Embryonic germline blastomeres (P1, P2, P3 and P4) fail to express many mRNAs expressed in somatic blastomeres (1). In contrast, rRNAs are expressed in both cell types (2). In principle, this difference could be due to the rapid degradation of newly transcribed mRNAs in germline blastomeres, or to a block in mRNA transcription in the germ lineage. To begin to distinguish between these two possibilities, we have characterized the distribution of phosphorylated isoforms of the large subunit of RNA polymerase II (RNAP II LS) in embryos. For this purpose, we used 2 monoclonal antibodies, H5 and H14, that recognize distinct phosphorylated epitopes on the carboxy terminal domain (CTD) of RNAP II LS (3). Phosphorylation of the CTD has been correlated with transcriptional elongation; indeed, we find that H5 and H14 start to stain somatic nuclei in the 4-cell stage, when these cells are known to begin mRNA transcription. However, no H5 staining, and only weak H14 staining, were detected in germline blastomeres. In contrast, an antibody (4) that recognizes both phosphorylated and unphosphorylated RNAP II showed no difference in staining between somatic and germline blastomeres. These results indicate that germline blastomeres lack the subpopulation of phosphorylated RNAP II recognized by H5, and have reduced levels of the subpopulation of phosphorylated RNAP II recognized by H14. These observations are consistent with a block in RNAP II activity in the early germ lineage. We have previously shown that the germline-specific factor PIE-1 is required to block mRNA production in the embryonic germ lineage (1). To test whether pie-1 is also required for the germline-specific patterns of H5 and H14 staining described here, we stained embryos derived from pie-1(zu154) mothers with H5 and H14. We found that in the absence of pie-1 activity, germ cells stain with H5 and H14 in a pattern identical to that of somatic cells. These results indicate that pie-1 activity is required for the soma-germline differences observed in H5 and H14 staining, and suggest that PIE-1 may function to repress, directly or indirectly, RNAP II activity in the germ lineage. We have also used the H5 and H14 antibodies to stain Drosophila embryos. Remarkably, we find that pole cells (the Drosophila embryonic germ cells) also fail to stain with H5, and stain only weakly with H14. These observations raise the possibility that a block in RNAP II activity may be part of an evolutionarily conserved mechanism that distinguishes germline from soma during early embryogenesis. We thank Miriam Golomb, Steve Warren and Jeff Corden for providing antibodies, and Debbie Andrew for advice on Drosophila handling. 1. Seydoux et al., 1996. Nature 382, 713-716. 2. Seydoux et al., 1996. East Coast Worm Meeting Abstracts, p 31. 3. Bregman et al., 1995. Journal of Cell Biology 129, 287-298; and M. Patturajan and J. Corden, pers. comm. 4. Sanford et al., 1985. Journal of Biological Chemistry 260, 8064-8069

Affiliation:
- Dept. of Molecular Biology and Genetics. Johns Hopkins University School of Medicine 725 N. Wolfe St./515 PCTB Baltimore, MD 21205