DESCRIPTION (provided by applicant): We identified Early Developmental Regulator 3 (EDR3), as a candidate tumor suppressor gene in a screen of a region of human chromosome 3q26 previously shown to undergo LoH in osteosarcoma. We found intragenic deletions in the EDR3 gene consistent with a tumor suppressor etiology in osteosarcoma primary tumors and tumor cell lines. Examination of the expression pattern of EDR3 revealed that both mRNA and protein showed a near ubiquitous pattern of expression in adult and embryonic tissues. EDR3 is a homolog of the Drosophila Polyhomeotic gene and a member of the Polycomb Group (PcG) family of proteins. The PcG family of proteins acts as long-term repressors by contributing to the formation and stable transmission of heterochromatin. By co-immunoprecipitation, we found that the EDR3 protein was bound to E2F6, Bmi1, YY1 and M33 but not pRB1 or CtBP. This suggested that EDR3 was part of a previously described human Polycomb repressive complex (hPRC-H), which contains E2F6 and is active in G0. Consistent with this observation, we found that as cells entered G0, EDR3 localization in the nucleus changed from a diffuse to a highly punctuate pattern. This shift coincided with the ability of the EDR3-containing complex to bind to YY1 and E2F DNA-binding sites and to the c-myc promoter. When we tested co-localization of EDR3 with E2F6 and Bmi1, we found that the shift to a punctuate pattern of localization as the cells entered G0, was shared by EDR3 and E2F6 but not Bmi1. We found that EDR3 was affected in >70% of osteosarcoma tumors. Examination of other tumors showed that EDR3 protein was absent in other tumors suggesting that this could be a checkpoint, which is commonly lost in tumorigenesis. Consistent with this model, we found that EDR3 could act as a growth suppressor when overexpressed in normal cells, or when inducibly expressed in tumor cells. Our hypothesis is that the EDR3 protein acts as part of a repressor complex to regulate long-term maintenance of G0 in developing osteoblasts and mesenchymal stem cells. A corollary hypothesis is that loss of EDR3 function in osteoblast progenitor cells could lead to a loss of the ability to maintain G0 and thus favor tumorigenesis. To test our hypotheses, we propose the following specific aims: 1) To examine the role of EDR3 in osteosarcoma tumorigenesis and osteoblastic differentiation; 2) To characterize the components of the PcG complex containing EDR3 and E2F6; and 3) To identify the targets of EDR3 regulation.