Novel Approaches to Allelic Series of Genomic Modifications in Mammals: Despite the long history of the mouse as a model system for mammalian genetics, as well as a decade of gene targeting experiments, mutations exist for only a small percentage of its genes. The high cost of animal husbandry and the difficulty of optimizing and monitoring mutation frequency represent major limitations when designing screens in mice. We have worked on an alternative approach that capitalizes on methods capable of systematically generating mutations in totipotent embryonic stem (ES) cells and then to derive mice from the mutagenized cells. We have begun large-scale production of mutagenized ES cells followed by high-throughput mutation detection. The immediate goal is to produce an allelic series of mutations in several genes of interest to the lab.

Mammalian Polycomb-group Complexes: In mammals, dosage compensation of X-linked genes is achieved by transcriptional silencing of one X chromosome in the female. This process, called X-inactivation, is usually random in the embryo proper. However, in marsupials and the extraembryonic region of the mouse, X-inactivation is imprinted: the paternal X chromosome is preferentially inactivated while the maternal X is always active. We have shown that eed, a member of the mouse Polycomb-group (Pc-G) of genes, is critical for imprinted X-inactivation in extraembryonic tissue. Results from mice carrying a paternally inherited X-linked GPF transgene implicate eed in the stable maintenance of imprinted X-inactivation in extraembryonic tissues. Based on the recent finding that Eed interacts with histone deacetylases, we hypothesize that this maintenance activity involves hypoacetylation of the inactivated paternal X in he extraembryonic tissues. We are also pursuing the hypothesis that eed is required for maintenance of autosomally imprinted genes.

Mammalian Swi/Snf Chromatin Remodeling Complexes: The mammalian SWI/SNF complexes utilize the energy of ATP hydrolysis to displace nucleosomes from promoter regions and enable numerous genes to be transcribed. Mammals have two catalytic subunits, brahma (Brm) and brahma-related gene 1 (Brg 1), that are 75% identical but found in distinct complexes. Brm was previously shown to be dispensable for embryonic development suggesting that Brm and Brg1 are functionally redundant. However, we have generated a null mutation of Brg1 and surprisingly, homozygotes die during the peri-implantation stage. In addition, Brg1 heterozygotes are predisposed to exencephaly and apocrine gland tumors suggesting that gene dosage is important. These results provide the first evidence that biochemically similar chromatin-remodeling complexes can have dramatically different functions during mammalian development.