We study the relationships between mammalian double strand break (DSB) repair, genomic instability and cancer. Two major pathways contribute to DSB repair: homologous recombination (HR) and non-homologous end joining (NHEJ). Each plays a key role in suppressing cancer and aging in mammals. Indeed, defects in HR, caused by germ line mutation of one of two major hereditary breast/ovarian cancer predisposition genes, BRCA1 and BRCA2, are significant contributors to cancer risk in the human population.
BRCA1 and BRCA2, together with interacting recombination proteins such as Rad51, mediate an error-free form of HR called “sister chromatid recombination” (SCR). Approximately twenty years ago, we proposed that BRCA genes function as tumor suppressors by controlling SCR at sites of replication fork stalling in replicating cells. We adapted the Escherichia coli Tus/Ter replication fork barrier for use in mammalian cells, enabling us stall the mammalian replisome at a defined chromosomal locus, following transient expression of the Tus polypeptide. This new tool enables us to address fundamental questions regarding the biology of the mammalian stalled fork response, ranging from biochemical analyses to the molecular characterization of repair processes triggered by fork stalling. Our work has revealed that the “rules” of repair at stalled forks differs substantially from those governing repair of a conventional chromosomal double strand break. We found that BRCA1 and BRCA2 suppress error-prone replicative responses at stalled forks—a process potentially analogous to “break-induced replication” in yeast. This finding suggested a new mechanism, specific to the stalled fork, by which loss of BRCA genes may promote genomic instability and cancer. Our most recent work has revealed the mechanism underlying a novel tandem duplicator phenotype in BRCA1 mutant cancers, which we find is also triggered specifically by fork stalling. These lines of research have enabled us to begin to model complex chromosomal rearrangements arising from defective processing of the stalled fork.
Many human BRCA variants are known to be either neutral or pathogenic. Others, termed BRCA “variants of uncertain significance” (VUS), are difficult to classify due to their scarcity in the human population. Since the cancer risk associated with VUS alleles is unknown, it is not clear what prophylactic interventions, if any, are appropriate for a woman who carries a BRCA VUS allele. Our underlying hypothesis is that the HR function of a given BRCA variant can be used to predict its status as a neutral or pathogenic variant. We developed a method to rapidly analyze the HR functions of individual BRCA1 variants. In collaboration with a number of cancer genetics labs, we are using this assay in an attempt to predict cancer risk associated with individual BRCA1 VUS alleles.
Division of Hematology-Oncology/Cancer Research Institute
Center for Life Sciences, CLS 438
3 Blackfan Circle
Boston, MA 02115