Cancer Genome-wide Association Studies Research Articles

Loss-of-function variants in ATM confer risk of gastric cancer.

Gastric cancer is a serious health problem worldwide, with particularly high prevalence in eastern Asia. Genome-wide association studies (GWAS) in Asian populations have identified several loci that associate with gastric cancer risk. Here we report a GWAS of gastric cancer in a European population, using information on 2,500 population-based gastric cancer cases and 205,652 controls. We found a new gastric cancer association with loss-of-function mutations in ATM (gene test, P = 8.0 × 10(-12); odds ratio (OR) = 4.74). The combination of the loss-of-function variants p.Gln852*, p.Ser644* and p.Tyr103* (combined minor allele frequency (MAF) = 0.3%) also associates with pancreatic and prostate cancers (OR = 3.81 and 2.18, respectively) and gives an indication of risk of breast and colorectal cancers (OR = 1.82 and 1.97, respectively). Cancers in those carrying loss-of-function ATM mutations are diagnosed at a significantly earlier age than in non-carriers. Our results confirm an association between gastric cancer in Europeans and three loci previously reported in Asians, MUC1, PRKAA1 and PSCA, refine the association signal at PRKAA1 and support a pathogenic role for the tandem repeat identified in MUC1.

Deciphering Signaling Pathway Networks to Understand the Molecular Mechanisms of Metformin Action.

A drug exerts its effects typically through a signal transduction cascade, which is non-linear and involves intertwined networks of multiple signaling pathways. Construction of such a signaling pathway network (SPNetwork) can enable identification of novel drug targets and deep understanding of drug action. However, it is challenging to synopsize critical components of these interwoven pathways into one network. To tackle this issue, we developed a novel computational framework, the Drug-specific Signaling Pathway Network (DSPathNet). The DSPathNet amalgamates the prior drug knowledge and drug-induced gene expression via random walk algorithms. Using the drug metformin, we illustrated this framework and obtained one metformin-specific SPNetwork containing 477 nodes and 1,366 edges. To evaluate this network, we performed the gene set enrichment analysis using the disease genes of type 2 diabetes (T2D) and cancer, one T2D genome-wide association study (GWAS) dataset, three cancer GWAS datasets, and one GWAS dataset of cancer patients with T2D on metformin. The results showed that the metformin network was significantly enriched with disease genes for both T2D and cancer, and that the network also included genes that may be associated with metformin-associated cancer survival. Furthermore, from the metformin SPNetwork and common genes to T2D and cancer, we generated a subnetwork to highlight the molecule crosstalk between T2D and cancer. The follow-up network analyses and literature mining revealed that seven genes (CDKN1A, ESR1, MAX, MYC, PPARGC1A, SP1, and STK11) and one novel MYC-centered pathway with CDKN1A, SP1, and STK11 might play important roles in metformin’s antidiabetic and anticancer effects. Some results are supported by previous studies. In summary, our study 1) develops a novel framework to construct drug-specific signal transduction networks; 2) provides insights into the molecular mode of metformin; 3) serves a model for exploring signaling pathways to facilitate understanding of drug action, disease pathogenesis, and identification of drug targets.

Common variation at 1q24.1 (ALDH9A1) is a potential risk factor for renal cancer.

So far six susceptibility loci for renal cell carcinoma (RCC) have been discovered by genome-wide association studies (GWAS). To identify additional RCC common risk loci, we performed a meta-analysis of published GWAS (totalling 2,215 cases and 8,566 controls of Western-European background) with imputation using 1000 Genomes Project and UK10K Project data as reference panels and followed up the most significant association signals [22 single nucleotide polymorphisms (SNPs) and 3 indels in eight genomic regions] in 383 cases and 2,189 controls from The Cancer Genome Atlas (TCGA). A combined analysis identified a promising susceptibility locus mapping to 1q24.1 marked by the imputed SNP rs3845536 (P combined =2.30x10-8). Specifically, the signal maps to intron 4 of the ALDH9A1 gene (aldehyde dehydrogenase 9 family, member A1). We further evaluated this potential signal in 2,461 cases and 5,081 controls from the International Agency for Research on Cancer (IARC) GWAS of RCC cases and controls from multiple European regions. In contrast to earlier findings no association was shown in the IARC series (P=0.94; P combined =2.73x10-5). While variation at 1q24.1 represents a potential risk locus for RCC, future replication analyses are required to substantiate our observation.

Deciphering Genome Environment Wide Interactions Using Exposed Subjects Only.

The recent successes of genome-wide association studies (GWAS) have renewed interest in genome environment wide interaction studies (GEWIS) to discover genetic factors that modulate penetrance of environmental exposures to human diseases. Indeed, gene-environment interactions (G × E), which have not been emphasized in the GWAS era, could be a source contributing to the missing heritability, a major bottleneck limiting continuing GWAS successes. In this manuscript, we describe a design and analytic strategy to focus on G × E using only exposed subjects, dubbed as e-GEWIS. Operationally, an e-GEWIS analysis is equivalent to a GWAS analysis on exposed subjects only, and it has actually been used in some earlier GWAS without being explicitly identified as such. Through both analytics and simulations, e-GEWIS has been shown better efficiency than the usual cross-product-based analysis of G × E interaction with both cases and controls (cc-GEWIS), and they have comparable efficiency to case-only analysis of G × E (c-GEWIS), with potentially smaller sample sizes. The formalization of e-GEWIS here provides a theoretical basis to legitimize this framework for routine investigation of G × E, for more efficient G × E study designs, and for improvement of reproducibility in replicating GEWIS findings. As an illustration, we apply e-GEWIS to a lung cancer GWAS data set to perform a GEWIS, focusing on gene and smoking interaction. The e-GEWIS analysis successfully uncovered positive genetic associations on chromosome 15 among current smokers, suggesting a gene-smoking interaction. Although this signal was detected earlier, the current finding here serves as a positive control in support of this e-GEWIS strategy.

A Recurrent Mutation in PARK2 Is Associated with Familial Lung Cancer

PARK2, a gene associated with Parkinson disease, is a tumor suppressor in human malignancies. Here, we show that c.823C>T (p.Arg275Trp), a germline mutation in PARK2, is present in a family with eight cases of lung cancer. The resulting amino acid change, p.Arg275Trp, is located in the highly conserved RING finger 1 domain of PARK2, which encodes an E3 ubiquitin ligase. Upon further analysis, the c.823C>T mutation was detected in three additional families affected by lung cancer. The effect size for PARK2 c.823C>T (odds ratio = 5.24) in white individuals was larger than those reported for variants from lung cancer genome-wide association studies. These data implicate this PARK2 germline mutation as a genetic susceptibility factor for lung cancer. Our results provide a rationale for further investigations of this specific mutation and gene for evaluation of the possibility of developing targeted therapies against lung cancer in individuals with PARK2 variants by compensating for the loss-of-function effect caused by the associated variation.

Addressing population-specific multiple testing burdens in genetic association studies.

The number of effectively independent tests performed in genome-wide association studies (GWAS) varies by population, making a universal P-value threshold inappropriate. We estimated the number of independent SNPs in Phase 3 HapMap samples by: (1) the LD-pruning function in PLINK, and (2) an autocorrelation-based approach. Autocorrelation was also used to estimate the number of independent SNPs in whole genome sequences from 1000 Genomes. Both approaches yielded consistent estimates of numbers of independent SNPs, which were used to calculate new population-specific thresholds for genome-wide significance. African populations had the most stringent thresholds (1.49 × 10(-7) for YRI at r(2) = 0.3), East Asian populations the least (3.75 × 10(-7) for JPT at r(2) = 0.3). We also assessed how using population-specific significance thresholds compared to using a single multiple testing threshold at the conventional 5 × 10(-8) cutoff. Applied to a previously published GWAS of melanoma in Caucasians, our approach identified two additional genes, both previously associated with the phenotype. In a Chinese breast cancer GWAS, our approach identified 48 additional genes, 19 of which were in or near genes previously associated with the phenotype. We conclude that the conventional genome-wide significance threshold generates an excess of Type 2 errors, particularly in GWAS performed on more recently founded populations.

Comparison of dimension reduction-based logistic regression models for case-control genome-wide association study: principal components analysis vs. partial least squares.

With recent advances in biotechnology, genome-wide association study (GWAS) has been widely used to identify genetic variants that underlie human complex diseases and traits. In case-control GWAS, typical statistical strategy is traditional logistical regression (LR) based on single-locus analysis. However, such a single-locus analysis leads to the well-known multiplicity problem, with a risk of inflating type I error and reducing power. Dimension reduction-based techniques, such as principal component-based logistic regression (PC-LR), partial least squares-based logistic regression (PLS-LR), have recently gained much attention in the analysis of high dimensional genomic data. However, the performance of these methods is still not clear, especially in GWAS. We conducted simulations and real data application to compare the type I error and power of PC-LR, PLS-LR and LR applicable to GWAS within a defined single nucleotide polymorphism (SNP) set region. We found that PC-LR and PLS can reasonably control type I error under null hypothesis. On contrast, LR, which is corrected by Bonferroni method, was more conserved in all simulation settings. In particular, we found that PC-LR and PLS-LR had comparable power and they both outperformed LR, especially when the causal SNP was in high linkage disequilibrium with genotyped ones and with a small effective size in simulation. Based on SNP set analysis, we applied all three methods to analyze non-small cell lung cancer GWAS data.

Evaluating methods for modeling epistasis networks with application to head and neck cancer.

Epistasis helps to explain how multiple single-nucleotide polymorphisms (SNPs) interact to cause disease. A variety of tools have been developed to detect epistasis. In this article, we explore the strengths and weaknesses of an information theory approach for detecting epistasis and compare it to the logistic regression approach through simulations. We consider several scenarios to simulate the involvement of SNPs in an epistasis network with respect to linkage disequilibrium patterns among them and the presence or absence of main and interaction effects. We conclude that the information theory approach more efficiently detects interaction effects when main effects are absent, whereas, in general, the logistic regression approach is appropriate in all scenarios but results in higher false positives. We compute epistasis networks for SNPs in the FSD1L gene using a two-phase head and neck cancer genome-wide association study involving 2,185 cases and 4,507 controls to demonstrate the practical application of the methods.

Genetic Susceptibility Markers of Multiple Myeloma in African-Americans

Multiple myeloma (MM) is 2-3 times more common among African-Americans compared to non-Hispanic whites. The 2-3-fold increased risk among family members of cases suggests a genetic contribution to risk. Genome-wide association studies (GWAS) in populations of European ancestry have identified seven novel risk loci at 2p23.3 (rs6746082), 3q26.2 (rs10936599), 3p22.1 (rs1052501), 6p21.32 (rs2285803), 7p15.3 (rs4487645), 17p11.2 (rs4273077) and 22q13.1 (rs877529) (Broderick, et al. Nat Genet, 2011, Chubb, et al. Nat Genet, 2013), three of which were replicated in another European series (Martino et al., Br J Haematol, 2012). Here we examined the index signals and conducted fine-mapping for each locus in a case-control study of 1,049 multiple myeloma cases and 7,084 controls of African ancestry to identify better markers of risk and novel independent loci in seven previously reported regions in this high risk population.Incident cases were recruited from 10 clinical centers and SEER cancer registries from 2011 to 2013 and genotyped using the Illumina HumanCore GWAS array. Control data were obtained from previous genome-wide studies of breast and prostate cancer, genotyped using the Illumina 1M-Duo in 4425 male controls from the African Ancestry Prostate Cancer Consortium (consisting of 14 independent studies) and 2632 female controls from a breast cancer GWAS of African-American women (consisting of 9 independent studies). Imputation to 1000 Genomes (March 2012 release) was conducted for regions around six of the previously identified single nucleotide polymorphisms [SNPs] (the HLA region harboring rs2285803 is still being imputed, results will be presented). A case-control analysis of SNPs/indels >1% frequency within 250 kb of each index variant was conducted using unconditional multivariable logistic regression adjusting for age, sex and five leading principal components. Region-specific alpha levels were determined through permutation tests. The minimum alpha level across the six regions was α=0.002.All previously reported risk variants were common in African-Americans (minor allele frequency [MAF] >0.05). For five of the six SNPs, we had ≥94% power to detect the same effect observed in non-Hispanic whites, and 64% power for the less common variant rs10936599 (MAF=0.07). We observed directionally consistent effects (odds ratio [OR]>1) for the six risk variants tested, with three replicating at p≤0.05 (7p15.3, p=1.4x10-7; 17p11.2, p=0.05; 22q13.1, p=0.02). For three of the six regions, we observed better markers of risk in African-Americans that were correlated with the index SNP in Europeans (7p15.3, rs56333627, p=1.5x10-5, r2=0.89; 17p11.2, rs34562254, p=2.9x10-3, r2=0.90; 22q13.1, rs2092410, p=1.1x10-4 r2=.71). The missense variant identified in the 17p11.2 region (rs34562254, Pro251Leu) is located in TNFRSF13B, which encodes the protein TACI, a B cell surface receptor which plays a role in B cell maturation, apoptosis and antibody production by inducing activation of transcription factors including NFAT and NFκβ. In addition, there is evidence suggesting that TACI is involved in MM pathogenesis.Our results demonstrate that many of the risk loci for MM found in European ancestry populations are also risk loci in men and women of African ancestry and that by fine-mapping, we are able to identify variants that better capture risk in populations of African ancestry. DisclosuresTerebelo:Celgene Corp: Membership on an entity’s Board of Directors or advisory committees. Lonial:Millennium: The Takeda Oncology Company: Consultancy, Research Funding; Celgene: Consultancy, Research Funding; Novartis: Consultancy, Research Funding; Bristol-Myers Squibb: Consultancy, Research Funding; Onyx Pharmaceuticals: Consultancy, Research Funding.

TGFβ receptor 1: an immune susceptibility gene in HPV-associated cancer.

Only a minority of those exposed to human papillomavirus (HPV) develop HPV-related cervical and oropharyngeal cancer. Because host immunity affects infection and progression to cancer, we tested the hypothesis that genetic variation in immune-related genes is a determinant of susceptibility to oropharyngeal cancer and other HPV-associated cancers by performing a multitier integrative computational analysis with oropharyngeal cancer data from a head and neck cancer genome-wide association study (GWAS). Independent analyses, including single-gene, gene-interconnectivity, protein-protein interaction, gene expression, and pathway analysis, identified immune genes and pathways significantly associated with oropharyngeal cancer. TGFβR1, which intersected all tiers of analysis and thus selected for validation, replicated significantly in the head and neck cancer GWAS limited to HPV-seropositive cases and an independent cervical cancer GWAS. The TGFβR1 containing p38-MAPK pathway was significantly associated with oropharyngeal cancer and cervical cancer, and TGFβR1 was overexpressed in oropharyngeal cancer, cervical cancer, and HPV(+) head and neck cancer tumors. These concordant analyses implicate TGFβR1 signaling as a process dysregulated across HPV-related cancers. This study demonstrates that genetic variation in immune-related genes is associated with susceptibility to oropharyngeal cancer and implicates TGFβR1/TGFβ signaling in the development of both oropharyngeal cancer and cervical cancer. Better understanding of the immunogenetic basis of susceptibility to HPV-associated cancers may provide insight into host/virus interactions and immune processes dysregulated in the minority of HPV-exposed individuals who progress to cancer.

GPA: a statistical approach to prioritizing GWAS results by integrating pleiotropy and annotation.

Results from Genome-Wide Association Studies (GWAS) have shown that complex diseases are often affected by many genetic variants with small or moderate effects. Identifications of these risk variants remain a very challenging problem. There is a need to develop more powerful statistical methods to leverage available information to improve upon traditional approaches that focus on a single GWAS dataset without incorporating additional data. In this paper, we propose a novel statistical approach, GPA (Genetic analysis incorporating Pleiotropy and Annotation), to increase statistical power to identify risk variants through joint analysis of multiple GWAS data sets and annotation information because: (1) accumulating evidence suggests that different complex diseases share common risk bases, i.e., pleiotropy; and (2) functionally annotated variants have been consistently demonstrated to be enriched among GWAS hits. GPA can integrate multiple GWAS datasets and functional annotations to seek association signals, and it can also perform hypothesis testing to test the presence of pleiotropy and enrichment of functional annotation. Statistical inference of the model parameters and SNP ranking is achieved through an EM algorithm that can handle genome-wide markers efficiently. When we applied GPA to jointly analyze five psychiatric disorders with annotation information, not only did GPA identify many weak signals missed by the traditional single phenotype analysis, but it also revealed relationships in the genetic architecture of these disorders. Using our hypothesis testing framework, statistically significant pleiotropic effects were detected among these psychiatric disorders, and the markers annotated in the central nervous system genes and eQTLs from the Genotype-Tissue Expression (GTEx) database were significantly enriched. We also applied GPA to a bladder cancer GWAS data set with the ENCODE DNase-seq data from 125 cell lines. GPA was able to detect cell lines that are biologically more relevant to bladder cancer. The R implementation of GPA is currently available at http://dongjunchung.github.io/GPA/.

Abstract B14: Population genetics analysis of prostate cancer GWAS SNPs to evaluate West African genetic ancestry as a risk factor

Abstract Background: Prostate cancer (Pca) is the most common cancer among men in the U.S., and African American (AA) men have higher incidence compared to European American men. Environmental and sociocultural factors may play a role for increasing Pca risk among AA men, but the causal relationship of any environmental and social/cultural factors with Pca risk is not well demonstrated. On the other hand, there is strong evidence for genetic factors that explain increased incidence in AAs. Admixture mapping identified loci at 8q24 region showing association between African genetic ancestry and Pca risk, and a genome-wide association study (GWAS) in AAs identified a novel locus at 17q21 associated with Pca. However, replication of variants identified in GWAS among European and Asian populations has been difficult in AAs. Here, we performed population genetics analyses of Pca GWAS identified single nucleotide polymorphisms (SNPs) to evaluate whether these SNPs are informative of genetic ancestry and whether West African genetic Ancestry (WAA) is a Pca risk factor. Samples and Methods: We accessed National Human Genome Research Institute GWAS catalog in December, 2012, and there were 80 SNPs reported in 17 studies. We downloaded genotype data of 76 SNPs in 1000 Genomes (1KG) Project Phase I May 2011 and 62 SNPs in Human Genome Diversity Project (HGDP) from SPSmart. We performed principal component analysis using Golden Helix SVS 7. STRUCTURE was also used to estimate genetic ancestry and examine population clustering patterns. Replication study subjects, 470 AA men from Washington, DC (220 Pca cases and 250 controls) were genotyped as a part of GWAS in AAs with Illumina Infinium 1M. Logistic regression analysis was performed adjusting for age and 5 principal components (PCs). Results: Mean heterozygosity of 76 Pca GWAS SNPs in the 1KG Project dataset was lower in African populations than in European populations, possibly due to ascertainment bias. The majority of Pca GWAS SNPs had FST less than 0.20, but rs9623117 on chromosome 22 was highly differentiated (FST=0.63). In the 1KG Project dataset, PC1 separated African populations from non-African populations, and PC2 reflects variation in non-African populations separating East Asian from European populations. AA (ASW) individuals fell between the African and European clusters. Fewer Pca GWAS SNPs were genotyped in HGDP dataset, but they successfully captured global human genome variation forming African and non-African population clusters. In the STRUCTURE analysis with the 1KG project dataset, K=2 separated African from non-African populations, and K=3 separated European from East Asian populations. WAA estimated using 76 Pca GWAS SNPs was significantly correlated with WAA estimated using ancestry informative markers. Only four Pca GWAS SNPs were replicated and three of them were independently associated (P<0.05) in our replication samples. The number of risk alleles were positively associated with Pca risk (PTREND=0.002), and AAs carrying ≥5 risk alleles had 3.5 times greater odds of Pca (95% C.I.; 1.36-9.17; P=0.009) compared to AAs with ≤ 2 risk alleles. Risk allele frequencies of the three independently associated SNPs were higher in West Africans than in Europeans and was positively correlated with WAA estimates (PANOVA<0.001). Conclusion: Results of our analyses provide evidence suggesting that WAA is a risk factor for Pca. Population genetics analyses show that Pca GWAS SNPs can successfully capture global human genome variation and are informative of genetic ancestry. AAs with high WAA are likely to carry more Pca risk alleles than AAs with low WAA. Citation Format: Ken Batai, Ebony Shah, Rick A. Kittles. Population genetics analysis of prostate cancer GWAS SNPs to evaluate West African genetic ancestry as a risk factor. [abstract]. In: Proceedings of the Sixth AACR Conference: The Science of Cancer Health Disparities; Dec 6–9, 2013; Atlanta, GA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2014;23(11 Suppl):Abstract nr B14. doi:10.1158/1538-7755.DISP13-B14

Abstract B11: Replication of previously identified breast cancer susceptibility loci in a breast cancer case-control study on women of African ancestry

Abstract To date, more than 100 single nucleotide polymorphisms (SNPs) have been found to be associated with breast cancer susceptibility in large genome-wide association studies (GWAS) conducted predominantly in women of European and Asian ancestries. To identify breast cancer susceptibility alleles in populations of African ancestry, we performed a GWAS in 3,686 subjects from Nigeria, Barbados and the United States (Baltimore, Chicago, Detroit, Philadelphia and the Southern Community Cohort Study), using the Illumina HumanOmni2.5 array. Using stringent quality control criteria, a total of 1,657 cases (777 with known estrogen receptor [ER] status) and 2,029 controls were successfully genotyped for 2,116,365 SNPs. Subsequently, imputation was conducted using reference panels from The 1000 Genomes Project and logistic regression models controlling age and global ancestry were applied to examine the association of 78 known breast cancer GWAS index SNPs and 44 iCOGS (Illumina iSelect genotyping array for Collaborative Oncological Gene-Environment Study) SNPs with breast cancer risk in our study population. Only 4 SNPs were statistically significant (unadjusted P<0.01) associated with breast cancer in women of African descent: rs3112612 (16q12.1, P=0.0021), rs3817198 (11p15.5, P=0.0049), rs10069690 (5p15.33, P=0.0056), and rs1978503 (18q21.2, P=0.0066). Additional five SNPs were found to be associated with breast cancer at the unadjusted P<0.05 significance level: rs458685 (12q21.3), rs361147 (4q31.3), rs4322600 (14q31.3), rs616488 (1p36.22), and rs17356907 (12q22). We also observed positive associations for ER-negative breast cancer: rs3817198 (11p15.5, P=0.0029), rs6762644 (3p26.1, P=0.0036), rs10069690 (5p15.33, P=0.0048), and rs3803662 (16q12.1, P=0.0050); and ER-positive breast cancer: rs12355688 (10q22.3, P=0.011), rs1978503 (18q21.2, P=0.020), rs3112612 (16q12.1, P=0.032), and rs4784227 (16q12.1, P=0.038). In conclusion, our observations highlight the necessity of validating previous breast cancer GWAS findings across diverse populations. Different environmental factors, allele frequencies and linkage disequilibrium patterns may influence the genetic risk profiles in different populations. Larger genetic studies in women of African ancestry hold promise to reveal additional risk variants for developing polygenic risk models for breast cancer susceptibility in this population. Citation Format: Yonglan Zheng, Dezheng Huo, Temidayo O. Ogundiran, Adeyinka G. Falusi, Oladosu Ojengbede, Clement Adebamowo, William J. Blot, Wei Zheng, Qiuyin Cai, Lisa B. Signorello, Katherine L. Nathanson, Susan M. Domchek, Timothy R. Rebbeck, Michael S. Simon, Anselm J.M. Hennis, Barbara Nemesure, Suh-Yuh Wu, Maria Cristina Leske, Stefan Ambs, Abayomi Odetunde, Imaria Anetor, Stella Akinleye, Qun Niu, Jing Zhang, Anna Pluzhnikov, Anuar Konkashbaev, Lin Chen, Eric R. Gamazon, Younghee Lee, Nancy J. Cox, Olufunmilayo O. Olopade. Replication of previously identified breast cancer susceptibility loci in a breast cancer case-control study on women of African ancestry. [abstract]. In: Proceedings of the Sixth AACR Conference: The Science of Cancer Health Disparities; Dec 6–9, 2013; Atlanta, GA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2014;23(11 Suppl):Abstract nr B11. doi:10.1158/1538-7755.DISP13-B11

Abstract C80: The interaction between menopausal hormone replacement therapy and the 2q35 (rs13387042) breast cancer risk variant explains the effect modification by genetic ancestry among Hispanic women

Abstract Background: Breast cancer genome-wide association studies published to date are mostly based on samples of European or Asian origin. There have been few published studies in African American women and none in Latinas. In a previous study we reported statistically significant interactions between genetic ancestry and genotypes for 3 out of 10 GWAS-discovered breast cancer risk variants, in 4,694 Hispanics from the U.S. and Mexico who participated in one of three population-based case-control studies of breast cancer (Single nucleotide polymorphisms (SNPs): rs13387042-2q35 region, rs17157903-RELN gene and rs7696175-TLR1 gene). For these three SNPs, the risk allele showed stronger associations among women of higher Indigenous American (IA) ancestry. In the present study, we investigated multiple know reproductive and lifestyle related breast cancer risk factors to evaluate if the previously observed heterogeneity could be due to the correlation between those factors and ancestry. Methods: Genetic ancestry was previously estimated with a set of 104 ancestry-informative markers (AIMs) selected to discriminate between IA and European ancestry. We used stratified logistic regression analyses to compare the associations with different genetic variants by proportion of IA ancestry (low, intermediate, high) and questionnaire based information on menopausal status, family history of breast cancer, number of full-term pregnancies, age at first full-term pregnancy, breast feeding, body mass index, height, alcohol intake, smoking, menopausal hormone replacement therapy, age at menarche, diabetes and use of oral contraceptives. Results: The three investigated SNPs showed suggestive interactions with the reproductive/lifestyle related breast cancer risk factors. For SNP rs13387042 the interaction with menopausal hormone replacement therapy explained the previously observed interaction with ancestry. This SNP was associated with breast cancer risk only among postmenopausal women who never used hormone replacement therapy, and this association was apparent in all ancestry categories [AA genotype vs. GG genotype OR 3.01, 95%CI 1.06-8.52 in the low Indigenous American ancestry category; OR 1.61, 95%CI 1.04-2.51 in the intermediate ancestry category; OR 3.37, 95%CI 1.34-8.46 in the high ancestry category, p value interaction 0.96]. For the other two SNPs, we saw interactions with non-genetic risk factors, but those did not explain the previously reported heterogeneity by ancestry. Conclusions: Our results showed that the correlation between genetic ancestry and a non-genetic factor might explain the previously reported interaction between a breast cancer risk variant and genetic ancestry in Latinas for at least one out of three SNPs. Future studies should investigate the possibility that for some variants that have been associated with breast cancer risk, the effect heterogeneity by ancestry might be due to differences between populations in the genetic architecture of the regions were the variants are located. Citation Format: Laura Fejerman, Mariana C. Stern, Elad Ziv, Esther M. John, Gabriela Torres-Mejia, Lisa M. Hines, Roger Wolff, Kathy B. Baumgartner, Anna Giuliano, Martha L. Slattery. The interaction between menopausal hormone replacement therapy and the 2q35 (rs13387042) breast cancer risk variant explains the effect modification by genetic ancestry among Hispanic women. [abstract]. In: Proceedings of the Sixth AACR Conference: The Science of Cancer Health Disparities; Dec 6–9, 2013; Atlanta, GA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2014;23(11 Suppl):Abstract nr C80. doi:10.1158/1538-7755.DISP13-C80

The 19q12 bladder cancer GWAS signal: association with cyclin E function and aggressive disease.

A genome-wide association study (GWAS) of bladder cancer identified a genetic marker rs8102137 within the 19q12 region as a novel susceptibility variant. This marker is located upstream of the CCNE1 gene, which encodes cyclin E, a cell-cycle protein. We performed genetic fine-mapping analysis of the CCNE1 region using data from two bladder cancer GWAS (5,942 cases and 10,857 controls). We found that the original GWAS marker rs8102137 represents a group of 47 linked SNPs (with r(2) ≥ 0.7) associated with increased bladder cancer risk. From this group, we selected a functional promoter variant rs7257330, which showed strong allele-specific binding of nuclear proteins in several cell lines. In both GWASs, rs7257330 was associated only with aggressive bladder cancer, with a combined per-allele OR = 1.18 [95% confidence interval (CI), 1.09-1.27, P = 4.67 × 10(-5)] versus OR = 1.01 (95% CI, 0.93-1.10, P = 0.79) for nonaggressive disease, with P = 0.0015 for case-only analysis. Cyclin E protein expression analyzed in 265 bladder tumors was increased in aggressive tumors (P = 0.013) and, independently, with each rs7257330-A risk allele (P(trend) = 0.024). Overexpression of recombinant cyclin E in cell lines caused significant acceleration of cell cycle. In conclusion, we defined the 19q12 signal as the first GWAS signal specific for aggressive bladder cancer. Molecular mechanisms of this genetic association may be related to cyclin E overexpression and alteration of cell cycle in carriers of CCNE1 risk variants. In combination with established bladder cancer risk factors and other somatic and germline genetic markers, the CCNE1 variants could be useful for inclusion into bladder cancer risk prediction models.

SNP characteristics predict replication success in association studies.

Successful independent replication is the most direct approach for distinguishing real genotype-disease associations from false discoveries in genome-wide association studies (GWAS). Selecting SNPs for replication has been primarily based on P values from the discovery stage, although additional characteristics of SNPs may be used to improve replication success. We used disease-associated SNPs from more than 2,000 published GWASs to identify predictors of SNP reproducibility. SNP reproducibility was defined as a proportion of successful replications among all replication attempts. The study reporting association for the first time was considered to be discovery and all consequent studies targeting the same phenotype replications. We found that -Log(P), where P is a P value from the discovery study, is the strongest predictor of the SNP reproducibility. Other significant predictors include type of the SNP (e.g., missense vs intronic SNPs) and minor allele frequency. Features of the genes linked to the disease-associated SNP also predict SNP reproducibility. Based on empirically defined rules, we developed a reproducibility score (RS) to predict SNP reproducibility independently of -Log(P). We used data from two lung cancer GWAS studies as well as recently reported disease-associated SNPs to validate RS. Minus Log(P) outperforms RS when the very top SNPs are selected, while RS works better with relaxed selection criteria. In conclusion, we propose an empirical model to predict SNP reproducibility, which can be used to select SNPs for validation and prioritization.

Abstract 4175: TGFBR1 and other immune-related genes modify susceptibility to HPV-associated head and neck cancer

Abstract OBJECTIVES: Human papilloma virus (HPV)-related cancer of the oropharynx (OPC) makes up an increasing proportion of head and neck cancers (HNC). Since viral infection and progression to cancer are both strongly influenced by the immune system, the host immune response is one potential determinant of susceptibility to HPV-related cancers. The overall goal of this project is to identify immune-genetic variants associated with an increased risk of developing virally-mediated HNC. STUDY DESIGN: Data from a HNC Genome Wide Association Study (GWAS) performed by the International Agency for Research on Cancer (IARC) and the Centre National de Gestion (CNG) was evaluated for associations between genetic variants in immune genes and the risk of developing OPC. METHODS: An association analysis using multivariate logistic regression was performed on the HNC GWAS data limited to OPC. We analyzed all SNPs intersecting immune genes based on Gene Ontology using a multi tier computational approach which included single gene analysis, gene-gene interaction (GGI) analysis, pathway analysis and protein-protein interaction (PPI) analysis. Results were validated in 2 in-silico replications: the HNC GWAS limited to individuals with serologic evidence of HPV infection, and a cervical cancer GWAS. RESULTS: The OPC GWAS found 17 immune SNPs to be associated with OPC at a significance level of p<10-4 -10-7. GGI analysis demonstrated a high degree of significant functional connection amongst the top ranked immune genes, and PPI analysis identified three significant functional clusters of immune genes related to TGFBR1 signaling, innate immunity, or Th1/Th2 immune balance. Pathway analysis identified multiple immune pathways significantly enriched in the OPC GWAS including 4 cell death, 5 innate immunity, 4 NFKB, 2 T-cell activation, and 4 TGFB signaling pathways. While several immune genes were found to overlap multiple modes of analysis, TGFBR1 intersected all tiers of analysis and was thus selected for validation. Replication in the GWAS dataset limited to serologically HPV+ HNC cases validated 14 out of the 15 significant genes including TGFBR1. Replication in a cervical cancer GWAS demonstrated TGFBR1 to be significantly associated with cancer risk and an identical TGFBR1 containing pathway to be highly significant in both cervical cancer and OPC GWAS. CONCLUSIONS: The high degree of concordance among multiple independent tiers of analysis, strongly implicates variations in immune-related genes as modulators of susceptibility to HPV-mediated HNC and highlights TGFB signaling as a process potentially dysregulated across HPV-related cancers. Thus data highlights the role of host immunogenetics as a modulator of virally-related cancer risk and the utility of multi-component computational strategies in testing biologically based hypothesis about mechanisms of disease susceptibility. Citation Format: Chaya Levovitz, Dan Chen, Emma Ivansson, Ulf Gyllensten, James McKay, Paul Brennan, Paolo Boffetta, Andrew Sikora. TGFBR1 and other immune-related genes modify susceptibility to HPV-associated head and neck cancer. [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 4175. doi:10.1158/1538-7445.AM2014-4175

Abstract 3285: Functional analysis of the 9p22 locus implicates the transcriptional regulation of BNC2 as a mechanism in ovarian cancer predisposition

Abstract An ovarian cancer Genome Wide Association Study (GWAS) identified 9p22.2 as a novel susceptibility locus with the most statistically significant single nucleotide polymorphisms (SNPs) located in an intergenic region near the Basonuclin 2 (BNC2) gene, which codes for a putative transcription regulator containing three pairs of zinc finger (ZF) domains. The minor alleles are protective in terms of ovarian cancer susceptibility. However, the molecular mechanisms by which these SNPs modify susceptibility remain largely unknown. The significant SNPs in the 9p22.2 locus lie in non-coding regions and therefore are hypothesized to affect the activity of regulatory elements and modify the expression of a target gene(s). In order to test this hypothesis we conducted fine mapping for the locus which delimited a ∼74kb region containing SNPs with a p-value less than 10-8. FAIRE-Seq and ChIP-Seq experiments for histone markers conducted in immortalized ovarian surface epithelial cells (IOSE) and fallopian tube epithelial cells (IFTE) were used to prioritize functional SNP candidates which overlap with regulatory elements. Luciferase assays tested the ability of these regulatory elements to activate transcription. Chromosome conformation capture (3C) experiments demonstrate a physical interaction between the BNC2 promoter and candidate regulatory elements containing risk-associated SNPs. Methylation Quantification at Trait Loci (mQTL) revealed an association between decreased methylation at the BNC2 promoter and the protective minor alleles. Expression analysis shows decreased expression of BNC2 in cancer versus normal tissue implicating tumor suppressor function of BNC2. Therefore the protective minor allele likely increases expression of BNC2 which in turn contains tumor suppressing properties that decrease ovarian cancer risk. BNC2 protein is in complex with transcriptional regulatory proteins, in particular the NURD complex components. Moreover, BNC2 acts as a transcriptional repressor in in vitro transfection assays indicating that it functions in transcriptional repression. We then used protein binding microarrays and CHIP-Seq experiments to identify its putative DNA binding motifs and its downstream target genes. Analysis of this data suggests that BNC2 functions in a regulatory transcription network that impacts on genes implicated in ovarian cancer with enrichment for genes in the TGF-beta response pathway. Citation Format: Melissa A. Buckley, Howard C. Shen, Gustavo A. Mendoza-Fandino, Nicholas T. Woods, Anxhela Gjyshi, Juliet French, Kate Lawrenson, Honglin Song, Jonathan Tyrer, Renato S. Carvalho, Alexandra Valle, Ann Chen, Sean Yoder, Gregory Bloom, Ya-Yu Tsai, Ally Yang, Timothy R. Hughes, Xiaotao Qu, Mine Cicek, Melissa Larson, Ellen Goode, Brooke Fridley, Susan Ramus, Georgia Chenevix-Trench, Paul Pharoah, Thomas A. Sellers, Simon Gayther, Alvaro N.A. Monteiro, Ovarian Cancer Association Consortium. Functional analysis of the 9p22 locus implicates the transcriptional regulation of BNC2 as a mechanism in ovarian cancer predisposition. [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 3285. doi:10.1158/1538-7445.AM2014-3285

Abstract 944: Translational implications of the 19q12 bladder cancer GWAS signal for aggressive bladder cancer

Abstract Background: A genome-wide association study (GWAS) of bladder cancer identified a genetic marker rs8102137 within the 19q12 region as a novel susceptibility variant. This marker is located upstream of the CCNE1 gene, which encodes cyclin E, a cell cycle protein. Methods: Genetic fine mapping analysis of the CCNE1 region was performed using data from two bladder cancer GWAS, which included a total of 5,942 cases and 10,857 controls. Cyclin E protein expression was evaluated by immunohistochemistry analysis in 265 bladder tumors in relation to CCNE1 genetic variants and tumor characteristics. Functional effects of cyclin E over-expression were evaluated with cell cycle assays. Findings: The original GWAS marker rs8102137 represents a group of 47 linked SNPs (with r2≥0.7) associated with increased bladder cancer risk. From this group we selected a functional promoter variant rs7257330, which showed strong allele-specific binding of nuclear proteins in several cell lines. In both GWAS, rs7257330 was associated only with aggressive bladder cancer (muscle-invasive or high-grade non-muscle-invasive cancer), with a per-allele odds ratio (OR) =1.18 (95%CI=1.09-1.27, p=4.67×10-5). Cyclin E protein expression was increased in aggressive bladder tumors (p=0.013) and, independently, with each rs7257330-A risk allele (ptrend=0.024). Over-expression of recombinant cyclin E in cell lines caused significant acceleration of cell cycle. Conclusions: Molecular mechanisms linking the CCNE1 GWAS signal and aggressive bladder cancer risk are related to cyclin E over-expression and cell cycle regulation. In combination with other genetic and clinical markers, CCNE1 genetic variants may be translationally useful for early prediction of aggressive bladder cancer. Citation Format: Yi-Ping Fu, Indu Kohaar, Lee Moore, Petra Lenz, Jonine D. Figueroa, Wei Tang, Patricia Porter-Gill, Stephen Chanock, Stephen M. Hewitt, Debra T. Silverman, Nathaniel Rothman, NCI-GWAS Bladder Cancer Consortium, Ludmila Prokunina-Olsson. Translational implications of the 19q12 bladder cancer GWAS signal for aggressive bladder cancer. [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 944. doi:10.1158/1538-7445.AM2014-944

Abstract 3269: QTLs in breast tumor and breast normal adjacent FFPE specimens from the Nurses’ Health Study

Abstract Rationale: Genome-wide association studies (GWAS) of breast cancer have identified 71 risk alleles, majority of which are located in intergenic or intronic regions. We used a network medicine approach to associate breast cancer risk single nucleotide polymorphisms (SNPs) with transcript abundance in formalin fixed paraffin embedded (FFPE) breast tissue. Identification of expression quantitative loci (eQTLs) may help to better understand the regulatory mechanisms by which these risk variants influence breast cancer susceptibility. Method: We identified invasive postmenopausal breast cancer cases in the Nurses’ Health Study (NHS) diagnosed from 1990-2004 with GWAS data and sufficient RNA for expression profiling in breast tumor and normal adjacent breast tissue. RNA was extracted using the Qiagen AllPrep RNA isolation kit, amplified using the NuGen Ovation FFPE WTA kit, and profiled using the Affymetrix Human Transcriptome Array (HTA 3.0v1). The HTA includes 6,892,960 features for measuring gene expression, alternative splicing, coding SNPs, and noncoding transcripts. After filtering and removing probes with a low dynamic range in our samples, we included 25,000 gene expression probes in our QTL analyses. We mapped these probes to genes using hg19. Results: Analyses were conducted separately for 376 breast tumor and 274 normal adjacent samples. We conducted quality control analyses, corrected for assay plate-to-plate variation and included patient's age at diagnosis and year of diagnosis as covariates in the multivariate linear regression model. We identified 11 trans eQTLs in normal adjacent breast tissue, 11 trans eQTLs in estrogen receptor (ER)+ breast tumor samples, and 12 trans eQTLs in ER- breast tumor samples (permutation adjusted p-value<0.05). We also developed a new method, functional quantitative trait loci (fQTL) analysis, to gain additional pathway insight into genetic associations important in breast cancer. In the fQTL analysis we tested for the association between SNPs and the expression of gene functional classes and pathways, evaluating the hypothesis that SNPs may also be associated with regulation of processes in addition to individual genes. Using a cutoff of false discovery rate <10%, we identified 2 SNPs associated with 2 Gene Ontology Molecular Functions in normal adjacent breast tissue, 1 SNP associated with 5 Molecular Functions in ER- tumor samples but no significant SNP associations in ER+ tumor samples. Integrated eQTL and fQTL analyses and variant annotation are ongoing. Conclusion: In summary, our results provide functional insights on the underlying biology of loci identified in breast cancer GWAS in the specimen type that is most impactful in translation to clinical practice. Identification of gene transcripts that can be measured in FFPE tissue and are associated with breast cancer risk loci is critical in understanding the mechanism by which these variants affect risk and mediate disease processes. Citation Format: Alejandro Quiroz-Zarate, Benjamin J. Harshfield, Rong Hu, Nick Knoblauch, Andrew H. Beck, Vincent Carey, Susan E. Hankinson, Rulla M. Tamimi, David J. Hunter, John Quackenbush, Aditi Hazra. QTLs in breast tumor and breast normal adjacent FFPE specimens from the Nurses’ Health Study. [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 3269. doi:10.1158/1538-7445.AM2014-3269