The genetic contributions to breast cancer development among Latinas are not well understood. Here, we carry out a genome-wide association study of breast cancer in Latinas and identify a genome-wide significant risk variant, located 5’ of the Estrogen Receptor 1 gene (ESR1) (6q25 region). The minor allele for this variant is strongly protective (rs140068132: OR 0.60, 95%CI 0.53-0.67, P=9×10−18), originates from Indigenous Americans, and is uncorrelated with previously reported risk variants at 6q25. The association is stronger for estrogen receptor negative disease (OR 0.34 95% CI 0.21-0.54) than estrogen receptor positive disease (OR 0.63 95% CI 0.49-0.80) (P heterogeneity=0.01) and is also associated with mammographic breast density, a strong risk factor for breast cancer (P=0.001). rs140068132 is located within several transcription factor binding sites and electrophoretic mobility shift assays with MCF-7 nuclear protein demonstrate differential binding of the G/A alleles at this locus. These results highlight the importance of conducting research in diverse populations.

Background: Multiple breast cancer genome-wide association studies (GWAS) have been conducted, with most samples analyzed being of European or Asian origin. No GWAS has been published to date on US Latinas or Latin American women. Previous studies have demonstrated that among Latina women, those with the highest Indigenous American ancestry are protected from breast cancer. We performed the first genome-wide association study of breast cancer in Latinas in a total of 1523 cases and 3221 controls and validated the strongest associations in an additional 977 Latina cases and 1158 controls. Methods: For the discovery phase, samples were genotyped with either the 6.0 Affymetrix array or Illumina Infinitum 660W/Omni 2.5 arrays. Additional genotypes were imputed from 1000 genomes and the final analysis included ∼7 million SNPs. SNPs for replication were genotyped using Sequenom. Logistic regression analyses included the first 10 principal components as covariates. Results: We identified two new variants, located 5’ of the ESR1 gene (6q25 region), which are strongly associated with breast cancer in this population. The minor allele for these variants is strongly protective (rs140068132: OR 0.61, 95%CI 0.48-0.75, p=3x10-13 & rs147157845: OR 0.63 95% CI 0.49-0.76; p =2x10-12). These alleles are present only in Latinos, correlate with Indigenous American ancestry, and are uncorrelated with previously reported risk variants in the ESR1 region. We found a strong association for both ER-positive and ER-negative breast cancer, but significantly more protection for ER-negative disease (p=0.014). The rs140068132 variant is also associated with breast density. The average breast density among women who are homozygous for the protective allele is approximately half of that of women homozygous for the common allele [Mean (sd) density for Homozygous GG = 9 (8); Homozygous AA = 17 (14), p 0.0009]. rs140068132, is located within a putative binding region for multiple transcription factors. MAPPER predicts disruption of putative transcription factor binding sites for the rare allele at rs140068132 but not rs147157845. Electromotility shift assays showed differential binding of the two alleles at the rs140068132 locus, with the protective allele (G) showing substantially reduced binding affinity compared to the common A allele. Conclusions: The variants at this locus represent, to our knowledge, the strongest genome wide effect size on breast cancer for a common variant (frequency >0.05 in any major population). These results may help in understanding the previous observations of lower breast cancer risk among Latina women with more Indigenous American ancestry and highlight the importance of conducting GWAS in diverse populations. Citation Format: Laura Fejerman, Nasim Ahmadiyeh, Donglei Hu, Scott Huntsman, Kenneth Beckman, Jennifer Caswell, Esther M. John, Gabriela Torres-Mejia, Luis Carvajal-Carmona, Magdalena Echeverry, Anna Marie Tuazon, Carolina Ramirez, COLUMBUS Consortium, Christopher Gignoux, Celeste Eng, Esteban Gonzalez-Burchard, Brian Henderson, Loic Le Marchand, Eliseo J. Perez-Stable, Christopher A. Haiman, Elad Ziv. Genome wide association study of breast cancer in Latinas identifies protective variants of Indigenous American origin on 6q25. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr LB-299. doi:10.1158/1538-7445.AM2014-LB-299

Background The BRCA1 c.3331_3334delCAAG founder mutation has been reported in hereditary breast and ovarian cancer families from multiple Hispanic groups. We aimed to evaluate BRCA1 c.3331_3334delCAAG haplotype diversity in cases of European, African, and Latin American ancestry. Methods BC mutation carrier cases from Colombia (n = 32), Spain (n = 13), Portugal (n = 2), Chile (n = 10), Africa (n = 1), and Brazil (n = 2) were genotyped with the genome-wide single nucleotide polymorphism (SNP) arrays to evaluate haplotype diversity around BRCA1 c.3331_3334delCAAG. Additional Portuguese (n = 13) and Brazilian (n = 18) BC mutation carriers were genotyped for 15 informative SNPs surrounding BRCA1. Data were phased using SHAPEIT2, and identical by descent regions were determined using BEAGLE and GERMLINE. DMLE+ was used to date the mutation in Colombia and Iberia. Results The haplotype reconstruction revealed a shared 264.4-kb region among carriers from all six countries. The estimated mutation age was ~ 100 generations in Iberia and that it was introduced to South America early during the European colonization period. Conclusions Our results suggest that this mutation originated in Iberia and later introduced to Colombia and South America at the time of Spanish colonization during the early 1500s. We also found that the Colombian mutation carriers had higher European ancestry, at the BRCA1 gene harboring chromosome 17, than controls, which further supported the European origin of the mutation. Understanding founder mutations in diverse populations has implications in implementing cost-effective, ancestry-informed screening.

Cancer is a leading cause of non-communicable morbidity and mortality worldwide. While many genes that predispose individuals for different cancers have been discovered, the population prevalence of mutations in these genes remains largely undetermined in many populations. Developing customized screening panels and screening for novel variants for a large population samples can be expensive and time consuming. We have developed a low-cost high-throughput pipeline and method to screen 480 customizable amplicons (∼20 genes, ∼144Kbp) for up to 384 samples per run. By combining a bioinformatics pipeline for design and analysis with Fluidigm microfluidics PCR and Illumina MiSeq, we can quickly screen the full coding sequence of multiple cancer panels at a high depth of coverage at a low cost per sample (∼$20-40/sample). The amplicon design pipeline provides ability to develop amplicon primer sets and pooling strategies. Libraries of 384 barcoded samples are run in a single MiSeq lane. Sequencing data is analyzed by an automated pipeline, which aligns using BWA-MEM, calls variants using VarScan 2, and annotates variants using multiple datasets using Annovar. We have applied these pipelines and methods to identify mutation in known cancer genes, such as APC, MSH2, MLH1, BRCA1, and BRCA2, in several hundred Hispanic individuals with familial and early-onset colon and breast cancer. While many of these mutations have been previously reported, we have identified several novel pathogenic changes that appear to have Amerindian origin. Of which, a few are shared among many individuals in isolated geographic locations, suggesting founder effects are common in some of these populations. In conclusion, we have developed a low-cost high-throughput method for screening customizable panels for known and novel mutations. These panels typically consist of 20 cancer genes and our group has already developed panels that are specific for breast cancer, colon cancer and thyroid cancer. Our study is an initial step to assess prevalence of known cancer causing genes and identify novel genes/mutations that contribute to different cancers in the Hispanic community. These discoveries provide a foundation for early detection, prevention, and treatment of familial cancers. Citation Format: Ruta Sahasrabudhe,, Paul Lott, Anna Marie Tuazon, John Williamson, Natalia Belter, Ana Estrada, Mabel Bohorquez, Rodrigo Prieto, Angel Criollo, Alejandro Velez, Jorge Castro, Gilbert Mateus, Magdalena Echhevery, Luis G. Carvajal-Carmona. Development of low-cost high-throughput screening methods for detecting germline mutations in multiple cancer genes. [abstract]. In: Proceedings of the AACR Special Conference: Cancer Susceptibility and Cancer Susceptibility Syndromes; Jan 29-Feb 1, 2014; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(23 Suppl):Abstract nr 42. doi:10.1158/1538-7445.CANSUSC14-42

Cancer is a leading cause of non-communicable morbidity and mortality worldwide. Hispanics are the largest and fastest growing ethnic group in the U.S. and the largest population group in the Americas. Due to health care disparities, this population remains understudied and under-screened for cancer causing gene mutations. While many genes that predispose individuals for different cancers have been discovered, the prevalence of mutations in these genes remains largely undetermined in many populations. Developing customized screening panels and screening for novel variants for large population samples can be expensive and time consuming. We have developed a low-cost, high-throughput pipeline and method to screen 480 customizable amplicons (∼20 genes, ∼144Kbp) for up to 384 samples per run. By combining a bioinformatics pipeline for design and analysis with Fluidigm microfluidics PCR and Illumina MiSeq, we can quickly screen multiple cancer panels at a high depth of coverage, at a low cost per sample. The amplicon design pipeline provides ability to develop amplicon primer sets and pooling strategies. Libraries of 384 barcoded samples are run in a single MiSeq lane. Sequencing data is analyzed by an automated pipeline, which aligns using BWA-MEM, calls variants using VarScan 2, and annotates variants using multiple datasets using Annovar. We have applied these pipelines and methods to identify mutation in known cancer genes, such as APC, MSH2, MLH1, BRCA1, and BRCA2, in several hundred Hispanic individuals with familial and early-onset colon and breast cancer. While many of the mutations have been previously reported, we have identified several novel mutations that appear to have Amerindian origin. Moreover, a few are shared among many individuals in isolated geographic locations, suggesting founder effects are common in some of these populations. Over 50% of the patients screened to date have no mutations in our cancer panels; we plan to use these samples to screen for novel genes using whole exome or genome sequencing in the next phase of our study. In conclusion, we have developed a low-cost, high-throughput pipeline for screening customizable panels for known and novel mutations. Our study is an initial step to assess prevalence of known cancer causing genes and identify novel genes/mutations that contribute to different cancers in the Hispanic community. These discoveries provide a foundation for early detection, prevention, and treatment of familial cancers in Hispanic populations. Citation Format: Paul Lott, Ruta Sahasrabudhe, Anna Marie Tuazon, John Williamson, Ana Estrada, Mabel Bohorquez, Rodrigo Prieto, Angel Criollo, Alejandro Velez, Jorge Castro, Gilbert Mateus, María Magdalena Echeverry, Luis Carvajal-Carmona. Development of low-cost high throughput screening pipeline for detecting germline cancer causing mutations in Hispanic populations. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 1520. doi:10.1158/1538-7445.AM2014-1520