The convergent action of Wnt and MAPK kinase signaling pathways results in unequal levels of the HMG box transcription factor POP-1 in sister cells that arise from anterior/posterior cell divisions: anti-POP-1 antibody stains less intensely in the nuclei of the posterior daughter cells as a result of these signals (Lin et al., 1995, 1998; Meneghini et al., 1999; Rocheleau et al., 1999). This 'POP-1 asymmetry' occurs reiteratively throughout embryogenesis. POP-1 asymmetry correlates with the generation of developmentally unequal daughters and POP-1 is required for these differences in many cases. The mechanisms by which this asymmetry is reiteratively generated are not well understood. The
end-1 and
end-3 genes are candidate zygotic targets of maternal POP-1 in the EMS lineage. These genes, which specify the fate of the E cell, are bound directly by the zygotically expressed MED-1 and -2 GATA factors in both the E and MS lineages; POP-1 activity represses the end genes in MS, thereby restricting endoderm fate to E. To examine in vivo whether this repression involves the direct action of POP-1, we expressed a GFP::POP-1 fusion protein under control of the zygotic
med-1 promoter. This fusion expresses in all EMS descendants until the 8E stage, and rescues the maternal-effect lethal phenotype of a
pop-1(
zu189) mutation. We found that in MS, and other anterior, but not posterior, nuclei in the EMS lineage, GFP::POP-1 binds to an extrachromosomal array containing the
end-3 promoter (based on the "nuclear spot assay"). These observations support the model that Wnt/MAPK signaling blocks the repressive function of POP-1 in E, permitting MED-1,2 to activate
end-1,3 and specify endoderm fate. However, our findings also demonstrate that the observed in vivo interaction of POP-1 with
end-3 is neither necessary nor sufficient for repression: we observe interaction of POP-1 in Ea, in which
end-3 is apparently fully expressed, and detect no interaction in posterior daughters of the MS lineage, in which
end-3 expression is fully repressed. We propose that POP-1 can initiate a repressive state only at the time that transcription of a gene is first established (e.g.,
end-3 in the E cell), and is not subsequently required to maintain this repressed state. Examination of GFP::POP-1 also demonstrates the dynamic nature of POP-1 asymmetry in vivo . The GFP expression recapitulates the asymmetry seen with anti-POP-1 staining, demonstrating that this asymmetry reflects bona fide differences in nuclear POP-1 levels. During development, the GFP signal becomes cytoplasmic and uniformly distributed in mitotic cells. However, nuclear GFP::POP-1 asymmetry is re-established almost immediately after reformation of daughter interphase nuclei. Because GFP::POP-1 asymmetry is recursively generated even in cells in which we cannot detect message from the fusion gene, we suggest that this asymmetry does not depend on asymmetric synthesis of the protein, but involves other mechanisms such as differential degradation or nuclear localization. Our observations also suggest that GFP::POP-1 differs qualitatively in anterior and posterior daughters. In anterior daughters only, we observe bright punctate fluorescence beginning immediately after mitosis; the punctate spots apparently coalesce into one or two large puncta by the end of interphase. We observe similar structures with myc-tagged POP-1, suggesting they may not simply be an artifact of the GFP fusion per se . We have begun to examine the structural and genetic requirements for this dynamic behavior. The asymmetry in both POP-1 levels and presence of puncta requires WRM-1 and LIT-1, but not CBP-1. Neither the first 44 amino acids of POP-1, implicated in a canonical ß-catenin-Lef interaction with the ß-catenin BAR-1, nor BAR-1 itself, are required for either asymmetry. Furthermore, the conserved HMG box, and the carboxyl portion of the protein downstream of the HMG box, are dispensable for asymmetry, implicating the ~145 residues between these two domains. It will be of interest to identify the specific structural elements responsible for these asymmetries and to assess further the biological relevance of the puncta observed in anterior nuclei specifically.