Search

In broccoli, post-harvest senescence is rapid, with loss of chlorophyll resulting in the yellowing of sepals from immature florets. Associated with this deterioration is an increase in ethylene production. In order to delay broccoli deterioration an antisense 1- aminocyclopropane- 1-carboxylic acid (ACC) oxidase gene from pTOMI3 was transferred into broccoli plants. To produce transgenic plants an improved protocol was developed for Agrobacterium rhizogenes-mediated transformation of broccoli. This procedure uses compounds that enhance the virulence of A. rhizogenes (l0 mM mannopine and 50 µM acetosyringone in the co-cultivation medium and 1 mM arginine in the selection medium) and a Brassica campestris feeder layer. Leaf explants or intact cotyledons of three broccoli cultivars: Green Beauty (Gy), Shogun (Sh) and Green Belt (Gb) were co-cultivated with A. rhizogenes strain A4T harbouring the binary vector pLN35. The T-DNA of this binary vector contains genes encoding antisense ACC oxidase (35S-ACC-5'7') and neomycin phosphotransferase II (NOS-NPTIT-NOS). As a control for the effect of transformation with the rol genes, pLN34 was used. This differs from pLN35 by having the ACC oxidase gene in the sense orientation with a cloning artefact at the 5' end, duplicating some of the polyA tail at the 3' end making the ACC oxidase gene non-functional. Two cultivars were successfully transformed, Shogun (Sh) and Green Beauty (Gy), with a transformation efficiency of 35% and 17%, respectively. Fertile plants were regenerated from kanamycin-resistant hairy roots by transfer to hormone containing media. Integration of the T-DNA was confirmed by the polymerase chain reaction (PCR) and Southern analyses. Transgenic lines obtained after transformation of Green Beauty (Gy), Dominator (D) and Shogun (Sh) broccoli plus the parental lines were evaluated for their changes in respiration, ethylene production and ACC oxidase activity using mature flowers. Seventeen 17 lines were analysed including 11 lines transformed by pLN35, 3 lines transformed by pLN34 (labelled with the suffix NF for non-functional), and 3 non-transgenic controls. Transformation and shoot regeneration procedures used to obtain the lines Sh/2, Sh/4, Sh/5, Sh/6, Sh/7 and Sh/8 were according to the improved protocol. The lines Sh/NF1, Sh/NF15, Gy/6, Gy/7, Gy/9, Gy/NF2, D/I and D/2 were obtained by the transformation and shoot regeneration procedures described by Christey et al. (1997). Regenerated plants were transferred to a contained greenhouse where fertile plants were obtained to supply flowers. Of the 11 transgenic lines evaluated, 10 lines showed a significant reduction in ethylene production relative to the control from 50 h after harvest. Green Beauty flowers showed a significant difference between the transgenics and control and demonstrated how ethylene levels could control the stable or climacteric-like increase in respiration. ACC oxidase activity was higher in transgenic plants, consistent with the initially higher ethylene production. ACC oxidase activity did not, however, reflect the increase in ethylene production found after 50 h for the controls. The results of this research suggested that two ethylene production systems may operate with only the second being inhibited by the antisense ACC oxidase used and that the later system was not detected by the ACC oxidase assay used. Agronomic characterisation was conducted on 17 lines, including 11 lines transformed by pLN35, 3 lines transformed by pLN34 and 3 non-transgenic controls. For morphological characterisation an extra line transformed by pLN35 was included giving a total of 18 lines analysed from three cultivars, Shogun, Green Beauty and Dominator. The transgenic plants showed evidence of Ri-induced morphological changes to varying degrees. The transgenic lines Gyl7, D/l and D/2, performed within the limits of acceptability for all head quality parameters (size, density, colour, shape and leafiness). The transgenic lines D/l and D/2 showed significant improvements in head colour relative to the control from 48 h after harvest. These results are consistent with the ethylene production patterns of these lines analysed previously also suggesting that two enzyme systems are involved in broccoli senescence, having two bursts of ethylene production, with the second burst inhibited by the antisense ACC oxidase gene. The results also show that post-harvest ethylene synthesis and therefore possibly broccoli senescence can be regulated by using an antisense ACC oxidase gene.

The similarities and differences between plants and other organisms can be studied by introducing non-plant regulatory sequences into the plant genome. Experiments were carried out to study gene expression under the regulation of an animal virus enhancer in faba bean (Vicia faba L. cv. Puma). A binary vector, pGA643-GUS-SV, was constructed carrying the cauliflower mosaic virus 35S (CaMV 35S) plant promoter, the Simian virus 40 (SV40) animal enhancer and the β-D glucuronidase (GUS) reporter gene. This plasmid was transferred to hairy root-inducing Agrobacterium tumefaciens strain A4T. Vicia faba L. cv. Puma plants were inoculated with A. tumefaciens A4T carrying pGA643-GUS-SV. Control V.faba L. cv. Puma plants were inoculated with A4T carrying the enhancer-less binary vector pGA643-GUS. The hairy roots were assayed for GUS expression and DNA content. Southern blots were carried out on these hairy roots to confirm GUS gene transfer and integration into V.faba L. cv. Puma, and to assess GUS copy numbers. The GUS activities showed high variability in both treatment pGA643-GUS and treatment pGA643-GUS-SV. This variability was reduced when the GUS activity values were adjusted by GUS copy numbers. The considerable amount of variability that remained after the adjustment was probably due to (i) differences in hairy root physiology and (ii) insert positioning effects in the plant genome. The mean GUS activity was slightly higher in treatment pGA643-GUS-SV than that in treatment pGA643-GUS. This difference, however, was not statistically significant at the 5% level, probably partly due to the variability observed in GUS activity values. The inability of the SV40 enhancer to significantly stimulate GUS gene expression in V.faba L. cv. Puma might have also been due to (i) the lack of an appropriate transcription factor or factors to bind to the enhancer, (ii) the nature of the intervening DNA between the enhancer and the promoter and (iii) possibly, to an unknown inhibiting factor present in V. jaba L. cv. Puma.