Current Studies


Our lab studies the interplay between genetic variants and somatic mutations, genetic variants important in cancer susceptibility in high-risk individuals and the general population, the utility of polygenic risk scores in risk prediction in breast cancer, colorectal cancer and cutaneous squamous cell carcinoma.  Descriptions of active studies are detailed below.

Germline variant by somatic mutation interactions

Somatic mutations in TP53, PIK3CA, and KRAS that drive breast and colorectal cancer (CRC) have unequal prevalence among racial and ethnic groups. Some of the factors leading to somatic mutations in specific genes can be attributed to socioeconomic and environmental exposures, but recent studies suggest that a large proportion of these may be due to inherited genetic variants. Emerging data from our lab and others suggest that germline genetic variants impact somatic events in tumors. We are using existing germline genotype data and tumor sequencing data to identify associations between germline genetic variants and mutations in genes showing racial-specific frequencies. Specifically we are looking for variants that associate with TP53- and PIK3CA- mutation status in breast cancer and KRAS mutation-status in CRC with the goals to understand the association between somatic mutations, germline genetic variants and tumorigenesis in racial disparities, and to determine the biological impact of GxM associations on cancer phenotypes using in vitro CRISPR knock-in and knock-out models.

Breast GxM:  Preliminary studies in breast cancer cases have identified 5 SNVs that associate with TP53 somatic mutations with p-values <1×10-6 and 34 with p-values <1×10-5. Some variants associated with the presence of any TP53 mutation; others were exclusively associated with loss-of-function or gain-of-function TP53 mutations. From the GxM analysis, we identified 44 SNVs that associate with PIK3CA mutations with p-values < 1×10-5.  Some variants were associated with the presence of any PIK3CA somatic mutation and others only with hotspot mutations (H1047R, E542K or E545K). We have chosen 192 variants using in silico tools for validation studies in multi-ancestral populations which are underway.

Colorectal GxM:  From the GxM study for variants associated with KRAS mutation status, we identified three SNVs with p-values less than 1×10-6, 50 SNVs with p-values less 1×10-5 and 6,371 SNVs with p-values <0.0001; top SNVs represent multiple independent loci.  Using in silico tools, we have selected 96 variants for validation.

OCA2 Variant Classification and non-UV protective roles in skin cancer

Pigmentation is very strongly associated with skin cancer, as dark pigment provides effective ultraviolet (UV) light protection. Variants in pigmentary genes have been associated with risk of basal cell carcinoma (BCC), cutaneous melanoma (CM) and cutaneous squamous cell carcinoma (cSCC) in candidate gene studies and genome-wide association studies (GWAS) Pathogenic variants in pigment genes, including OCA2 and TYR, are associated with autosomal recessive albinism and hereditary melanoma. In previous studies, we found alleles associated with blue eye color and skin tone at the OCA2 gene locus to be associated with decreased time to first cSCC in organ transplant recipients. It is widely thought that the increased cancer risk in patients with blue eye color alleles is due to the loss of UV light protection in the skin by pigmentation.  However, ocular melanomas rarely exhibit UV mutational signatures and blue eye color is a risk factor for ocular melanoma. Emerging studies suggest that deficiency of OCA2 function may also impact other phenotypes including angiogenesis.  To explore potential non-UV protective role(s) of OCA2 in skin cancer we are collaborating with Brad Blaser’s laboratory and have developed an oca2 knock-out fish which will be crossed to melanoma fish models.

OCA2 variants associated with decreased time to first cSCC post-transplant

A related aim is to use the zebrafish model to screen tens of missense changes of uncertain significance (variants of uncertain significance) in the OCA2 gene found upon clinical testing for effects on pigment phenotypes.  OCA2 pathogenic variants can increase the risk of skin cancer and can cause Oculocutaneous Albinism type 2 when inherited on both alleles.

This project will test the hypothesis that zebrafish can be used as a model to rapidly screen missense variants in the human OCA2 gene for impacts on skin cancer development, pigment and non-pigment related phenotypes and that oca2 loss enhances melanoma latency in a zebrafish model of melanoma in a non-UV setting.

Polygenic Risk Score RCT in BRCA Pathogenic Variant Carriers.

Genome-wide association studies (GWAS) have uncovered hundreds of variants that modify risk of complex diseases. What hasn’t been fully addressed is whether this information has utility for clinical management, and, if so, how to incorporate genetic modifiers into patient clinical management. Pathogenic variants in BRCA1 and BRCA2 are associated with a high lifetime risk of developing breast and ovarian cancer. Currently individuals found to carry a mutation in these genes are given a large range of lifetime risk estimates during result disclosure. Some of the variation in cancer risk in BRCA mutation carriers is due to genetic modifiers.  Studies by others and those we have been involved with have shown that polygenic risk scores that incorporate these genetic risk modifiers can be used to more precisely define breast cancer risk in BRCA1 and BRCA2 pathogenic variant carriers. We are conducting a randomized controlled trial to determine if a more personalized risk score has benefit for BRCA1/2 pathogenic variant carriers in decision-making processes for cancer prevention management.

Drivers for Aggressive cSCC

Cutaneous SCC is the second most common form of cancer with over 700,000 SCCs treated in the United States each year.  The majority of SCCs are readily treated through surgery; however, 2-3% of SCCs metastasize leading to over 3000 deaths per year.  Until recently, metastatic SCC was associated with extremely poor survival rate with only 56% of individuals with distant metastases surviving.  With newer immunotherapies, it is expected that the mortality rate will decrease.  However, some individuals with high risk cSCC, such as organ transplant recipients, may not be good candidates for immunotherapy. Thus, new therapeutic strategies are needed. In our previous exome studies we found that somatic mutations in KMT2D, encoding a lysine methyltransferase, occurred at nearly twice the frequency (62%) in metastatic or aggressive cSCCs compared to primary cSCCs that did not metastasize.  On-going studies are evaluating the phenotypic consequences of loss of KMT2D in isogenic cSCC cell lines.  Our data suggest that loss of KMT2D increases migration and progression through the cell cycle. As KMT2D is important in regulation, future studies will evaluate the pathways and regulatory perturbations of KMT2D somatic mutations in cSCCs.

Migration of wildtype COLO-16 cells compared to two KMT2D CRISPR-knock out lines, X18314 and X18449.  Knock-out lines show faster migration.

Consortium/Collaborative Studies

Together with genetic counselors in the Division of Human Genetics, we contribute data and biospecimens from consented individuals with pathogenic variants and variants of uncertain significance (VUS) in high risk cancer susceptibility genes to the CIMBA and ENIGMA consortia in order to identify genetic and lifestyle modifiers of penetrance of high risk genes, to understand outcomes and prognostic markers and to classify VUS.