JMG Online First, published on November 18, 2017 as 10.1136/jmedgenet-2017-104922 Telomere biology J Med Genet: first published as 10.1136/jmedgenet-2017-104922 on 18 November 2017. Downloaded from ORIGINAL ARTICLE Genome-wide association study of telomere length among South Asians identifies a second RTEL1 association signal Dayana A Delgado,1 Chenan Zhang,1,2 Lin S Chen,1 Jianjun Gao,3 Shantanu Roy,1,4 Justin Shinkle,1 Mekala Sabarinathan,1 Maria Argos,5 Lin Tong,1 Alauddin Ahmed,6 Tariqul Islam,6 Muhammad Rakibuz-Zaman,6 Golam Sarwar,6 Hasan Shahriar,6 Mahfuzar Rahman,7 Mohammad Yunus,8 Farzana Jasmine,1 Muhammad G Kibriya,1 Habibul Ahsan,1,9,10,11 Brandon L Pierce1,9,10 ► Additional material is ABSTRact telomeres is carried out by the holoenzyme, telo- published online only. To view Background Leucocyte telomere length (TL) is a merase, a ribonucleic protein consisting of a reverse please visit the journal online transcriptase (TERT) and RNA template (TERC), (http:// dx. doi. org/ 10. 1136/ potential biomarker of ageing and risk for age-related jmedgenet- 2017- 104922). disease. Leucocyte TL is heritable and shows substantial which enable the elongation of the telomere repeat differences by race/ethnicity. Recent genome-wide sequence, conferring replicative immortality.3–5 For numbered affiliations see association studies (GWAS) report ~10 loci harbouring Leucocyte telomere length (TL) shortens with age, end of article. SNPs associated with leucocyte TL, but these studies and associations between short TL and increased focus primarily on populations of European ancestry. risk for numerous age-related diseases have been Correspondence to reported (ie, cardiovascular disease, hyperten- Dr Brandon L Pierce, Objective This study aims to enhance our Department of Public Health understanding of genetic determinants of TL across sion, liver disorders, diabetes, atherosclerosis Sciences, The University of populations. and mortality); consequently, TL is regarded as a Chicago, 5841 South Maryland Methods We performed a GWAS of TL using data on surrogate marker for biological ageing.6–10 Addi- Avenue, MC2000 Chicago, IL 5075 Bangladeshi adults. We measured TL using one of tionally, there is increasing evidence suggesting 60637, USA; brandonpierce@ uchicago. edu two technologies (qPCR or a Luminex-based method) more complex relationships among TL, cellular and used standardised variables as TL phenotypes. senescence and human health, with longer TL Received 13 July 2017 Results Our results replicate previously reported confering increased risk for several cancer types, Revised 13 September 2017 associations in the TERC and TERT regions (P=2.2×10−8 including melanoma,11 glioma,12 neuroblastoma,13 Accepted 26 September 2017 and P=6.4×10−6, respectively). We observed a novel lung adenocarcinoma14 and chronic lymphocytic association signal in the RTEL1 gene (intronic SNP leukaemia.15 rs2297439; P=2.82×10−7) that is independent of Leucocyte TL is influenced by inherited genetic http://jmg.bmj.com/ previously reported TL-associated SNPs in this region. factors.16–19 Twin studies report TL heritability esti- The minor allele for rs2297439 is common in South mates as high as 80%,19 while other family studies esti- Asian populations (≥0.25) but at lower frequencies in mate a heritability ranging between 34% and 50%.18 other populations (eg, 0.07 in Northern Europeans). Genome-wide association studies (GWAS) have Among the eight other previously reported association shown that leucocyte TL is influenced by inherited signals, all were directionally consistent with our study, genetic polymorphisms. These studies have identified loci harbouring common genetic variants associated but only rs8105767 (ZNF208) was nominally significant on September 30, 2021 by guest. Protected copyright. (P=0.003). SNP-based heritability estimates were as high with leucocyte TL, including: ACYP2 (2p16.2), TERC as 44% when analysing close relatives but much lower (3p26), TERT (5p15.33), NAF1 (10q24.33), OBFC1 when analysing distant relatives only. (10q24.33), DCAF4 (14q24.2), CTC1 (17p13.1), Conclusions In this first GWAS of TL in a South Asian ZNF208 (19p12), ZNF676 (19p12) and RTEL1 population, we replicate some, but not all, of the loci (20q13.3).16 20–29 These variants have been discovered reported in prior GWAS of individuals of European and replicated in studies of individuals of primarily ancestry, and we identify a novel second association European ancestry. Considering the strong evidence signal at the RTEL1 locus. of population differences in TL, which may be due in part to polygenic adaptation,30 genetic studies of TL in non-European populations are warranted.31 In this study, we undertook S of TL among Bangladeshi indi- INTRODUCTION viduals to identify new TL-associated variants and Telomeres are DNA–protein complexes that protect confirm the effects of loci reported in prior studies. chromosome ends.1 In differentiated cells, the To cite: Delgado DA, progressive attrition of the DNA component of Zhang C, Chen LS, et al. telomeres, a tandem repeat sequence (TTAGGG ), EXPERIMENtaL PROCEDURES J Med Genet Published n Online First: [please include with each round of mitotic division ultimately leads Study participants Day Month Year]. to loss of telomere protection, triggering cell senes- The Health Effects of Arsenic Longitudinal Study doi:10.1136/ cence or apoptosis.1 2 In germ cells, stem cells and (HEALS) is a prospective cohort study in Araihazar, jmedgenet-2017-104922 most cancer cells, synthesis and maintenance of Bangladesh, designed to assess the health effects of Delgado DA, et al. J Med Genet 2017;0:1–8. doi:10.1136/jmedgenet-2017-104922 1 Copyright Article author (or their employer) 2017. Produced by BMJ Publishing Group Ltd under licence. Telomere biology J Med Genet: first published as 10.1136/jmedgenet-2017-104922 on 18 November 2017. Downloaded from arsenic exposure through drinking water. Details of the HEALS (Qiagen) and 25 ng of template DNA. Both plate designs used the have been described previously.32 In brief, 11 746 participants of following primer sequences: 5′ GGT TTTT GAGG GTGA GGGT ages 18–75 years old were recruited between October 2000 and GAGG GTGA GGGT GAGGGT3′ (forward), 5′ TCC CGAC TATC May 2002. An expansion of the HEALS cohort (ACE) by 8287 CCTA TCCC TATC CCTA TCCC TATCCCTA3′ (reverse) for participants occurred between 2006 and 2008. The Bangla- telomere sequence amplification and 5′ CAG CAAG TGGGAAGG desh Vitamin E and Selenium Trial (BEST) is a 2×2 factorial, TGTAATCC3′ (forward), 5′ CCC ATTC TATC ATCA ACGG double-blind, randomised chemoprevention trial designed to GTACAA3′ (reverse) for 36B4 sequence amplification. assess the effect of vitamin E and selenium supplementation on For plate design 1, we used a protocol previously described non-melanoma skin cancer risk among 7000 adults. BEST by Ehrlenbach et al.36 The qPCR reaction was performed under participants were recruited from Araihazar, Matlab and other the following conditions: 95°C for 15 min to activate Taq poly- surrounding districts of Bangladesh. Details of the BEST study merase, 30 cycles at 95°C for 15 s, 54°C for 2 min and 72°C for methods have been described previously.33 1 min, followed by 72°C for 4 min. For plate design 2, we used a protocol previously described by Cawthon.37 The telomere reac- tion used the following protocol: 95°C for 15 min to activate DNA extraction and genotyping Taq polymerase, 25 cycles at 95°C for 15 s and 54°C for 2 min, Blood collection, DNA extraction, genotyping and quality while the RPLP0 (reference) reaction ran at 95°C for 15 min to control have been described previously.34 35 To summarise, in activate Taq polymerase, 30 cycles at 95°C for 15 mins and 58°C each study, study physicians collected blood from all participants for 1 min. at baseline. Genomic DNA for HEALS samples was extracted Threshold cycle (C ), is the threshold that represents a level from clot blood using the Flexigene DNA kit (Cat #51204) from t above background (noise) fluorescence and is the number of Qiagen (Valencia, California, USA). For BEST samples, DNA was cycles necessary to detect a real signal from the sample. C values extracted from whole blood using the QIAamp 96 DNA Blood t were averaged across replicates for each sample. The relative Kit (Cat #51161) from Qiagen. A total of 5499 DNA samples TL was calculated as a ratio of telomere repeat copy number to were genotyped on HumanCytoSNP-12 v2.1 chips with 299 140 single-gene (RPLP0) copy number, that is, T/S ratio (telomere to SNPs measured. Genotyping quality control was performed single gene reference), according to the following equation: using PLINK. Individuals with call rates of <97% (n=13), 2Cttelomere (Reference)/2Cttelomere (Sample) gender mismatches (n=79) and technical replicate samples or T/S = 2CtRPLP0 (Reference)/2CtRPLP0 (Sample) (1) duplicates (n=53) were removed. SNPs that had low call rates, that is (<95%) (n=20), were monomorphic (n=39 798), did The coefficients of variation (CV) were calculated for each not have rs identifiers (n=941) and had hardy-weinberg equilib- set of replicates (one for each study sample triplicate and one rium (HWE) P values of <10−10 (n=634), were excluded. This for each reference sample sextuplicate). The CV was calculated quality control resulted in 5354 individuals and 257 747 geno- as the SD of the sample replicates divided by the mean of the typed SNPs. We generated imputed genotypes using the 1000 replicates: 2 Genomes Project reference panel. We excluded SNPs that had an ′ 2 −10 x x imputation r <0.3, were triallelic, had an HWE of P<10 and − ′ CV=100 √ ( ) x had minor allele frequency (MAF) ≤0.01. HWE testing in an × ∑ n ÷ (2) unrelated set of participants confirms that no SNPs used in the To confirm reproducibility of the measured T/S ratio for plate analysis have an HWE P value of <10−7.
Details
-
File Typepdf
-
Upload Time-
-
Content LanguagesEnglish
-
Upload UserAnonymous/Not logged-in
-
File Pages8 Page
-
File Size-