Penelope A C Sparrow, Philip J Dale, Judith A Irwin.


John Innes Centre, Colney Lane, Norwich, NR4 7UH, UK




Previous work over many years has led to the development of protocols to allow selected genotypes to be transformed by Agrobacterium-mediated transformation. These methods may be tailor-made to transform a particular genotype of interest. More often, a genotype amenable to transformation is used and the introduced gene transferred to the line of interest by cross-pollination.† In a winter-sown crop with a vernalisation requirement such as Brassica oleracea, to transfer the desired gene in this way and produce a new variety may take 10-15 years.


Agrobacterium-mediated transformation can be split into three main stages: Agrobacterium infection, selection of transformed cells and shoot regeneration. The aim of this study is to move away from the empirical variation of plant tissue culture methods and to begin to understand the genetic control of three principle factors that affect the transformation process. In this paper, we describe experiments to identify genetic loci associated with shoot and root regeneration.† Work is also ongoing to identify loci associated with background antibiotic resistance (for the selection stage) and tissue sensitivity to Agrobacterium infection.† Preliminary findings will be discussed.


KEYWORDS:† Regeneration, Tissue Culture, Variation, Genetic Control, Genotype.



Transformation has been reported for all the main Brassica species, B.rapa (Radke et al. 1992), B. oleracea (De Block et al. 1989, Christey et al. 1997), B. nigra (Gupta et al. 1993), B.juncea (Barfield et al. 1991), B.carinata (Narasimhulu et al. 1992) and B. napus (Moloney et al. 1989).† Protocols vary not only between the Brassica species but also within the species, with some genotypes being particularly recalcitrant to transformation.† Transformation is therefore, highly genotype dependent. Factors that potentially make a genotype amenable to transformation are:


      Susceptibility to Agrobacterium: Plant genotypes respond in a diverse way depending on the type and strain of Agrobacterium used (Ramsay and Kumar 1990, Lindsey and Gallois 1990). One reason for this is that some genotypes elicit a hypersensitive response in the presence of Agrobacteria. In extreme cases this can result in necrosis of the tissue at the infection site, hindering transformation events (unpublished data).


      Background antibiotic resistance.The level of antibiotic required to select for transformed cells varies between species.† It is thought that genotypes differ in their natural background resistance to these antibiotics, and hence different levels of selection will be required for different genotypes.† Brassica oleracea is the most sensitive of the diploid Brassicas to amino glycoside antibiotics and variations are also seen within this species (unpublished data).†


      Regeneration ability.† Again the ability to regenerate whole plants in vitro is highly genotype dependent, and is therefore likely to be under genetic control.†† Murata and Orton (1987) noted variation in regeneration ability between the Brassica species, with B. oleracea (CC) being the most responsive diploid followed by B. nigra (BB); whilst B. rapa (AA) was fairly recalcitrant to in vitro regeneration.† The interspecific hybrids elicit an intermediate response to that of the parents i.e. AA<AACC<CC.† Variation is also seen within the species Christey et al. (1991) and Irwin et al. (paper 551 of this session, 1999).†††† The genetic control of a number of tissue culture responses has been investigated through quantitative genetic studies, Mendelian genetic analysis and gene mapping (reviewed by Henry et al. 1994).† QTL analysis has progressed further in cereals, such as wheat (Ben Amer et al. 1997) Barley (Mano et al. 1996) and Rice (Taguchi-Shiobara et al. 1997), where a number of QTLís associated to tissue culture response have been identified.† The genetic control of plant regeneration in Brassica is however less clear.† Studies have been carried out into anther culturability (Aslam et al. 1990) where the genetic control was described as Ďextremely complexí.†† Diallel analysis for shoot regeneration in B. napus (Ono et al. 1994) concluded that Ďshoot regeneration ability, was controlled by major gene(s) with an additive and dominant effect, and that these genes could be transferred from high responsive genotypes to low or unresponsive genotypes by sexual crossingí.


It is likely to be a combination of these 3 factors that will determine the transformability of individual genotype.In this paper we describe a study to investigate the genetic control of shoot and root regeneration in B. oleracea.




Plant materials.


A doubled haploid (DH) mapping population derived from a cross between B. oleracea ssp. alboglabra (A12) and B. oleracea ssp. italica †(Green Duke) and a detailed RFLP map associated to this population was supplied by Derek Lydiate for use in this programme. This material was generated by Bohuon (1995) and full map details can be found in Bohuon et al. (1996).


Regeneration procedure.


Preliminary studies were carried out on 30 DH lines from the above-mentioned mapping population.† Mature seeds were surface sterilised and germinated on full strength MS medium supplemented with 3% sucrose and vitamins.†† Cotyledonary petioles and hypocotyl explants were excised from 5-day-old seedlings and cultured on regeneration medium (as above plus 2mg/l BAP).† Regeneration frequencies (number of explants with shoots (or roots) / total number of explants) were scored after 16, 23 and 44 days.† Approximately 60 explants were established per explant type per DH line.


Statistical analysis


Analysis of variance (ANOVAs) for the arcsin ÷ % transformed means was used to detect differences in regeneration frequencies between the doubled haploid lines.†† The MAPMAKERQTL programme (Paterson et al. 1988) was used to detect putative QTLís.





Variation in both shoot and root regeneration was noted between the two parents and their doubled haploid F1 lines.† The final score data is presented (after 44days).


One-way analysis of variance (ANOVA) performed on transformed data (arcsin ÷% means), showed there to be a significant difference in the regeneration responses between the DH lines (p values of less than 0.0001 noted for shoot regeneration from both cotyledonary and hypocotyl explants). The distribution patterns for shoot regeneration, from cotyledonary explants after 44 days in culture, did not approximate to a normal distribution (Figure 1).† A 12x12 diallel is currently being carried out to determine the number of genes associated to this trait.



Figure 1: ††††† Frequency distribution for shoot regeneration from cotyledonary explants, † ††††††††††† after 44 days in culture.


Frequency distributions for shoot regeneration from hypocotyl explants approximated to a normal distribution indicating a polygenic character (Figure 2a).† Shoot regeneration data was entered into MAPMAKERQTL and compared against the B. oleracea linkage map.† Putative QTLís were identified on linkage group O5.† The LOD profile of this QTL can be seen in Figure 2b.


(2a)†††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††††† (2b)


Figure 2a: ††† Histogram showing frequency distribution of shoot regeneration from †††††††††† ††††††††††† hypocotyl explants after 44 days in culture, among the DH lines.

Figure 2b:†††† The LOD profile, of the putative shooting QTL associated to regeneration from hypocotyl explants.†† QTLís are assigned to regions with LODís of 2.0 and above.


Root regeneration from cotyledonary explants, was markedly different between the two parents (44% A12 and 0% Green Duke). One-way analysis of variance (ANOVA) performed on arcsin ÷% transformed means, showed there to be significant differences in root regeneration between the DH lines (p= < 0.0001).† Distribution patterns for root regeneration can be seen in Figure 3.† Putative QTLís were noted on linkage groups 01, 08 and 09.† The LOD profile of the QTL associated on linkage group O9 is shown in Figure 4.




Figure 3: ††††† Histogram showing frequency distribution of root regeneration from ††††††††††††† ††††††††††† cotyledonary explants after 44 days in culture, among the DH lines.

††††††††††††††††††††††† * Data transformed mean of the % root regeneration

Figure 4:†††††† The LOD profile associated to rooting from cotyledonary petioles, on linkage group O9.†† QTLís are assigned to regions with LODís of 2.0 and above.




The amenability of a plant to in vitro regeneration is influenced by the plant genotype, which is thus of major importance in the transformability of a cultivar.† The differences observed between different cultivars during in vitro tissue culture, result from quantitative or qualitative genetic differences.


The ability to form roots in vitro, from cotyledonary explants, appears to be controlled by many genes.† Putative QTLís have been located on linkage groups O1, O8 and O9 of B. oleracea.† The ability to form roots in vitro is of considerable importance in the establishment of commercially viable transgenic plants.


Shoot regeneration from cotyledonary and hypocotyl explants appears to be under different genetic control.† The formation of shoots from cotyledonary explants seems to be controlled by relatively few genes.† Work is currently in progress to determine the exact number of genes involved, with the aid of a large diallel crossing programme, where high, intermediate, low and zero responding lines are being crossed.† Shooting from hypocotyl explants appears to be controlled by a larger number of genes and putative QTLís have been located on linkage group O5 of B. oleracea.




We would like to thank Derek Lydiate (Agriculture and, Agrifood Canada) and Graham King† (HRI, Wellesbourne, UK) for supplying the seed stocks and mapping details for use in this study.† This work is funded by the UK Ministry of Agriculture Fisheries and Food (MAFF Contract No. HH0909SFV).




Ben Amer-IM, Korzun-V, Worland-AJ and Borner-A . Genetic mapping of QTL controlling tissue† culture†

response on chromosome 2B of wheat (Triticum aestivum L.) in relation to major genes and RFLP markers.† Theor. Appl. Genet. 1997, 94,1047-1052.

Bohuon-EJR.† A genetic analysis of Brassica oleracea. Ph.D. Thesis University of Birmingham, UK. 1995.

Bohuon-EJR, Keith-DJ, Parkin-IAP, Sharpe-A, Lydiate-DJ (1996).† Alignment of the conserved C genomes of

Brassica oleracea and Brassica napus.† Theor. Appl. Genet 93: 841-847.

Christey-MC and Earle-ED. Regeneration of Brassica oleracea from peduncle explants.† †††††††††† HortScience 1991,

26(8): 1069-1072.

Henry-Y, Vain-P and De Buyser-J.† Genetic analysis of in vitro plant tissue culture responses and †††††††††† regeneration capacities.† Euphytica. 1994, 79: 45-58.

Lindsey-K, Gallois-P.† Transformation of Sugarbeet (Beta vulgaris) by Agrobacterium tumefaciens. ††††† Journal of Exp.Bot. 1990, 41:226, 529-536.

Mano-Y, Takahashi-H, Sato-K and Takeda-K.† Mapping genes for callus growth and shoot regeneration in

barley.† Breeding Sci. 1996, 46; 137-142.

Murata-M, Orton-TJ. Callus initiation and regeneration capacities in Brassica species.† Plant Cell Tiss Organ

Cult 1987,11: 111-123.

Ono-Y,† Takahata-Y and Kaizuma-N. Diallel analysis of shoot regeneration in Brassica † napus.

Cruciferae Newsletter. 1994, 16: 57-58.

Paterson-AH, Lander-ES, Hewitt-JD, Peterson-S, Lincoln-SE and Tanksley-SD.† Resolution of ††† quantitative

traits into Mendelian factors by using a complete linkage map of restriction fragment length polymorhisms.† Nature. 1988, 335: 721-726.

Ramsay-G and Kumar-A. Transformation of Vicia faba cotyledon and stem tissues by †††† Agrobacterium rhizogenes: infectivity and cytological studies. Journal of Experimental Botany. ††††††††† 1990, 41: 228, 841-847.

Taguchi-Shiobara-F, Lin-SY, Tanno-K, Komatsuda-T, Yano-M, Sasaki-T and Oka-S.† Mapping †† quantitative trait loci associated with regeneration ability of seed callus in rice, Oryza sativa L. Theor.Appl.Genet. 1997, 95: 828-833.