CO-SUPPRESSION OF NITRATE REDUCTASE HOST GENES AND TRANSGENES IN TRANSGENIC TOBACCO PLANTS

Constructs carrying the entire or part of the tobacco nitrate reductase cDNA (NIA) cloned between the promoter and terminator sequences of the 35S RNA of the cauliflower mosaic virus were introduced into tobacco, in an attempt to improve nitrate assimilation. Several transgenic plants that had elevated NIA mRNA and nitrate reductase (NR) activity were obtained. In addition, a few plants that exhibited a chlorotic phenotype characteristic of NR-deficient mutants were also obtained. One of these plants contained no NIA mRNA, no NR activity and accumulated nitrate. This phenotype was therefore assumed to result from co-suppression of 35S-NIA transgenes and host NIA genes. NR-deficient plants were also found among the progeny of transformants overexpressing NIA mRNA. Genetic analyses indicated that these NR-deficient plants were homozygous for the 35S-NIA transgene, although not all homozygous plants were deficient for NR. The ratio of normal to NR-deficient plants in the progeny of homozygous plants remained constant at each generation, irrespective of the state of expression of the NIA genes (active or inactive) in the previous generation. This ratio also remained unchanged when field trials were performed in two areas of France: Versailles and Bergerac. The analysis of homozygous plants revealed that co-suppression was reversible at some stage of sexual reproduction. Indeed, host genes and transgenes reactivated at each generation, and co-suppression always appeared after a lag period of normal growth, suggesting that the phenomenon is developmentally regulated. We observed that the triggering of co-suppression was delayed when plants were initially grown under limited light and/or watered with limited nitrate supply (light and nitrate both being required for the expression of the host NIA genes). However, this delay did not affect the final ratio between normal and NR-deficient plants after transfer to nitrate-fertilized fields. Independent transformants exhibited either different co-suppression ratios or no co-suppression at all, irrespective of the transgene copy number, suggesting that genomic sequences surrounding the transgene might play a role in determining co-suppression.