Mammalian genomes encode several proteins, which have domains that can bind to DNA containing 5- methylcytosines at CpG dinucleotides (methyI-CpG binding domains or MBDs; present for example in the proteins MeCP2, MBD1, MBD2, MBD3, and MBD4). Two of these proteins, MBD2 and MBD3, are closely related and MBD2 is a known transcriptional repressor. Recently, we have identified and cloned two new homologues of MBD2 and MBD3, which we have named MBD3L1 and MBD3L2 (for """"""""MBD3-1ike 1 and 2""""""""). Both MBD3L1 and MBD3L2 lack the MBD domains. MBD3L1 has so far been characterized as a testisspecific transcriptional repressor. In this application, we propose to analyze these new MBD3-1ike proteins in detail. Our hypothesis is that MBD3L1 and MBD3L2 are transcriptional regulators that function in conjunction with MBD2 and/or MBD3 to modulate gene activity. We propose four Specific Aims: 1) to characterize the transcriptional repression functions of MBD3L1 and MBD3L2 and its dependence on CpG methylation, 2) to identify and characterize the in vivo binding partners of MBD3L1 and MBD3L2 (preliminary data suggest that these proteins can form heterodimers with MBD2 and MBD3), 3) to determine if MBD3L1 and MBD3L2 are components of a larger complex in vivo (e.g., the Mi-2/NuRD complex), and 4) to derive knockout mice for Mbd311 and Mbd312. We will test the hypothesis that Mbd311 knockout mice will have a male sterility phenotype. Since both proteins are likely functioning as transcriptional regulators, it will be important to determine if there are changes in gene expression patterns in Mbd311 and Mbd312 (-/-) cells.