In addition to previously identified roles for RNA, several new ways in which RNA serves as a regulator of gene expression have recently been described. RNA molecules are involved in both transcriptional and post-transcriptional forms of regulation, sometimes heritably affecting gene activity. Whereas most previously characterized regulatory roles of RNA involve downregulation, I describe a role for maternal transcripts of a gene in promoting zygotic activity of that gene, which I term the licensing of genetic activity. This regulation occurs in the germ line, a tissue notable for its abundance of genetic surveillance mechanisms. The maternal-effect regulation described here was identified using alleles of a sex-determining gene in Caenorhabditis elegans called fem-1. Females homozygous for fem-1 deletions produce heterozygous offspring that exhibit germ-line feminization and have reduced fem-1 activity and transcript accumulation. This phenotype can be rescued by injecting fem-1 RNA into the maternal germ line. The reduction in activity of the zygotic fem-1 locus is heritable, suggesting that the gene is becoming epigenetically silenced. Thus the maternal fem-1 RNA licenses the activity of the zygotic locus by preventing its silencing. By restricting germ-line activity to genes that were expressed in the germ line of the previous generation, this process may contribute to protecting the identity and integrity of the germ line. I performed an RNAi screen of candidate genes to ask whether they are required for maternal-effect silencing or licensing. Several enhancers and suppressors of germ-line feminization in the descendants of fem-1 deficiency homozygotes were identified. Chromatin regulation may be involved, and small-RNA pathways are important for both the silencing and licensing components of fem-1 regulation. Based on my characterization of this phenomenon, I proposed models of how maternal-effect regulation of fem-1 may be mediated. To test predictions of certain models, I examined whether specific characteristics of fem-1 make it susceptible to this silencing. Results of these experiments limit the possible models of maternal-effect regulation and suggest directions for future investigation.

In addition to previously identified roles for RNA, several new ways in which RNA serves as a regulator of gene expression have recently been described. RNA molecules are involved in both transcriptional and post-transcriptional forms of regulation, sometimes heritably affecting gene activity. Whereas most previously characterized regulatory roles of RNA involve downregulation, I describe a role for maternal transcripts of a gene in promoting zygotic activity of that gene, which I term the licensing of genetic activity. This regulation occurs in the germ line, a tissue notable for its abundance of genetic surveillance mechanisms. The maternal-effect regulation described here was identified using alleles of a sex-determining gene in Caenorhabditis elegans called fem-1. Females homozygous for fem-1 deletions produce heterozygous offspring that exhibit germ-line feminization and have reduced fem-1 activity and transcript accumulation. This phenotype can be rescued by injecting fem-1 RNA into the maternal germ line. The reduction in activity of the zygotic fem-1 locus is heritable, suggesting that the gene is becoming epigenetically silenced. Thus the maternal fem-1 RNA licenses the activity of the zygotic locus by preventing its silencing. By restricting germ-line activity to genes that were expressed in the germ line of the previous generation, this process may contribute to protecting the identity and integrity of the germ line. I performed an RNAi screen of candidate genes to ask whether they are required for maternal-effect silencing or licensing. Several enhancers and suppressors of germ-line feminization in the descendants of fem-1 deficiency homozygotes were identified. Chromatin regulation may be involved, and small-RNA pathways are important for both the silencing and licensing components of fem-1 regulation. Based on my characterization of this phenomenon, I proposed models of how maternal-effect regulation of fem-1 may be mediated. To test predictions of certain models, I examined whether specific characteristics of fem-1 make it susceptible to this silencing. Results of these experiments limit the possible models of maternal-effect regulation and suggest directions for future investigation.