Transformation of a large number of potato varieties: genotype-dependent variation in efficiency and somaclonal variability

Transformation of potato is a genotype dependent process as was shown by experiments conducted with 16 varieties. Not all genotypes could be transformed with a single procedure hence two different procedures were attempted for all 16 varieties in a pilot experiment. Large differences in regeneration capacity of putative transformants were observed with the two protocols. Regeneration capacity and transformation efficiency were not correlated. All varieties were transformed with the same construct, composed of a kanamycin resistance gene and an antisense gene coding for granule-bound starch synthase. This led to different percentages of plants with the desired maximum effect (i.e. amylose-free starch) ranging from less than 1 percent to 23.3 percent. It was shown that variety-dependent phenotypic variation occurred ranging from 1 to 21%. Field experiments, conducted over a number of years, using plants with different degrees of antisense effect (from no detectable effect to maximum effect) showed that most transformants would have a decreased yield and starch content as determined by specific gravity measurements. However, these negative effects can be overcome by selecting the proper transgenic plants. Molecular characterisation of transformants using PCR, showed that 90% of the analysed transgenic plants, belonging to all effect classes, contained vector DNA sequences since they contained either the NPTIII gene or the trfA gene. The other 10% of the transgenic plants had no insertion of vector DNA.