Sense Cosuppression of Flower Color Genes: Metastable Morphology-based Phenotypes and the Prepattern-threshold Hypothesis

Abstract

In flowering plants, transformation with a transgene whose coding sequence is homologous to an endogenous plant gene, and whose expression is driven in the sense orientation by a strong promoter, often produces the opposite outcome of that which was intended. Instead of over-expression of the gene product, transcripts from both the endogenous gene and the transgene are found to be “cosuppressed,” resulting in a null phenotype. Nuclear run-on experiments have shown that the cosuppressed state is posttranscriptional; experiments that alter transgene dosage suggest that the induction of cosuppression is threshold-dependent; and cytoplasmic RNA viruses that are homologous to a strongly expressed nuclear gene are subject to cosuppression, suggesting that it occurs in the cytoplasm (de Carvalho et al. 1992de Carvalho et al. 1995; Lindbo et al. 1993; Smith et al. 1994; Mueller et al. 1995; Goodwin et al. 1996). A preferred interpretation of these observations, presented in detail by Smith et al. (1994), is that the production of nuclear transcripts to levels exceeding some threshold concentration somehow activates a sequence-specific turnover process that acts in the cytoplasm. This phenomenon is referred to variously as “sense suppression,” “sense cosuppression,” or simply “cosuppression.” It can be distinguished from promoter-homology-dependent suppression, a gene silencing phenomenon that occurs at the transcriptional level, requires promoter homology, and is thought to be a consequence of ectopic pairing (for review, see Matzke and Matzke 1995; see also Vaucheret et al.; Meyer; both this volume). Sense cosuppression also can be distinguished from suppression caused by antisense transgenes (Jorgensen et al. 1996).