The sex-determining gene
fem-1 is necessary for all aspects of male development, including differentiation of the male soma and spermatogenesis in both sexes. XX and XO
fem-1 animals descended from
fem-1 mothers develop as true females.
fem-1 exhibits two types of maternal effect. The first effect is maternal rescue whereby a product of the maternal wild type allele partially rescues male development in
fem-1 progeny of a
fem-1 heterozygote. Three deletions that remove all or part of
fem-1 exhibit a second maternal effect. When a wild type male is crossed to a female homozygous for one of these alleles, 15-90 percent of their progeny, depending on the deletion allele in question, exhibit germline feminization. This observation suggests that the three deficiencies eliminate a product that is maternally required for spermatogenesis.
fem-1 is the only gene partially or wholly deleted in the three alleles, yet other
fem-1 alleles, including nonsense alleles, do not exhibit this effect: the heterozygous progeny of a female homozygous for a non-deficiency
fem-1 null allele are wild type XX hermaphrodites and XO males. We suggest on this basis that the maternally required product is unlikely to be protein and may instead be an RNA derived from the
fem-1 locus. The maternal absence effect persists in later generations, suggesting that the maternal product epigenetically modifies the activity of a zygotic target gene. Two kinds of evidence suggest that
fem-1 germline transcription may be the zygotic target affected by the maternal product. First, the heterozygous progeny of
fem-1 deficiency homozygote females, as well as exhibiting germline feminization, contribute reduced
fem-1 maternal rescuing activity to their XO progeny. Second, in situ hybridization experiments show that
fem-1 transcript levels are reduced in descendants of deficiency homozygotes. We suggest that reduced
fem-1 transcription in the germline of heterozygous progeny of deficiency homozygotes accounts for reduced
fem-1 transcript accumulation in these animals, leading to their own germline feminization and impaired ability to rescue male development in their
fem-1 progeny. We propose that a product of the
fem-1 locus, probably an RNA, is maternally necessary for zygotic germline expression of the paternally contributed wild type
fem-1 allele. Perhaps this is a mechanism for licensing germline transcription of
fem-1; in the absence of maternal germline expression, a gene may be silenced in the zygotic germline. We are currently testing genes required in other silencing phenomena for involvement in the germline silencing of
fem-1 in the progeny of deficiency homozygotes.