We are interested in the mechanisms that distinguish between soma and germline in the early embryo. As reported previously, we have uncovered a striking difference between somatic and germline blastomeres: starting in the four-cell stage, several newly transcribed mRNAs can be detected in somatic blastomeres, but none to date have been detected in germline blastomeres (ref. 1,2). Our initial studies suggest that this asymmetry is essential for the formation of the germline, and depends on mechanisms that inhibit gene expression in the germ lineage (ref. 2). We hypothesize that the main function of this inhibition may be to protect germline blastomeres from the activity of transcription factors that would otherwise promote somatic development in these cells. So far, our analysis of embryonic transcription has been restricted to the limited number of messenger RNAs known to be transcribed in the early embryo (13 examined to date). Is the production of all RNAs inhibited in germline blastomeres? To begin to answer this question, we have characterized the distribution of maternal SL1 RNA, and that of newly transcribed rRNAs, by in situ hybridization. We find that the pattern of maternal SL1 RNA turn over in early embryos is consistent with a general lack of mRNA transcription in germline blastomeres. In contrast, using probes specific to the precursor form of ribosomal RNAs, initial results indicate that these RNAs are transcribed in both somatic and germline blastomeres. These data suggest that mechanisms that inhibit gene expression in germline blastomeres may be specific to mRNAs. We are currently testing this hypothesis further by looking at the expression pattern of other classes of RNAs. What factors inhibit gene expression in germ cells? By surveying previously identified mutations for their effects on gene expression in embryos, we have identified one potential candidate for such a factor: the germline-specific protein PIE-1 (ref. 3). In the absence of
pie-1 activity, newly transcribed mRNAs (such as the
pes-10 RNA) are detected in both somatic and germline blastomeres (ref. 2). Curiously, however, we find that beta-galactosidase expression from a
pes-10::lacZ fusion remains restricted to somatic lineages even in the absence of
pie-1 activity. This result suggests that factors other than
pie-1 can function to repress gene expression in the germ lineage. To identify these factors, we have started a genetic screen for mutations that result in the inappropriate activation of a
pes-10::lacZ fusion in embryonic germ cells. Special thanks to Andy Fire in whose laboratory this work was started. 1. Seydoux, G. and Fire, A. 1994. Development 120, 2823-2834 2. Seydoux, G. and Fire, A. 1995. C. elegans Meeting Abstracts, 29. 3. Mello, C. C. et al. 1992. Cell 70, 163-176. Mello, C. C. et al., 1995. C. elegans Meeting Abstracts, 30.