Common Variants in CRP and LEPR Influence High Sensitivity C-Reactive Protein Levels in North Indians

Abstract

Background

High sensitivity C-reactive protein (hsCRP) levels are shown to be influenced by genetic variants in Europeans; however, little is explored in Indian population.

Methods

Herein, we comprehensively evaluated association of all previously reported genetic determinants of hsCRP levels, including 18 cis (proximal to CRP gene) and 73 trans-acting (distal to CRP gene) variants in 4,200 North Indians of Indo-European ethnicity. First, we evaluated association of 91 variants from 12 candidate loci with hsCRP levels in 2,115 North Indians (1,042 non-diabetic subjects and 1,073 patients with type 2 diabetes). Then, cis and trans-acting variants contributing maximally to hsCRP level variation were further replicated in an independent 2,085 North Indians (1,047 patients with type 2 diabetes and 1,038 non-diabetic subjects).

Results

We found association of 12 variants from CRP, LEPR, IL1A, IL6, and IL6R with hsCRP levels in non-diabetic subjects. However, only rs3093059-CRP [β = 0.33, P = 9.6×10−5] and the haplotype harboring rs3093059 risk allele [β = 0.32 µg/mL, P = 1.4×10−4/Pperm = 9.0×10−4] retained significance after correcting for multiple testing. The cis-acting variant rs3093059-CRP had maximum contribution to the variance in hsCRP levels (1.14%). Among, trans-acting variants, rs1892534-LEPR was observed to contribute maximally to hsCRP level variance (0.59%). Associations of rs3093059-CRP and rs1892534-LEPR were confirmed by replication and attained higher significance after meta-analysis [βmeta = 0.26/0.22; Pmeta = 4.3×10−7/7.4×10−3 and βmeta = −0.15/−0.12; Pmeta = 2.0×10−6/1.6×10−6 for rs3093059 and rs1892534, respectively in non-diabetic subjects and all subjects taken together].

Conclusion

In conclusion, we identified rs3093059 in CRP and rs1892534 in LEPR as major cis and trans-acting contributor respectively, to the variance in hsCRP levels in North Indian population.

About the Authors:

Anubha Mahajan

Affiliation: Genomics and Molecular Medicine Unit, CSIR-Institute of Genomics and Integrative Biology, Delhi, India

Rubina Tabassum

Affiliation: Genomics and Molecular Medicine Unit, CSIR-Institute of Genomics and Integrative Biology, Delhi, India

Sreenivas Chavali

Affiliation: Genomics and Molecular Medicine Unit, CSIR-Institute of Genomics and Integrative Biology, Delhi, India

Om Prakash Dwivedi

Affiliation: Genomics and Molecular Medicine Unit, CSIR-Institute of Genomics and Integrative Biology, Delhi, India

Ganesh Chauhan

Affiliation: Genomics and Molecular Medicine Unit, CSIR-Institute of Genomics and Integrative Biology, Delhi, India

Saurabh Ghosh

Affiliation: Human Genetics Unit, Indian Statistical Institute, Kolkata, India

Nikhil Tandon

* E-mail: [email protected] (DB); [email protected] (NT)

Affiliation: Department of Endocrinology, All India Institute of Medical Sciences, New Delhi, India

Dwaipayan Bharadwaj

* E-mail: [email protected] (DB); [email protected] (NT)

Affiliation: Genomics and Molecular Medicine Unit, CSIR-Institute of Genomics and Integrative Biology, Delhi, India

Introduction

High sensitivity C-reactive protein (hsCRP), is a key component of non-specific host defense in humans. Elevated basal hsCRP levels have been found in low-grade systemic inflammation and are used as a marker of risk for various pathological conditions including type 2 diabetes (T2D) [1], [2]. These levels are genetically determined with ∼30–50% heritability [3].

Both cis- (proximal to CRP gene) and trans-acting (distal to CRP gene) genetic variants have been identified to contribute to hsCRP levels in different ethnic cohorts [4], [5], [6], [7], [8], [9], [10]. Recent genome-wide association studies (GWAS) have confirmed the involvement of several previously identified variants and additionally recognized new variants, providing important insights into hsCRP level regulation [11], [12]. However, very little is explored in this regard in Indian population, which is a high risk group for T2D. Elevated levels of hsCRP were found to be associated with increased risk of T2D in North Indian population in our previous study [13]. Moreover, our recent analysis of TNF-LTA locus identified a variant (rs2229094) influencing hsCRP levels in this population [14]. Thus, a more thorough examination of genetic determinants of hsCRP levels in this population is warranted.

Herein, we comprehensively evaluated association of all previously reported genetic determinants of hsCRP levels in North Indians.

Materials and Methods

Ethics Statement

The study was carried out in accordance with the principles of Helsinki Declaration and was approved by Human Ethics Committee of Institute of Genomics and Integrative Biology (CSIR) and All India Institute of Medical Sciences Research Ethics Committee. Informed written consent was obtained from all the participants.

Study population

The initial investigation included 2,115 unrelated Indo-Europeans (1,073 T2D patients and 1,042 non-diabetic subjects) from North India residing in and around Delhi. Details of subject recruitment have been described previously [13], [14], [15]. An independent set of 2,085 samples (1,047 T2D patients and 1,038 non-diabetic subjects), recruited similar to that in the initial study, was used for further replication. Plasma hsCRP levels were measured using ELISA kit (Biocheck Inc., CA, USA) and Cobas Integra 400 Plus (Roche Diagnostics, GmbH, Mannheim, Germany).

Variant selection and genotyping

We selected all the index variants shown to be associated with hsCRP levels till July 2009 or their proxies. Further, to capture other variants in the selected genes, we also included additional common variants that were prioritized based on minor allele frequency (MAF >0.05) in at least two HapMap populations, tagging and linkage disequilibrium (LD) pattern information, and spacing between the variants to cover the gene. The selected variants comprised of 18 cis variants from CRP and 73 from 11 trans-acting loci (APOE, LEPR, IL1A, IL1B, IL6, IL6R, FAM13C1, GCKR, SLC1A3, HNF1A, and a gene-desert region of chromosome 12q23.2) (Table S1).

Genotyping of 47 variants was done using Illumina GoldenGate assay [16]. Samples (n = 90) with a poor call rate (<90%) and variants (n = 3) failing quality control were excluded from the analysis. Average call rate for successful assays was 99.6% with >99.9% concordance in 7.2% duplicates. The remaining 44 variants and one that failed Illumina assay were genotyped using iPLEX (Sequenom, San Diego, USA). The average genotyping success rate of the quality control passed variants (n = 41) was 97.7%, with 99.9% consensus in 4% duplicates. Four rare variants (MAF<0.05) and three that deviated from Hardy-Weinberg equilibrium (P<5.9×10−4; 0.05/85) were excluded, and thus 78 variants were considered for association analyses.

Further genotyping of two variants was done in an independent sample set on iPLEX platform for replication. Average genotyping success rate for these was 98.3% with 99.9% concordance in 5% duplicates.

Statistical analysis

Linear-regression was performed assuming an additive model to determine association of variants with hsCRP levels. Samples with hsCRP level >10 µg/mL were excluded from the analysis. hsCRP concentrations were natural logarithm transformed to ensure a normal distribution. To assess independent contribution of each variant, stepwise multivariate linear regression models with age, sex, and body mass index (BMI) as covariates were constructed. Variants with a P<0.05 were retained in the forward stepwise-selection models. Analyses were first carried out in non-diabetic subjects not taking lipid-lowering drugs, adjusting for age, sex, BMI, and smoking/tobacco chewing and then in T2D patients and all subjects, additionally adjusting for diseases status and medication as appropriate.

Bonferroni correction was used to account for multiple testing and a P<8.6×10−4 was considered significant after correcting for number of independent variants α = 0.05/58). Pairwise linkage disequilibrium (LD) between the SNPs for all the genes was determined using Haploview 4.0 (Figure S1). Association of haplotypes with frequency >0.1 was examined through sliding window approach using PLINK by adjusting for age, sex, BMI, smoking/tobacco chewing, diseases status, and medication, as appropriate, based on 10,000 permutations. Fixed and random effect meta-analysis was performed for the two replicated variants, in non-diabetic subjects, T2D patients and all subjects.

Statistical power of the study was evaluated using QUANTO version 1.2 [17]. Our study had 78% and 92% power to detect association for a variant with MAF = 0.2 and effect size of 0.3 µg/mL for additive model for non-diabetic subjects and for all subjects analyzed together respectively. Statistical analyses were performed using SPSS version 17.0 (SPSS, Chicago, IL) and PLINK [18].

Results

A brief description of study populations is presented in Table S2.

Cis-acting effects

CRP variants rs3093059 and rs3093077, that were in strong LD (r2 = 0.89), showed strongest associations with hsCRP in non-diabetic subjects [β = 0.32 µg/mL, P = 1.6×10−4 and β = 0.26, P = 1.3×10−3 respectively; Table 1]. Nominal associations were observed for variants rs3116654, rs4131568, and rs1205. A total of 2.42% of the residual variation in hsCRP could be explained by variants rs3093059, rs3116654, and rs4131568. Variant rs3093059 contributed maximally (1.14%) to hsCRP level variance and also retained significant association with hsCRP levels after accounting for multiple testing. The haplotype GATCTTCG encompassing rs3093059 risk allele also showed association with hsCRP levels [β = 0.32, P = 1.4×10−4/Pperm = 9.0×10−4; Table S3]. We observed similar results when separate analyses for T2D patients and all the subjects were performed (Table 1). Further, we tested association of haplotypes of different lengths to detect a signal that could not be observed by single marker analysis. The β and P values for haplotype associations were found to be similar as the individual associated SNPs (Figure S2). We did not find any haplotype that could detect region or signal in addition to single marker analysis.

[Figure omitted. See PDF.]

Table 1. Variants associated with hsCRP levels.

https://doi.org/10.1371/journal.pone.0024645.t001

Trans-acting effects

Three LEPR, two IL6R, one IL6, and one IL1A variants showed association with hsCRP levels (P<0.05; Table 1). All three LEPR and two IL6R variants (rs4129267, and rs4845622) were in strong LD (r2 = 0.91–0.96 and 0.96 respectively). The most significant association was detected for LEPR-rs1892534 [β = −0.16, P = 5.5×10−3], however, none of the associations retained significance after accounting for multiple testing. These loci accounted for a total of 1.63% hsCRP level variance, with 0.59%, 0.56%, and 0.48% variance attributed to rs1892534-LEPR, rs1800796-IL6, and rs4129267-IL6R respectively. LEPR, IL6 and IL6R also showed association with hsCRP levels when all subjects were analyzed together. The most significant association was observed for rs769449-APOE [β = −0.27, P = 1.2×10−3]. However, none of these retained significance after correcting for multiple testing. Similar results were obtained when analyses were performed separately in T2D patients (Table 1). The haplotype GGGCCA-IL6 was associated with hsCRP levels when all the subjects were analyzed together [β = −0.16, P = 3.9×10−4/Pperm = 1.2×10−3]. We did not find any haplotype that could detect region or signal in addition to single marker analysis (Figure S2).

Follow up and Meta-analysis

The rs3093059-CRP and rs1892534-LEPR, strongest associated cis and trans-acting variants respectively, were successfully replicated in an independent sample of North Indians (Table 2). Meta-analysis further strengthened the associations in non-diabetic subjects [βmeta = 0.26, Pmeta = 4.3×10−7 and βmeta = −0.15, Pmeta = 2.0×10−6 for rs3093059 and rs1892534, respectively, for random effect]. No heterogeneity was detected between two study populations (I2 = 0). Similarly, significant associations were observed when all subjects were analyzed together, however only nominal association of rs1892534-LEPR was observed when analysis was performed separately in T2D patients (Table 2).

[Figure omitted. See PDF.]

Table 2. Association of rs3093059-CRP and rs1892534-LEPR with hsCRP levels.

https://doi.org/10.1371/journal.pone.0024645.t002

Discussion

Recent GWAS and other previous candidate gene based investigations have demonstrated role of cis-acting variants of CRP and trans-acting variants from genes including LEPR, IL6R, IL6 etc in regulation of hsCRP levels. Our study demonstrated strong association of a cis-acting variant rs3093059-CRP and a trans-acting variant rs1892534-LEPR with hsCRP levels in North Indians. These two variants contributed maximally to the variance in hsCRP levels.

Several studies have shown strong association of variants at CRP locus with plasma levels of hsCRP [7], [19]–[22]. Our study also demonstrates cis acting variants to greatly influence plasma hsCRP levels. Among them, rs3093059 showed strongest association with hsCRP levels that was confirmed in an independent study population. The results in the combined analysis of T2D patients and non-diabetic subjects were comparable to that seen in non-diabetic subjects alone. Our observations are consistent with previous reports demonstrating that both CRP variants and haplotypes are associated with hsCRP levels [7], [19].

Among the trans acting variants, strongest association was observed at rs1892534 of LEPR. Leptin receptor signaling has been linked to appetite control, weight regulation, glucose homeostasis, vascular risk, and hsCRP levels [8], [11]. Previous reports provided evidence for the association of rs1892534 with hsCRP levels in European populations [8], [11]. Thus, our results support the hypothesis that leptin receptor influence inflammatory traits through genetic modulation of hsCRP levels.

We found nominal association of other trans-acting loci namely IL1A, IL6, IL6R, and APOE with hsCRP levels in non-diabetic North Indians. All these variants contributed to ∼2% of hsCRP level variance. The analysis of all subjects taken together essentially substantiated the results seen in non-diabetic subjects. IL6R and APOE variants were shown to be associated with hsCRP levels in apparently healthy Caucasian women in an agnostic approach [11]. This is perhaps the first study to replicate the association of rs1800587-IL1A variant [5], though it could not attain the statistical significance after multiple testing corrections. The effects of all associations were directionally consistent and the effect sizes were almost comparable to earlier studies [5], [8], [11]. Hence, our study validates the observations in European populations, suggesting involvement of overlapping genetic mechanisms in regulating hsCRP levels in Europeans and North Indians.

Though we replicated the association of CRP, LEPR, IL1A, IL6, IL6R, and APOE, we could not detect association of the most widely implicated tri-allelic CRP variant rs3091244. The tri-allelic variant rs3091244 has been repeatedly shown to be associated with plasma hsCRP levels and an in vitro study suggested its functional role in transcription factor binding [20]. There was no difference in allele frequencies between Caucasians and North Indians at this locus but we observed differences in LD pattern around this variant in the CRP gene between the two populations (Figure S1). A recent GWAS study for hsCRP plasma levels also indicated the differences in genetic background in CRP between different ethnic groups [21]. Thus the ethnic differences in genetic architecture at this locus could be the reason for difference in association results.

Moreover recent studies have discovered GCKR as a novel hsCRP modulating candidate gene and have provided confirmatory evidence for the association of HNF1A [9], [11]. However, variants from both these loci were not associated in our study. Allele frequencies at these loci differed significantly between the Caucasians and North Indians and could have potentially limited the power to detect association, if any. Differences in LD patterns between the populations could potentially lead to differences in results but the comparison of LD around these genes showed similar patterns in Indians and Europeans (Figure S1) and thus do not account for the observed differences in association results.

Limited statistical power and population stratification might also lead to the disparity in the association results. There is a likelihood of false negative observations for variants with small effect sizes, as our study is sufficiently powered to capture only large effects (>0.25 µg/mL) of the common variants with frequencies more than 0.20. Our study is, however, unlikely to be influenced by population stratification, because the study population is highly homogeneous, as previously demonstrated [15], [23]. We have included only North Indian subjects belonging to Indo-European ethnicity from Delhi and surrounding regions that form a homogenous cluster as reported by the Indian Genome Variation Consortium [23].

Further, we would like to mention that although we attempted to cover entire gene and investigate all the reported associated variants, we could not type few of them due to assay failures. Moreover, the ascertainment procedure of the subjects that recruited subjects based on the presence or absence of T2D could potentially affect the estimation of effect sizes. We could not determine the population level effect sizes that could have provided added information from the study. However, with the major focus of the study to determine the genetic variants influencing hsCRP levels in Indian population, this is not essentially a limitation of the study.

Epidemiological studies have consistently demonstrated that the risk of T2D and cardio-vascular disease (CVD) increase with higher hsCRP levels. In a study conducted on samples inclusive of those studied here, we earlier demonstrated that hsCRP levels were associated with T2D in a BMI independent manner [13]. In the light of hsCRP being a potential biomarker for CVD and T2D, the search for genetic variants influencing its expression is of great importance. With the realization that variant frequencies and haplotype structures vary considerably between ethnic groups, the present study provides a clear picture of genetic factors associated with plasma hsCRP levels in Indian population. Identification of genetic regulators of hsCRP levels might help in the development of prognostic markers for detecting individuals at risk for CVD and T2D.

In conclusion, we identified rs3093059 in CRP and rs1892534 in LEPR as the major cis and trans-acting contributor respectively, to the variance in hsCRP levels in North Indian population. Still further studies investigating variants from CRP and its regulatory genes are warranted to explain the total genetic control of circulating hsCRP levels.

Supporting Information

[Figure omitted. See PDF.]

Figure S1.

Pairwise linkage disequilibrium (LD) between the selected SNPs in the genes investigated in the study. Left panel shows the plots with D' values and right panel shows plot with r2values. LD plots for North Indians were drawn using the genotype data from the present study whereas LD plots for Europeans were made from genotype data from HapMap.

https://doi.org/10.1371/journal.pone.0024645.s001

(DOC)

Figure S2.

Association of individual associated SNPs and their haplotypes with hsCRP levels. A: CRP and B: IL6. The −log10 P values are plotted against the respective SNPs and haplotype combinations formed by that SNP and all the SNPs before. The trend lines of −log10 P values for individual SNPs and highest −log10 P values for haplotypes are drawn for the comparison of association of SNPs and haplotypes.

https://doi.org/10.1371/journal.pone.0024645.s002

(DOC)

Table S1.

Details of the polymorphisms selected for this study. MAF: Minor Allele Frequency; HWE: Hardy-Weinberg equilibrium. * MAF, Genotype distribution and HWE estimations have been done in non-diabetic subjects alone. † Tri-allelic variant. The two less-common alleles (T and A) were pooled. ‡ MAF<0.05. §Assay failure. ∥ Significant deviation from HWE. Chromosome positions are relative to NCBI Build 36.3 assembly, dbSNP b126.

https://doi.org/10.1371/journal.pone.0024645.s003

(DOC)

Table S2.

Descriptive characteristics of the study populations. Data are presented as median values (inter-quartile ranges). N: Number of subjects; TC: total cholesterol; LDL-C: low-density lipoprotein cholesterol; HDL-C: high-density lipoprotein cholesterol; TG: triglyceride; hsCRP: high sensitivity C-reactive protein.

https://doi.org/10.1371/journal.pone.0024645.s004

(DOC)

Table S3.

Haplotype association with hsCRP levels. β: Beta. β and P values presented were adjusted for age, sex, BMI, and smoking/tobacco chewing for analysis done in non-diabetic subjects; age, sex, BMI, smoking/tobacco chewing, and medication for diabetic subjects; and age, sex, BMI, smoking/tobacco chewing, disease status, and medication for all subjects taken together. Pperm: P value obtained after 10,000 permutations.

https://doi.org/10.1371/journal.pone.0024645.s005

(DOC)

Acknowledgments

We thank all the subjects for their participation. We also thank Dr. Abhay Sharma and Dr. Arijit Mukhopadhyay (Institute of Genomics and Integrative Biology, CSIR) for critical reading of the manuscript.

Author Contributions

Conceived and designed the experiments: DB NT. Performed the experiments: OPD GC. Analyzed the data: AM RT SC SG. Wrote the paper: AM RT SC.

Citation: Mahajan A, Tabassum R, Chavali S, Dwivedi OP, Chauhan G, Ghosh S, et al. (2011) Common Variants in CRP and LEPR Influence High Sensitivity C-Reactive Protein Levels in North Indians. PLoS ONE6(9): e24645. https://doi.org/10.1371/journal.pone.0024645

References

1. Pradhan AD, Manson JE, Rifai N, Buring JE, Ridker PM (2001) C-reactive protein, interleukin 6, and risk of developing type 2 diabetes mellitus. JAMA 286: 327–334.AD PradhanJE MansonN. RifaiJE BuringPM Ridker2001C-reactive protein, interleukin 6, and risk of developing type 2 diabetes mellitus.JAMA286327334

2. Ridker PM (2009) C-reactive protein: eighty years from discovery to emergence as a major risk marker for cardiovascular disease. Clin Chem 55: 209–215.PM Ridker2009C-reactive protein: eighty years from discovery to emergence as a major risk marker for cardiovascular disease.Clin Chem55209215

3. MacGregor AJ, Gallimore JR, Spector TD, Pepys MB (2004) Genetic effects on baseline values of C-reactive protein and serum amyloid a protein: a comparison of monozygotic and dizygotic twins. Clin Chem 50: 130–134.AJ MacGregorJR GallimoreTD SpectorMB Pepys2004Genetic effects on baseline values of C-reactive protein and serum amyloid a protein: a comparison of monozygotic and dizygotic twins.Clin Chem50130134

4. Vickers MA, Green FR, Terry C, Mayosi BM, Julier C, et al. (2002) Genotype at a promoter polymorphism of the interleukin-6 gene is associated with baseline levels of plasma C-reactive protein. Cardiovasc Res 53: 1029–1034.MA VickersFR GreenC. TerryBM MayosiC. Julier2002Genotype at a promoter polymorphism of the interleukin-6 gene is associated with baseline levels of plasma C-reactive protein.Cardiovasc Res5310291034

5. Sainz J, Perez E, Gomez-Lopera S, Jurado M (2008) IL1 gene cluster polymorphisms and its haplotypes may predict the risk to develop invasive pulmonary aspergillosis and modulate C-reactive protein level. J Clin Immunol 28: 473–485.J. SainzE. PerezS. Gomez-LoperaM. Jurado2008IL1 gene cluster polymorphisms and its haplotypes may predict the risk to develop invasive pulmonary aspergillosis and modulate C-reactive protein level.J Clin Immunol28473485

6. Crawford DC, Sanders CL, Qin X, Smith JD, Shephard C, et al. (2006) Genetic variation is associated with C-reactive protein levels in the Third National Health and Nutrition Examination Survey. Circulation 114: 2458–2465.DC CrawfordCL SandersX. QinJD SmithC. Shephard2006Genetic variation is associated with C-reactive protein levels in the Third National Health and Nutrition Examination Survey.Circulation11424582465

7. Kathiresan S, Larson MG, Vasan RS, Guo CY, Gona P, et al. (2006) Contribution of clinical correlates and 13 C-reactive protein gene polymorphisms to interindividual variability in serum C-reactive protein level. Circulation 113: 1415–1423.S. KathiresanMG LarsonRS VasanCY GuoP. Gona2006Contribution of clinical correlates and 13 C-reactive protein gene polymorphisms to interindividual variability in serum C-reactive protein level.Circulation11314151423

8. Zhang YY, Gottardo L, Mlynarski W, Frazier W, Nolan D, et al. (2007) Genetic variability at the leptin receptor (LEPR) locus is a determinant of plasma fibrinogen and C-reactive protein levels. Atherosclerosis 191: 121–127.YY ZhangL. GottardoW. MlynarskiW. FrazierD. Nolan2007Genetic variability at the leptin receptor (LEPR) locus is a determinant of plasma fibrinogen and C-reactive protein levels.Atherosclerosis191121127

9. Reiner AP, Barber MJ, Guan Y, Ridker PM, Lange LA, et al. (2008) Polymorphisms of the HNF1A gene encoding hepatocyte nuclear factor-1 alpha are associated with C-reactive protein. Am J Hum Genet 82: 1193–1201.AP ReinerMJ BarberY. GuanPM RidkerLA Lange2008Polymorphisms of the HNF1A gene encoding hepatocyte nuclear factor-1 alpha are associated with C-reactive protein.Am J Hum Genet8211931201

10. Qi L, Rifai N, Hu FB (2009) Interleukin-6 receptor gene, plasma C-reactive protein, and diabetes risk in women. Diabetes 58: 275–278.L. QiN. RifaiFB Hu2009Interleukin-6 receptor gene, plasma C-reactive protein, and diabetes risk in women.Diabetes58275278

11. Ridker PM, Pare G, Parker A, Zee RY, Danik JS, et al. (2008) Loci related to metabolic-syndrome pathways including LEPR,HNF1A, IL6R, and GCKR associate with plasma C-reactive protein: the Women's Genome Health Study. Am J Hum Genet 82: 1185–1192.PM RidkerG. PareA. ParkerRY ZeeJS Danik2008Loci related to metabolic-syndrome pathways including LEPR,HNF1A, IL6R, and GCKR associate with plasma C-reactive protein: the Women's Genome Health Study.Am J Hum Genet8211851192

12. Elliott P, Chambers JC, Zhang W, Clarke R, Hopewell JC, et al. (2009) Genetic Loci associated with C-reactive protein levels and risk of coronary heart disease. JAMA 302: 37–48.P. ElliottJC ChambersW. ZhangR. ClarkeJC Hopewell2009Genetic Loci associated with C-reactive protein levels and risk of coronary heart disease.JAMA3023748

13. Mahajan A, Tabassum R, Chavali S, Dwivedi OP, Bharadwaj M, et al. (2009) High-sensitivity C-reactive protein levels and type 2 diabetes in urban North Indians. J Clin Endocrinol Metab 94: 2123–2127.A. MahajanR. TabassumS. ChavaliOP DwivediM. Bharadwaj2009High-sensitivity C-reactive protein levels and type 2 diabetes in urban North Indians.J Clin Endocrinol Metab9421232127

14. Tabassum R, Mahajan A, Chauhan G, Dwivedi OP, Ghosh S, et al. (2010) Evaluation of DOK5 as a susceptibility gene for type 2 diabetes and obesity in North Indian population. BMC Med Genet 11: 35.R. TabassumA. MahajanG. ChauhanOP DwivediS. Ghosh2010Evaluation of DOK5 as a susceptibility gene for type 2 diabetes and obesity in North Indian population.BMC Med Genet1135

15. Mahajan A, Tabassum R, Chavali S, Dwivedi OP, Chauhan G, et al. (2010) Obesity-dependent association of TNF-LTA locus with type 2 diabetes in North Indians. J Mol Med 88: 515–522.A. MahajanR. TabassumS. ChavaliOP DwivediG. Chauhan2010Obesity-dependent association of TNF-LTA locus with type 2 diabetes in North Indians.J Mol Med88515522

16. Shen R, Fan JB, Campbell D, Chang W, Chen J, et al. (2005) High-throughput SNP genotyping on universal bead arrays. Mutat Res 573: 70–82.R. ShenJB FanD. CampbellW. ChangJ. Chen2005High-throughput SNP genotyping on universal bead arrays.Mutat Res5737082

17. Gauderman WJ (2003) Candidate gene association analysis for a quantitative trait, using parent-offspring trios. Genet Epidemiol 25: 327–338.WJ Gauderman2003Candidate gene association analysis for a quantitative trait, using parent-offspring trios.Genet Epidemiol25327338

18. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, et al. (2007) PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 81: 559–575.S. PurcellB. NealeK. Todd-BrownL. ThomasMA Ferreira2007PLINK: a tool set for whole-genome association and population-based linkage analyses.Am J Hum Genet81559575

19. Carlson CS, Aldred SF, Lee PK, Tracy RP, Schwartz SM, et al. (2005) Polymorphisms within the C-reactive protein (CRP) promoter region are associated with plasma CRP levels. Am J Hum Genet 77: 64–77.CS CarlsonSF AldredPK LeeRP TracySM Schwartz2005Polymorphisms within the C-reactive protein (CRP) promoter region are associated with plasma CRP levels.Am J Hum Genet776477

20. Szalai AJ, Wu J, Lange EM, McCrory MA, Langefeld CD, et al. (2005) Single-nucleotide polymorphisms in the C-reactive protein (CRP) gene promoter that affect transcription factor binding, alter transcriptional activity, and associate with differences in baseline serum CRP level. J Mol Med 83: 440–447.AJ SzalaiJ. WuEM LangeMA McCroryCD Langefeld2005Single-nucleotide polymorphisms in the C-reactive protein (CRP) gene promoter that affect transcription factor binding, alter transcriptional activity, and associate with differences in baseline serum CRP level.J Mol Med83440447

21. Wu Y, McDade TW, Kuzawa CW, Borja J, Li Y, et al. (2011) 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. Y. WuTW McDadeCW KuzawaJ. BorjaY. Li2011Genome-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

22. Dehghan A, Dupuis J, Barbalic M, Bis JC, Eiriksdottir G, et al. (2011) Meta-analysis of genome-wide association studies in >80 000 subjects identifies multiple loci for C-reactive protein levels. Circulation 123: 731–738.A. DehghanJ. DupuisM. BarbalicJC BisG. Eiriksdottir2011Meta-analysis of genome-wide association studies in >80 000 subjects identifies multiple loci for C-reactive protein levels.Circulation123731738

23. IndianGenomeVariationConsortium (2008) Genetic landscape of the people of India: a canvas for disease gene exploration. J Genet 87: 3–20.IndianGenomeVariationConsortium2008Genetic landscape of the people of India: a canvas for disease gene exploration.J Genet87320