Genetic analysis of anther and leaf disc culture in two clones of Solanum chacoense Bitt. and their reciprocal hybrids

Two diploid clones of self-incompatible Solanum chacoense Bitt. with androgenetic ability were tested for anther and leaf disc culture response together with eight of their reciprocal F1 hybrids. Large differences were found among genotypes in frequency of anther induction as well as in the phase of plant regeneration. Anthers harvested in June showed a significantly higher percentage of response (17.5%) at the induction phase than those collected in July (13.8%) or August (12.7%). The lowest induction frequency was observed in May (7.3%). By contrast, plant regeneration from induced anthers did not vary during this time. Genotypic differences were also observed in leaf disc response. The two parental clones and two of their hybrids failed to produced any shoots. Among the remaining genotypes, two had only sporadic occurrence of shoot formation, two gave an intermediate response (15% and 24% of their discs carried shoots), whereas the discs of the two remaining genotypes responded well to culture (68% and 77%). The genetic analysis performed on the reciprocal hybrids revealed that a positive significant correlation existed between anther induction and leaf disc response (Spearman's r=0.82; p=0.01). This suggests that, under our conditions, these two aspects of tissue culture might share a common system of genetic control. Estimates of broad sense heritabilities, for leaf disc culture, 83% were obtained and the number of effective factors involved in the control of tissue culture response, indicated a relatively simple genetic control. Finally, considering the potentialities opened by the use of RFLP analysis, it might be possible to find probes that are linked with genes involved in tissue culture competence.     

Sugarbeet leaf disc culture: an improved procedure for inducing morphogenesis

In preparation for gene transfer experiments we investigated factors that might affect the production of shoots and somatic embryos from the wound callus of cultured sugarbeet leaf discs. A complex interaction was found between the leaf disc plating density, the disc culture medium, the source-shoot culture medium and the frequency of disc transfer to fresh medium. The most productive protocol utilized: source shoots maintained on MS medium containing 0.25 mg 1^-1 BA; multiple leaf discs (ten 4-mm discs/plate) plated onto an enriched modification of MS medium (RV) containing 1.0 mg 1^-1 BA and solidified with 0.3% Gelrite (not permitted to dry during hardening); and transfer of the discs to fresh medium every two weeks during the first month. This standard protocol produced more callus per plate and higher rates of morphogenesis per unit dry weight of callus than did the one-step method of Saunders and Doley. Water availability considerations were found to be critical to obtaining high morphogenic rates. Root induction frequency and quality was superior on shoots transplanted to MS medium containing 1 mg 1^-1 NAA as the sole growth regulator compared to IAA at the same concentration.Abbreviations BA N⁶-benzyladenine - IAA indole-3-acetic acid - NAA -naphthaleneacetic acid     

Genetic control of in vitro shoot regeneration from leaf explants inSolanum chacoense Bitt.

Although the heritable nature of plant tissue culture responses is now well documented in many species, only a few studies have been conducted to elucidate complete inheritance patterns. Genetic control of in vitro shoot regeneration from leaf explants was investigated inSolanum chacoense using parental, F₁ and F₂ generations. Broad-sense heritability estimates were high for frequency (percentage) of responsive leaf explants (61–83%) and number of shoots regenerated per responsive explant (53–75%). Consistent with high heritability estimates, a hypothesis involving three genes could be formulated to explain the variability in the response observed in this study. This model implies that homozygous recessive alleles at any two (out of three) loci are required for the highest response, i.e., more than two shoots per explant in more than 40% of the explants. The presence of homozygous recessive alleles at any one of the three loci produces an intermediate response, i.e., fewer than 40% of the explants regenerating fewer than two shoots per explant, and a dominant allele at all the three loci results in non-responsiveness. Additional minor modifier genes, each with a small effect, would also be required to account for the variable intensity of regeneration within groups. Such a relatively simple genetic control of in vitro regenerability suggests that incorporation of this trait should be easy in potato improvement programmes.     

Genetic studies of corn (Zea mays L.) anther culture response

Summary Anthers of two maize (Zea mays L.) inbred lines, DBTS (P₁) and B73 (P₂), their F₁, F₂ and first backcross generations — F₁ x DBTS (B₁), and F₁ x B73 (B₂) — were float cultured in YP medium to study the inheritance of corn anther culturability using generation mean analysis. Significant effects of generation were observed for the three traits measured: anther response (%), frequency of embryos (%) and anther productivity. Variation among the generations was similar for anther response and frequency of embryos: no significant differences were found among the P₁, F₁, F₂ and B₁ means, but the means of P₂ and B₂ were significantly lower than those of the other generations. For anther productivity, the F₂ generation tended to have a slightly higher tendency for multiple embryo formation. A simple additive-dominance model was adequate in explaining the inheritance of anther response and frequency of embryos, but digenic epistasis (additive x dominance) was involved in the inheritance of anther productivity. Additive genetic variance was higher than non-additive genetic variance for all the traits; however, only environmental variance was significant. Narrow-sense heritability estimates were 65% and 75% for anther response and frequency of embryos, respectively. Significant inter-plant variation was observed within generations, even for the inbred line DBTS, but isozymic analysis involving five enzyme loci did not reveal any genotypic variability within the inbred lines DBTS and B73.     

Acquisition of competence for shoot regeneration in leaf discs ofSaintpaufla ionantha xconfusa hybrids (African violet) cultured in vitro

Leaf discs fromSaintpaulia ionantha xconfusa hybrids (African violet) were transferred between basal medium (BM) containing no hormones and shoot-inducing medium (SIM) containing 2.0 mg 1^–1 indole acetic acid and 0.08 mg l^–1 6-benzylaminopurine to determine whether there is a window of competence for shoot regeneration. Leaf discs precultured on BM prior to transfer to SIM formed buds 3 days earlier than the controls (leaf discs not precultured) regardless of whether the discs were placed upside down or right side up on the medium. This suggests that cultured leaf cells were not competent for shoot induction during the first 3 days of culture. Leaf discs cultured right side up (abaxial surface to the medium) did not form buds on BM alone, unlike discs cultured upside down. Leaf disc survival was affected by a delay in hormonal exposure, but surviving leaf discs produced as many shoots as control leaf discs. This suggests that in the absence of exogenous plant hormones, cellular competence to regenerate shoots is not lost in excised leaf discs of African violet.Abbreviations BM Basal medium - SIM_Sho Shoot-inducing medium     

Accelerated production and identification of fertile,homozygous transgenic wheat lines by anther culture

Anther culture has been developed in the winter wheat cultivar Florida to achieve accelerated production and identification of homozygous transgenic lines. With untransformed, seed-derived plants to develop the culture system, it was shown that cold pre-treatment of spikes excised from donor plants and addition of 2,4-dichlorophenoxyacetic acid together with either kinetin or 6-benzylaminopurine in the callus induction medium improves the anther culture response. The procedure developed allowed production of fertile homozygous lines within 8–9 months, which includes an 8-week vernalisation period. With transgenic wheat plants produced by particle bombardment as donors, we show that the system can be used to produce homozygous transgenics, requiring one generation cycle. Both T₀ tissue culture-derived plants and their T₁ seed-derived descendents serve as suitable donors. We show that an anther culture response comparable to that of untransformed, seed-derived plants can be achieved with T₀ tissue culture-derived plants. PCR and Southern molecular analyses of anther culture-derived transgenics show that the transgenes are stably inherited; there are no perturbations at the chromosomal level around the sites of transgene integration as a result of in vitro chromosome manipulation during anther culture.     

Gametic selection in anther culture of rice (Oryza sauva L.)

Summary The segregation and recombination of heterozygous isozyme markers have been monitored in anther culture derivatives (i.e., six nonmorphogenic microspore-derived callus [NMC] populations and two anther culture plant [ACP] populations) and F₂ plants generated from six F₁ hybrids of rice, including five japonica upland/improved indica tropical hybrids. The alleles in excess at some loci displaying skewed segregations in the F₂s were consistently overrepresented in the NMC populations. These alleles were also generally found to be overabundant in the two ACP populations except for certain loci that contrastingly segregated in a 11 ratio. Additional distortions were found to be specific to AC derivatives indicating the existence of in vitro gametic selection. Overall, however, the gametic selection in the ACP materials was neutral with regard to the indica and japonica differentiation. Estimates of linkages between markers borne by chromosome 6 using AC-derivative data were consistent with those noted in the F₂s and with current knowledge of the isozyme locus linkage map. Given the average neutrality of gametic selection and the consistency of linkage relationships in the ACPs, their further use as rice molecular mapping and gene tagging populations can be investigated with confidence.Joint contribution with Research Institute for Tropical Food Crops, Department of the International Centre of Agronomical Research for Development (IRAT-CIRAD), 45bis avenue de la Belle Gabrielle, F-94736 Nogent s/Marne Cédex, France and International Rice Research Institute (IRRI), P.O.Box 933, Manila, Philippines     

Genetic studies of rice (Oryza sativa L.) anther culture response

Anthers of two rice (Oryza sativa L.) varieties, Lunhui 422 (P₁) and Jinzao 5 (P₂), their F₁, F₂ and first backcross generation-F₁ x Lunhui 422 (B₁), and F₁ x Jinzao 5 (B₂)-were cultured in L8 medium to study the inheritance of rice anther culturability using generation mean analysis. Significant effects of generation were observed for the four traits measured: anther response (%), frequency of callus induction (%), frequency of callus differentiation (%) and culture efficiency (%). Variation among the generations was similar for all traits: significant differences were found among six generations and the means of the P₂ and B₂ were significantly lower than those of the other generations. The frequency of callus differentiation showed a nonsignificant difference among the P₁, F₁, F₂ and B₁ generations which had slightly high values than the P₂ and B₂. Additive genetic variance (V_A) was higher than non-additive genetic variance (V_D) for anther response and frequency of callus induction. However, A_V was lower than V_D for frequence of callus differentiation and culture efficiency, V_A was significant for all traits except for the culture efficiency, and V_D was nonsignificant for all traits except for the frequency of callus differentiation. On the other hand, environmental variation (V_E) was significant for the 4 traits. Narrow-sense heritability estimates were 95.52%, 82.19% and 13.54% for anther response, frequency of callus induction and culture efficiency, respectively.Abbreviations 2,4-d 2,4-dichlorophenoxyacetic acid - IAA indole-3-acetic acid     

The response of anther culture to culture temperature in Triticum aestivum

Summary The response of anther culture to culture temperature was studied in detail using many varieties, F₁ hybrids and pollen-derived lines of wheat (Triticum aestivum) as materials. The suitable culture temperature for inducing pollen callus (or embryoids) in wheat anther culture ranged from 26 °C to 30 °C, varying with genotypes. But for the great majority of wheat genotypes the suitable culture temperatures lay between 28 °C and 30°C. The most significant genotypic variation in the response to culture temperature was observed in the comparison between the culture at 33 °C for eight days followed by culture at 25 °C (or 26 °C) and the continuous culture at 25 °C (or 26 °C). This genotypic variation in the response to culture temperature is a heritable character which may be controlled by multiple genes. The effect of culture at 30 °C for eight days followed by culture at 26 °C was similar to, or in some cases, better than that of continuous culture at 28 °C, and the effect of culture at 32 °C for eight days followed by culture at 28 °C was similar to that of continuous culture at 30 °C. In the range from 26 °C to 32 °C, the overwhelming majority of pollen calli emerged before the 40th day after anther inoculation, and the higher the culture temperature, the earlier and more concentrated the emerging period of the pollen callus. The pollen callus obtained at high temperatures above 28 °C should be transferred in time onto the regeneration medium at 25°–27°C to induce shoots.     

Shoot regeneration of mesophyll protoplasts transformed by Agrobacterium tumefaciens,not achievable with untransformed protoplasts

Summary Alternative methods for shoot regeneration in protoplast derived cultures were developed in Nicotiana paniculata and Physalis minima. In both species protoplast derived callus is not regeneratable to shoots by conventional methods, e.g. hormone treatment. Leaf discs and stem segments of N. paniculata and P. minima were incubated with Agrobacterium tumefaciens shooter strains harbouring pGV 2215 or pGV 2298 or wildtype strain B6S3. After 36 h of co-incubation protoplasts were prepared. (Leaf disc and stem segment cloning). Co-cultivation experiments were also undertaken with protoplasts of both species. Transformed clones, characterized by their hormone independent growth and octopine production, could be isolated after about two months. Transformation frequencies of leaf disc and stem segment cloning and co-cultivation experiments varied from 5×10^–3 to 5×10^–5. After about one year of cultivation on hormone-free culture medium, shoots could be recovered from colonies of N. paniculata, transformed by the strain harbouring pGV 2298. In protoplast derived colonies of P. minima, shoot induction was obtained only after transformation by bacteria carrying pGV 2215. This demonstrates the importance of the particular shooter mutant, as well as the response of the host plant. Transformed shoots of P. minima produced octopine, whereas octopine production in transformed shoots and callus of N. paniculata was undetectable after one year of cultivation, though T-DNA was still present in the plant genome. Transformed shoots of N. paniculata and P. minima do not produce any roots. Shoots of N. paniculata have an especially tumerous phenotype. Shoots of both species were successfully grafted to normal donor plants of N. tabacum.Abbreviations B5-h Gamborg medium without hormones (Gamborg 1968) - V47 protoplast medium (Binding 1974) - D2a protoplast medium (Li et al. 1980) - MS-h Murashige and Skoog medium without hormones (Murashige and Skoog 1962) Dedicated to Professor Dr. G. Melchers in occasion of his 80th birthday     

Inheritance and mapping of embryo sac abortion in hybrid between Indica and Japonica rice (Oryza sativa L.)

A doubled haploid (DH) population derived from anther culture of F₁’s between an Indica var. Zhai-Ye-Qing 8 and a Japonica var. Jing-Xi17 as well as two backcross populations derived from this DH population were used to investigate inheritance of the embryo sac abortion at early megagametogenesis occurring in Indica/Japonica rice crosses. Two major loci, dominant and complementaryesa-1 andesa-2, located on chromosomes 6 and 12 respectively, were detected. Genetic analysis indicated that embryo sac fertility is mainly regulated by the gametophytic genotype at these two loci.     

Somatic embryogenesis and plant regeneration from isolated protoplasts of Lavatera thuringiaca

Embryogenic cell suspensions of Lavatera thuringiaca L. were established from leaf petiole and shoot regeneration was achieved when cells were plated on medium without growth regulators. We tested three methods for protoplast culture, isolated from a one-year old embryogenic cell suspension, to determine the best conditions for L. thuringiaca protoplast culture and shoot regeneration. The highest protoplast plating efficiency was obtained with the agaroseembedded method, reaching 30%, while the nursing culture method gave 5% when the protoplasts were plated over Whatman paper No. 2. However, the same nursing culture failed to produce protoplast-derived microcalluses when the protoplasts were plated on a nitrocellulose filter. The liquid thin layer method gave the lowest plating efficiency with only 0.5%. Shoot regeneration from protoplast-derived microcalluses was achieved in two steps; first, globular embryo development was favored in medium low in auxin (2,4-d and BA at 0.01 and 0.05 mg 1^-1, respectively), second, the globular embryos further differentiate into shoots in medium without growth regulators or in medium containing GA₃ (0.5 to 1.0 mg 1^-1).Abbreviations 2,4-d 2,4-dichlorophenoxyacetic acid - BA benzyladenine - GA₃ gibberellic acid - NAA -naphthaleneacetic acid - IBA indole-3-butyric acid     

Effect of genome, induction medium and temperature pretreatment on green plant production in durum ( Triticum turgidum var. durum ) x bread wheat ( Triticum aestivum L. em Thell) crosses

The recalcitrancy of durum wheat (Triticum turgidum var. durum) to anther culture, was attempted to be overcome by transferring the responsible genes form bread wheat B-genome to the respective on durum wheat, determining an appropriate induction medium and clarifying the necessity of cold pretreatment. For this, three durum wheat cultivars were crossed to two bread wheat (Triticum aestivum L. em Thell) cultivars. The resulting F₁ plants and their original cultivars were grown in the field and anthers at the appropriate microspore stage were cultured on potato-2 and W₁₄ media with and without low temperature pretreatment. No green plants were produced from the parental durum wheat cultivars. In contrast, green plants were produced from the F₁ plants. The best results in three of the four F₁ hybrids were recorded when potato-2 was used as induction medium. A more variable response of the examined genotypes was noticed with respect to temperature pretreatment. Regarding green plant production, a negative effect of cold pretreatment was observed in two of the F₁ hybrids when they were cultured on potato-2. Chromosome counts on root tips from the resulting green plants revealed that they all carried D-genome chromosomes. The last observation could suggest that D-genome chromosomes are necessary for anther culture response in wheat. Yet, the production of one green plant with 15 chromosomes may indicate that the development of extracted durum genotypes from bread wheat genotypes with good response to in vitro anther culture might be possible. Further work however, is needed for this to be verified.     

Adventitions shoot formation on excised leaves of in vitro grown shoots of apple cultivars

Leaves taken from micropropagated shoots of several apple (Malus domestica Borkh.) cultivars were cultured in vitro on Linsmaier & Skoog (LS) medium or the rice anther culture medium of Chu et al. (N6) containing various concentrations of either benzyladenine (BA) or thidiazuron (TDZ) plus naphthaleneacetic acid (NAA). Of the TDZ concentrations tested, 10 M was most effective and it was equivalent to, or better than, 22 M BA for both the percentage of leaves regenerating shoots and number of shoots formed per regenerating leaf in almost every experiment. Lower concentrations of NAA (1.1 and 5.4 M) gave best results with both BA and TDZ. N6 medium gave consistently better results than LS. Lowering total salt concentration or total N concentration of LS to that of N6 did not improve the response nor did changing the NO₃:NH₄ ratio. The 3–4 leaves on the most distal part of the shoot were most responsive and tended to form the most adventitious shoots. Placing the leaf cultures in the dark for the first 2–3 weeks of the culture period produced the best results. Optimum results were obtained by culturing leaves from the distal part of the shoot in the dark for 2 weeks on N6 medium containing 10 M TDZ and 1.1 or 5.4 M NAA, then moving the cultures to 16 h daylight at a photon flux of 60 mol s^-1m^-2.     

Pollen callus culture in Triticum aestivum

Summary Pollen shed between 4–8 d from anthers of Triticum aestivum cultured in liquid medium gave rise to calluses. Tillers were harvested at the mid-to late-unicellular pollen stages and chilled for 8 d at 4–5 °C before the anthers were dissected out. Pollen cultures gave about 6 times as many calluses on a per anther basis as anthers cultured on solid medium. With the most productive of 5 cultivars tested, pollen culture results in roughly one callus for each anther used, though the calluses formed by pollen culture were less productive for the regeneration of shoots than calluses derived from anthers cultured on solid medium. The ratio of green to albino shoots is roughly 1 1 for anther cultures but considerably less for pollen cultures.     

Genetic and non-genetic factors affecting anther culture of Brussels sprouts (Brassica oleracea var. gemmifera)

Summary Eight inbred lines of Brussels sprouts and ten F₁ hybrids derived from them were tested for their response to anther culture. From 5–19 plants per genotype were tested, and each plant was tested on 3–6 separate occasions. Results from the inbred lines were broadly similar to those from the F₁ hybrids, despite the inbreds producing fewer buds and having a higher frequency of anther deformities. The maximum embryo yield from an inbred line was 215 embryos per 100 anthers, and from a hybrid was 275. From estimation of the variance components it was calculated that, for both inbreds and hybrids, about half the total variation was genetic whereas variation due to plants within genotypes and to occasions within plants were each about 13% of the total. The narrow sense heritability of responsiveness to anther culture (estimated by the proportion of variation between inbred lines which was genetic) was 0.48, and there was partial dominance for this character. In three cases the hybrid outyielded the better inbred, and this heterosis may well be due to dispersed dominant genes.     

Interspecific hybrids between Brassica napus L. and B. oleracea L. developed by embryo culture

Summary Interspecific hybrids between Brassica napus and B. oleracea are difficult to produce, and previous attempts to transfer economic characters from one species to the other have largely been unsuccessful. In these studies, oilseed rape cv. Tower (2n38) (B. napus) was crossed with broccoli and kale (2n18) (B. oleracea), and hybrid plants were developed from embryos in culture by either organogenesis or somatic embryogenesis. In rape × broccoli, F₁ plants were regenerated from hybrid embryos and the plants produced viable selfed seeds. F₅ plants (2n38) homozygous for white flower colour were selected for high oil content (47%) and Line 15; a selection from these plants produced fertile hybrids with rape, broccoli and kale without embryo culture. In reciprocal crosses between oilseed rape cv. Tower and an aphid resistant diploid kale, 28 and 56 chromosome F₁ hybrid plants were regenerated from somatic embryos. The 56 chromosome plants were self-fertile and it was concluded from F₂ segregation ratios that a single dominant gene controls resistance to cabbage aphid in kale. The 28 chromosome F₁'s were self-sterile, but these and the 56 chromosome F₁'s could be backcrossed to rape and kale. A cross between the F₁ (2n56) and a forage rape resulted in the selection of a cabbage aphid (Brevicoryne brassicae L.) resistant line (Line 3). Both Line 15 and Line 3 can serve as bridges for gene interchange between B. campestris, B. napus and B. oleracea, which has not been possible hitherto. Hybridisations between rape and tetraploid kale produced F₁ plants with 37 chromosomes. One F₂ plant possessed coronal scales and the inheritance was shown to be controlled by a single recessive gene unlinked to petal colour.This paper is dedicated to Mr. T. P. Palmer, a colleague and close friend who retired from the DSIR as Assistant Director of the Crop Research Division in September 1984     

Protoplast isolation,culture and regeneration from Egyptian cultures of Pea and Beans

Abstract

A protocol of protoplast isolation from Egyptian varieties of pea and bean is reported. Protoplast cultures were established from apical shoots of pea (Pisum sativum) and suspension cultures of bean (Phaseolus vulgaris). To isolate protoplasts of pea, apical shoot tissues were digested for 10 h using enzyme solution containing 1% pectinase, 0.5% cellulase, 0.5% hemicellulase, 10% mannitol and 0.1% CaCl₂-2H₂O. For protoplast isolation from suspension culture of bean, collected cells were incubated for 6 h in digestion solution containing 0.5% pectinase, 0.25% of each of cellulase and hemicellulase, 10% mannitol and 0.1% CaCl₂-2H₂O. Purified protoplasts were cultured in liquid culture medium. Microcalli were obtained after 30 days of culture. Calli colonies with a diameter of about 5 mm were developed after one month of culturing on solid B5 medium containing 2% sucrose, 2 g/l casein hydrolysate, 0.7% agar and supplemented with either 1 mg/l of each 2,4-D and kin in case of pea or 2 mg/l 2,4-D+0.5 mg/l kin in case of bean. Protoplast derived callus of pea was successfully differentiated into shoot and root, and highest frequency of shoot organogenesis was recorded on medium containing 0.5 mg/l NAA+2 mg/l BA. Protoplast derived callus of bean, on the other hand, gave rise to a high frequency of root formation when cultured on medium containing 1 mg/l NAA, but attempts to regenerate shoots from this callus was unsuccessfull.     

Heterozygous diploid plants regenerated from anther culture of F1 rice plants

Summary Plants derived from anther culture are theoretically haploid, but diploid plants are also known to arise. Anther culture-derived diploid plants are usually homozygous and are believed to be due to spontaneous doubling of chromosomes in either microsporocytes or callus cells during the culture process. However, heterozygous diploid regenerants may also arise from a) regeneration from cultured somatic cells, b) mutation occurring during or after a spontaneous doubling event, c) fusion of unlike haploid cells in chimeric callus, and d) regeneration from diploid microsporocytes resulting from aberrant meioses. This study was designed to elucidate the frequency and origin of diploid regenerants from rice anther culture. Regenerants were obtained from 11 F₁ genotypes. Progeny testing detected heterozygosity in 7 out of 211 regenerants. Each of the heterozygous regenerants were from ‘Calrose 76’/waxy ‘M-101’, Half of the diploid regenerants from this cross were heterozygous. No heterozygous regenerants arose from the other 10 F₁ genotypes. Progeny testing indicated that two of the heterozygous regenerants were as heterozygous as the F₁ plants for three parental characters. The other five regenerants exhibited decreased levels of heterozygosity. One of the heterozygous regenerants exhibited evidence of mutation for a non-parental character. However, mutation is an unlikely cause of the observed high levels of parental-type heterozygosity. No evidence for the occurrence of chimeric callus was detected, making this an unlikely cause as well. The most likely origin of the observed partial heterozygosity is regeneration from diploid microspores, which could also produce plants exhibiting complete parental-type heterozygosity.     

Genetic analysis of protoplast regeneration ability in Brassica oleracea

In Brassica oleracea L., plant regeneration from protoplasts is genotype-dependent and colony formation can be obtained routinely. In order to identify genes for regenerability, we performed a genetic analysis of the characteristic in the F₂ generation of a cross between two accessions selected for high and low regenerability. Regeneration frequencies were obtained from protoplast culture of 248 individual F₂ plants after 5, 10, and 15 weeks of culture on regeneration medium. Broad-sense heritability estimate was 0.492 at the early stage and lower (0.046–0.149) at advanced stages. The frequency distribution observed during short-term culture can be explained by two independent loci with duplicate dominant genes controlling regeneration. In long-term culture, one additional dominant gene can confer regeneration; i.e., three independent loci are responsible for regenerability. The results suggest that selection for high regeneration response may be efficient at early stages because of the lower environmental influence on the characteristic, and because fewer genes are involved at this stage than at advanced stages. The control of regenerability by a few dominant genes facilitates incorporation of the trait into breeding material. Molecular markers linked to the genes may assist in the selection of genotypes with high regeneration percentage in the future. This revised version was published online in June 2006 with corrections to the Cover Date.