Hepatocyte gene expression and DNA methylation as ancestry-dependent mechanisms in African Americans

African Americans (AAs) are an admixed population with widely varying proportion of West African ancestry (WAA). Here we report the correlation of WAA to gene expression and DNA methylation in AA-derived hepatocytes, a cell type important in disease and drug response. We perform mediation analysis to test whether methylation is a mediator of the effect of ancestry on expression. GTEx samples and a second cohort are used as validation. One hundred and thirty-one genes are associated with WAA (FDR < 0.10), 28 of which replicate and represent 220 GWAS phenotypes. Among PharmGKB pharmacogenes, VDR, PTGIS, ALDH1A1, CYP2C19, and P2RY1 nominally associate with WAA (p < 0.05). We find 1037 WAA-associated, differentially methylated regions (FDR < 0.05), with hypomethylated genes enriched in drug-response pathways. In conclusion, WAA contributes to variability in hepatocyte expression and DNA methylation with identified genes previously implicated for diseases disproportionately affecting AAs, including cardiovascular (PTGIS, PLAT) and renal (APOL1) disease, and drug response (CYP2C19).

Integrated analysis of genetic variation and gene expression reveals novel variant for increased warfarin dose requirement in African Americans

Essentials Genetic variants controlling gene regulation have not been explored in pharmacogenomics. We tested liver expression quantitative trait loci for association with warfarin dose response. A novel predictor for increased warfarin dose response in African Americans was identified. Precision medicine must take into account population-specific variation in gene regulation.

SUMMARY

Background Warfarin is commonly used to control and prevent thromboembolic disorders. However, because of warfarin's complex dose-requirement relationship, safe and effective use is challenging. Pharmacogenomics-guided warfarin dosing algorithms that include the well-established VKORC1 and CYP2C9 polymorphisms explain only a small proportion of inter-individual variability in African Americans (AAs). Objectives We aimed to assess whether transcriptomic analyses could be used to identify regulatory variants associated with warfarin dose response in AAs. Patients/Methods We identified a total of 56 expression quantitative trait loci (eQTLs) for CYP2C9, VKORC1 and CALU derived from human livers and evaluated their association with warfarin dose response in two independent AA warfarin patient cohorts. Results We found that rs4889606, a strong cis-eQTL for VKORC1 (log10 Bayes Factor = 12.02), is significantly associated with increased warfarin daily dose requirement (β = 1.1; 95% confidence interval [CI] 0.46 to 1.8) in the discovery cohort (n = 305) and in the replication cohort (β = 1.04; 95% CI 0.33 -1.7; n = 141) after conditioning on relevant covariates and the VKORC1 -1639G>A (rs9923231) variant. Inclusion of rs4889606 genotypes, along with CYP2C9 alleles, rs9923231 genotypes and clinical variables, explained 31% of the inter-patient variability in warfarin dose requirement. We demonstrate different linkage disequilibrium patterns in the region encompassing rs4889606 and rs9923231 between AAs and European Americans, which may explain the increased dose requirement found in AAs. Conclusion Our approach of interrogating eQTLs identified in liver has revealed a novel predictor of warfarin dose response in AAs. Our work highlights the utility of leveraging information from regulatory variants mapped in the liver to uncover novel variants associated with drug response and the importance of population-specific research.

Abstract P1-05-11: Comprehensive comparison of breast cancer molecular portraits by African and European ancestry in the cancer genome atlas

Background: African American breast cancer patients have worse survival rates than European American patients. Although racial differences in the distribution of breast cancer intrinsic subtype are known, it is unclear if there are other inherent genomic differences contributing to this racial outcome disparity.

Methods: We defined patient race based on genomic ancestry and compared multiple molecular features of breast cancer between 154 black and 776 white patients in The Cancer Genome Atlas (TCGA). We examined the contribution of these molecular features to survival outcomes using Cox proportional hazards models. We also estimated the heritability of breast cancer subtypes using a mixed effect model.

Results: Compared to whites, black patients had higher odds of basal-like (odds ratio=3.80, p&lt;0.001) and HER2-enriched (odds ratio=2.22, p=0.027) breast cancers in reference to luminal A subtype. Beyond differences in relative frequency of intrinsic subtypes, black and white patients had distinct gene expression, protein expression, and somatic mutation landscapes. However, the majority of these molecular differences were eliminated after adjusting for subtype; in the subtype-adjusted models, we found 142 genes, 16 methylation probes, 4 copy number segments, 1 protein, and no somatic mutation were differentially expressed or present between black and white patients. Using the top 40 differentially expressed genes, we built a race-enriched gene signature, which had excellent capacity of distinguishing breast tumors from black versus white patients (c-index=0.852 in the validation dataset). We also estimated the heritability of breast cancer subtype (basal vs. non-basal) to be 0.436 (p=1.5x10-14) and showed that two genetic variants (rs1078806 in FGFR2, rs34084277 in BABAM1) were associated with intrinsic subtype and can partially explain racial differences in subtype frequencies.

Conclusion: On the molecular level, once intrinsic subtype frequency differences are accounted for, there are few genomic or proteomic differences observed between blacks and whites. More than 40% of breast cancer subtype frequency differences may be due to genetic ancestry. These results suggest that future studies are warranted to investigate genetic and non-genetic factors that contribute to the development and progression of breast cancer subtypes in order to reduce racial disparity.

Citation Format: Huo D, Hu H, Rhie SK, Gamazon ER, Cherniack AD, Liu J, Yoshimatsu TF, Pitt JJ, Hoadley KA, Troester M, Ru Y, Lichtenberg T, Sturtz LA, Shelley CS, Mills GB, Laird PW, Shriver CD, Perou CM, Olopade OI. Comprehensive comparison of breast cancer molecular portraits by African and European ancestry in the cancer genome atlas [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P1-05-11.

Genome-wide association study identifies pharmacogenomic loci linked with specific antihypertensive drug treatment and new-onset diabetes

We conducted a discovery genome-wide association study with expression quantitative trait loci (eQTL) annotation of new-onset diabetes (NOD) among European Americans, who were exposed to a calcium channel blocker-based strategy (CCB strategy) or a β-blocker-based strategy (β-blocker strategy) in the INternational VErapamil SR Trandolapril STudy. Replication of the top signal from the SNP*treatment interaction analysis was attempted in Hispanic and African Americans, and a joint meta-analysis was performed (total 334 NOD cases and 806 matched controls). PLEKHH2 rs11124945 at 2p21 interacted with antihypertensive exposure for NOD (meta-analysis p=5.3×10⁻⁸). rs11124945 G allele carriers had lower odds for NOD when exposed to the β-blocker strategy compared with the CCB strategy [OR=0.38 (0.24-0.60), p=4.0×10⁻⁵], while A/A homozygotes exposed to the β-blocker strategy had increased odds for NOD compared with the CCB strategy [OR=2.02 (1.39-2.92), p=2.0×10⁻⁴]. eQTL annotation of the 2p21 locus provides functional support for regulating gene expression.

A genome-wide sib-pair scan for quantitative language traits reveals linkage to chromosomes 10 and 13

Although there is considerable evidence that individual differences in language development are highly heritable, there have been few genome-wide scans to locate genes associated with the trait. Previous analyses of language impairment have yielded replicable evidence for linkage to regions on chromosomes 16q, 19q, 13q (within lab) and at 13q (between labs). Here we report the first linkage study to screen the continuum of language ability, from normal to disordered, as found in the general population. 383 children from 147 sib-ships (214 sib-pairs) were genotyped on the Illumina(®) Linkage IVb Marker Panel using three composite language-related phenotypes and a measure of phonological memory (PM). Two regions (10q23.33; 13q33.3) yielded genome-wide significant peaks for linkage with PM. A peak suggestive of linkage was also found at 17q12 for the overall language composite. This study presents two novel genetic loci for the study of language development and disorders, but fails to replicate findings by previous groups. Possible reasons for this are discussed.

Ethnicity-specific pharmacogenetics: the case of warfarin in African Americans

Using a derivation cohort (N=349), we developed the first warfarin dosing algorithm that includes recently discovered polymorphisms in VKORC1 and CYP2C9 associated with warfarin dose requirement in African Americans (AAs). We tested our novel algorithm in an independent cohort of 129 AAs and compared the dose prediction to the International Warfarin Pharmacogenetics Consortium (IWPC) dosing algorithms. Our algorithm explains more of the phenotypic variation (R(2)=0.27) than the IWPC pharmacogenomics (R(2)=0.15) or clinical (R(2)=0.16) algorithms. Among high-dose patients, our algorithm predicted a higher proportion of patients within 20% of stable warfarin dose (45% vs 29% and 2% in the IWPC pharmacogenomics and clinical algorithms, respectively). In contrast to our novel algorithm, a significant inverse correlation between predicted dose and percent West African ancestry was observed for the IWPC pharmacogenomics algorithm among patients requiring ⩾60 mg per week (β=-2.04, P=0.02).

Genome-wide discovery of genetic variants affecting tamoxifen sensitivity and their clinical and functional validation

BACKGROUND

Beyond estrogen receptor (ER), there are no validated predictors for tamoxifen (TAM) efficacy and toxicity. We utilized a genome-wide cell-based model to comprehensively evaluate genetic variants for their contribution to cellular sensitivity to TAM.

DESIGN

Our discovery model incorporates multidimensional datasets, including genome-wide genotype, gene expression, and endoxifen-induced cellular growth inhibition in the International HapMap lymphoblastoid cell lines (LCLs). Genome-wide findings were further evaluated in NCI60 cancer cell lines. Gene knock-down experiments were performed in four breast cancer cell lines. Genetic variants identified in the cell-based model were examined in 245 Caucasian breast cancer patients who underwent TAM treatment.

RESULTS

We identified seven novel single-nucleotide polymorphisms (SNPs) associated with endoxifen sensitivity through the expression of 10 genes using the genome-wide integrative analysis. All 10 genes identified in LCLs were associated with TAM sensitivity in NCI60 cancer cell lines, including USP7. USP7 knock-down resulted in increasing resistance to TAM in four breast cancer cell lines tested, which is consistent with the finding in LCLs and in the NCI60 cells. Furthermore, we identified SNPs that were associated with TAM-induced toxicities in breast cancer patients, after adjusting for other clinical factors.

CONCLUSION

Our work demonstrates the utility of a cell-based model in genome-wide identification of pharmacogenomic markers.

Enrichment of cis-regulatory gene expression SNPs and methylation quantitative trait loci among bipolar disorder susceptibility variants

We conducted a systematic study of top susceptibility variants from a genome-wide association (GWA) study of bipolar disorder to gain insight into the functional consequences of genetic variation influencing disease risk. We report here the results of experiments to explore the effects of these susceptibility variants on DNA methylation and mRNA expression in human cerebellum samples. Among the top susceptibility variants, we identified an enrichment of cis regulatory loci on mRNA expression (eQTLs), and a significant excess of quantitative trait loci for DNA CpG methylation, hereafter referred to as methylation quantitative trait loci (mQTLs). Bipolar disorder susceptibility variants that cis regulate both cerebellar expression and methylation of the same gene are a very small proportion of bipolar disorder susceptibility variants. This finding suggests that mQTLs and eQTLs provide orthogonal ways of functionally annotating genetic variation within the context of studies of pathophysiology in brain. No lymphocyte mQTL enrichment was found, suggesting that mQTL enrichment was specific to the cerebellum, in contrast to eQTLs. Separately, we found that using mQTL information to restrict the number of single-nucleotide polymorphisms studied enhances our ability to detect a significant association. With this restriction a priori informed by the observed functional enrichment, we identified a significant association (rs12618769, P(bonferroni)<0.05) from two other GWA studies (TGen+GAIN; 2191 cases and 1434 controls) of bipolar disorder, which we replicated in an independent GWA study (WTCCC). Collectively, our findings highlight the importance of integrating functional annotation of genetic variants for gene expression and DNA methylation to advance the biological understanding of bipolar disorder.

Clinical translation of cell-based pharmacogenomic discovery

The use of cell-based models has emerged as a promising means to discover and validate pharmacologic phenotype-genotype relationships. The availability of large-scale genome studies in both human and model systems is now allowing us an unprecedented opportunity to understand how well cell-based models identify clinically relevant genetic variants associated with drug response and toxicity. Here we review these studies and the emerging translational information.

Genome-wide meta-analysis identifies variants associated with platinating agent susceptibility across populations

Platinating agents are used in the treatment of many cancers, yet they can induce toxicities and resistance that limit their utility. Using previously published and additional world population panels of diverse ancestry totaling 608 lymphoblastoid cell lines (LCLs), we performed meta-analyses of over 3 million SNPs for both carboplatin- and cisplatin-induced cytotoxicity. The most significant SNP in the carboplatin meta-analysis is located in an intron of NBAS (p = 5.1 × 10⁻⁷). The most significant SNP in the cisplatin meta-analysis is upstream of KRT16P2 (p = 5.8 × 10⁻⁷). We also show that cisplatin-susceptibility SNPs are enriched for carboplatin-susceptibility SNPs. Most of the variants that associate with platinum-induced cytotoxicity are polymorphic across multiple world populations; therefore, they could be tested in follow-up studies in diverse clinical populations. Seven genes previously implicated in platinating agent response, including BCL2, GSTM1, GSTT1, ERCC2, and ERCC6 were also implicated in our meta-analyses.

Genome-wide association and meta-analysis in populations from Starr County, Texas, and Mexico City identify type 2 diabetes susceptibility loci and enrichment for expression quantitative trait loci in top signals

AIMS/HYPOTHESIS

We conducted genome-wide association studies (GWASs) and expression quantitative trait loci (eQTL) analyses to identify and characterise risk loci for type 2 diabetes in Mexican-Americans from Starr County, TX, USA.

METHOD

Using 1.8 million directly interrogated and imputed genotypes in 837 unrelated type 2 diabetes cases and 436 normoglycaemic controls, we conducted Armitage trend tests. To improve power in this population with high disease rates, we also performed ordinal regression including an intermediate class with impaired fasting glucose and/or glucose tolerance. These analyses were followed by meta-analysis with a study of 967 type 2 diabetes cases and 343 normoglycaemic controls from Mexico City, Mexico.

RESULT

The top signals (unadjusted p value <1 × 10(-5)) included 49 single nucleotide polymorphisms (SNPs) in eight gene regions (PER3, PARD3B, EPHA4, TOMM7, PTPRD, HNT [also known as RREB1], LOC729993 and IL34) and six intergenic regions. Among these was a missense polymorphism (rs10462020; Gly639Val) in the clock gene PER3, a system recently implicated in diabetes. We also report a second signal (minimum p value 1.52 × 10(-6)) within PTPRD, independent of the previously implicated SNP, in a population of Han Chinese. Top meta-analysis signals included known regions HNF1A and KCNQ1. Annotation of top association signals in both studies revealed a marked excess of trans-acting eQTL in both adipose and muscle tissues.

CONCLUSIONS/INTERPRETATION

In the largest study of type 2 diabetes in Mexican populations to date, we identified modest associations of novel and previously reported SNPs. In addition, in our top signals we report significant excess of SNPs that predict transcript levels in muscle and adipose tissues.