Transgenic Expression of a Toxin-Coding Killer Virus of the Yeast Zygosaccharomyces bailii in Saccharomyces cerevisiae: Genetic Evidence for a Possible Function of “Cryptic” Mycoviruses in the Evolution of Their Hosts

Abstract

The killer toxin-secreting yeast Zygosaccharomyces bailii 412 contains two cytoplasmically inherited double-stranded RNA (dsRNA) viruses (ZbV-L, ZbV-M) responsible for the expression of a killer phenotype in its infected host. While ZbV-L functions as a classical helpervirus by providing capsid (cap) and RNA polymerase functions (cap/pol) necessary for packaging and replication of both viruses, M-dsRNA-containing killer viruses (ZbV-M) are satellites of ZbV-L that contain the genetic information for toxin production only. Both viruses were shown to be sufficient to confer the Z. bailii killer phenotype upon transfected spheroplasts of a S. cerevisiae non-killer strain, resulting in toxin-secreting transfectants. Genetic analysis of transfected S. cerevisiae strains indicated that the heterologous expression of a Z. bailii killer phenotype depends on the expression of a single chromosomal gene: KEX3. Thus, kex3 mutants of S. cerevisiae are phenotypically non-killers showing normal, i.e. wild-type levels of the toxin-coding M-dsRNA. In the genetic background of a kex3 mutation, the stably maintained ZbV-M killer viruses would be regarded as “cryptic”, because of the lack of any dsRNA-associated phenotype. Cytoductive transfer of ZbV-M from “cryptic” non-killers into KEX3 non-killer recipient strains resulted in heterozygous (kex3/KEX3) diploids that expressed a strong killer phenotype. However, meiotic segregants derived from dissection of kex3/KEX3 diploids showed (a) a loss in cell viability and (b) a significant selection of a certain genotype among the surviving haploid progenies. We describe a model in which Zb-M killer viruses lead to a suicidal phenotype of haploid segregants under conditions in which Kex3p is fully expressed and an additional mutation within either of two chromosomally encoded toxin-sensitivity genes (TOS1 and/or TOS2) is present. Survival of these strains is only possible in the presence of a kex3 mutation, leading to stably maintained ZbV-M viruses that, because of the lack of any detectable phenotype, would be regarded as “cryptic”. In this respect we discuss a more general model postulating a remarkable effect of such “cryptic” mycoviruses on the evolution of the virus-infected host.