Euphytica (2005) 144: 109–117 DOI: 10.1007/s10681-005-5038-x
C
Springer 2005
Genotypic differences in callus formation and regeneration of somatic embryos in Cyclamen persicum Mill. Traud Winkelmann∗ & Margrethe Serek Institute of Floriculture, Tree Nursery Science and Plant Breeding, Section of Floriculture, University of Hannover, Herrenhaeuser Strasse 2, D-30419 Hannover, Germany (∗ author for correspondence: e-mail:
[email protected]) Received 12 January 2005; accepted 6 April 2005
Key words: inheritance, regeneration ability, somatic embryogenesis
Summary This study was performed in order to check the applicability of a protocol for somatic embryogenesis in Cyclamen persicum and to compare the in vitro response of cultivars belonging to different cyclamen growth types and flower colours obtained from two breeders. One hundred ovules per plant were prepared from 32 tested F1 hybrid cultivars represented by five plants each. Callus induction monitored after 8 weeks varied between 42 and 85%. Variation occurred between replications and single plants of a cultivar. Consistency and colour of the developing calluses were diverse. Only in 1 cultivar out of the 32 tested no regeneration of somatic embryos was obtained. The regeneration rates varied from 0 to 65% among cultivars. In the group of midi-type cultivars the regeneration response was lower than that in mini and giganteum types. High regeneration rates were observed in cultivars with flamed flowers. Abbreviations: 2,4-D, 2,4-dichlorophenoxyacetic acid; 2iP, 6-(γ ,γ -dimethylallylamino)purine Introduction In cyclamen (Cyclamen persicum Mill.), an ornamental species with great economic importance mainly in Europe and Japan, vegetative propagation is limited to in vitro culture techniques. Cyclamen breeders are interested in two main applications of vegetative propagation, (i) clonal propagation of breeding lines being parents of F1 hybrids and (ii) mass propagation of selected single plants that carry excellent characters. Besides reports on organogenesis (reviewed by Takamura & Miyajima, 1997; Dillen et al., 1996), efficient protocols for somatic embryogenesis have been published (Wicart et al., 1984; Fukui et al., 1988; Otani & Shimada, 1991; Kiviharju et al., 1992; Kreuger et al., 1995; Takamura et al., 1995; Schwenkel & Winkelmann, 1998). While some protocols included seedling tissues as starting material (Kreuger et al., 1995; Takamura et al., 1995), we have concentrated on the method starting from somatic tissue in ovules
(Schwenkel & Winkelmann, 1998). By applying this method, mass propagation was achievable, especially if liquid culture systems and bioreactors were used (Winkelmann et al., 1998; Hohe et al., 1999a, b, 2001). Stability of plants regenerated via somatic embryogenesis using this protocol was high (Schwenkel & Winkelmann, 1998). About 97% of the regenerants were found to be true-to type. In addition, attempts to develop artificial seeds were undertaken by studying desiccation (Winkelmann et al., 2004a) and encapsulation of somatic embryos (Winkelmann et al., 2004b). However, all these investigations were conducted with one or a few embryogenic cell lines, which can be regarded as being derived from ovules of model genotypes. These embryogenic callus lines were established in the period between 1995 and 1997, and based on cultivars developed in the late 1980s. Since then, cyclamen breeders have developed new cultivars with improved characters, especially with improved uniformity and shortened culture time. Therefore, it is necessary to undertake tissue culture studies with these new
110 cultivars to ascertain their potential as material for mass clonal propagation. An hypothesis for the inheritance of the regeneration ability in cyclamen has been proposed following detailed genetic analyses (P¨uschel et al., 2003). It was postulated that two dominant major genes are responsible for the regeneration ability, so that only genotypes carrying two recessive alleles at both loci are not able to regenerate somatic embryos. This means that a high percentage of genotypes can be expected to be able to develop somatic embryos. However, P¨uschel et al. (2003) reported extended variation in both regeneration rate (i.e. the number of calluses producing somatic embryos) and regeneration intensity (i.e. the number of somatic embryos developed by one ovulederived callus). Also, the number of genotypes forming soft and suspendable embryogenic callus, which can be maintained stably and transferred to liquid culture, is very low. Therefore, data on these traits from diverse genotypes generate understanding of the genetic basis influencing these characters, allowing them to be used more efficiently. The present study was undertaken to analyse the ability to form callus and somatic embryos in 32 F1 hybrid cyclamen cultivars of current commercial importance. Therefore, cultivars were collected from two cyclamen breeders, which covered a wide range of different growth habits and flower colours. Up to 70% of the cultivars grown today are F1 hybrids (Bongartz, 1999). Objectives were to determine the rates of callus formation and regeneration of somatic embryos from these calluses. Results were evaluated in relation to regeneration ability of different genetic pools of two breeders that were reflected in different growth types and different flower colours. Variation in the in vitro response between single plants within the cultivars and between different experiments was studied also.
Materials and methods Plant material Thirty-two F1 cultivars were obtained from two cyclamen breeders, Goldsmith Seeds Europe BV, Andijk, The Netherlands, and S.A.S. Morel Diffusion, Frejus, France, representing the diversity of existing growth types and flower colours of cyclamen (Table 1). Seeds were sown in July 2003 and germinated in growth chambers at 20 ◦ C and 95% relative humidity in the dark. After 4 weeks, seedlings were transferred
to greenhouses and grown at 18 ◦ C/18 ◦ C day/night temperatures until the first flower buds became visible. Thereafter, the night temperature was reduced to 16 ◦ C for flower development. Plants were fertilised according to usual production practice. If plant protection sprays were applied, ovules were not harvested for ten days after the treatment in order to avoid any variation induced by insecticides or fungicides. Preparation and culture of ovules Three to four days before anthesis, unopened flower buds were harvested and surface sterilized as described before (Schwenkel & Winkelmann, 1998). Ovules were prepared and 25 each were incubated at 24 ◦ C in the dark in Petri dishes (6 cm) with callus induction medium consisting of half strength macro- and micro nutrients according to Murashige & Skoog (1962), Fe EDTA full strength, 30 g l−1 sucrose, 2 g l−1 glucose, 250 mg l−1 peptone, 0.5 mg l−1 nicotinic acid, 0.1 mg l−1 thiamine-HCl, 0.5 mg l−1 pyridoxine-HCl, 100 mg l−1 myo-inositol, 9.05 µM 2,4-D (2,4-dichlorophenoxyacetic acid) and 3.94 µM 2iP (6-(γ ,γ -dimethylallylamino)purine) (Schwenkel & Winkelmann, 1998). The developing calluses were sub-cultured to fresh callus induction medium in 150 ml glass vessels containing 25 ml medium and cultured for an additional 8 weeks, after which the calluses were transferred to hormone-free medium in vessels of the same size for differentiation of somatic embryos. After 8 weeks of culture on this medium the final evaluation was performed. Experimental set-up and evaluations Five plants from each cultivar were selected randomly for in vitro screening. One hundred ovules, which were derived from two flower buds, were prepared from each plant. In the first experiment, environmental effects were studied by comparing the in vitro response of ovules prepared at different dates. Ten plants, five each from two cultivars 16 and 17 (Table 1) were used for this experiment. Evaluation was conducted as described below. For this experiment, analyses of variance were performed to test differences between the replications (Table 2). The aim of the second experiment was to investigate the in vitro response of a wide spectrum of cultivars. All 32 cultivars available were included in this experiment. After the initial culture passage over
111 Table 1. Characterization of cultivars included in the studies and regeneration of somatic embryos in the primary culture on 2,4-D containing medium Regeneration of somatic embryos (%) after
Internal cultivar no.
Series name
Cultivar name/flower colour
Growth type
Breeder
8 weeks
16 weeks
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
Robusta Sierra Sierra Sierra Sierra Laser Laser Laser Laser Laser Miracle Midori Silverado Silverado Miracle Miracle Laser Metis Metis Metis Metis Metis Halios Halios Halios Halios Halios Latinia Latinia Salmon Latinia Latinia Latinia
Salmon White Purple Scarlet (red) Fuchsia Orchid White Rose Purple Flame Scarlet (red) Deep Salmon Purple White Scarlet (red) White&eye Purple Purple Salmon Rose Scarlet Red Purple Evolution Deep Magenta Pure White Deep Rose Pure White Compact Magenta Flame Purple Flame Scarlet Salmon Scarlet Red Salmon Wine Red Pure White Compact ‘Salmon Flame’
Giganteum Giganteum Giganteum Giganteum Giganteum Midi Midi Midi Midi Midi Mini Mini Mini Mini Mini Mini Midi Mini Mini Mini Mini Mini Giganteum Giganteum Giganteum Giganteum Giganteum Midi Midi Midi Midi Midi
Goldsmith Seeds Goldsmith Seeds Goldsmith Seeds Goldsmith Seeds Goldsmith Seeds Goldsmith Seeds Goldsmith Seeds Goldsmith Seeds Goldsmith Seeds Goldsmith Seeds Goldsmith Seeds Goldsmith Seeds Goldsmith Seeds Goldsmith Seeds Goldsmith Seeds Goldsmith Seeds Goldsmith Seeds Morel Morel Morel Morel Morel Morel Morel Morel Morel Morel Morel Morel Morel Morel Morel
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.2 0 0 1.1 0 0 0 0 0 0 0.3 0 0 0 0
7.8 0 0 0 0 0 5.2 0 8.6 1.9 1.7 0.3 0.7 2.2 2.2 38.9 48.8 21.2 0 1.5 1.9 13.6 0.4 5.6 1.1 6.7 33.6 13.4 1.6 0 0 9.3
Table 2. Comparison of two replications regarding the in vitro response of two cultivars (16 and 17) Cultivar no.
In vitro response
Mean square
F-value
P-value
16
Callus formation Regeneration of somatic embryos
115.6 176.4
0.564 0.676
0.474 0.435
17
Callus formation Regeneration of somatic embryos
1464.1 122.5
1.495 0.133
0.256 0.723
Results of ANOVA including data of five plants per cultivar.
8 weeks, rates of contamination, callus formation and regeneration of somatic embryos were recorded. The percentage of callus formation was calculated based on the number of contamination-free ovules. Since pronounced differences in callus size were observed, two sizes, calluses of 2–5 mm diameter and those bigger than 5 mm in diameter were distinguished. Regeneration of somatic embryos on the callus induction medium was evaluated after 8 and 16 weeks. At the final evaluation after 24 weeks, the regeneration rate was calculated from the number of calluses which regenerated somatic embryos based as a proportion
112 of the total number of calluses. In order to show variation in regeneration intensity, calluses with 10 or less somatic embryos and calluses with more than 10 somatic embryos are presented separately (Figure 1b). For statistical analyses cultivars were grouped into six classes according to growth types (giganteum, midi, mini) and breeder (Goldsmith, Morel). Within each of these six classes, ANOVA and pairwise comparisons of means were performed taking the five individual genotypes of each cultivar as replicates. Means were compared using Tukey’s test at α ≤ 0.05 using SAS software. Results Variation in the in vitro response between replications Ovules of 10 plants were taken into culture in two replications prepared at different times in order to test the reproducibility of callus induction and somatic embryo formation. There was no significant difference in in vitro response between replications (Figure 1, Table 2). Regarding callus formation, the results of both replications were corresponding well (Figure 1 a), only in plant 17/1 there was a difference between the first (64% callus forming ovules) and the second replication (23%). The ratio of big to smaller calluses was similar in both replications, although larger variations occurred in some plants (16/3, 17/1,17/2, Figure 1a). Although there was no significant difference in regeneration of somatic embryos between replications (Table 2), there was a large difference between plants ranging from 4% in replicate 1 of plant 16/2 to 100% in replicate 2 of plant 16/5 (Figure 1b). Nevertheless, all genotypes showed an ability to regenerate somatic embryos and the ranking of the different genotypes was similar for both replicates. In conclusion, differences between replicates were observed, but these had no significant effect on callus formation and regeneration. On the other hand, considerable differences in the in vitro response were recorded between single plants within a cultivar. Therefore, a high number of plants per cultivar were used in the subsequent screening. Thus, in the following experiment 32 cultivars represented by 5 plants each were analysed. Variation in callus formation between cultivars Cultivars were grouped by growth type and breeder (Figures 2 and 3). All 160 tested plants showed callus
formation (Figure 2). Particularly low rates of callus formation were found in the cultivar ‘Latinia Wine Red’. But many ovules of this cultivar were contaminated by bacteria and were therefore not included in subsequent calculations. This high contamination rate was unusual, for most other plants few if any ovules had to be discarded for this reason. Mean callus induction rates for the giganteum-type cultivars from Goldsmith and Morel were 61 and 49%, respectively. There was considerable variation in callus initiation among plants of one cultivar within the giganteum group (see standard deviation in Figure 2), one exception was cultivar ‘Sierra Scarlet’. Overall, callus formation on ovules of plants from the midi cultivars was slightly higher than in the giganteum group and there was less variation in initiation (Figure 2). For Goldsmith cultivars, an average of 65% callus formation was calculated, and for Morel cultivars it was 66% if the severely contaminated cultivar ‘Latinia Wine Red’ was not included. Ovules from mini-type cultivars produced callus at high rates in many cultivars (Figure 2) with little difference between plants from breeders (69 and 64% in average for cultivars bred by Goldsmith and Morel, respectively). Some variation existed in callus initiation between single plants of most cultivars, except for the cultivar ‘Metis Purple Evolution’ which showed a uniform response. In summary, high rates of callus formation were observed for most cultivars in all growth type groups and gene pools of both breeders. Calluses varied in colour and consistency. Of special importance was the formation of soft and embryogenic callus, which is necessary for establishing liquid cultures and mass propagation. Since this evaluation of callus types needs several subcultures, these investigations are in progress. Soft embryogenic callus lines were obtained from ovules of eight cultivars so far, but their maintenance and embryogenic potential have yet to be determined. Variation in the ability to regenerate somatic embryos among cultivars Somatic embryo formation was rarely observed in three genotypes during primary culture on callus induction medium (Table 1), but calluses of 23 cultivars formed somatic embryos in the second subculture in medium containing 2,4-D (Table 1). Most cultivars showed low regeneration rates, except for cultivars 16, 17, and 30 with regeneration rates of 38.9, 48.8 and 33.6%, respectively.
113
Figure 1. Effect of different times of ovule preparation (replication 1 and replication 2) on callus formation (a) and regeneration of somatic embryos (b) of 10 plants of 2 cultivars (‘Miracle Purple’ = 16/1–16/5, and ‘Laser Purple’ = 17/1–17/5).
After an additional culture passage on hormonefree medium regeneration of somatic embryos was analysed in more detail. Taking at first only the ability for somatic embryogenesis into account, all 32 tested
cultivars except one, ‘Laser Rose’, produced somatic embryos on ovule-derived calluses in all five tested plants (Figure 3). Single plants out of several other cultivars also did not show formation of somatic embryos,
114
Figure 2. Callus formation from ovules of 32 cyclamen cultivars (bars indicate means of five genotypes, whiskers indicate standard deviations. Within one group of cultivars defined by growth type and breeder, comparisons of means were carried out, bars with the same letters do not differ significantly in Tukey’s test at α = 0.05).
but from the 160 plants tested in total, 135 were able to regenerate. In giganteum cultivars, regeneration of somatic embryos was observed in most plants, but differences existed between cultivars in the regeneration rates. Cultivars from Goldsmith expressed relatively low regeneration rates (average was 17%) and intensities (number of somatic embryos per callus), the best responding being ‘Robusta Salmon’ and ‘Sierra Scarlet’ (Figure 3). In contrast, the giganteum cultivars of Morel had an average regeneration rate of 39% with cultivars ‘Halios Magenta Flame’ and ‘Halios Purple Flame’ having significantly higher regeneration rates of 53 and 49%, respectively, than the remaining cultivars in this group. In the group of midi types regeneration rates were much lower than in the giganteum group, with the exception of cultivar ‘Laser Purple’ (Figure 3). Mean regeneration rates were 16 and 17% for cultivars from Goldsmith and Morel, respectively. Variation in regeneration rates existed between and within cultivars of mini cyclamen (Figure 3) with
higher rates recorded for plants from cultivar ‘Metis Salmon Rose’ (Figure 3). Goldsmith mini cultivars had an average regeneration rate of 15% compared to 30% for Morel cultivars. Callus formation was homogenous and not influenced strongly by flower colours (Table 3). However, regeneration of somatic embryos was not evenly distributed over the colour groups (Table 3). Although considerable variation occurred, flamed cultivars tended to have the highest ability to produce somatic embryos (mean regeneration rate of 36%), while rose and fuchsia coloured cultivars had the lowest regeneration rates of 7 and 2%, respectively. Discussion The different reaction of genotypes of one species to in vitro culture conditions has been described frequently and is thought to be due to different endogenous phytohormone levels or differing ability to react to exogenously applied growth regulators. In the present
115
Figure 3. Regeneration of somatic embryos on ovule-derived calluses of 32 cyclamen cultivars (bars indicate means of 5 genotypes, whiskers indicate standard deviations. Within one group of cultivars defined by growth type and breeder, comparisons of means were carried out, bars with the same letters do not differ significantly in Tukey’s test at α = 0.05).
study, genotypic differences in the in vitro response were verified for different cyclamen cultivars, and for single plants within F1 hybrid cultivars. Part of this variation has to be assigned to environmental effects, Table 3. Analysis of callus formation and regeneration of somatic embryos for 31 cyclamen cultivars grouped by their flower colour Callus formation [%]
Regeneration of somatic embryos [%]
Flower colour group
Number of cultivars in Standard Standard this group Mean deviation Mean deviation
White Rose Salmon Scarlet Purple Flamed Fuchsia Magenta
7 3 4 6 5 4 1 1
66.6 57.3 57.5 61.0 68.6 53.3 65 64
8.5 9.2 8.6 13.7 5.4 14.0 0 0
18.0 6.7 26.0 21.7 23.0 36.0 2.0 20.0
12.2 5.3 24.6 12.0 17.2 15.7 0 0
as was shown by the comparison of two replications (Figure 1). In previous studies, environmental effects have been analysed in more detail (P¨uschel, 2000). Working with the same protocol, P¨uschel (2000) reported on seasonal differences as well as additional factors influencing callus induction and somatic embryo formation. Since many parameters have an effect of plant growth in vitro, high deviations between replications are the rule in these types of studies, as described for somatic embryogenesis in anther cultures of Brassica oleracea var. gemmifera (Ockendon, 1985; Ockendon & Sutherland, 1987). All genotypes of cyclamen formed callus on the ovules, although the consistency and the regeneration ability of these calluses varied (Figures 2 and 3). These results, as well as the low losses due to contaminations, confirm the high suitability of ovules as starting material for establishing in vitro cultures in cyclamen. In many genotypes, differentiation of somatic embryos was observed in the presence of 2,4-D on callus induction medium (Table 1). These genotypes
116 did not require the transfer to hormone-free medium for undergoing the realisation phase of somatic embryogenesis. Possible explanations are, that 2,4-D was depleted from the medium during the 8 weeks of culture, or that these genotypes were less sensitive to 2,4-D. No correlation was detected between callus formation and somatic embryo regeneration, similar to the results of Schween & Schwenkel (2003) who did not find a correlation between callus formation and organogenesis in Primula genotypes. Most important from the point of plant propagation was the evaluation of the regeneration ability. Thirtyone out of 32 cultivars were able to form somatic embryos, a trait that seems to be widely present in cyclamen with different genetic backgrounds. Corresponding results were found in previous studies with breeding lines and cultivars of cyclamen, where 29 out of 30 tested single plants were classified as being able to regenerate (Schwenkel & Winkelmann, 1998). On the other hand, Takamura & Tanaka (1996) observed the differentiation of somatic embryos in only 5 out of 13 cyclamen cultivars, but they used a different regeneration protocol starting from etiolated petioles. If the hypothesis for inheritance of regeneration ability via somatic embryogenesis in cyclamen is taken into account (P¨uschel et al., 2003, see Introduction), then it would be expected that only a relatively low percentage of genotypes would not undergo somatic embryogenesis. The present results should assist in the development and testing of genetic markers that are currently under investigation in the segregating populations characterized by P¨uschel et al. (2003). Such markers and a more detailed understanding of the physiological factors leading to high regeneration response will improve the applicability of the protocol. Besides the classification into plus- and minustypes the data show significant differences between cultivars in their regeneration rate. Mini and giganteum cultivars had a tendency to have higher regeneration rates than the midis, but these differences were not significant. There was no affect in regeneration capability among cultivars derived from different breeders programmes (ANOVA: p = 0.051), mainly because of the variability between cultivars belonging to the same series. These cyclamen series are selected by phenotype rather than by their genetic background. It should be noted that regeneration rates in vitro do not correlate with any important horticultural traits; all cultivars from both breeders included in this study were of high standard in germination, growth and flowering.
When cultivars were grouped by flower colour differences in the regeneration response became obvious, with high rates found in cultivars with flamed flowers. This result is particularly interesting, since the model genotype used in most previous studies (Hohe et al., 1999a, b, 2001; Winkelmann et al., 2004a, b) was also a cultivar with purple flamed flowers. The reasons for the high regeneration response of cultivars with flamed flowers remain unclear and have to be verified in future analyses with cultivars of different genetic background. Acknowledgments This laborious screening would not have been realized without the excellent technical assistance of Annette Steding and Wiltrud von Oertzen, which is highly appreciated. The authors are grateful to the breeders Goldsmith Seeds Europe BV and S.A.S. Morel Diffusion for providing the seeds. The authors would like to thank Dr. Sridevy Sriskandarajah (KVL Copenhagen, Denmark) and Prof. emeritus E.W. Hewett (Massey University, Palmerston North, New Zealand) for critical review of the manuscript. References Bongartz, W., 1999. Cyclamen. Thalacker-Medien, Braunschweig, pp. 42–50. Dillen, W., I. Dijkstra & J. Oud, 1996. Shoot regeneration in longterm callus cultures derived from mature flowering plants of Cyclamen persicum Mill. Plant Cell Rep 15: 545–548. Fukui, H., T. Yamamoto, T. Asano & M. Nakamura, 1988. Effect of plant growth regulators on in vitro organogenesis of cyclamen (Cyclamen persicum Mill.). Res Bull Fac Agric Gifu Univ 53: 139–145. Hohe, A., T. Winkelmann & H.-G. Schwenkel, 1999a. CO2 accumulation in bioreactor suspension cultures of Cyclamen persicum Mill. and its effect on cell growth and regeneration of somatic embryos. Plant Cell Rep 18: 863–867. Hohe, A., T. Winkelmann & H.-G. Schwenkel, 1999b. The effect of oxygen partial pressure in bioreactors on cell proliferation and subsequent differentiation of somatic embryos of Cyclamen persicum Mill. Plant Cell Tiss Org Cult 59: 39–45. Hohe, A., T. Winkelmann & H.-G. Schwenkel, 2001. Development of somatic embryos of Cyclamen persicum Mill. in liquid culture. Gartenbauwissenschaft 66: 219–224. Kiviharju, E., U. Tuominen & T. T¨orm¨al¨a, 1992. The effect of explant material on somatic embryogenesis of Cyclamen persicum Mill. Plant Cell Tiss Org Cult 28: 187–194. Kreuger, M., E. Postma, Y. Brouwer & G.-J. van Holst, 1995. Somatic embryogenesis of Cyclamen persicum in liquid medium. Physiol Plant 94: 605–612. Murashige, T. & F. Skoog, 1962. A revised medium for rapid growth and bioassay with tobacco tissue cultures. Physiol Plant 15: 473– 497.
117 Ockendon, D.J., 1985. Anther culture in Brussels sprouts (Brassica oleracea var.gemmifera) II. Effect of genotype on embryo yields. Ann Appl Biol 107: 101–104. Ockendon, D.J. & R.A. Sutherland, 1987. Genetic and non-genetic factors affecting anther culture of Brussels sprouts (Brassica oleracea var.gemmifera) Theor Appl Genet 74: 566–570. Otani, M. & T. Shimada, 1991. Somatic embryogenesis and plant regeneration from Cyclamen persicum MILL. leaf cultures. Plant Tiss Cult Lett 8: 121–123. P¨uschel, A.K., 2000. Zur genetischen Analyse der Kallusbildung und somatischen Embryogenese in vitro bei Cyclamen persicum Mill. PhD Thesis, University of Hannover. P¨uschel, A.K., H.-G. Schwenkel & T. Winkelmann, 2003. Inheritance of the ability for regeneration via somatic embryogenesis in Cyclamen persicum Mill. Plant Cell Tiss Org Cult 72: 43– 51. Schween, G. & H.-G. Schwenkel, 2003. Effect of genotype on callus induction, shoot regeneration, and phenotypic stability of regenerated plants in the greenhouse of Primula spec. Plant Cell Tiss Org Cult 72: 53–61. Schwenkel, H.-G. & T. Winkelmann, 1998. Plant regeneration via somatic embryogenesis from ovules of Cyclamen persicum Mill. Plant Tiss Cult Biotechnol 4: 28–34.
Takamura, T. & M. Tanaka, 1996. Somatic embryogenesis from the etiolated petiole of cyclamen (Cyclamen persicum Mill.). Plant Tiss Cult Lett 13: 43–48. Takamura, T. & I. Miyajima, 1997. Micropropagation of Cyclamen persicum Mill. In: Y.P.S. Bajaj (Ed.), Biotechnology in Agriculture and Forestry, Vol. 40, High-tech and Micropropagation IV. Springer, Berlin Heidelberg, pp. 96–112. Takamura, T., I. Miyajima & E. Matsuo, 1995. Somatic embryogenesis of Cyclamen persicum MILL. ‘Anneke’ from aseptic seedlings. Plant Cell Rep 15: 22–25. Wicart, G., A. Muras & A. Lutz, 1984. Histological study of organogenesis and embryogenesis in Cyclamen persicum Mill. tissue cultures: Evidence for a single organogenic pattern. Protoplasma 119: 159–167. Winkelmann, T., A. Hohe & H.-G. Schwenkel, 1998. Establishing embryogenic suspension cultures in Cyclamen persicum ‘Purple Flamed’. Adv Hort Sci 12: 25–30. Winkelmann, T., L. Meyer & M. Serek, 2004a. Desiccation of somatic embryos of Cyclamen persicum Mill. J Horticult Sci Biotechnol 79(3): 479–483. Winkelmann, T., L. Meyer & M. Serek, 2004b. Germination of encapsulated somatic embryos of Cyclamen persicum. HortScience 39: 1093–1097.