Common Variation in the WNK1 Gene and Blood Pressure in Childhood
The Avon Longitudinal Study of Parents and Children
WNK1 gene variants have been associated with adult blood pressure. We aimed to investigate relationships between WNK1 variants and blood pressure, as well as blood pressure change with age, in a longitudinal childhood study. Associations between single nucleotide polymorphisms in WNK1 and blood pressure and the rate of blood pressure change between 7 and 11 years were examined in the Avon Longitudinal Study of Parent and Children Study (n=5326 for systolic blood pressure at 11 years). We observed associations (P<0.05) with diastolic blood pressure gradient with age for 33 of 82 typed and imputed polymorphisms, including polymorphisms in exons 4, 10, and 11 (rs10774466, rs1012729, and rs9804992). The minor allele (G) of rs1012729 (frequency: 25.6%) was associated with a gender-adjusted change in a diastolic blood pressure gradient of −0.11 mm Hg/y (95% CI: −0.20 to −0.03 mm Hg/y; P=0.0054). No associations were shown with the systolic blood pressure gradient. At age 11 years, 30 polymorphisms showed association (P<0.05) with systolic blood pressure, including variants in exons 4 and 10 (rs10774466 and rs1012729). Only 3 polymorphisms were associated with diastolic blood pressure at 11 years. In exploration of polymorphism-dietary cation interactions on systolic blood pressure at 11 years, 59 reached significance (P<0.05; 12.3 expected by chance), mostly (n=33) related to dietary calcium. The findings show that common intronic and exonic WNK1 variants are associated with diastolic blood pressure gradient from 7 to 11 years and with systolic blood pressure at 11 years. Our study suggests that previously reported effects of WNK1 variants on blood pressure are mediated via effects on the gradient of blood pressure change with age.
Blood pressure (BP) is a major determinant of cardiovascular health, and modest changes in BP impact substantially on the risk of stroke and coronary artery disease.1 Improved understanding of mechanisms of BP regulation should facilitate advances in the prevention and treatment of hypertension. Because BP is a heritable trait,2 efforts are underway to identify genetic variants that have a role in BP regulation.3
Genes underlying rare monogenic forms of hypertension provide obvious candidates for genetic association studies of BP.4 Mutations in the WNK1 gene, which codes for a WNK (“with no lysine” [K]) protein, cause pseudohypoaldosteronism type II (Gordon’s syndrome), an autosomal dominant condition characterized by hypertension and hyperkaliemia.5 Evidence from 2 studies implicated common WNK1 gene variants in the regulation of BP in adults.6,7 WNK1 has an integral role in pathways regulating renal tubular sodium transport, and a third study showed that WNK1 variants were associated with BP response to thiazide diuretics.8
Only one of the above studies was undertaken in a general population sample, and none were undertaken in children. Our aim was to quantify the effect of common WNK1 variants on baseline BP and on the rate of change of BP with age in a large population-based sample of children: the Avon Longitudinal Study of Parent and Children (ALSPAC). Our secondary aim was to explore possible interactions between dietary sodium, calcium, and potassium and WNK1 gene variants.
Subjects and Phenotyping
The ALSPAC is a prospective cohort study investigating the health and development of children. Details of methods have been published previously.9 Briefly, 14 541 pregnant women who were registered in 3 Bristol-based health districts with an expected date of delivery between April 1991 and December 1992 were enrolled on entry to antenatal care. Children were invited to attend for annual clinic-based assessments between the ages of 7 and 11 years. At the ages of 7, 9, and 11 years, clinic measures included height, weight, and BP. BP was measured with a Dinamap 9301 Vital Signs Monitor, and the mean of 2 readings was taken.10
Dietary sodium, calcium, potassium, and energy intake11 were derived from a 3-day unweighed food record at age 10 years, checked by a nutritionist in the clinic, with the child and usually a parent. In previous work using this method of dietary assessment, in a 10% subsample of this cohort, results have been shown to compare well with those of national surveys.12 Sodium was assessed from food sources only, with vegetables coded as cooked in unsalted water and no account taken of salt added at the table.
The data set included only white singleton children of European descent, and, where the data set included siblings born to the same mother, only the firstborn was included. Parents gave written consent for children in this study. Ethical approval for the study was obtained from the ALSPAC Law and Ethics Committee and the local research ethics committees.
Genotyping and Quality Assurance
Choice of Single Nucleotide Polymorphisms
We aimed to tag common single nucleotide polymorphisms ([SNPs]; minor allele frequency: ≥10% in the HapMap CEU population) in WNK1 and a 5-kb region 5′ to WNK1 (chromosome 12 region 727762 to 888218, HapMap release #20). Tag SNPs were selected using an aggressive (2- and 3-marker haplotype) algorithm used in Tagger,13 specifying r2=0.8 (between a tag SNP and a putative untyped SNP). We forced in rs1468326, rs765250, rs2301880, and rs2277869 to enhance comparability of our findings with earlier reports.6–8 The 13 SNPs we selected showed mean r2=0.963 and minimum r2=0.812 (between tag and untyped SNPs).
DNA Extraction and Genotyping Method
DNA, from cord blood or peripheral blood, was extracted and processed as described previously.14 All of the genotyping was performed by KBioscience (http://www.kbioscience.co.uk). SNPs were genotyped using the KASPar chemistry, which is a competitive allele-specific PCR SNP genotyping system using FRET quencher cassette oligos (http://www.kbioscience.co.uk/genotyping/genotyping_chemistry.htm). Blind duplicates and Hardy-Weinberg equilibrium tests were used as quality control tests.
Genotyping Quality Control and Imputation of Missing Genotypes
Eleven of the 13 SNPs were successfully genotyped (rs11064546 was monomorphic, and the assay for rs11064519 failed). The genotype error rate was assessed from nonmissing genotype calls by repeating the genotyping of 226 samples per SNP. Over the 11 SNPs, the number of genotype comparisons where both samples in the pair had a called genotype was 2322, and 14 of these comparisons showed different genotypes (an error rate of 0.6%).
We used information from the 11 successfully genotyped SNPs and from the pattern of haplotype variation in the HapMap CEU samples to impute missing genotypes. By using extra information from a suitable reference population, such as HapMap, imputation improves the power to detect associations that are attributable to variants that have not themselves been typed.15 We undertook these analyses using Markov Chain Haplotyping Software (MaCH 1.0: http://www.sph.umich.edu/csg/abecasis/MaCH/).16,17 Imputation, using this program16–19 and a related approach,20,21 has found use in recent genetic association studies, because it greatly eases the comparison of findings between studies that have genotyped correlated but nonidentical SNPs.
Our association tests were based on SNPs that were genotyped or that could be imputed with relatively high confidence. We assessed the latter using a quality score defined by MaCH 1.0 as the predicted r2 between true and imputed genotypes.17 We refer to this quality measure as r2_QC to distinguish it from r2 representing the extent of linkage disequilibrium between 2 genotyped SNPs.
We excluded from association tests SNPs with an r2_QC value of ≤0.30 in any imputation run. We also rejected any SNPs with marked departure from Hardy-Weinberg equilibrium (P<0.001). After imputation, data were available on 97 SNPs in the region of interest (11 genotyped and 86 imputed). We excluded 15 SNPs (all imputed) because of low r2 values and/or departure from Hardy-Weinberg equilibrium (Table S1, available online at http://hyper.ahajournals.org). Association tests were, therefore, based on 82 SNPs (11 genotyped and 71 imputed).
To comprehensively model the longitudinal BP data, we used a generalized linear mixed-modeling variance components approach, implemented using Gibbs sampling in WinBUGS.22 This approach not only takes appropriate account of the covariance between longitudinal BP measures within individuals, but it also allows us to model the effects of covariates of interest (including a given SNP) on the baseline BP and also on the gradient of BP increase with age. Age is centered for the analysis (subtracting mean age) so that baseline BP approximates to BP at age 9 years. The model (described fully in the Appendix in the online data supplement) incorporates individual-specific random effects for baseline BP and for the gradient of BP increase with age. It represents a simplification of our previously developed longitudinal models for family data.23
Our association tests focused on an additive genetic model without any formal correction for multiple testing given the previous evidence for BP associations with WNK1 gene variants. We did not adjust for a wide range of covariates, because genetic association studies are not susceptible to confounding by the lifestyle and phenotypic characteristics that typically confound nongenetic observational studies.24
We also undertook cross-sectional analyses at age 11 years. This enabled us to model the effect of the SNP on BP at age 11 years and also to examine the interaction between an SNP of interest and tertiles of dietary sodium, calcium, or potassium, because the latter were measured at 10 years of age. These cross-sectional analyses were undertaken using Stata (release 9.2, Stata Corp), incorporating gender as a covariate. To test for interaction, we tested for departure from additivity of the effects of the minor allele on BP across tertiles of the relevant cation. These interaction tests were interpreted as exploratory (see secondary aim above).
Systolic and diastolic BPs gradually increased between 7 and 11 years of age in boys and in girls (Table 1). At 11 years of age, both systolic and diastolic BPs were higher in girls than in boys (P<0.0001 for both tests). We included in our longitudinal analyses those individuals with BP measures at all 3 of the ages (7, 9, and 11 years), ie, 4495 individuals for systolic BP and 4494 individuals for diastolic BP. Dietary cation intake was recorded at age 10 years in 4940 of these children, with mean values (SDs) as follows: calcium, 798 mg (293 mg); potassium, 2419 mg (620 mg); and sodium, 2585 mg (702 mg).
Association tests were based on 82 SNPs (11 genotyped and 71 imputed). A pattern of strong linkage disequilibrium was observed between the SNPs, as shown in the Figure, which illustrates the SNP positions, and their relationship to the linkage disequilibrium structure in HapMap. The rs numbers, genomic position, allele names, minor allele frequencies, r2_QC values for imputation, and tests of deviation from Hardy-Weinberg equilibrium are included in Tables 1 and 2⇓ and Tables S2 through S4. The r2_QC values of the imputed SNPs included in the analysis ranged from 0.35 to 0.78 (median: 0.66).
Association Tests: Longitudinal Analyses
WNK1 SNPs showed associations with the rate of change of diastolic BP with increasing age (Table 2) reaching nominal significance (P<0.05) in 4 of the 11 genotyped SNPs (Table 2) and in 33 of the 82 SNPs in the complete set of typed and imputed SNPs (Table S2). No associations were shown between any WNK1 SNPs and baseline systolic or diastolic BP or with the rate of change of systolic BP with increasing age (Table 2 and Table S2). Recall that age was centered for the analysis such that baseline BP approximates to BP at age 9 years.
The strongest associations with the rate of change of diastolic BP were in imputed SNPs (Table S2). The gender-adjusted gradient in DBP was 0.64 mm Hg per year of increase in age (SD: 0.60 mm Hg). After adjustment for gender, each copy of the minor allele (G) of intronic SNP rs4388955 (frequency: 25.8%) was associated with a change in the gradient of diastolic BP of −0.12 mm Hg/y (95% CI: −0.20 to −0.04 mm Hg/y; P=0.0038), and each copy of the minor allele (G) of the exon 10 SNP, rs1012729 (frequency: 25.6%), was associated with a change in the gradient of diastolic BP of −0.11 mm Hg/y (95% CI: −0.20 to −0.03 mm Hg/y; P=0.0054). The effect sizes of minor alleles at each of these SNPs equate to one-sixth of a SD of the diastolic BP gradient. In all, 3 of the SNPs associated with the gradient of diastolic BP were in coding regions rs10774466 (exon 4), rs1012729 (exon 10), and rs9804992 (exon 11).
Association Tests: Cross-Section Analyses at Age 11 Years and Interactions
Associations with systolic BP at 11 years of age reached nominal statistical significance (P<0.05) for 2 of the 11 genotyped SNPs (Table 3) and for 30 of the 82 SNPs in the complete set of genotyped and imputed SNPs (Table S3). Fewer SNPs were associated with diastolic BP at 11 years of age (1 of 11 typed SNPs and 3 of 82 SNPs in the complete set). The strongest association with systolic BP at age 11 years was for intronic SNP rs6489752; each copy of the minor allele (A, frequency: 25.1%) was associated with a gender-adjusted systolic BP difference of −0.52 mm Hg (95% CI −0.94 to −0.10 mm Hg; P=0.014; Table S3). Two of the 3 exonic SNPs were associated with systolic BP at age 11: rs10774466 (exon 4) and rs1012729 (exon 10). To assess whether the effect of WNK1 variants on systolic BP by age 11 differed substantially between males and females, we tested for evidence of interaction between sex and 5 of the above SNPs (genotyped SNPs rs2301880 and rs2277869 and imputed intronic SNP rs6489752 and imputed exonic SNPs rs10774466 and rs1012729). No interactions were found (P>0.05).
Exploratory analyses of all of the possible interactions between all of the SNPs and dietary sodium, calcium, or potassium on systolic and diastolic BP are shown in Table S4. Of the typed SNPs, for systolic BP, 8 nominally significant (P<0.05) interactions were observed (versus 1.65 expected by chance), whereas for diastolic BP, no interactions were observed. In the set of imputed and typed SNPs, for systolic BP, 59 of 246 possible interactions reached nominal significance (versus 12.3 expected by chance). For diastolic BP, no interactions were observed. Of the 59 interactions observed for systolic BP in these exploratory analyses, 33 related to dietary calcium, 19 to dietary potassium, and 7 to dietary sodium. We also repeated these tests of interaction with additional adjustment for dietary energy intake, with similar findings (results not shown).
Common WNK1 polymorphisms were associated with the rate of change of diastolic BP between ages 7 and 11 years and with systolic BP at 11 years of age in ALSPAC. Although SNPs such as rs11064536 also showed association with diastolic BP at 11 years, many more SNPs, such as rs2301880, had a stronger association with the rate of change of diastolic BP than with baseline diastolic BP, and these SNPs did not show association with diastolic BP at 11 years. WNK1 variants appear to exert a more modest effect on BP in childhood than reported previously in adulthood.6 Our findings suggest that the reported effects of WNK1 variants on BP in adulthood occur, at least in part, via effects on the rate of change of BP with increasing age. Furthermore, our exploratory analyses of interactions suggest that dietary calcium may modify the BP response to WNK1 variants. This is the first study to examine the associations between WNK1 variants and BP in childhood and also the first to show that WNK1 variants are associated with the rate of change of BP with increasing age.
We imputed untyped SNPs to augment the information available from typed SNPs. This approach is an important precursor to large-scale efforts to evaluate selected signals in additional independent samples.15 As a result, we undertook a substantial number of highly correlated statistical tests. A particularly useful statistic is the false-positive report probability (FPRP), which cannot be interpreted without taking account of multiple testing.25,26 The FPRP uses information from the power of the study and the prior probability of a true association; FPRP values <0.2 tend to be described as noteworthy.25 For the association of rs4388955 with the gradient of DBP, the FPRP was 7.9%, assuming that 1 in 20 tested SNPs would be causally associated (P<0.05) with DBP gradient (31.0%, assuming that only 1 in 100 truly associated SNPs). The compelling evidence for the involvement of WNK1 in monogenic forms of hypertension5 together with published evidence of multiple common variants in WNK1 showing association with BP6,7 and the BP response to diuretics8 would argue for a high prior probability. Thus, the associations that we report are unlikely to be attributable to chance effects but, given the nonnegligible FPRP, caution in the interpretation of these associations is required until they are replicated.
Particular strengths of the ALSPAC include a large sample size and rigorous approaches applied to minimize error. As with all genetic epidemiology studies, we expect measurement error to have impacted on BP measures, dietary measures, and genotyped and imputed SNPs. For example, an error rate of 1.5% has been reported for imputation based on the Illumina HumanHap300 BeadChip array.17 These sources of error tend to attenuate associations and interactions, so our reported effect sizes may be underestimates. Nevertheless, spurious associations can and do occur in genetic association studies (eg, because of chance or artifacts associated with genotyping or imputation errors) and replication of our findings, and comparison of our findings with earlier studies will be important. Our findings are consistent with those of previous studies. The minor alleles for rs765250 and rs2301880 (both typed SNPs), which were associated with a reduced diastolic BP gradient in ALSPAC, were associated with lower adult systolic and diastolic BPs in a previous study.6 Furthermore, rs1159744, rs2277869, and rs2107614, which also showed associations with the diastolic BP gradient in this study, have been associated with BP response to diuretics.8 It is interesting to note that we found stronger associations between WNK1 variants with the diastolic BP gradient than diastolic BP, given that WNK1 mutations do not always lead to hypertension in children and young adults.27
Previous studies also implicated SNPs that showed association with systolic BP at age 11 years in our study. These include consistent directions of estimated effects on systolic BP for rs23018806 and associations with systolic BP response to diuretics for rs2107614 and rs1159744.8 Our analyses of interactions with dietary cations were exploratory, and, to date, evidence from other genetic association studies in humans is lacking. We demonstrated more interactions than would be expected by chance, mainly relating to calcium and potassium intake. Evidence from animal models highlights the potential importance of dietary potassium28,29 on WNK1 activity, but these studies did not examine the influence of dietary calcium. Therefore, our preliminary findings would suggest the need to investigate in further studies whether either dietary calcium or potassium could modify the effects of WNK1 variants on BP.
The associations that we observed with exonic SNPs were related to exons 4, 10, and 11. WNK1, spanning 156 kb of genomic DNA, has ≥28 exons producing multiple transcripts because of alternative splicing involving exons 4, 9, 11, and 12.30 WNK1 is a very large protein with an autoinhibitory domain and coiled-coil domains that is involved with protein-protein interactions,31 including those with WNK432 in the distal nephron. A full-length form of WNK1 (L-WNK1, which is ubiquitously expressed) inhibits WNK4, and WNK4, in turn, inhibits the thiazide-sensitive Na-Cl cotransporter (NCC) in the apical membrane of epithelial cells lining the distal convoluted tubule.33 A short form of WNK1 is transcribed from an alternative fourth exon. This kidney-specific form, KS-WNK1, is identical to L-WNK1 from exon 5 but lacks the kinase domain and inhibits the action of L-WNK1.31,33 Recent work in Xenopus oocytes highlighted an interaction of KS-WNK1 with ROMK (also called Kir1.1, coded for by KCNJ1).29 Available evidence is consistent with the theory that there are multiple variants in different regions of the WNK1 gene that might plausibly influence expression, affect splicing, or alter function via coding changes. Additional work is required to develop more detailed knowledge of the relevant pathways.
The observation of genetic variants associated with differential rates of change in BP raises the question of whether there are modifiable environmental and dietary exposures that might also alter the rate of BP elevation as children develop. Additional studies could usefully investigate the effect of WNK1 variants on BP and the rate of BP change at different ages and between males and females, BP responses to thiazide diuretics, and also the potential interaction with dietary cations, especially calcium.
Common variants in the WNK1 gene contribute to BP in children. The findings add to the weight of epidemiological evidence suggesting that WNK1 exerts real, but modest, effects on BP.
We are extremely grateful to all of the families who took part in this study, the midwives for their help in recruiting them, and the whole ALSPAC team, which includes interviewers, computer and laboratory technicians, clerical workers, research scientists, volunteers, managers, receptionists, and nurses.
Sources of Funding
The United Kingdom Medical Research Council, the Wellcome Trust, and the University of Bristol provide core support for ALSPAC. This research was specifically funded by the Sir Jules Thorn Charitable Trust (reference 06SC/02A). N.J.S. holds a British Heart Foundation Chair in Cardiology. M.D.T. holds a Medical Research Council Clinician Scientist Fellowship (G0501942).
- Received June 18, 2008.
- Revision received July 23, 2008.
- Accepted August 28, 2008.
Levy D, DeStefano AL, Larson MG, O'Donnell CJ, Lifton RP, Gavras H, Cupples LA, Myers RH. Evidence for a gene influencing blood pressure on chromosome 17. Genome scan linkage results for longitudinal blood pressure phenotypes in subjects from the Framingham Heart Study. Hypertension. 2000; 36: 477–483.
Tobin MD, Sheehan NA, Samani NJ, Burton PR. The Genetic Epidemiology of Hypertension in Volume on Genetics of Hypertension, Handbook of Hypertension. Edinburgh, United Kingdom: Elsevier; 2007.
Wilson FH, Disse-Nicodeme S, Choate KA, Ishikawa K, Nelson-Williams C, Desitter I, Gunel M, Milford DV, Lipkin GW, Achard JM, Feely MP, Dussol B, Berland Y, Unwin RJ, Mayan H, Simon DB, Farfel Z, Jeunemaitre X, Lifton RP. Human hypertension caused by mutations in WNK kinases. Science. 2001; 293: 1107–1112.
Tobin MD, Raleigh SM, Newhouse S, Braund P, Bodycote C, Ogleby J, Cross D, Gracey J, Hayes S, Smith T, Ridge C, Caulfield M, Sheehan NA, Munroe PB, Burton PR, Samani NJ. Association of WNK1 gene polymorphisms and haplotypes with ambulatory blood pressure in the general population. Circulation. 2005; 112: 3423–3429.
Newhouse SJ, Wallace C, Dobson R, Mein C, Pembroke J, Farrall M, Clayton D, Brown M, Samani N, Dominiczak A, Connell JM, Webster J, Lathrop GM, Caulfield M, Munroe PB. Haplotypes of the WNK1 gene associate with blood pressure variation in a severely hypertensive population from the British Genetics of Hypertension (BRIGHT) Study. Hum Mol Genet. 2005; 14: 1805–1814.
Turner ST, Schwartz GL, Chapman AB, Boerwinkle E. WNK1 kinase polymorphism and blood pressure response to a thiazide diuretic. Hypertension. 2005; 46: 758–765.
Martin RM, Ness AR, Gunnell D, Emmett P, Davey Smith G. Does breast-feeding in infancy lower blood pressure in childhood? The Avon Longitudinal Study of Parents and Children (ALSPAC). Circulation. 2004; 109: 1259–1266.
Holland B, Welch AA, Unwin JD, Buss DH, Paul AA, Southgate DAT. In: McCance, Widdowson, eds. The Composition of Foods. London, United Kingdom: HMSO; 1991.
Scott LJ, Mohlke KL, Bonnycastle LL, Willer CJ, Li Y, Duren WL, Erdos MR, Stringham HM, Chines PS, Jackson AU, Prokunina-Olsson L, Ding CJ, Swift AJ, Narisu N, Hu T, Pruim R, Xiao R, Li XY, Conneely KN, Riebow NL, Sprau AG, Tong M, White PP, Hetrick KN, Barnhart MW, Bark CW, Goldstein JL, Watkins L, Xiang F, Saramies J, Buchanan TA, Watanabe RM, Valle TT, Kinnunen L, Abecasis GR, Pugh EW, Doheny KF, Bergman RN, Tuomilehto J, Collins FS, Boehnke M. A genome-wide association study of type 2 diabetes in Finns detects multiple susceptibility variants. Science. 2007; 316: 1341–1345.
Willer CJ, Sanna S, Jackson AU, Scuteri A, Bonnycastle LL, Clarke R, Heath SC, Timpson NJ, Najjar SS, Stringham HM, Strait J, Duren WL, Maschio A, Busonero F, Mulas A, Albai G, Swift AJ, Morken MA, Narisu N, Bennett D, Parish S, Shen H, Galan P, Meneton P, Hercberg S, Zelenika D, Chen WM, Li Y, Scott LJ, Scheet PA, Sundvall J, Watanabe RM, Nagaraja R, Ebrahim S, Lawlor DA, Ben-Shlomo Y, Davey-Smith G, Shuldiner AR, Collins R, Bergman RN, Uda M, Tuomilehto J, Cao A, Collins FS, Lakatta E, Lathrop GM, Boehnke M, Schlessinger D, Mohlke KL, Abecasis GR. Newly identified loci that influence lipid concentrations and risk of coronary artery disease. Nat Genet. 2008; 40: 161–169.
Sanna S, Jackson AU, Nagaraja R, Willer CJ, Chen WM, Bonnycastle LL, Shen H, Timpson N, Lettre G, Usala G, Chines PS, Stringham HM, Scott LJ, Dei M, Lai S, Albai G, Crisponi L, Naitza S, Doheny KF, Pugh EW, Ben-Shlomo Y, Ebrahim S, Lawlor DA, Bergman RN, Watanabe RM, Uda M, Tuomilehto J, Coresh J, Hirschhorn JN, Shuldiner AR, Schlessinger D, Collins FS, Davey Smith G, Boerwinkle E, Cao A, Boehnke M, Abecasis GR, Mohlke KL. Common variants in the GDF5-UQCC region are associated with variation in human height. Nat Genet. 2008; 40: 198–203.
Kathiresan S, Melander O, Guiducci C, Surti A, Burtt NP, Rieder MJ, Cooper GM, Roos C, Voight BF, Havulinna AS, Wahlstrand B, Hedner T, Corella D, Tai ES, Ordovas JM, Berglund G, Vartiainen E, Jousilahti P, Hedblad B, Taskinen MR, Newton-Cheh C, Salomaa V, Peltonen L, Groop L, Altshuler DM, Orho-Melander M. Six new loci associated with blood low-density lipoprotein cholesterol, high-density lipoprotein cholesterol or triglycerides in humans. Nat Genet. 2008; 40: 189–197.
Spiegelhalter D, Thomas A, Best N. WinBUGS Version 1.4 - User Manual. Cambridge, United Kingdom: MRC Biostatistics Unit; 2003.
Burton PR, Scurrah KJ, Tobin MD, Palmer LJ. Covariance components models for longitudinal family data. Int J Epidemiol. 2005; 34: 1063–1077.
Wacholder S, Chanock S, Garcia-Closas M, El Ghormli L, Rothman N. Assessing the probability that a positive report is false: an approach for molecular epidemiology studies. J Natl Cancer Inst. 2004; 96: 434–442.
Tobin MD, Tomaszewski M, Braund PS, Hajat C, Raleigh SM, Palmer TM, Caulfield M, Burton PR, Samani NJ. Common variants in genes underlying monogenic hypertension and hypotension and blood pressure in the general population. Hypertension. 2008; 51: 1658–1664.
Lazrak A, Liu Z, Huang CL. Antagonistic regulation of ROMK by long and kidney-specific WNK1 isoforms. Proc Natl Acad Sci U S A. 2006; 103: 1615–1620.
Wade JB, Fang L, Liu J, Li D, Yang CL, Subramanya AR, Maouyo D, Mason A, Ellison DH, Welling PA. WNK1 kinase isoform switch regulates renal potassium excretion. Proc Natl Acad Sci U S A. 2006; 103: 8558–8563.
Delaloy C, Lu J, Houot AM, Disse-Nicodeme S, Gasc JM, Corvol P, Jeunemaitre X. Multiple promoters in the WNK1 gene: one controls expression of a kidney-specific kinase-defective isoform. Mol Cell Biol. 2003; 23: 9208–9221.
Yu AS. WNK signaling in the distal tubule: an inhibitory cascade regulating salt transport. Am J Physiol Renal Physiol. 2006; 290: F617–F618.