A variant of the myosin light chain kinase gene is associated with severe asthma in African Americans

Asthma is a complex phenotype influenced by environmental and genetic factors for which severe irreversible structural airway alterations are more frequently observed in African Americans. In addition to a multitude of factors contributing to its
of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Genetic Epidemiology 31: 296–305 (2007) A Variant of the Myosin Light Chain Kinase Gene is Associatedwith Severe Asthma in African Americans Carlos Flores, 1 y Shwu-Fan Ma, 1 y Karen Maresso, 1 Carole Ober, 2 and Joe G. N. Garcia 1  1 Department of Medicine, University of Chicago, Chicago, Illinois 2 Department of Human Genetics, University of Chicago, Chicago, Illinois Asthma is a complex phenotype influenced by environmental and genetic factors for which severe irreversible structuralairway alterations are more frequently observed in African Americans. In addition to a multitude of factors contributing toits pathobiology, increased amounts of myosin light chain kinase (MLCK), the central regulator of cellular contraction, have been found in airway smooth muscle from asthmatics. The gene encoding MLCK (  MYLK  ) is located in 3q21.1, a regionnoted by a number of genome-wide studies to show linkage with asthma and asthma-related phenotypes. We studied 17  MYLK   genetic variants in European and African Americans with asthma and severe asthma and identified a single non-synonymous polymorphism (Pro147Ser) that was almost entirely restricted to African populations and which wasassociated with severe asthma in African Americans. These results remained highly significant after adjusting forproportions of ancestry estimated using 30 unlinked microsatellites (adjusted odds ratio: 1.76 [95% confidence interval, CI:1.17–2.65],  p 5 0.005). Since all common HapMap polymorphisms in   500kb contiguous regions have low-to-moderatelinkage disequilibrium with Pro147Ser, we speculate that this polymorphism is causally related to the severe asthmaphenotype in African Americans. The association of this polymorphism, located in the N-terminal region of the non-muscleMLCK isoform, emphasizes the potential importance of the vascular endothelium, a tissue in which MLCK is centrallyinvolved in multiple aspects of the inflammatory response, in the pathogenesis of severe asthma. This finding also offersa possible genetic explanation for some of the more severe asthma phenotype observed in African American asthmatics. Genet. Epidemiol.  31:296–305, 2007.  r 2007 Wiley-Liss, Inc. Key words: MLCK; SNP; population stratification; endothelium; cytoskeleton  Abbreviations : ADAM33, A disintegrin and metalloprotease domain 33 gene; ADRB2, Beta-2 adrenergic receptor gene; ALOX5,Arachidonate 5-lipoxygenase gene; BHR, Bronchial hyper-responsiveness; bp, Base pairs; Ca2 1 , Calcium; CCDC14, Coiled-coil domaincontaining 14 gene; CD14, CD14 antigen gene; CSGA, Collaborative study on the genetics of asthma; CTLA4, Cytotoxic T-lymphocyte-associated protein 4 gene; FEV 1 , Forced expiratory volume in one second; HWE, Hardy–Weinberg equilibrium; IgE, Immunoglobulin E;IgG, Immunoglobulin G; IL4, Interleukin 4 gene; IL12B, Interleukin 12B gene; kb, Kilobases; kDa, Kilodalton; MAF, Minor allelefrequency; Mb, Megabases; MLCs, Myosin light chains; NCBI, National Center for Biotechnology Information; ng, Nanograms;nmMLCK, Non-muscle myosin light chain kinase; Pro147Ser, Proline to serine change at amino acid position 147; PTPLB, Proteintyrosine phosphatase-like member b gene; RIPK2, Receptor-interacting serine-threonine kinase 2 gene; ROPN1, Ropporin, rhophilinassociated protein 1 gene; smMLCK, Smooth muscle myosin light chain kinase; TGFB1, Transforming growth factor beta 1 gene.Contract grant numbers: RO1 HL72414, VO1 HL49596 and HL58064. y Carlos Flores and Shwu-Fan Ma contributed equally to this work.  Correspondence to: Joe G. N. Garcia, Section of Pulmonary and Critical Care Medicine, Department of Medicine, University of Chicago,Chicago. E-mail: jgarcia@medicine.bsd.uchicago.eduReceived 12 December 2006; Accepted 31 December 2006Published online 31 January 2007 in Wiley InterScience ( 10.1002/gepi.20210 INTRODUCTION Asthma is a common complex inflammatorycondition affecting 7% of the US population[Mannino et al., 2002] and associated with bronchial hyper-responsiveness (BHR), airflowobstruction and atopy (IgE-mediated sensitiza-tion). Atopy and asthma are closely related entitieswith their individual influence on clinical pheno-typeexpression oftendifficulttodistinguish[Weiss,1996]. In fact, familial and individual clustering of asthma and atopy, total serum IgE levels and thecorrelation of the positive skin prick tests with thedevelopment of asthma symptoms, suggest causaloverlap for these conditions [Lester et al., 2001]. Inthe most severe form, asthma is associated with r 2007 Wiley-Liss, Inc.  irreversible structural alterations in the bronchialwall, resulting from abnormal repair and airwayremodeling, characterized by smooth muscle hy-pertrophy and hyperplasia [Moore and Peters,2006]. Functional alterations in airway smoothmuscle of asthmatics have been associated withincreased amounts of cytoskeletal proteins involvedin the contractile machinery, including the smoothmuscle and non-muscle myosin light chain kinase(MLCK) [Benayoun et al., 2003].MLCK is a Ca 2 1 /calmodulin-dependent kinasewith a key role in contractile activities of non-muscle (endothelium) tissue [Garcia et al., 1995]and smooth muscle tissues via phosphorylation of myosin light chains (MLCs), which promotesmyosin interaction with cytoskeletal actin fila-ments [Garcia et al., 1997a]. In humans, the MLCKis encoded by the  MYLK   gene spanning 217kb on3q21.1 (two new non-coding exons have beenadded to the 5 0 end of the gene in the NCBI build35, extending the gene to 272kb). At least eighttranscripts are transcribed by  MYLK   resulting inthree types of proteins: a   210kDa non-muscleMLCK (nmMLCK) initially isolated from humanendothelium [Garcia et al., 1997a; Lazar andGarcia, 1999], a shorter   108kDa smooth muscleMLCK (smMLCK,) preferentially expressed insmooth muscle tissues [Garcia et al., 1997a; Lazarand Garcia, 1999], and a   19kDa telokin for actinfilament stabilization. Both smMLCK andnmMLCK isoforms exhibit MLC kinase activityand robust phosphorylation of MLCs. A majordistinguishing feature of the two isoforms is thatnmMLCK contains an additional 922-amino-acidstretch at the N-terminus not present in smMLCK[Garcia et al., 1997a], providing distinct cellularfunctions through unique interactions withother contractile proteins [Garcia et al., 1997b] assuggested by studies utilizing the nmMLCKknockout mice [Wainwright et al., 2003]. Ourlab has demonstrated that the nmMLCK isoformis a key participant in the inflammatory response by virtue of its ability to regulate vascularintegrity (through the interplay of cell contractileforces and the cell-cell/cell-matrix contacts) andleukocyte influx into the lung tissue [Garcia et al.,1998]. The nmMLCK is activated by differentpro-inflammatory stimuli leading to an increase invascular leakage and edema formation, as wellas leukocyte recruitment to airways, all of themcharacteristic features of asthma contributing toairway obstruction and BHR [Lee et al., 2006].Genome-wide linkage screens (for a recentreview see Wills-Karp and Ewart [2004]) haverepeatedly identified over a dozen genomicregions conferring susceptibility to asthma and/or related phenotypes. However, due to thediverse location of linkage signals among studies,these results also evidence the complex etiologyof the disease [Blumenthal et al., 2004]. In fact,candidate gene association studies have also failedto show a single gene contributing to risk in allpopulations studied [Ober and Hoffjan, 2006].Interestingly, several studies have reported link-age to asthma or related phenotypes in chromo-some 3q [Dizier et al., 2000; Lee et al., 2000;Haagerup et al., 2002, 2004; Koppelman et al.,2002; Blumenthal et al., 2004; Kurz et al., 2005;Brasch-Andersen et al., 2006; Pillai et al., 2006],with some of the linkage signals confined to 3q21,approximately 5Mb from the  MYLK   gene [Dizieret al., 2000; Lee et al., 2000; Blumenthal et al., 2004;Kurz et al., 2005].We recently re-sequenced the entire  MYLK   codingregion (including at least 200bp of exon-intron boundaries) as well as 2kb of both 5 0 and 3 0 flankingregions in European and African Americans andidentified more than 50  MYLK   polymorphisms.After exploring the linkage disequilibrium (LD)patterns, and testing the association of   MYLK  variants with susceptibility to acute lung injury(ALI), a common devastating illness characterized by a profound inflammation and alveolar flooding,we identified several variants associated with ALI[Gao et al., 2006]. On the basis of functional[Benayoun et al., 2003] and positional [Dizier et al.,2000; Lee et al., 2000; Blumenthal et al., 2004; Kurzet al., 2005] evidence, we examined  MYLK   as acandidate gene for susceptibility to asthma. Weevaluated the association of 17  MYLK   variants intwo ethnically diverse populations (European andAfrican Americans) collected as part of the Colla- borativeStudyontheGeneticsofAsthma(CSGA)inChicago [CSGA, 1997]. We report here an associa-tion of a non-synonymous (Pro147Ser) singlenucleotide polymorphism (SNP) with susceptibilityto severe asthma in African Americans. MATERIALS AND METHODS STUDY POPULATIONS AND CLINICALEVALUATION A total of 302 unrelated asthma cases (110European Americans and 192 African Americans)were recruited in the Chicago area as part of theCSGA protocol [CSGA, 1997]. A brief descriptionof asthma and atopy diagnoses in these samples Genet. Epidemiol.  DOI 10.1002/gepi 297  MYLK and Severe Asthma in African Americans  has been previously reported [Ober et al., 2000;Lester et al., 2001]. Controls (209 EuropeanAmericans and 193 African Americans) consistedof adult individuals with negative personal andfirst-degree relative family history for asthma.Cases and controls reported at least three grand-parents who were either of European or AfricanAmerican ancestry. Twenty seven EuropeanAmericans and 103 African Americans asthmaticswere diagnosed with severe asthma on the basisof night time symptoms, prescribed use of eitherinhaled or oral steroids, a forced expiratoryvolume measurements (FEV 1 )  o 60% predictedvalue at any time and bronchodilator reversibility( Z 15% increase in baseline FEV 1  after bronchodi-lator use). This study was approved by theInstitutional Review Board. Written informedconsent was obtained from all participants. SNP SELECTION AND GENOTYPING To study the variation in  MYLK   gene, wesequenced the coding and potential regulatoryregions in 36 individuals from two different ethnicgroups (18 European and African Americans,respectively) [Gao et al., 2006]. ComplementarySNPs available from Assays-on-Demand (AppliedBiosystems, Foster City, CA) were selected to fillin intronic regions and reach a density of one SNPper   8kb (25 and 24 for European and AfricanAmericans, respectively). This project precededthe publicly available data from HapMap [TheInternational HapMap Consortium, 2005]. Toreduce the redundancy of the initial SNP genotyp-ing set, we extracted three random sub-samples of 50 control individuals from the genotype dataset(independently for European and African Amer-icans) and used them to select, through anexhaustive search, the smallest possible subset of informative SNPs satisfying a minimum perfor-mance (haplotype  r 2 >0.90) using TagIT 3.03 soft-ware [Weale et al., 2003]. A SNP-dropping-with-re-sampling method [Ahmadi et al., 2005] wasused to further evaluate the expected properties of this new SNP set. This newly selected SNP set wassufficient to capture SNPs not genotyped from theinitial set with  r 2 >0.85 in any of the three samplesubsets. A total of 17 SNPs (13 and 11 SNPs forEuropean and African Americans, respectively)were used for genotyping. Genotyping wasperformed by TaqMan TM allelic discriminationassays on a 7900HT Fast Real-Time PCR System(Applied Biosystems, Foster City, CA) using  10ng of whole genome amplified DNA (GEHealthcare, UK). Genotyping was blind to caseand control status of samples. Approximately 10%of the samples were genotyped in duplicate tomonitor genotyping quality. STATISTICAL ANALYSIS Departures from Hardy–Weinberg equilibrium(HWE) were tested by means of the Pearson’sgoodness-of-fit  w 2 test. To adjust for populationstratification, individual ancestries (European,African and Asian) were estimated using Struc-ture 2.1 [Pritchard et al., 2000] based on 30unlinked ancestry-informative microsatellites aspreviously described [Amundadottir et al., 2006].SNP associations were tested by means of anadditive model using the Armitage trend test. Aconservative correction for multiple testing basedon the method described by Nyholt [2004] wasadopted; the adjusted  a  level was obtained bydividing  a 5 0.05 by the effective number of independent tests, which were estimated basedon pairwise LD between genotyped SNPs for bothpopulations independently. We did not correct the p -values for the two conditions tested because of the non-independent nature of both phenotypes inthe samples. SPSS 14.0 (SPSS, Inc., Chicago, IL)was used for multiple logistic regression analysis,including use of individual ancestries as covari-ates, to estimate the odds ratio (OR) with therespective 95% confidence interval (CI). To testhaplotype associations and adjust for multipletesting, the reconstructed haplotypes from the bestaverage goodness-of-fit output of five PHASE 2.1runs [Stephens et al., 2001; Stephens and Donnelly,2003] was used for the multilocus test implemen-ted by HapVLMC with 1,000 permutations[Browning, 2006]. Briefly, this analysis first buildsa tree graph model of haplotype clusters, bygrouping haplotypes with low counts and adapt-ing to the underlying LD between SNPs, so thateach haplotype defines a path in the graph between the initial and the terminal SNP. Then,the haplotype clusters are tested for association bya sliding window in the graph from the initial tothe terminal SNP. The patterns of LD wereexplored using Haploview 3.32 [Barrett et al.,2005] and the haplotype blocks were defined asdescribed in Gabriel et al. [2002]. RESULTS Ancestry estimates for all individuals, based on30 microsatellite multilocus genotypes, demon- 298 Flores et al. Genet. Epidemiol.  DOI 10.1002/gepi  strated that African or Asian admixture in Eur-opean American samples was low (98.7 7 8.7%European, 1.1 7 8.6% African, and 0.2 7 1.5%Asian). In contrast, substantial European admix-ture was estimated for African Americans(24.1 7 12.1% European, 72.7 7 12.2% African, and3.2 7 3.1% Asian). However, these are in the rangeof previous estimates of European (18–26%) orAsian (  2.6%, if Native Americans are consideredas a proxy for Asians) contributions to AfricanAmericans from Chicago [Destro-Bisol et al., 1999;Smith et al., 2004; Lind et al., 2007]. In line withthese results, the European Americans showedmore homogeneous individual European ances-tries (96% of individuals had more than 0.95European ancestry) than the African Americans,in which only 66% of individuals had more than0.70 of African ancestry (Fig. 1), consistent with theresults reported by Smith et al. [2004]. Thisexploration revealed a few outlier individuals(Fig. 1) with extremely high (>0.75) Africanancestries in the European Americans (Fig. 1A;three controls) or extremely high (>0.75) Europeanancestries in the African Americans (Fig. 1B; twocases). These five individuals were excluded fromour analysis.None of the  MYLK   SNPs showed significantdepartures from HWE (data not shown). Minorallele frequencies by population, SNP information,and their  p -values of association with asthma andsevere asthma are shown in Table I. In EuropeanAmericans, significant associations were foundfor three of the 13 SNPs genotyped in the  MYLK  gene ( p o 0.05): rs4678047 with asthma; rs34261801and rs4678062 with severe asthma. In AfricanAmericans, only 1 (rs9840993) of the 11 SNPsgenotyped showed a significant association withsevere asthma ( p 5 0.004). However, based on theeffective number of independent tests (10 and 9 inEuropean and African Americans, respectively),only the rs9840993 C/T SNP (predicting a Pro147-Ser change), associated with severe asthma inAfrican Americans, remained significant afteradjustment for multiple tests (adjusted  p 5 0.037).Although the estimated individual ancestrieswere not significantly different between AfricanAmerican controls and severe asthmatics( p >0.120 for all three ancestries), these wereincluded as covariates in a multiple logisticregression model to test the independent associa-tion of rs9840993 (additive model) with severeasthma. This confirmed that the ancestral (bycomparison to the chimp sequence) allele C(147Pro) at rs9840993 was a risk factor for severeasthma in African Americans even after adjust-ment for proportions of ancestry (OR: 1.76, 95%CI: Fig. 1. Proportion of individual ancestries estimated on the basis of 30 microsatellites. Histograms represent European ancestryproportions for European Americans (A), and African ancestry proportions for African Americans (B). Arrows indicate the outlier individuals that were not considered for the analysis. Numbers in  x -axes denote the upper bound for each interval. 299 MYLK and Severe Asthma in African Americans Genet. Epidemiol.  DOI 10.1002/gepi  TABLE I. Genotyped SNPs, MAFs and unadjusted  P  -values for association with asthma and severe asthma in the two populations European Americans African AmericansMAF   P -value MAF   P -valueSNP a Location  b (effect) rs ] Position(Build 35)Controls( N  5 206) c Asthma( N  5 110)Severe( N  5 27) CA d CS e Controls( N  5 193)Asthma( N  5 190) c Severe( N  5 102) c CA d CS e MYLK _ 037 Intron 31 rs860224 124820104 0.19 0.23 0.30 0.476 0.239 0.24 0.28 0.27 0.260 0.506AOD3 Intron 28 rs820447 124830869 0.18 0.23 0.28 0.154 0.093 — — — — —AOD9 Intron 19 rs820325 124868367 0.14 0.11 0.09 0.449 0.333 0.27 0.23 0.23 0.262 0.413AOD12 Intron 17 rs33264 124892478 0.21 0.23 0.22 0.643 0.901 — — — — —MYLK _ 025 Intron 17 — 124893795 0.10 0.12 0.09 0.535 0.790 — — — — —AOD15 Intron 16 rs820336 124898471 0.21 0.22 0.21 0.717 0.997 0.20 0.18 0.21 0.522 0.912MYLK _ 004 Exon 8(Thr335Thr)rs4678047 124935528 0.38 0.28 0.30  0.041  0.331 — — — — —MYLK_003 Exon 8(Val261Ala)rs3796164 124935751 — — — — — 0.42 0.41 0.49 0.797 0.071MYLK_002 Exon 5(Ser147Pro)rs9840993 124940583 — — — — — 0.31 0.36 0.43 0.121  0.004 AOD20 Intron 3 rs11718105 124946398 0.38 0.44 0.48 0.256 0.076 0.08 0.06 0.05 0.196 0.170AOD22 Intron 2 rs34261801 124962475 0.46 0.47 0.35 0.179  0.019  0.08 0.10 0.11 0.352 0.237AOD23 Intron 2 rs1869863 124980346 — — — — — 0.24 0.23 0.29 0.791 0.191AOD24 Intron 2 rs11707609 124986114 — — — — — 0.09 0.10 0.10 0.649 0.956MYLK_007 Exon 2(His21Pro)rs28497577 124995317 0.08 0.06 0.08 0.341 0.983 0.36 0.36 0.37 0.960 0.933AOD27 Intron 1 rs4678062 125011798 0.24 0.29 0.38 0.316  0.038  — — — — —AOD29 5 0 locusregionrs36025624 125024094 0.09 0.11 0.10 0.603 0.830 0.08 0.08 0.09 0.846 0.793MYLK_016 5 0 locusregionrs2682211 125034828 0.19 0.17 0.25 0.722 0.287 — — — — — a Designation as in Gao et al. [2006].  b Based on the Garcia et al. [1997a,b] description of the gene. c Counts and MAF excluding the individuals with outlier inferred ancestries as revealed from 30 unlinked microsatellite loci. d Control vs. asthma comparison. e Control vs. severe asthma comparison. Statistically significant  P -values in bold.SNP, single nucleotide polymorphism; MAF, minor allele frequency.  3   0   0   F   l     o r  e  s  e  t    a l     .  G e  n e  t   . E   p i    d   e  m i    o  l    .  D O I    1   0   . 1   0   0   2   /   g e  p i   
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks