Portal de Periódicos da CAPES

Allelic Heterogeneity at the CRP Locus Identified by Whole-Genome Sequencing in Multi-ancestry Cohorts

2019; Elsevier BV; Volume: 106; Issue: 1 Linguagem: Inglês

10.1016/j.ajhg.2019.12.002

1537-6605

Laura M. Raffield, Apoorva K. Iyengar, Biqi Wang, Sheila M. Gaynor, Cassandra N. Spracklen, Xue Zhong, Madeline H. Kowalski, Shabnam Salimi, Linda M. Polfus, Emelia J. Benjamin, Joshua C. Bis, Russell P. Bowler, Brian E. Cade, Won Jung Choi, Alejandro P. Comellas, Adolfo Correa, Pedro Cruz, HarshaVardhan Doddapaneni, Peter Durda, Stephanie M. Gogarten, Deepti Jain, Ryan W. Kim, B Král, Leslie A. Lange, Martin G. Larson, Cecelia Laurie, Jiwon Lee, Seonwook Lee, Joshua P. Lewis, Ginger Metcalf, Braxton D. Mitchell, Zeineen Momin, Donna M. Muzny, Nathan Pankratz, Cheol Joo Park, Stephen S. Rich, Jerome I. Rotter, Kathleen A. Ryan, Daekwan Seo, Russell P. Tracy, Karine A. Viaud‐Martinez, Lisa R. Yanek, Lue Ping Zhao, Xihong Lin, Bingshan Li, Yun Li, Josée Dupuis, Alexander P. Reiner, Karen L. Mohlke, Paul L. Auer,

Genetic Neurodegenerative Diseases

Whole-genome sequencing (WGS) can improve assessment of low-frequency and rare variants, particularly in non-European populations that have been underrepresented in existing genomic studies. The genetic determinants of C-reactive protein (CRP), a biomarker of chronic inflammation, have been extensively studied, with existing genome-wide association studies (GWASs) conducted in >200,000 individuals of European ancestry. In order to discover novel loci associated with CRP levels, we examined a multi-ancestry population (n = 23,279) with WGS (∼38× coverage) from the Trans-Omics for Precision Medicine (TOPMed) program. We found evidence for eight distinct associations at the CRP locus, including two variants that have not been identified previously (rs11265259 and rs181704186), both of which are non-coding and more common in individuals of African ancestry (∼10% and ∼1% minor allele frequency, respectively, and rare or monomorphic in 1000 Genomes populations of East Asian, South Asian, and European ancestry). We show that the minor (G) allele of rs181704186 is associated with lower CRP levels and decreased transcriptional activity and protein binding in vitro, providing a plausible molecular mechanism for this African ancestry-specific signal. The individuals homozygous for rs181704186-G have a mean CRP level of 0.23 mg/L, in contrast to individuals heterozygous for rs181704186 with mean CRP of 2.97 mg/L and major allele homozygotes with mean CRP of 4.11 mg/L. This study demonstrates the utility of WGS in multi-ethnic populations to drive discovery of complex trait associations of large effect and to identify functional alleles in noncoding regulatory regions. Whole-genome sequencing (WGS) can improve assessment of low-frequency and rare variants, particularly in non-European populations that have been underrepresented in existing genomic studies. The genetic determinants of C-reactive protein (CRP), a biomarker of chronic inflammation, have been extensively studied, with existing genome-wide association studies (GWASs) conducted in >200,000 individuals of European ancestry. In order to discover novel loci associated with CRP levels, we examined a multi-ancestry population (n = 23,279) with WGS (∼38× coverage) from the Trans-Omics for Precision Medicine (TOPMed) program. We found evidence for eight distinct associations at the CRP locus, including two variants that have not been identified previously (rs11265259 and rs181704186), both of which are non-coding and more common in individuals of African ancestry (∼10% and ∼1% minor allele frequency, respectively, and rare or monomorphic in 1000 Genomes populations of East Asian, South Asian, and European ancestry). We show that the minor (G) allele of rs181704186 is associated with lower CRP levels and decreased transcriptional activity and protein binding in vitro, providing a plausible molecular mechanism for this African ancestry-specific signal. The individuals homozygous for rs181704186-G have a mean CRP level of 0.23 mg/L, in contrast to individuals heterozygous for rs181704186 with mean CRP of 2.97 mg/L and major allele homozygotes with mean CRP of 4.11 mg/L. This study demonstrates the utility of WGS in multi-ethnic populations to drive discovery of complex trait associations of large effect and to identify functional alleles in noncoding regulatory regions. Whole-genome sequencing (WGS) data are being rapidly generated in deeply phenotyped cohorts or case-referent samples of complex disorders by projects such as the United Kingdom’s 100,000 Genomes Project,1The NIHR BioResource on behalf of the 100000 Genomes Project Whole-genome sequencing of rare disease patients in a national healthcare system.bioRxiv. 2019; https://doi.org/10.1101/507244Crossref Google Scholar the National Institute of Mental Health’s Whole Genome Sequencing for Psychiatric Disorders Consortium,2Sanders S.J. Neale B.M. Huang H. Werling D.M. An J.-Y. Dong S. Abecasis G. Arguello P.A. Blangero J. Boehnke M. et al.Whole Genome Sequencing for Psychiatric Disorders (WGSPD)Whole genome sequencing in psychiatric disorders: the WGSPD consortium.Nat. Neurosci. 2017; 20: 1661-1668Crossref PubMed Scopus (57) Google Scholar the National Human Genome Research Institute’s Centers for Common Disease Genomics (CCDG) project (see Web Resources), and the National Heart, Lung, and Blood Institute’s Trans-Omics for Precision Medicine (TOPMed) Program.3Taliun D. Harris D.N. Kessler M.D. Carlson J. Szpiech Z.A. Torres R. Taliun S.A.G. Corvelo A. Gogarten S.M. Kang H.M. et al.Sequencing of 53,831 diverse genomes from the NHLBI TOPMed Program.bioRxiv. 2019; https://doi.org/10.1101/563866Crossref Scopus (0) Google Scholar WGS resources can improve interrogation of low-frequency and rare variation associated with quantitative traits or clinical outcomes4Lappalainen T. Scott A.J. Brandt M. Hall I.M. Genomic Analysis in the Age of Human Genome Sequencing.Cell. 2019; 177: 70-84Abstract Full Text Full Text PDF PubMed Scopus (119) Google Scholar compared to genotyping array-based studies. However, sample sizes remain modest compared to large-scale genome-wide association studies (GWASs). WGS-based analysis may offer particular advantages for non-European populations currently underrepresented in GWASs, with ∼95% of GWAS participants being of European or East Asian ancestry.5Popejoy A.B. Fullerton S.M. Genomics is failing on diversity.Nature. 2016; 538: 161-164Crossref PubMed Scopus (805) Google Scholar WGS can assess population-specific variants which are at very low frequency or absent in large European GWASs, including variants that are often poorly imputed with standard reference panels and genotyping arrays. Current imputation reference panels for non-European populations (notably 1000 Genomes phase 3, n = 5,008 haplotypes across 26 mostly non-European populations6The 1000 Genomes Project Consortium A global reference for human genetic variation.Nature. 2015; 526: 68-74Crossref PubMed Scopus (8419) Google Scholar) are also much smaller than resources like the Haplotype Reference Consortium (HRC) for European populations (n = 64,976 haplotypes),7McCarthy S. Das S. Kretzschmar W. Delaneau O. Wood A.R. Teumer A. Kang H.M. Fuchsberger C. Danecek P. Sharp K. et al.Haplotype Reference ConsortiumA reference panel of 64,976 haplotypes for genotype imputation.Nat. Genet. 2016; 48: 1279-1283Crossref PubMed Scopus (1386) Google Scholar making imputation of low-frequency variants more difficult. Along with discrepancies in imputation reference panel size, many genotyping arrays have poor genomic coverage in non-European populations.8Wojcik G.L. Fuchsberger C. Taliun D. Welch R. Martin A.R. Shringarpure S. Carlson C.S. Abecasis G. Kang H.M. Boehnke M. et al.Imputation-Aware Tag SNP Selection To Improve Power for Large-Scale, Multi-ethnic Association Studies.G3 (Bethesda). 2018; 8: 3255-3267Crossref PubMed Scopus (18) Google Scholar Because WGS assesses the entire genome of each individual, the limitations of genotyping arrays and imputation reference panels are easily overcome, allowing better understanding of the genetic architecture of complex traits in non-European populations. Based on previous success in identifying novel coding low-frequency or population-specific variants for inflammatory biomarkers in sequencing-based analyses,9Polfus L.M. Raffield L.M. Wheeler M.M. Tracy R.P. Lange L.A. Lettre G. Miller A. Correa A. Bowler R.P. Bis J.C. et al.Whole genome sequence association with E-selectin levels reveals Loss-of-function variant in African Americans.Hum. Mol. Genet. 2019; 28: 515-523Crossref PubMed Scopus (12) Google Scholar,10Schick U.M. Auer P.L. Bis J.C. Lin H. Wei P. Pankratz N. Lange L.A. Brody J. Stitziel N.O. Kim D.S. et al.Cohorts for Heart and Aging Research in Genomic EpidemiologyNational Heart, Lung, and Blood Institute GO Exome Sequencing ProjectAssociation of exome sequences with plasma C-reactive protein levels in >9000 participants.Hum. Mol. Genet. 2015; 24: 559-571Crossref PubMed Scopus (30) Google Scholar we evaluated the ability of WGS to identify additional high-impact non-coding variation for commonly assessed inflammation biomarker C-reactive protein (CRP). CRP is an acute-phase protein synthesized in the liver and is often used as a biomarker for chronic low-grade inflammation. As such, its relationship to cardiovascular disease (CVD) has been well established by numerous epidemiological studies, though current analyses do not point to a causal relationship with CVD.11Prins B.P. Abbasi A. Wong A. Vaez A. Nolte I. Franceschini N. Stuart P.E. Guterriez Achury J. Mistry V. Bradfield J.P. et al.PAGE ConsortiumInternational Stroke Genetics ConsortiumSystemic Sclerosis consortiumTreat OA consortiumDIAGRAM ConsortiumCARDIoGRAMplusC4D ConsortiumALS consortiumInternational Parkinson’s Disease Genomics ConsortiumAutism Spectrum Disorder Working Group of the Psychiatric Genomics ConsortiumCKDGen consortiumGERAD1 ConsortiumInternational Consortium for Blood PressureSchizophrenia Working Group of the Psychiatric Genomics ConsortiumInflammation Working Group of the CHARGE ConsortiumInvestigating the Causal Relationship of C-Reactive Protein with 32 Complex Somatic and Psychiatric Outcomes: A Large-Scale Cross-Consortium Mendelian Randomization Study.PLoS Med. 2016; 13: e1001976Crossref PubMed Scopus (108) Google Scholar,12Ligthart S. Vaez A. Võsa U. Stathopoulou M.G. de Vries P.S. Prins B.P. Van der Most P.J. Tanaka T. Naderi E. Rose L.M. et al.LifeLines Cohort StudyCHARGE Inflammation Working GroupGenome Analyses of >200,000 Individuals Identify 58 Loci for Chronic Inflammation and Highlight Pathways that Link Inflammation and Complex Disorders.Am. J. Hum. Genet. 2018; 103: 691-706Abstract Full Text Full Text PDF PubMed Scopus (159) Google Scholar CRP has also been associated with inflammatory disorders,13Markatseli T.E. Voulgari P.V. Alamanos Y. Drosos A.A. Prognostic factors of radiological damage in rheumatoid arthritis: a 10-year retrospective study.J. Rheumatol. 2011; 38: 44-52Crossref PubMed Scopus (35) Google Scholar,14Gaitonde S. Samols D. Kushner I. C-reactive protein and systemic lupus erythematosus.Arthritis Rheum. 2008; 59: 1814-1820Crossref PubMed Scopus (73) Google Scholar type 2 diabetes,15Wang X. Bao W. Liu J. Ouyang Y.-Y. Wang D. Rong S. Xiao X. Shan Z.-L. Zhang Y. Yao P. Liu L.G. Inflammatory markers and risk of type 2 diabetes: a systematic review and meta-analysis.Diabetes Care. 2013; 36: 166-175Crossref PubMed Scopus (505) Google Scholar and overall mortality,16Zacho J. Tybjaerg-Hansen A. Nordestgaard B.G. C-reactive protein and all-cause mortality--the Copenhagen City Heart Study.Eur. Heart J. 2010; 31: 1624-1632Crossref PubMed Scopus (89) Google Scholar and recent Mendelian randomization studies have pointed to a potential causal role in bipolar disorder and schizophrenia.12Ligthart S. Vaez A. Võsa U. Stathopoulou M.G. de Vries P.S. Prins B.P. Van der Most P.J. Tanaka T. Naderi E. Rose L.M. et al.LifeLines Cohort StudyCHARGE Inflammation Working GroupGenome Analyses of >200,000 Individuals Identify 58 Loci for Chronic Inflammation and Highlight Pathways that Link Inflammation and Complex Disorders.Am. J. Hum. Genet. 2018; 103: 691-706Abstract Full Text Full Text PDF PubMed Scopus (159) Google Scholar CRP demonstrates substantial heritability in family-based studies (∼30% in East Asians,17Austin M.A. Zhang C. Humphries S.E. Chandler W.L. Talmud P.J. Edwards K.L. Leonetti D.L. McNeely M.J. Fujimoto W.Y. Heritability of C-reactive protein and association with apolipoprotein E genotypes in Japanese Americans.Ann. Hum. Genet. 2004; 68: 179-188Crossref PubMed Scopus (50) Google Scholar ∼30%–40% in Europeans,18Pankow J.S. Folsom A.R. Cushman M. Borecki I.B. Hopkins P.N. Eckfeldt J.H. Tracy R.P. Familial and genetic determinants of systemic markers of inflammation: the NHLBI family heart study.Atherosclerosis. 2001; 154: 681-689Abstract Full Text Full Text PDF PubMed Scopus (325) Google Scholar, 19Vickers M.A. Green F.R. Terry C. Mayosi B.M. Julier C. Lathrop M. Ratcliffe P.J. Watkins H.C. Keavney B. Genotype at a promoter polymorphism of the interleukin-6 gene is associated with baseline levels of plasma C-reactive protein.Cardiovasc. Res. 2002; 53: 1029-1034Crossref PubMed Scopus (234) Google Scholar, 20Schnabel R.B. Lunetta K.L. Larson M.G. Dupuis J. Lipinska I. Rong J. Chen M.-H. Zhao Z. Yamamoto J.F. Meigs J.B. et al.The relation of genetic and environmental factors to systemic inflammatory biomarker concentrations.Circ Cardiovasc Genet. 2009; 2: 229-237Crossref PubMed Scopus (45) Google Scholar ∼45% in African Americans21Fox E.R. Benjamin E.J. Sarpong D.F. Rotimi C.N. Wilson J.G. Steffes M.W. Chen G. Adeyemo A. Taylor J.K. Samdarshi T.E. Taylor Jr., H.A. Epidemiology, heritability, and genetic linkage of C-reactive protein in African Americans (from the Jackson Heart Study).Am. J. Cardiol. 2008; 102: 835-841Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar). CRP levels vary by race/ethnicity group with higher levels observed in individuals of African ancestry compared to European or East Asian ancestry.22Khera A. McGuire D.K. Murphy S.A. Stanek H.G. Das S.R. Vongpatanasin W. Wians Jr., F.H. Grundy S.M. de Lemos J.A. Race and gender differences in C-reactive protein levels.J. Am. Coll. Cardiol. 2005; 46: 464-469Crossref PubMed Scopus (551) Google Scholar,23Lakoski S.G. Cushman M. Palmas W. Blumenthal R. D’Agostino Jr., R.B. Herrington D.M. The relationship between blood pressure and C-reactive protein in the Multi-Ethnic Study of Atherosclerosis (MESA).J. Am. Coll. Cardiol. 2005; 46: 1869-1874Crossref PubMed Scopus (117) Google Scholar The genetic architecture of CRP has been investigated in diverse populations by whole-exome sequencing (WES),10Schick U.M. Auer P.L. Bis J.C. Lin H. Wei P. Pankratz N. Lange L.A. Brody J. Stitziel N.O. Kim D.S. et al.Cohorts for Heart and Aging Research in Genomic EpidemiologyNational Heart, Lung, and Blood Institute GO Exome Sequencing ProjectAssociation of exome sequences with plasma C-reactive protein levels in >9000 participants.Hum. Mol. Genet. 2015; 24: 559-571Crossref PubMed Scopus (30) Google Scholar genome-wide association,24Wu Y. McDade T.W. Kuzawa C.W. Borja J. Li Y. Adair L.S. Mohlke K.L. Lange L.A. Genome-wide association with C-reactive protein levels in CLHNS: evidence for the CRP and HNF1A loci and their interaction with exposure to a pathogenic environment.Inflammation. 2012; 35: 574-583Crossref PubMed Scopus (43) Google Scholar, 25Okada Y. Takahashi A. Ohmiya H. Kumasaka N. Kamatani Y. Hosono N. Tsunoda T. Matsuda K. Tanaka T. Kubo M. et al.Genome-wide association study for C-reactive protein levels identified pleiotropic associations in the IL6 locus.Hum. Mol. Genet. 2011; 20: 1224-1231Crossref PubMed Scopus (70) Google Scholar, 26Wojcik G.L. Graff M. Nishimura K.K. Tao R. Haessler J. Gignoux C.R. Highland H.M. Patel Y.M. Sorokin E.P. Avery C.L. et al.Genetic analyses of diverse populations improves discovery for complex traits.Nature. 2019; 570: 514-518Crossref PubMed Scopus (334) Google Scholar and fine-mapping studies imputed to various reference panels27Reiner A.P. Beleza S. Franceschini N. Auer P.L. Robinson J.G. Kooperberg C. Peters U. Tang H. Genome-wide association and population genetic analysis of C-reactive protein in African American and Hispanic American women.Am. J. Hum. Genet. 2012; 91: 502-512Abstract Full Text Full Text PDF PubMed Scopus (85) Google Scholar,28Kocarnik J.M. Richard M. Graff M. Haessler J. Bien S. Carlson C. Carty C.L. Reiner A.P. Avery C.L. Ballantyne C.M. et al.Discovery, fine-mapping, and conditional analyses of genetic variants associated with C-reactive protein in multiethnic populations using the Metabochip in the Population Architecture using Genomics and Epidemiology (PAGE) study.Hum. Mol. Genet. 2018; 27: 2940-2953Crossref PubMed Scopus (11) Google Scholar in tens of thousands of samples. Most recently, the largest GWAS was conducted in up to 204,402 individuals of European ancestry, identifying 58 loci and explaining 7% of the trait variance.12Ligthart S. Vaez A. Võsa U. Stathopoulou M.G. de Vries P.S. Prins B.P. Van der Most P.J. Tanaka T. Naderi E. Rose L.M. et al.LifeLines Cohort StudyCHARGE Inflammation Working GroupGenome Analyses of >200,000 Individuals Identify 58 Loci for Chronic Inflammation and Highlight Pathways that Link Inflammation and Complex Disorders.Am. J. Hum. Genet. 2018; 103: 691-706Abstract Full Text Full Text PDF PubMed Scopus (159) Google Scholar Some studies have also reported population-specific variants associated with CRP levels.27Reiner A.P. Beleza S. Franceschini N. Auer P.L. Robinson J.G. Kooperberg C. Peters U. Tang H. Genome-wide association and population genetic analysis of C-reactive protein in African American and Hispanic American women.Am. J. Hum. Genet. 2012; 91: 502-512Abstract Full Text Full Text PDF PubMed Scopus (85) Google Scholar Among reported loci, the locus surrounding the CRP (MIM: 123260) gene itself on chromosome 1 explains the largest portion of phenotypic variance (1.4%12Ligthart S. Vaez A. Võsa U. Stathopoulou M.G. de Vries P.S. Prins B.P. Van der Most P.J. Tanaka T. Naderi E. Rose L.M. et al.LifeLines Cohort StudyCHARGE Inflammation Working GroupGenome Analyses of >200,000 Individuals Identify 58 Loci for Chronic Inflammation and Highlight Pathways that Link Inflammation and Complex Disorders.Am. J. Hum. Genet. 2018; 103: 691-706Abstract Full Text Full Text PDF PubMed Scopus (159) Google Scholar), with multiple distinct signals reported and clear evidence of allelic heterogeneity across populations.27Reiner A.P. Beleza S. Franceschini N. Auer P.L. Robinson J.G. Kooperberg C. Peters U. Tang H. Genome-wide association and population genetic analysis of C-reactive protein in African American and Hispanic American women.Am. J. Hum. Genet. 2012; 91: 502-512Abstract Full Text Full Text PDF PubMed Scopus (85) Google Scholar,28Kocarnik J.M. Richard M. Graff M. Haessler J. Bien S. Carlson C. Carty C.L. Reiner A.P. Avery C.L. Ballantyne C.M. et al.Discovery, fine-mapping, and conditional analyses of genetic variants associated with C-reactive protein in multiethnic populations using the Metabochip in the Population Architecture using Genomics and Epidemiology (PAGE) study.Hum. Mol. Genet. 2018; 27: 2940-2953Crossref PubMed Scopus (11) Google Scholar For example, using approximate conditional analysis, the most recent European GWAS analysis reported 13 signals at the CRP locus (including rs149520992, an intergenic variant with a minor allele frequency [MAF] of 1% in Europeans and rare in other populations),12Ligthart S. Vaez A. Võsa U. Stathopoulou M.G. de Vries P.S. Prins B.P. Van der Most P.J. Tanaka T. Naderi E. Rose L.M. et al.LifeLines Cohort StudyCHARGE Inflammation Working GroupGenome Analyses of >200,000 Individuals Identify 58 Loci for Chronic Inflammation and Highlight Pathways that Link Inflammation and Complex Disorders.Am. J. Hum. Genet. 2018; 103: 691-706Abstract Full Text Full Text PDF PubMed Scopus (159) Google Scholar and four distinct signals (shared across ancestry groups) were reported in the multi-ethnic fine-mapping effort from the Population Architecture using Genomics and Epidemiology (PAGE) study.28Kocarnik J.M. Richard M. Graff M. Haessler J. Bien S. Carlson C. Carty C.L. Reiner A.P. Avery C.L. Ballantyne C.M. et al.Discovery, fine-mapping, and conditional analyses of genetic variants associated with C-reactive protein in multiethnic populations using the Metabochip in the Population Architecture using Genomics and Epidemiology (PAGE) study.Hum. Mol. Genet. 2018; 27: 2940-2953Crossref PubMed Scopus (11) Google Scholar African-specific variant rs726640 or variants in linkage disequilibrium (LD) with it have also been reported in several previous studies.26Wojcik G.L. Graff M. Nishimura K.K. Tao R. Haessler J. Gignoux C.R. Highland H.M. Patel Y.M. Sorokin E.P. Avery C.L. et al.Genetic analyses of diverse populations improves discovery for complex traits.Nature. 2019; 570: 514-518Crossref PubMed Scopus (334) Google Scholar,27Reiner A.P. Beleza S. Franceschini N. Auer P.L. Robinson J.G. Kooperberg C. Peters U. Tang H. Genome-wide association and population genetic analysis of C-reactive protein in African American and Hispanic American women.Am. J. Hum. Genet. 2012; 91: 502-512Abstract Full Text Full Text PDF PubMed Scopus (85) Google Scholar,29Doumatey A.P. Chen G. Tekola Ayele F. Zhou J. Erdos M. Shriner D. Huang H. Adeleye J. Balogun W. Fasanmade O. et al.C-reactive protein (CRP) promoter polymorphisms influence circulating CRP levels in a genome-wide association study of African Americans.Hum. Mol. Genet. 2012; 21: 3063-3072Crossref PubMed Scopus (27) Google Scholar Using data from the NHLBI TOPMed WGS project, we sought to investigate the additional value of WGS (beyond whole-exome sequencing and imputed GWAS) for single-variant analysis in a set of 23,279 individuals predominantly of self-reported European, African American, East Asian, and Hispanic/Latino ancestry with measured CRP levels (Table S1). We identified association with CRP levels at eight known loci (CRP, APOE [MIM: 107741], HNF1A [MIM: 142410], LEPR [MIM: 601007], GCKR [MIM: 600842], IL6R [MIM: 147880], IL1F10 [MIM: 615296], and NLRP3 [MIM: 606416]) with p < 1 × 10−9 in an ancestry-pooled genome-wide single-variant analysis (Table 1, Figure S1). We also examined these eight CRP-associated loci separately in African American (n = 6,545) and European American (n = 15,065) participants (Table S2). In the European American analysis, at least one variant at each locus met the locus-wide significance threshold for association with CRP levels with the exception of the NLRP3 locus. The African American analysis also demonstrated at least one locus-wide significant variant at all loci except GCKR and LEPR.Table 1Eight Loci Significantly Associated (p < 1 × 10−9) with C-Reactive Protein Levels in TOPMedLocusLead VariantAnnotationp ValueBetaEffect AlleleTOPMed EAF OverallTOPMedAfrican American EAFTOPMedEuropean American EAFAfter Conditioning on Lead VariantNew Lead Variantp Value2nd Signal ThresholdTotal # SignalsLEPRrs7516341intronic1.9E−19−0.09C0.430.540.37rs726831294.7E−054.7E−061IL6Rrs4129267intronic5.0E−12−0.07T0.330.140.40rs1494177742.7E−046.3E−061CRPrs7551731intergenic1.1E−65−0.18C0.300.220.33rs730247951.2E−422.4E−068NLRP3rs56188865intronic2.6E−11−0.06C0.420.520.38rs1156950521.6E−054.5E−061GCKRrs1260326missense, p.Leu446Pro (GCKR)1.9E−13−0.08C0.660.850.58rs1836286274.7E−046.7E−061IL1F10rs6734238intergenic8.4E−120.07G0.410.450.41rs1484983914.1E−046.2E−061HNF1Ars2243458intronic1.5E−33−0.13T0.270.120.33rs5447597083.3E−064.3E−062APOErs429358missense, p.Cys130Arg (APOE4)1.1E−65−0.22C0.150.210.13rs1864720691.6E−054.7E−061Significance threshold for identification of second signals calculated as p = (0.05/tested variants). EAF, effect allele frequency, for those in TOPMed CRP analysis. Open table in a new tab Significance threshold for identification of second signals calculated as p = (0.05/tested variants). EAF, effect allele frequency, for those in TOPMed CRP analysis. We performed stepwise conditional analyses at each of the eight loci by conditioning on the lead variant at each locus and then sequentially conditioning on each new lead variant until no variants met our locus-wide significance thresholds (Table 1). Stepwise conditional analyses were performed in ancestry pooled and stratified (self-reported European American- and African American-specific) analyses. We identified two conditionally distinct signals at HNF1A and eight at the CRP locus (Table 2, Figures 1, S2, and S3). The presence of multiple association signals at both CRP and HNF1A has been reported in previous studies, with at least two signals identified at both loci in a recent multi-ethnic fine-mapping effort (four signals at CRP, two signals at HNF1A)28Kocarnik J.M. Richard M. Graff M. Haessler J. Bien S. Carlson C. Carty C.L. Reiner A.P. Avery C.L. Ballantyne C.M. et al.Discovery, fine-mapping, and conditional analyses of genetic variants associated with C-reactive protein in multiethnic populations using the Metabochip in the Population Architecture using Genomics and Epidemiology (PAGE) study.Hum. Mol. Genet. 2018; 27: 2940-2953Crossref PubMed Scopus (11) Google Scholar and in the largest European meta-analysis (13 approximate conditional signals at CRP and 2 at HNF1A).12Ligthart S. Vaez A. Võsa U. Stathopoulou M.G. de Vries P.S. Prins B.P. Van der Most P.J. Tanaka T. Naderi E. Rose L.M. et al.LifeLines Cohort StudyCHARGE Inflammation Working GroupGenome Analyses of >200,000 Individuals Identify 58 Loci for Chronic Inflammation and Highlight Pathways that Link Inflammation and Complex Disorders.Am. J. Hum. Genet. 2018; 103: 691-706Abstract Full Text Full Text PDF PubMed Scopus (159) Google Scholar The eight identified signals at the CRP locus include low-frequency, exonic variants (rs1800947 [p.Leu184Leu] and rs553202904, a noncoding proxy for rs77832441 [p.Thr59Met]) and noncoding variants with much higher MAF in African ancestry individuals. These African American-driven signals include both known (rs73024795) and previously unreported (rs11265259, rs181704186) associations. In an unrelated subset (n = 17,371), these eight conditionally distinct signals explained 4.2% of variance in natural log transformed CRP (2.6% in European Americans, 6.0% in African Americans). When performing stepwise conditional analyses at the CRP locus separately by ancestry, five conditionally distinct signals were identified in African Americans alone and four conditionally distinct signals were identified in European Americans. Based on these results and with consideration of population-specific allele frequencies, four signals at CRP were driven primarily by African American individuals (rs73024795, rs11265259, rs181704186, rs2211321) and two by European Americans (rs553202904, rs12734907) (Table S3). The other two signals (rs7551731 and rs1800947) were shared between African Americans and European Americans.Table 2Eight Conditionally Distinct Signals Associated with C-Reactive Protein Were Identified at the CRP Locus in TOPMedSignalVariantAnnotationBetap ValueEffect AlleleTOPMed Overall EAFTOPMedAfrican American EAFTOPMedEuropean American EAF1000 Genomes AFR EAF1000 Genomes EUR EAFSequential Conditional p ValueArs7551731intergenic−0.181.1E−65C0.300.220.330.200.31–Brs73024795intergenic0.365.0E−54T0.050.164.98E−040.18N/A1.2E−42Crs2211321intergenic−0.020.05C0.700.650.710.640.713.1E−27Drs553202904aProxy variant is missense, Thr59Met (r2 = 0.98 in analyzed TOPMed samples)intergenic−0.701.4E−12G0.0023.82E-040.003N/A0.0038.8E−17Ers11265259intergenic−0.188.9E−09C0.030.094.31E−040.10N/A9.3E−12Frs1800947synonymous, p.Leu184Leu−0.245.8E−26G0.050.010.060.0020.059.2E−09Grs12734907intergenic0.081.5E−12T0.260.080.340.020.377.9E−10Hrs181704186intergenic−0.613.9E−12G0.0030.0099.96E−050.01N/A1.0E−07Abbreviations: AFR, African; EUR, European; N/A, not applicable (monomorphic). Letters correspond to the signals displayed in the LocusZoom plot in Figure 1. Beta, p value, and overall effect allele frequency are from TOPMed pooled ancestry analysis. EAF, effect allele frequency, for those in TOPMed CRP analysis.a Proxy variant is missense, Thr59Met (r2 = 0.98 in analyzed TOPMed samples) Open table in a new tab Abbreviations: AFR, African; EUR, European; N/A, not applicable (monomorphic). Letters correspond to the signals displayed in the LocusZoom plot in Figure 1. Beta, p value, and overall effect allele frequency are from TOPMed pooled ancestry analysis. EAF, effect allele frequency, for those in TOPMed CRP analysis. To determine whether the association signals we observed at the CRP or HNF1A loci were tagging previously reported associations, we performed a separate conditional analysis by which we adjusted for all variants associated with CRP levels at the CRP or HNF1A loci in prior GWAS, fine-mapping, or exome-sequencing efforts (Tables S4 and S5). In this analysis, two African American-driven signals at CRP remained locus-wide significant including rs11265259 (signal “E”; β = −0.32, p = 7.3 × 10−18; African American MAF = 0.10) and rs181704186 (signal “H”; β = −0.46, p = 3.0 × 10−7; African American MAF = 0.01); both are rare or monomorphic in other ancestry populations, with no copies of the minor allele for either variant found in 1000 Genomes European, East Asian, or South Asian populations. We also note the unusually large effect size for rs181704186, with major allele homozygotes having mean CRP levels of 4.11 mg/L (similar to the overall TOPMed mean of 4.10 mg/L), heterozygotes, 2.97 mg/L, and minor allele homozygotes, 0.23 mg/L, respectively (Figure 2A). By contrast, the more common variant, rs11265259, has mean CRP levels of 4.10, 4.36, and 3.04 mg/L, respectively. LD in African Americans from TOPMed between rs11265259 and rs181704186 and known signals is listed in Table S6. After adjusting for known variants at the HNF1A locus (Table S5), both association signals were attenuated below the locus-wide significance threshold. We thus carried forward the two conditionally distinct CRP signals, and not the secondary signal at HNF1A, for further follow-up. As both remaining CRP variant associations appeared to be distinct from any previously identified CRP locus variant association, we attempted to replicate these two signals using CRP measurements in African American women from the Women’s Health Initiative (WHI) study (n = 7,108). The WHI participants had genotype data from an Affymetrix 6.0 array imputed to the TOPMed reference panel (freeze 5b, Michigan Imputation Server) but were not whole genome s