Transgenic onion plants with an antisense version of the bulb alliinase gene under CaMV35s promotional control have been produced. Biochemical pyruvate analysis showed that some transgenic plants, containing this gene construct, had significantly reduced alliinase activity. More detailed analysis of individual clones indicated that the reduced alliinase activity was only apparent in some plants suggesting that either chimeric plants were produced; escapes were regenerated along with transformants; differential gene silencing occurred; or there was a huge environmental variation in allilinase expression.The biochemical data was supported by multiplex RT-PCR that indicated a reduction in alliinase transcript in some lines. In at least one line, an approximate 10-fold reduction in alliinase transcript was detected. This corresponded to bulb material that was shown to have no measurable alliinase activity. In one further test, SNuPE analysis of the alliinase transcript present in the RNA was performed to determine whether a particular alliinase gene member had been silenced. The results indicated that all alliinase transcripts were equally silenced.Molecular fingerprinting techniques are now being developed to match the plants sacrificed for this analysis to their 'real' clones and offspring so that phenotypic assessment of reduced alliinase plants can be undertaken.

SummaryTransgenic onion plants (Allium cepa) containing the Cauliflower mosaic virus 35s promoter (CaMV35s) and gfp gene construct encoding the visual green fluorescent reporter protein from pBINm gfp ER and the CaMV35s‐bar gene construct encoding resistance to the herbicide phosphinothricin from pCAMBlA3301 were produced by Agrobacterium‐mediated transformation. These plants weregrown to maturity and selfed in order to determine the expression and inheritance of the transgenes. CaMV35s regulation in onion, as observed by GFP expression, was essentially constitutive, and profiles of regulation were typical of those observed in dicotyledonous plants. Inhibition of CaMV35s regulated gene expression was only observed in one transformant. Both the expression of GFP and tolerance to phosphinothricin appeared to be inherited in a Mendelian fashion. Levels of expression in F1 offspring varied, presumably due to environmental and genetic factors. However, it appeared that copy number did strongly influence GFP protein production and expression. In the majority of plants there were no obvious detrimental phenotypic effects caused by the transgene, the integration event, or Somaclonal variation due to the need to perform tissue culture.

Transgenic onion plants (Allium cepa) tolerant to herbicides containing active ingredients glyphosate and phosphinothricin were recovered from immature embryos of open pollinated and hybrid parent onion lines at a maximum transformation frequency of 0.9%. Transformants of different onion cultivars, grown on different selective agents and confirmed by Southern analysis, thrived with no apparent ill effects when sprayed with the respective herbicides at double the recommended field dosage for weed eradication. This study demonstrates that the transformation process described previously can be used with different selective agents and is cultivar independent.

Immature cotyledons from six cultivars and breeding lines of Pisum sativum were cocultivated with Agrobacterium tumefaciens and selected on a medium containing kanamycin. Transformed plants were recovered with a success rate of between 0.8 and 3.4% of explants treated. The plasmids used were based on pGA643 and have a nos promoter driving an nptII gene. Progeny from the primary transformants showed stable inheritance of the nos-nptII gene.