Breast cancer remains the most commonly diagnosed malignancy and is the second leading cause of cancer-related death in American women. It is estimated that there were more than 266,000 new cases of breast cancer in the United States in 2018, with 41,000 patients succumbing to the disease. Although the 5-year survival approaches 100% for those patients diagnosed with localized disease, the 5-year survival rate for patients diagnosed with distant metastatic disease in only 27%, highlighting the critical importance of the metastatic process in patient mortality. For those patients diagnosed with localized disease, the widespread application of adjuvant chemotherapy has reduced late relapse and long-term mortality by an estimated 19%. However, despite these efforts, 25% of patients receiving adjuvant chemotherapy still progress to metastatic disease. Despite changes in therapeutic strategies little improvement in the survival of these patients has been observed . At present it is estimated that 155,000 women are currently living with metastatic disease in the United States, highlighting the significant public health burden of this disease. It is therefore critically important to obtain a comprehensive understanding of the etiology and biology of metastases to develop more effective clinical interventions to further reduce the morbidity and mortality of advanced breast cancer. My laboratory pioneered the use of mouse meiotic genetic screens to gain additional insights into the etiology of metastatic progression in breast cancer. To implement this strategy my laboratory has exploited the highly metastatic FVB/N-Tg(MMTV-PyMT)634Mul genetically engineered mouse model of mammary carcinogenesis. This model expresses the mouse polyoma middle-T antigen (PyMT), expressed using mouse mammary tumor virus (MMTV) enhancer and promoter. This results in the activation of the PI-3-Kinase pathway, modeling one of the most frequently mutated pathways in human breast cancer. As a result, nulliparous female PyMT mice develop synchronous palpable, multifocal mammary tumors by 60 days of age. Detailed histopathology analysis suggests that this model has a high degree of similarity with human breast cancer and gene expression profiling suggest that it most closely resembles the luminal subtype of human breast cancer. By 100 days of age, in our colonies, greater than 85% of the animals develop macroscopic pulmonary metastases on the FVB/NJ background. To demonstrate that inherited allelic variants significantly contribute to metastatic progression, as a proof-of-principle experiment, we bred male PyMT mice to female mice from more than 25 different inbred mouse strains, representing different branches of the mouse phylogenetic tree. The oncogenic transgene was therefore on different genetic backgrounds in the resulting F1 progeny due to the contributions of the dams from each inbred strain. The F1 females were permitted to age for tumor and potential metastasis development, then euthanized and different tumor phenotypes measured. Significant variation in tumor latency, growth and pulmonary metastatic colonization were observed. Analysis of the metastatic capacity of the tumors across the different genetic backgrounds revealed a continuum rather than discrete classes of low or high metastatic burden, indicative of the presence of multiple polymorphic genes contributing to the phenotype rather than one or two dominant loci. No alterations in transgene expression were found, consistent with a role for polymorphic effects on these tumor phenotypes. Strains were identified that suppressed the metastatic capacity of the PyMT tumors, without altering tumor growth rate or latency. Chromosomal substitution strains and genetic experimental crosses were then generated to identify regions of the genome harboring metastasis susceptibility genes. Haplotype mapping across multiple experimental crosses restricted a candidate region on mouse chromosome 19 to a 110 kb interval containing only 5 genes. Sequencing analysis of the genes in the interval across high- and low-metastatic genetic backgrounds revealed a single amino acid substitution in Sipa1 that segregated with metastatic capacity. Functional analysis revealed that this substitution reduced the RAP-GAP activity of SIPA1 in the low metastatic strains. Subsequent modeling of this reduction by shRNA knockdown or overexpression of Sipa1 in a highly metastatic mammary tumor mouse allograft model demonstrated that Sipa1 transcript levels correlated with pulmonary metastatic burden. Epidemiological investigations by our lab and others found associations between SIPA1 polymorphisms and metastatic progression in human patients, consistent with SIPA1 as a metastasis susceptibility gene in both mice and humans. Population-based epidemiology further demonstrated the presence of inherited metastasis predisposition genes in breast cancer as well as other human cancers. These studies therefore validated our original hypothesis that inherited susceptibility is a significant component of cancer progression. In addition, these results suggest that our mouse-based genetic mapping strategy is a viable approach to identify human-relevant genes for further characterization of the metastatic cascade. Subsequently, my laboratory has increasingly incorporated genomic technologies to investigate how inherited polymorphisms influence metastatic disease. Incorporation of expression technologies, first Affymetrix chip-based expression analysis and more recently RNA-sequencing methods, with the genetic analysis have helped accelerate candidate gene identification for validation and analysis. In addition, we have also incorporated more sophisticated genetic mapping resources, primarily the Diversity Outbred mouse mapping panel, to improve the genetic mapping resolution and to generate experimental crosses that more closely resemble the genetic diversity observed across human populations. Moreover, we have utilized the resources generated from these projects to demonstrate the contribution of genetic background to clinically relevant factors such as the generation of breast cancer molecular subtype classifiers like the PAM50 and prognostic gene signatures such as MammaPrint. To date, my laboratory has published an additional 16 metastasis-associated genes and six microRNAs since the identification of Sipa1. In addition, we have validated an additional 11 genes as metastasis modifiers; these genes are in varying stages of characterization or manuscript preparation and submission.
Showing the most recent 10 out of 33 publications
