Disentangling the effect of Black/African Ancestry from confounding variables on colorectal cancer biology

SUMMARY There is a higher incidence and mortality of colorectal cancer (CRC) in individuals self-identifying as Black race compared to all other races and ethnicities in the US. Previous studies comparing self-identified Black vs self-identified White CRC frozen samples have detected significant differences in gene expression and indicate

that self-identified race is significantly associated with differences in CRC biology. However, the list of DEGs generated independently from each individual study shows limited overlap. This is likely because of the small sample sizes for each dataset, with the largest set being the TCGA data set with ~60 self-identified Black

samples. Complex environmental and germline genomic factors contribute to the DEGs and other biological results. However, a major barrier to disentangling these factors is accessing large numbers of CRC tissues annotated for potentially confounding co-variables linked to self-identified racial categories. To conduct such a

study we propose to demonstrate the feasibility of assembling a multi-disciplinary research team to analyze a large (>1000) self-identified Black/ African Ancestry (AA) vs. White/European Ancestry (EA) formalin fixed paraffin embedded (FFPE) tissue samples archived at Stony Brook University Hospital (SBUH), Kings County

Hospital Center (KCHC/NYCH+H) and Henry Ford Hospital (HFH). CXCL10 will be the initial DEG target since this is one of the few DEGs identified by more than one RNA-sequence dataset. CXCL10 is an interferon gamma inducible protein that is a T-cell attractant and has been positively associated with T-cell infiltration in CRC and

with post-operative progress free survival. The reduced expression of CXCL10 in self-identified Black CRC vs. White CRC could therefore potentially link self-identified racial differences in CRC immunologic transcriptional profiles and CRC outcomes. Data from the Broad single cell (sc) RNA sequence (seq) data portal revealed that

myeloid cells and particularly macrophage cell subtypes are a major source of CXCL10 expression in CRC tumors. However, no information on racial categories was made available for this scRNA dataset. A recent analysis of the TCGA data revealed that macrophage signature transcripts are also reduced in self-identified

Black vs. White CRC (10) but it remains to be demonstrated whether macrophage signature transcripts are colocalized with the CXCL10 transcripts in the same cell. In Aim 1, spatial profiling of CXCL10 expression in CRC FFPE sections using RNAScopeTM in situ hybridization will be piloted to evaluate the hypothesis that

CXCL10 expressing macrophages are reduced in self-identified Black vs. White CRC. In Aim 2, annotation of CRC FFPE tissues with respect to potentially confounding co-variables, (including estimates of ancestry admixture and mismatch repair- immunohistochemistry results) will be harmonized, with the overarching goal of

conducting multivariable analyses of racial effects on CRC cancer biology.