(Fig. 1C). Unfortunately, only a small percentage of the clei (Fig. 2A). Continuation of sporophytic divisions pro-
microspores successfully developed into embryos, with the duced multicellular microspores with relatively large nuclei
maximum number of healthy plants obtained being only (Fig. 2B), which developed into proembryos that were first
0.1 per original flower bud. Despite our initial enthusiasm visible after 3 weeks of culture (Fig. 2C). The dark back-
ground of the activated charcoal layer made it difficult to
detect the early proembryos with the dissection microscope
(Fig. 3A), but from week 4 of culture onward, the obser-
vations became more informative. Globular to heart-shape
stage embryos were visible after 4 weeks of culture, and
had reached the torpedo to cotyledon shape after 5 weeks
of culture (Fig. 3B). The first germination of embryos was
observed 1 week later (Fig. 3C).

Microspore embryos differed in quality

Fig. 1 Early development of hot pepper (Capsicum annuum)
‘Tombak’ microspores in isolated microspore culture, as observed
with DAPI staining. A Freshly isolated late-unicellular microspores. Germination of microspore-derived embryos began after 6–
B Microspores with 6–8 nuclei after 1 week in B medium and 8 weeks of culture on embryogenesis initiation media, after
10 days in rich medium with 6% sucrose. The microspores show which embryos were subcultured on germination medium.
precocious bursting. C Microspores with 10–14 nuclei (some nuclei
out of focus) after 4 h prefeeding in rich medium with 6% sucrose, The germination rate was dependant on embryo quality.
6 days in B medium, then 10 days in rich medium with 6% maltose. Actually, only approximately 20% of the embryos was nor-
Bars = 20 µm for A–C mal and complete, that is consisting of radicle, hypocotyl,
 5

Table 1 Effect of temperature treatment during the first week of

culture on embryo yield in shed-microspore culture of hot pepper
(Capsicum annuum L.) ‘Tombak’. After the temperature treatments,
cultures were continued at 28◦ C. The experiment was repeated three
times. Means within a column followed by the same letter are not
significantly different at P=0.05
Temperature (◦ C) Average yield of embryos per bud
Total no. of No. of normal-looking
embryos produced embryos
4 13.7 ab 6.3 ab
9 22.9 a 7.8 a
28 6.1 b 3.2 b
32 12.1 b 4.4 ab
35 8.7 b 3.0 b

at 28◦ C gave the lowest yield of both total and normal-

looking embryos. ‘Tombak’ donor plants in the experiment
visibly varied in size, in leaf and fruit shapes, and in the
amount of anthocyanin on the stems. ‘Tombak’ is an old
Fig. 3 Time-course of events during shed-microspore culture of hot
pepper (Capsicum annuum) ‘Tombak’. A: Dehisced anther accom- landrace that is not genetically pure. The heterogeneity
panied with proembryos in the medium after 3 weeks of culture. B of the plant material was also observed in culture, giving
and C Embryos developed after 5 and 6 weeks of culture, respec- rise to relatively high standard errors of the treatment
tively. D–F Three categories of embryos obtained after 7 weeks of means.
culture: complete embryos with visible cotyledons, embryos missing
cotyledons, and embryos consisting of radicles only. G Pin-shaped In addition to the 9◦ C treatment that gave the best results
seedlings developed from embryos without cotyledons. H Normal on embryo yield (see Table 1), the results obtained with
seedlings developed from complete embryos. I Plant formation on 4◦ C were also interesting in giving high total embryo yield
vermiculite medium. Bars = 1 mm for photographs A–B; 4 mm for as well as high yield of normal-looking embryos. There-
C; 3 mm for D–F; 8 mm for G–H; 1.5 cm for I fore, we tried to exploit 4◦ C as cold pretreatment of the
flower buds, which is a common pretreatment in tobacco
and cotyledons (Fig. 3D). Approximately 30% of the em- anther culture. The results on the effect of different periods
bryos was without cotyledons and apparently missing a (0–7 days) of 4◦ C cold pretreatment on shed-microspore
shoot apical meristem (Fig. 3E), and the remaining 50% culture of ‘Cemeti’, ‘Jatilaba’ and ‘Tombak’ are shown in
seemed to consist of a radicle only (Fig. 3F). The latter Fig. 4. A high percentage (60–90%) of responsive buds
group of embryos was the first to germinate, even before was observed in all the three varieties. A positive effect of
subculture to germination medium, but they were incapable flower bud cold pretreatment was found for the total em-
of producing complete seedlings. The intermediate group, bryo production, but it was most evident for the number of
lacking visible cotyledons, formed pin-shaped seedlings normal-looking embryos. The duration of 1 day 4◦ C cold
(Fig. 3G), of which only about 5% gave rise to normal pretreatment appeared to be optimal, even though the ef-
plants, while seedling formation from the complete em- fect was not always statistically significant, due to relatively
bryos with two cotyledons was almost 100% (Fig. 3H). high standard errors, as caused by heterogeneity within the
Well developing seedlings easily continued growth after donor plant populations.
subculture to vermiculite medium (Fig. 3I), and there- In a cytological study, we examined using DAPI how
after were successfully transferred to soil. Future exper- 1 day 4◦ C cold pretreatment of the buds affected the early
iments focused on improving the total embryo yield, as stages of microspore embryogenesis. The percentage of
well as the yield of normal-looking embryos (complete em- microspores containing various numbers of nuclei were
bryos with two cotyledons and embryos with one visible counted during the first 3 weeks of culture and the data
cotyledon). are presented in Table 2. In each treatment, the results for
five buds were followed for different periods. During the
first 2 weeks of culture, microspores were sampled from
Temperature stress improved embryo yield the anthers, while later after anther dehiscence, a represen-
tative part of the shed microspores was also included in the
The results presented in Table 1 on the effect of temper- sampling. When compared to the non-pretreated culture,
ature demonstrate that low temperatures during the first cold-pretreatment gave a decreased percentage of empty
week of culture increased embryo yield, i.e. the yield of microspores, but increased the percentage of microspores
total and normal-looking embryos. The incubation for containing more than two nuclei (sporophytically divid-
1 week at 9◦ C gave the highest embryo yield, and therefore ing microspores) after 2 weeks of culture. After 3 weeks
this temperature treatment was subsequently included as of culture, only pro-embryos (>20 nuclei) were found in
a standard treatment in our protocol. Continuous culture pretreated cultures (Table 2).
6

100 40 8
35 7

Yield of normal-looking
Percentage of responsive

Total embryo yield

80 30 6

embryos per bud

25 5

buds (%)
60

per bud
20 4
40 15 3
10 2
20
5 1
0 0 0
0 1 3 7 0 1 3 7 0 1 3 7
Days Days Days
Fig. 4 Effect of different periods of 4◦ C cold pretreatment (0, 1, 3, 9◦ C and then at 28◦ C. Bud responsiveness, total embryo yield, and
and 7 days) of the flower buds on shed-microspore culture of hot pep- the yield of normal-looking embryos were analysed. Data are the
per (Capsicum annuum) ‘Cemeti’ (•), ‘Jatilaba’ (
), and ‘Tombak’ means (+ or − standard errors) of three replicated experiments
(). After the bud pretreatment, the anthers were kept for 1 week at

Table 2 Percentages of microspores containing various numbers of culture was started from flower buds with or without pretreatment
nuclei counted after 0, 1, 2, and 3 weeks in shed-microspore culture for 1 day at 4◦ C. w: week
of hot pepper (Capsicum annuum L.) ‘Tombak’. Shed-microspore
No. of nuclei per microspore Percentages of microspores containing nuclei
In non-pretreated culture after In 4◦ C pretreated culture after
0w 1w 2w 3w 0w 1w 2w 3w
0 0.2 2.1 42.2 75.3 0.3 1.2 23.1 77.0
1 88.7 21.6 5.7 2.7 96.3 20.9 3.4 0.2
2 11.1 75.6 43.9 17.3 3.4 77.3 58.2 16.1
3–4 0 0.7 7.9 3.8 0 0.6 14.8 5.6
5–6 0 0 0.3 0.7 0 0 0.4 0.6
7–10 0 0 0 0.1 0 0 0.1 0.2
11–20 0 0 0 0.1 0 0 0 0.3
>20 0 0 0 0 0 0 0 0.1

Table 3 Performance of six genotypes from large type and four germinated plants were analysed. The experiment was repeated three
genotypes from curly type Indonesian hot pepper (Capsicum annuum times, per genotype using each time the anthers of five buds that were
L.) in shed-microspore culture. Bud responsiveness, total embryo pretreated for 1 day at 4◦ C
yield, yield of normal-looking embryos, and number of successfully
Hot pepper genotypes Percentage of Average yield per incubated bud
(types and accessions) responsive buds Total no. of embryos No. of normal-looking embryos No. of mature plants∗
produced
Large type
Galaxy 93 a∗∗ 31.5 a 10.8 a 7.1
Jatilaba 91 a 8.5 bc 2.1 c 1.3
LV-2319 74 cd 11.2 bc 2.8 bc nt
LV-2323 65 cde 4.6 c 1.7 c nt
Tit-Super 77 bc 12.8 abc 2.2 c nt
Tombak 90 ab 26.2 ab 5.8 b 4.3
Curly type
Cemeti 61 de 3.4 c 1.1 c 0.7
Laris 67 cde 3.8 c 1.3 c 1.0
Tornado 57 ef 3.3 c 1.9 bc 1.1
Typhoon 43 f 0.9 c 0.4 c nt
∗
Based on germination from at least 50 normal-looking embryos.
∗∗
Means within a column followed by the same letter are not significantly different at P=0.05.
nt: Not tested.
 7

Table 4 Effect of activated Concentration of activated Average yield of embryos per bud
charcoal in the solid under layer
medium on embryo yield in charcoal (% w/v) Total no. of embryos produced Normal-looking embryos
shed-microspore culture of hot No. Percentage
pepper (Capsicum annuum L.)
‘Tombak’. Data are the means 0.00 0.2 c 0.0 c 0.0
of three replicated experiments. 0.25 8.7 b 1.0 c 11.5
Means within a column 0.50 16.6 ab 4.2 b 25.3
followed by the same letter are 1.00 19.3 a 6.6 ab 34.2
not significantly different at
P=0.05 2.00 22.7 a 7.1 a 31.3

Table 5 Effect of the combined addition of zeatin (Zea) and IAA in Activated charcoal and growth regulators
the liquid upper layer medium on embryo yield in shed-microspore
culture of hot pepper (Capsicum annuum L.) ‘Galaxy’. Data are
the means of two replicated experiments. Means within a column The results on the effect of various concentrations of ac-
followed by the same letter are not significantly different at P=0.05. tivated charcoal on shed-microspore culture in Indonesian
Control: Culture without exogenous zeatin or IAA hot pepper ‘Tombak’ are presented in Table 4. Charcoal
Treatments (µM) Average yield of embryos per bud was needed, as embryo formation almost completely failed
Total no. of No. of normal-looking without its addition. Both the total embryo yield and the
embryos produced embryos yield of normal-looking embryos increased with increas-
ing concentrations of charcoal from 0 to 2%. However, the
Control 25.2 b 13.0 b effects of increasing concentrations of charcoal on both pa-
Zea 2.5+IAA 5.0 40.8 a 22.6 a rameters did not run in parallel, and no sign of a statistical
Zea 2.5+IAA 25.0 26.0 b 17.6 ab difference was observed between 1 and 2% charcoal. Two
Zea 10.0+IAA 5.0 28.9 b 17.8 ab percent charcoal impaired embryonic shoot development.
Zea 10.0+IAA 25.0 20.9 b 12.8 b Consequently, activated charcoal at a concentration of 1%
was chosen as standard for further experiments.
Influence of genotype on shed-microspore culture The effect of the addition of zeatin (2.5 to 25 µM) or
response IAA (2.5 to 50 µM) to the liquid upper layer was studied
in preliminary experiments with 0.5% activated charcoal.
The results of experiments on 1 day bud pretreatment at Improved embryogenesis was found with 2.5 or 10 µM
4◦ C with ten Indonesian cayenne pepper genotypes are zeatin as well as with 5 or 25 µM IAA. This investigation
shown in Table 3. All the genotypes appeared responsive, was continued with the medium containing charcoal at 1%,
but they differed in embryo yield and the production of focussing the attention more on the effect of different com-
mature plants. From the six genotypes tested for germi- binations of the two growth regulators. The results obtained
nation, two gave a high yield of more than four plants are shown in Table 5. The combination of 2.5 µM zeatin
per bud, while the plant recovery rate for the other geno- and 5 µM IAA was very promising, as it outperformed the
types was about one plant per bud. Notably, the general control treatment (i.e. without exogenous growth regula-
performance of curly type hot pepper was lower than tors) both in total embryo yield and in the yield of normal-
that of the large types. Generally, 60–75% of the normal- looking embryos.
looking embryos gave rise to plants. The embryos with two
cotyledons as well as the embryos with one cotyledon were
considered as normal-looking embryos. The embryos with Ploidy level of regenerated plants
two cotyledons produced viable seedlings for almost 100%,
while those with one cotyledon gave viable seedlings only Ploidy levels of plants regenerated from shed-microspore-
in about 30–50% of the cases. Of the total number of em- derived embryos in six accessions were determined us-
bryos in different genotypes, only 15–33% yielded com- ing flow cytometry (Table 6). The majority of the plants
plete seedlings. were haploid, indicating their origin to be from haploid

Table 6 Ploidy analysis of Accessions No. of plants analysed No. of plants with various ploidy levels
plants regenerated from
shed-microspore culture-derived 1× 2× 3× 4×
embryos in various Indonesian Galaxy 29 20 9 0 0
hot pepper (Capsicum annuum
L.) accessions Jatilaba 23 17 6 0 0
Tombak 39 19 20 0 0
Cemeti 32 17 15 0 0
Laris 24 15 8 1 0
Tornado 21 16 3 1 1
8

microspores. The frequency of diploid plants was rela- taining 6% sucrose prior to starvation. This finding contra-
tively low (14–33%) in four of the six accessions tested, dicts the generally accepted rule for embryogenesis induc-
whereas the other two showed 47% and 51% of sponta- tion from tobacco microspores, namely that microspores
neous diploidization. The selfed progeny of diploid plants during starvation undergo a characteristic cell cycle arrest,
of ‘Cemeti’ and ‘Tombak’ were found to be highly ho- leading to a dramatic reorganisation of cytoplasm and cy-
mogeneous, whereas the original donor populations were toskeleton in preparation for the first sporophytic division.
heterogeneous. This indicated homozygosity of the diploid Sporophytic division usually takes place in tobacco only
(doubled haploid) plants derived from shed-microspore cul- 5–7 days after subculture to rich medium (Touraev et al.
ture. 1996, 1997). Thus, the induction of embryogenesis by star-
vation stress might also occur by other ways or mecha-
nisms, as we found for the hot pepper.
Discussion A common drawback in in vitro haploid procedures with
pepper is the low germination success of the embryos. The
In this study, we developed an efficient shed-microspore problem is further increased by the transition from solid to
culture protocol for Indonesian cayenne type hot pepper liquid culture media, e.g. see Dumas de Vaulx et al. (1981)
that outperforms all the previously reported methods of and Mitykó et al. (1995) versus Morrison et al. (1986) and
haploid production in the entire pepper genus. The protocol Dolcet-Sanjuan et al. (1997). In our experiments, we also
comprises several parameters that were found to be impor- encountered this problem, which is clearly related to the oc-
tant in earlier published in vitro haploid production systems currence of apical shoot abnormalities. Only fully normal-
with pepper (e.g. Sibi et al. 1979; Morrison et al. 1986; looking embryos with two symmetrical cotyledons were
Vagera and Havránek 1985; Maheswary and Mak 1993; able to germinate at high frequencies into viable seedlings.
Mythili and Thomas 1995; Dolcet-Sanjuan et al. 1997; Certain parameters were found to improve embryo quality
Gémesné et al. 1998), but this is the first time that these in culture, namely a 1 day 4◦ C flower bud pretreatment,
parameters have been appropriately combined together to an increase in charcoal concentration or the addition of
achieve haploid production effectively. However, it is also zeatin and IAA in the medium. However, these treatments
important to consider the role of the germplasm in obtain- only brought about a slight quantitative improvement and
ing high embryo production and regenerants. Chillies or did not lead to qualitative changes in shoot formation. Re-
hot pepper, which are native to Central and South America, cently, it became evident that defective shoot formation is
were introduced to South-East Asia early in the sixteenth also a major shortcoming in other tissue culture systems
century (Berke and Shieh 2000), and in Indonesia they with pepper, including somatic embryogenesis (Steinitz et
likely evolved further in quite isolation from the rest of the al. 2003) and organogenesis (Ochoa-Alejo and Ramı́rez-
genus. There are no reports in the literature on haploid pro- Malagón 2001; Wolf et al. 2001). For a long time, the
duction from Indonesian hot pepper, and since we did not seriousness of the problem was not apparent as earlier re-
test other peppers of different origins or types, it is possible ports claiming successful somatic embryogenesis avoided
that a natural genetic capacity of Indonesian hot pepper for clearly to mention the shoot abnormalities (e.g. Harini and
microspore embryogenesis also contributed to the success Lakshmi-Sita 1993; Büyükalaca and Mavituma 1996; Jo et
of our protocol. al. 1996). So far, no conditions have been found in pepper
In addition to the promising shed-microspore culture tissue culture research that effectively promote formation
system, we obtained interesting results with isolated mi- of embryos without shoot abnormalities. We have partly
crospore culture of hot pepper. As compared to shed- solved this problem, as we obtained a higher germination
microspore culture, the change in microspore developmen- percentage than in earlier procedures of haploid production
tal fate from gametophytic to sporophytic development with pepper. Our future research will focus more on improv-
was much more evident in isolated microspores. Unfor- ing the embryo quality in shed-microspore culture. Obvi-
tunately, hardly any of the microspores developed further ously, the individual embryos produced in shed-microspore
into embryos. Multicellular microspores in isolated mi- culture are a better experimental material to study the de-
crospore cultures always contained rather small nuclei, fective shoot problem than the clusters of somatic embryos
while nuclear sizes were generally large in the shed-culture with fused basal parts from tissue culture (Steinitz et al.
multicellular microspores. Large nuclei with decondensed 2003).
chromatin are usually indicative of active gene transcrip- The results obtained in the present study clearly show
tion, while cell groups with small nuclei are considered that shed-microspore culture can be an attractive system
to be less active transcriptionally (Cooper and Hausman for haploid production in Indonesian hot pepper genotypes.
2003). Therefore, we presume that the shed-culture-derived Generally, however we observed high variation within treat-
multicellular microspores with their large nuclei are more ments, likely due to genetic differences between donor
amenable to embryogenesis than the ones in the isolated plants providing the anthers. The lack of uniformity be-
microspore cultures, which likely have low or no gene tween donor plants from a given accession was indicative
activity. of genetic heterogeneity. The aim of our study has been
We observed in the isolated microspore cultures that the establishment of a haploid production procedure for
sporophytic divisions already started during the starvation Indonesian hot pepper types, and therefore we were com-
period, if the microspores were prefed with medium con- mitted to use heterozygous local varieties, at least in the
 9

initial experiments. However, future work using homozy- Dolcet-Sanjuan R, Claveria E, Huerta A (1997) Androgenesis in
gous doubled haploid lines produced in our laboratory as Capsicum annuum L.-Effects of carbohydrate and carbon
donor plants will allow us to further improve our protocol. dioxide enrichment. J Am Soc Hort Sci 122:468–475
Dumas de Vaulx R, Chambonnet D, Pochard E (1981) Culture in
To conclude, in comparison with haploid production vitro d’anthères du piment (Capsicum annuum L.): amélioration
methods published earlier for pepper, we have established des taux d’obtention de plantes chez différents génotypes par
an efficient shed-microspore system for Indonesian hot pep- des traitements à +35◦ C. Agronomie 1:859–864
per genotypes. This protocol can be used as a tool for pro- Dumas de Vaulx R, Pochard E (1986) Parthénogenèse et androgenèse
chez le piment. Role actuel dans les programmes de sélection.
ducing doubled haploid plants and it will be implemented Le Sélectionneur Francais 36:3–16
for routine application in breeding programs in Indonesia. George L, Narayanaswamy S (1973) Haploid Capsicum through
We anticipate that the use of haploid technology in local experimental androgenesis. Protoplasma 78:467–470
breeding programs will greatly facilitate hot pepper genetic Gémesné JA, Sági ZS, Salamon P, Somogyi N, Zatykó L, Venzcel
G (1998) Experiences and results of in vitro haploid methods
improvement in Indonesia in the near future. In addition, application in pepper breeding programme. In: Proceedings of
shed-microspore culture can be used as a model system to the Xth Meeting on Genetics and Breeding of Capsicum and
study the widespread problem of defective shoot formation Eggplant, Avignon, France, Sept. 7–11, 1988, pp 201–205
in pepper tissue culture. Gyulai G, Gémesné JA, Sági ZS, Venczel G, Pintér P, Kristóf
Z, Törjék O, Heszky L, Bottka S, Kriss J, Zatykó L (2000)
Doubled haploid development and PCR-analysis of F1 hybrid
Acknowledgements The authors thank Dr K.S. Ramulu, Dr derived DH-R2 paprika (Capsicum annuum L.) lines. J Plant
A.H.M. van der Geest and Dr K.A. Boutilier for critically reading Physiol 156:168–174
the manuscript, Mr J.H.W. Bergervoet for flow cytometry, Dr Harini I, Lakshmi-Sita G (1993) Direct somatic embryogenesis and
P.K. Agarwal for useful suggestions during the initial period plant regeneration from immature embryos of chilli (Capsicum
of investigations, Mr A.H.J. Hermsen and Mr A. Kooijman for annuum L.). Plant Sci 89:107–112
plant care. This work was supported by the research program Hendy H, Pochard E, Dalmasso A (1985) Transmission de la
on “Biotechnology Research Indonesia-Netherlands (BIORIN)”, résistance aux nématodes Meloidogyne chitwood (Tylenchida)
with financial aid from the Royal Netherlands Academy of Arts portée par 2 lignées de Capsicum annuum L.: Étude de
and Sciences (KNAW) and the fellowship program on “Quality descendances homozygotes issues d’androgenèse. Agronomie
for Undergraduate Education Project (QUE)”, Bogor Agricultural 5:93–100
University (IPB), Bogor, Indonesia (IBRD LOAN No.4193-IND) Jo JY, Choi EY, Choi D, Lee KW (1996) Somatic embryogenesis
and plant regeneration from immature zygotic embryo culture
in pepper (Capsicum annuum L.). J Plant Biol 39:127–135
Johansson L, Andersson B, Eriksson T (1982) Improvement of
References anther culture technique: activated charcoal bound in agar
medium in combination with liquid medium and elevated CO2
concentration. Physiol Plant 54:24–30
Abak K, Pochard E, Dumas de Vaulx R (1982) Transmission of Kristiansen K, Andersen SB (1993) Effect of donor plant tem-
resistance to Phytophthora capsici on roots and stems of pepper perature, photoperiod, and age on anther culture response of
plants: study of doubled haploid lines issued from the cross Capsicum annuum L. Euphytica 67:105–109
‘PM217’ × ‘Yolo Wonder’ through anther culture. Capsicum Kyo M, Harada H (1986) Control of the developmental pathway of
Nwsl 1:62–63 tobacco pollen in vitro . Planta 168:427–432
Anonymous (2000a) Survey Pertanian, Produksi Sayuran dan Buah- Lanteri S, Portis E, Bergervoet HW, Groot SPC (2000) Molecular
buahan di Indonesia, 1999 (Agriculture Survey, Production markers for the priming of pepper seeds (Capsicum annuum
of Season Vegetables and Fruit, 1999). Badan Pusat Statistik L.). J Hort Sci Biotechnol 75:607–611
(BPS/Statistics Indonesia), Jakarta Ltifi A, Wenzel G (1994) Anther culture of hot and sweet pepper
Anonymous (2000b) Food and Agriculture Organization of the (Capsicum annuum L.): Influence of genotype and plant growth
United Nations (FAO) Yearbook: Production 1999, vol 53. temperature. Capsicum Eggplant Nwsl 13:74–77
FAO, Rome Maheswary V, Mak C (1993) The influence of genotypes and
Berke T, Shieh SC (2000) Chilli peppers in Asia. Capsicum Eggplant environments on induction of pollen plants for anther culture of
Nwsl 19:38–41 Capsicum annuum L. AsPac J Mol Biol Biotechnol 1:43–50
Büyükalaca S, Mavituma F (1996) Somatic embryogenesis and Mitykó J, Andrásfalvy A, Csilléry G, Fári M (1995) Anther-culture
plant regeneration of pepper in liquid media. Plant Cell Tiss response in different genotypes and F1 hybrids of pepper
Org Cult 46:227–235 (Capsicum annuum L.). Plant Breed 114:78–80
Caranta C, Palloix A, Gebre-Selassie K, Lefebvre V, Moury Morrison RA, Koning RE, Evans DA (1986) Anther culture of an
B, Daubèze AM (1996) A complementation of two genes interspecific hybrid of Capsicum. J Plant Physiol 126:1–9
originating from susceptible Capsicum annuum lines confers Munyon IP, Hubstenberger JF, Phillips GC (1989) Origin of plantlets
a new and complete resistance to pepper veinal mottle virus. and callus obtained from chile anther cultures. In Vitro Cell
Phytopathology 86:739–743 Dev Biol 25:293–296
Christensen HM, Bamford R (1943) Haploids in twin seedlings of Murashige T, Skoog F (1962) A revised medium for rapid growth and
pepper, Capsicum annuum L. J Heredity 34:99–104 bio-assays with tobacco tissue culture. Physiol Plant 15:473–
Cooper GM, Hausman RE (2003) The cell: a molecular approach, 497
3rd edn. Sinauer Associates, Sunderland USA, pp 739 Mythili JB, Thomas P (1995) Some factors influencing the in vitro
Custers JBM, Snepvangers SCHJ, Jansen HJ, Zhang L, Van establishment and callusing of anthers in Capsicum (Capsicum
Lookeren Campagne MM (1999) The 35S-CaMV promoter annuum L. var. Grossum Sendt). Indian J Plant Physiol 38:126–
is silent during early embryogenesis but activated during 130
nonembryogenic sporophytic development in microspore Nitsch JP, Nitsch C (1969) Haploid plants from pollen grains.
culture. Protoplasma 208:257–264 Science 163:85–85
Daubèze AM, Palloix A, Pochard E (1990) Resistance of androge- Ochoa-Alejo N, Ramı́rez-Malagón R (2001) In vitro chili pepper
netic autodiploid lines of pepper to Phytophthora capsici and biotechnology. In Vitro Cell Dev Biol Plant 37:701–729
tobacco mosaic virus under high temperature. Capsicum Nwsl Pechan PM, Keller WA (1988) Identification of potentially embryo-
8–9:47–48 genic microspores in Brassica napus. Physiol Plant 74:377–384
10

Pochard E, Dumas de Vaulx R (1979) Haploid parthenogenesis in Touraev A, Pfosser M, Vicente O, Heberle-Bors E (1996) Stress
Capsicum annuum L. In: Hawkes JG, Lester RN, Skelding AD as the major signal controlling the developmental fate
(eds) The biology and taxonomy of the Solanaceae. Academic tobacco microspores: towards a unified model of induc-
Press, London, pp 455–472 tion of microspore/pollen embryogenesis. Planta 200:144–
Pochard E, Selassié KG, Marchoux G (1983) Oligogenic resistance 152
to potato virus Y pathotype 1-2 in the line ‘Perennial’. Capsicum Touraev A, Vicente O, Heberle-Bors E (1997) Initiation of mi-
Nwsl 2:137–138 crospore embryogenesis by stress. Trend Plant Sci 2:297–
Polsoni L, Kott L, Beversdorf WD (1987) Large scale microspore 302
culture technique for mutation-selection studies in Brassica Touraev A, Heberle-Bors E (1999) Microspore embryogenesis and
napus. Can J Bot 66:1681–1685 in vitro pollen maturation in tobacco. In: Hall RD (ed.) Methods
Qin X, Rotino GL (1993) Anther culture of several sweet and hot in molecular biology, vol.111: Plant cell culture protocols.
pepper genotypes. Capsicum Eggplant Nwsl 12:59–62 Humana Press Inc. Totowa, NJ, pp 281–291
Regner F (1994) Microspore culture of Capsicum annuum. Capsicum Vagera J, Havránek P (1985) In vitro induction of androgenesis
Eggplant Nwsl 13:72–73 in Capsicum annuum L. and its genetic aspects. Biol Planta
Regner F (1996) Anther and microspore culture in Capsicum. In: 27:10–21
Jain SM, Sopory SK, Veilleux RE (eds) In vitro haploid pro- Wang YY, Sun CS, Wang CC, Chien NF (1973) The induction of the
duction in higher plants, vol. 3. Kluwer Academic Publishers, pollen plantlets of triticale and Capsicum annuum from anther
Dordrecht, pp 77–89 culture. Scientia Sinica 16:147–151
Sibi M, Dumas de Vaulx R, Chambonnet D (1979) Obtention de Wolf D, Matzevitch T, Steinitz B, Zelcer A (2001) Why is it difficult
plantes haploı̈des par androgenèse in vitro chez le piment to obtain transgenic pepper plants? Acta Hortic 560:229–
(Capsicum annuum L.). Ann Amélior Plantes 29:583–606 233
Steinitz B, Küsek M, Tabib Y, Paran I, Zelcer A (2003) Pepper Ziauddin A, Simion E, Kasha KJ (1990) Improved plant regeneration
(Capsicum annuum L.) regenerants obtained by direct somatic from shed microspore culture in barley (Hordeum vulgare L.)
embryogenesis fail to develop a shoot. In Vitro Cell Dev Biol cv. Igri. Plant Cell Rep 9:69–72
Plant 39:296–303