Common Variants in PCSK1 Influence Blood Pressure and Body Mass Index

Home , Proprotein convertase, Proprotein convertase 1, Subtilisin

ORIGINAL ARTICLE Common variants in PCSK1 inﬂuence blood pressure and body mass index

QGu1,2, M Yazdanpanah1, M van Hoek1, A Hofman3, X Gao2, FWM de Rooij1 and EJG Sijbrands1

Proprotein convertase subtilisin/kexin-type 1 (PCSK1) activates precursors pro-opiomelanocortin (POMC), proinsulin and prorenin. We investigated if common variants in the PCSK1 gene influence blood pressure and risk of hypertension. Additionally, we investigated the risk of obesity and type 2 diabetes (T2D). In the Rotterdam Study (RS1), a prospective, population-based cohort (n ¼ 5974), four single-nucleotide polymorphisms (rs10515237, rs6232, rs436321 and rs3792747) in PCSK1 were studied. Linear and Cox regression models served to analyze associations between variants and end points. Replication was performed in the Rotterdam Study Plus1 (RSPlus1, n ¼ 1895). Rs436321 was significantly associated with systolic and diastolic blood pressure and risk of hypertension (odds ratio (OR): 1.1–1.3; Po0.05 in both populations). Rs6232 was associated with body mass index (BMI) (P ¼ 0.007 and P ¼ 0.04 in RS1 and RSPlus1, respectively). In RSPlus1, heterozygotes for rs6232 had 1.5 times higher risk of obesity (OR: 1.46; 95% confidence interval: 1.04–2.03; P ¼ 0.03). We did not find significant associations of PCSK1 with fasting insulin levels and T2D. We found an association of genetic variation in the PCSK1 gene with blood pressure and hypertension. Furthermore, we replicated the association of PCSK1 with BMI and obesity. No relationship was found between PCSK1 variants and fasting insulin levels and T2D. Our findings suggest that genetic variation in PCSK1 may contribute to, at least, some of these interrelated disorders.

Journal of Human Hypertension (2015) 29, 82–86; doi:10.1038/jhh.2014.59; published online 17 July 2014

INTRODUCTION populations.12,23,24 Up to now, genetic variants have not been Proprotein convertase subtilisin/kexin-type 1 (PCSK1) has a major investigated in adult Caucasians in relation to blood pressure and role in the proteolytic processing of large prohormones into hypertension. mature bioactive proteins, including pro-opiomelanocortin In the present study, we investigated whether common (POMC), proinsulin, proglucagon, pro-islet amyloid polypeptide variants in the PCSK1 gene contribute to blood pressure (proIAPP) and prorenin.1–3 These end products are widely involved variation and hypertension in two prospective, population-based, in energy balance, glucose metabolism and blood pressure Caucasian cohorts. Furthermore, we sought to replicate the regulation. However, the effect of PCSK1 genetic variants on previously found associations between PCSK1 and obesity, T2D blood pressure has so far not been studied in Caucasian and related traits. populations. Rare mutations in PCSK1 lead to obesity and impaired glucose homeostasis.2,4–8 A large European study showed that common MATERIALS AND METHODS variants in PCSK1 are associated with obesity.9 This was Study design also shown in Mexicans, Asians, multiethnic Americans and a This study was conducted within the frame work of the Rotterdam Study Greek population.10–15 However, subsequent studies in (RS1), an ongoing prospective, population-based cohort study among 7983 Caucasians, including a large-scale genome-wide association subjects aged 55 years or older living in a suburb of Rotterdam. The design meta-analysis, provided inconsistent results.16–18 The association of the study has been described elsewhere. Baseline data were collected between common human PCSK1 variants and type 2 diabetes from March 1990 to July 1993, and follow-up was conducted at 1- to 3-year intervals.25 (T2D) has also not been firmly established: a study in an extended 19 Significant findings in the initial population (RS1) were replicated in the French family suggested linkage between PCSK1 and T2D. Rotterdam Study Plus1 (RSPlus1), which is a population consisting of 3011 Also, an association was reported in Czech adolescents subjects in the study district that became 55 years since the start of the 20 between a PCSK1 variant and blood glucose levels, but this RS1 or those of 55 years or over who migrated into the study district.25 The was not supported by subsequent studies, including large-scale medical ethics committee of Erasmus University has approved the study meta-analyses.17,21 and written informed consent was obtained from all participants. The relationship of PCSK1 with blood pressure has received less In both study populations, information concerning age, gender and medication use was obtained. Height and weight were measured and attention. In the adrenal medulla, PCSK1 carries out the specific À 2 cleavage of prorenin to convert it into active renin.22 Renin is an body mass index (BMI) (in kg m ) was computed. A trained research assistant measured systolic blood pressure and diastolic blood pressure important component of the renin–angiotensin–aldosterone (DBP) two times with a random-zero sphygmomanometer after a 5-min system, making it an excellent candidate gene for blood pressure 22 rest in seated position, and the two values were averaged. variation and hypertension. Recently, associations were found Blood samples were drawn after an overnight fast and fasting insulin between a PCSK1 polymorphism and hypertension in Asian level was measured by a commercially available assay (IRMA; Medgenix

1Erasmus Medical Center, Department of Internal Medicine, University Medical Center Rotterdam, Rotterdam, The Netherlands; 2Department of Endocrinology, Zhongshan Hospital Fudan University, Shanghai, China and 3Erasmus Medical Center, Department of Epidemiology and Biostatistics, Rotterdam, The Netherlands. Correspondence: Dr EJG Sijbrands, Erasmus Medical Center, Department of Internal Medicine–D435, University Medical Center Rotterdam, PO Box 2040, 3000 CA Rotterdam, The Netherlands. E-mail: [email protected] Received 3 March 2014; revised 2 June 2014; accepted 11 June 2014; published online 17 July 2014 PCSK1 inﬂuences blood pressure and BMI QGuet al 83 Diagnostics, Brussels, Belgium) with an intra- and interassay variation of 3– 6% and 5–12%, respectively. Fasting or non-fasting plasma glucose level Table 1. Basic characteristics of study populations were measured by the glucose hexokinase method (Medgenix Diagnostics) Characteristics Initial Replication with an intra-assay variation of o2.5% and 6.0%, respectively. population population

Definition of obesity, type 2 diabetes and hypertension Total number 5974 1895 Obesity was defined as a BMI of 30 kg m À 2 or higher. In accordance with Age at study entry (years) 69.4±9.1 64.7±8.2 the guidelines of the American Diabetes Association and WHO, T2D was Male (%) 40.6 45.5 À 2 diagnosed at a fasting plasma glucose level X7.0 mmol l À 1 or a random Body mass index (kg m ) 26.3±3.7 27.2±4.0 (non-fasting) plasma glucose level X11.0 mmol l À 1 26,27 or use of oral Obesity (%) 14.3 20.4 antidiabetic medications or insulin, and registered by a general practitioner Systolic blood pressure (mm Hg) 139.2±22.3 142.9±20.9 as having diabetes. In the initial population, the mean follow-up time for Diastolic blood pressure 73.7±11.5 78.8±10.8 incidence of T2D was 10.1 years (s.d. ¼ 4.56 years). Prevalent cases of T2D (mm Hg) at baseline and patients with type 1 diabetes were excluded from the Antihypertensive medication (%) 32.5 28.6 analyses. Hypertension (%) 56.2 60.6 Hypertension was defined as a systolic blood pressure X140 mm Hg Type 2 diabetes (%) 9.2 (incident) 10.9 and/or diastolic blood pressure X90 mm Hg and/or use of antihyperten- (prevalent) sive medication. Antidiabetic medication 4.4 4.7 Fasting serum insulina 66.69±37.34 — (pmol l À 1)(N ¼ 3674) Selection and genotyping of polymorphisms rs10515237 (AA/AG/GG) 2924/2566/470 — We selected four tagging single-nucleotide polymorphism (SNPs) of this (N ¼ 5960) gene (rs10515237, rs6232 (encoding N221D), rs436321 and rs3792747) rs6232 (AA/AG/GG) (N ¼ 5974) 5199/749/26 1678/207/10 from the HapMap database (Phase II) with minor allele frequencies above rs436321 (CC/CA/AA) (N ¼ 5963) 2654/2669/640 842/841/212 5% and r2 above 0.99 in the Caucasian (CEU) population. Microarray rs3792747 (AA/AG/GG) 3606/2077/289 — genotyping was available for the complete RS1 and RS Plus1 using the (N ¼ 5972) Infinium II HumanHap550K Genotyping BeadChip version 3 (Illumina, San Continuous values were expressed as means plus/minus s.d. aFasting Diego, CA, USA). Individuals with a sample call rate o98%, missing DNA, gender mismatch, excess autosomal heterozygosity, duplicates of family serum insulin was log transformed. relations with irritable bowel syndrome 497% and ethnic outliers (irritable bowel syndrome 44 s.d.) were excluded. SNPs with minor allele frequency of o1%, call rate o95% and Hardy–Weinberg equilibrium (HWE) Po10–6 were excluded. A total of 5974 and 1895 participants of the initial and the 9.2% were diagnosed having T2D during follow-up and 56.2% had replication population, whose DNA met the quality control criteria, and hypertension. had at least one tagging SNP successfully genotyped. The genotype call Obesity, T2D and hypertension were present in RSPlus1 in rates for rs10515237, rs6232, rs436321 and rs3792747 are 99.8%, 100%, 20.4%, 10.9% and 60.6%, respectively. 99.8% and 99.9%, respectively, in RS1 and 100% for both rs6232 and The rs10515237 A4G polymorphism was not in HWE rs436321 in RSPlus. (w2 ¼ 8.096; d.f. ¼ 1; P ¼ 0.004). The MAF of rs10515237 was 0.29 identical to HapMap-CEU, suggesting that the apparent deviation Statistical methods from HWE is due to random sampling variation rather than Deviation from Hardy–Weinberg equilibrium (HWE) was assessed by genotyping error. The three other polymorphisms were in HWE in 2 means of w test. both populations (w2o1.8; d.f. ¼ 1; P40.19). Univariate and linear regression models were used to assess contribu- tions of PCSK1 genotypes to BMI, fasting insulin level and blood pressure values. We performed logarithmic transformation of fasting insulin levels, Association of PCSK1 genotypes with continuous outcomes: blood because they were not normally distributed. In RS1, Cox proportional pressure, BMI and fasting insulin levels hazard models were used to examine PCSK1 polymorphisms with incident For the continuous outcomes table, see Supplementary Table 1. In T2D and hazard ratios (HRs) with 95% confidence intervals (CIs) were computed. Analyses of obesity, hypertension and T2D in RSplus1 were RS1, within subjects who were not using blood pressure lowering performed with logistic regression analyses and by calculating odds ratios medication, AA carriers for rs436321 SNP had 2.5 mm Hg higher (ORs) with 95% CIs. systolic and 1.40 mm Hg higher diastolic blood pressure levels We adjusted for year of birth and sex. In additional models, we also than CC carriers (P ¼ 0.047 and 0.02, respectively). adjusted for BMI in the causal path of T2D and hypertension to estimate the Univariate analysis showed that homozygosity for the minor effects explained by BMI. For analyses of blood pressures and fasting insulin allele of the rs6232 polymorphism was associated with higher BMI level, we excluded all subjects who used antihypertensive or antidiabetic (GG vs AA 1.97±0.73 kg m À 2, P ¼ 0.007). None of the polymorph- medications. In addition, we also performed the analyses adjusted for isms were associated with fasting insulin level (data not shown). antihypertensive and antidiabetic medication. The most common genotype The associations remained unchanged after further adjustment for was used as a reference group. Analyses were performed with SPSS software version 14.0 (IBM Software, Armonk, NY, USA). BMI. Rs10515237 and rs3792747 did not show significant Power was calculated using Quanto version 1.2.28 For hypertension, we associations with continuous outcomes in our discovery popula- had 85% power to detect an effect size of HR 1.2 for genetic variants with tion and were not taken forward for replication. an allele frequency of 10% in RS1. For obesity, we had 82% power to detect In RSPlus1, rs436321 was associated with blood pressure. AC an effect size of HR 1.2 for genetic variants with an allele frequency of 20% carriers of rs436321 had increased systolic blood pressure in RS1. For T2D, we had 93% power to detect an effect size of HR 1.3 for compared with CC carriers (mean difference ¼ 1.7 mm Hg, genetic variants with an allele frequency of 20% in RS1. P ¼ 0.03). Diastolic blood pressure did not significantly differ between any of the genotypes. In RSPlus1, rs6232 was also associated with BMI. Heterozygotes RESULTS for rs6232 had higher BMI compared with AA carriers (mean Baseline characteristics difference ¼ 0.60 kg m À 2, P ¼ 0.04) Table 1 presents the basic characteristics of two study Instead of excluding people on blood pressure lowering populations. Overall mean age was 69 years in RS1 and 65 in medication, we also performed analyses with additional adjust- RSPlus1. The populations consisted of 40.6% and 45.5% males, ment of antihypertensive medications or antidiabetic medications respectively. In RS1, obesity was found in 14.3% of participants, and this yielded identical results (data not shown).

& 2015 Macmillan Publishers Limited Journal of Human Hypertension (2015) 82 – 86 PCSK1 inﬂuences blood pressure and BMI QGuet al 84 Table 2. Risk of hypertension, obesity and type 2 diabetes by PCSK1 genotypes

Genotype N Hypertension Obesity Type 2 diabetes

OR (95% CI) P-value for trend OR (95% CI) P-value for trend HR/ORa (95% CI) P-value for trend

RS1 Rs10515237 AA 2812 1.00 (Reference) 0.61 1.00 (Reference) 0.51 1.00 (Reference) 0.46 AG 2475 1.05 (0.94–1.18) 1.02 (0.87–1.19) 1.08 (0.91–1.29) GG 456 0.85 (0.68–1.05) 0.84 (0.62–1.14) 1.06 (0.76–1.46) Rs6232 AA 4998 1.00 (Reference) 0.24 1.00 (Reference) 0.88 1.00 (Reference) 0.85 AG 721 0.91 (0.77–1.07) 0.96 (0.77–1.21) 0.95 (0.74–1.23) GG 25 0.88 (0.38–2.05) 1.82 (0.71–4.65) 1.93 (0.80–4.66) Rs436321 CC 2545 1.00 (Reference) 0.003 1.00 (Reference) 0.69 1.00 (Reference) 0.10 AC 2566 1.13 (1.01–1.26)* 0.98 (0.83–1.15) 0.90 (0.76–1.09) AA 622 1.28 (1.07–1.53)** 1.10 (0.85–1.41) 0.78 (0.58–1.06) Rs3792747 AA 3459 1.00 (Reference) 0.65 1.00 (Reference) 1.00 1.00 (Reference) 0.89 AG 2004 0.95 (0.85–1.07) 1.00 (0.85–1.17) 1.05 (0.88–1.26) GG 279 1.03 (0.79–1.33) 1.00 (0.72–1.42) 0.91 (0.60–1.37)

Replication RSPlus1 Rs6232 AA 1675 1.00 (Reference) 0.04 1.00 (Reference) 0.17 1.00 (Reference) 0.62 AG 207 1.15 (0.85–1.57) 1.46 (1.04–2.03)* 1.25 (0.81–1.94) GG 10 0.91 (0.23–3.62) — - Rs436321 CC 840 1.00 (Reference) 0.30 1.00 (Reference) 0.59 1.00 (Reference) 0.76 AC 840 1.23 (1.01–1.51)* 1.002 (0.79–1.27) 0.78 (0.57–1.07) AA 212 1.02 (0.74–1.40) 1.15 (0.79–1.65) 1.13 (0.72–1.77) Abbreviations: CI, conﬁdence interval; HR, hazard ratio; OR, odds ratio; PCSK1, proprotein convertase subtilisin/kexin-type 1. Adjusted for birth year and sex. Each genotype was compared with the common genotype. Signiﬁcant results (Po0.05) were shown in boldface types. *Po0.05, **Po0.01. aHR for intial population and OR for replication population.

Association of PCSK1 genotypes with categorical outcomes: risk of deficiency, the morbid obesity is preceded by early failure to grow, obesity, T2D and hypertension in knockout mice and in children with complete PCSK1 The rs436321 polymorphism was significantly associated with deficiency.7,30 Among many effects on neuroendocrine increased risk of hypertension in RS1 (Table 2). AC and AA processing, mutations in PCSK1 gene are thought to disrupt carriers of rs436321 had ORs of 1.13 (95% CI: 1.01–1.26) and POMC processing and reduce melanocortin signaling in the 1.28 (95% CI: 1.07–1.53), respectively, for hypertension compared hypothalamus, which is supposed to lead to hyperphagia and a with CC carriers (P ¼ 0.04 and 0.008, respectively). Heterozygosity positive energy imbalance.2 for the rs436321 was also associated with increased risk In addition to energy metabolism, PCSK1 is a prorenin of hypertension in the RSPlus1 (OR: 1.23 (95% CI: 1.01–1.51), processing enzyme and may therefore be involved in the P ¼ 0.04). pathogenesis of hypertension.22 However, little has been In RS1, homozygotes of the minor allele of rs6232 had 1.82 published on PCSK1 variants in relation to blood pressure and higher point estimate of the risk of obesity but this was not risk of hypertension. We found that the PCSK1 polymorphism significant (95% CI: 0.71–4.65, P ¼ 0.21). In contrast, heterozygotes rs436321 is associated with high blood pressure and increased risk of rs6232 in the RSPlus1 showed a trend toward increased risk of of hypertension. This is partly in line with a recent Chinese and obesity (OR: 1.46 (95% CI: 1.04–2.03); P ¼ 0.03). Rs10515237 and Taiwanese studies in which the rs6235 polymorphism was rs3792747 did not show significant associations with categorical associated with hypertension.23,24 We did not study the rs6235 outcomes in our discovery population and were not taken forward polymorphism, but it was tagged by the rs10515237 poly- for replication. morphism, which did not show an association in our study. The The polymorphisms were not related to risk of T2D (Table 2). associated rs436321 captures many other variants in the gene Further adjustment for BMI did not change the results (data not through the linkage disequilibrium structure. The effect on shown). hypertension may therefore be explained through linkage of rs436321 to other, perhaps functional, variants in the gene. Regarding the function of PCSK1 processing renin, the under- DISCUSSION lying mechanism of this association could be through the renin– In the present study, we found that the rs436321 variant in PCSK1 angiotensin–aldosterone system and through activation of the was significantly associated with blood pressure and hyperten- renin receptor. An effect on blood pressure through BMI is sion. In addition, we confirmed the association of rs6232 with BMI. unlikely, as our results remained unchanged after adjustment for We did not observe significant associations between PCSK1 BMI. Of course, confirmation of our findings in other studies is polymorphisms, fasting insulin level and risk of T2D. warranted. To the best of our knowledge, our report is the first to From late 1990s, PCSK1 was reported to cause monogenetic investigate PCSK1 genetic variants in relation to hypertension in obesity in both humans2,4,5,29 and N222D mice. In complete PCSK1 Caucasians.

Journal of Human Hypertension (2015) 82 – 86 & 2015 Macmillan Publishers Limited PCSK1 influences blood pressure and BMI QGuet al 85 We found that the rs6232 polymorphism in PCSK1 was ACKNOWLEDGEMENTS associated with higher BMI and possibly increased risk of obesity. The Rotterdam Study is supported by the Erasmus Medical Center and Erasmus Recently, epidemiologic studies have shown that PCSK1 may also University Rotterdam; the Netherlands Organization for Scientific Research; the contribute to polygenic obesity at a population level in Netherlands Organization for Health Research and Development (ZonMw); the Caucasians.9,13,16,18 However, these findings have been followed Research Institute for Diseases in the Elderly; the Ministry of Education, Culture and by inconsistent replication results.9 Our findings are in line with Science; the Ministry of Health, Welfare and Sports; the European Commission; and the original findings of Benzinou et al.9 Functional data the Municipality of Rotterdam. demonstrated that the A to G substitution at rs6232, which causes an amino-acid change at position 221 (N221D), results in 9 an impaired catalytic activity of PCSK1 protein. This may decrease REFERENCES the processing of POMC to ACTH, which may ultimately lead to 1 Benjannet S, Reudelhuber T, Mercure C, Rondeau N, Chretien M, Seidah NG. altered levels of a-MSH (processed by PCSK2) and may disturb Proprotein conversion is determined by a multiplicity of factors including con- energy intake and expenditure. vertase processing, substrate specificity, and intracellular environment. Cell type- Furthermore, PCSK1 is a key enzyme in the processing of specific processing of human prorenin by the convertase PC1. J Biol Chem 1992; proinsulin to insulin.31 A linkage study in an extended French 267(16): 11417–11423. family suggested that the region containing PCSK1 is a strong 2 Farooqi IS, Volders K, Stanhope R, Heuschkel R, White A, Lank E et al. Hyperphagia candidate for T2D susceptibility.19 Renstrom et al.17 found a and early-onset obesity due to a novel homozygous missense mutation in pro- modest but significant protective effect of PCSK1 on T2D risk after hormone convertase 1/3. J Clin Endocrinol Metab 2007; 92(9): 3369–3373. adjustment for BMI. However, other studies could not confirm the 3 Scamuffa N, Calvo F, Chretien M, Seidah NG, Khatib AM. Proprotein convertases: 21,32 lessons from knockouts. FASEB J 2006; 20(12): 1954–1963. associations. We did not find an association with fasting 4 Jackson RS, Creemers JW, Farooqi IS, Raffin-Sanson ML, Varro A, Dockray GJ et al. insulin levels or with T2D risk, which is in line with the findings of Small-intestinal dysfunction accompanies the complex endocrinopathy of human 32 Strawbridge et al. in a large genome-wide association meta- proprotein convertase 1 deficiency. J Clin Invest 2003; 112(10): 1550–1560. analysis. They found that a PCSK1 variant is associated with 5 Jackson RS, Creemers JW, Ohagi S, Raffin-Sanson ML, Sanders L, Montague CT proinsulin levels, but not with fasting insulin levels or T2D. Of note, et al. Obesity and impaired prohormone processing associated with mutations in in our study, power was limited for detecting associations of the the human prohormone convertase 1 gene. Nat Genet 1997; 16(3): 303–306. variants with T2D. 6 Creemers JW, Choquet H, Stijnen P, Vatin V, Pigeyre M, Beckers S et al. Hetero- Strengths of our study include the prospective study design of zygous mutations causing partial prohormone convertase 1 deficiency contribute RS1, and the availability of detailed information on obesity, T2D to human obesity. Diabetes 2012; 61(2): 383–390. 7 Martin MG, Lindberg I, Solorzano-Vargas RS, Wang J, Avitzur Y, Bandsma R et al. and hypertension and their associated continuous traits, allowing Congenital proprotein convertase 1/3 deficiency causes malabsorptive diarrhea to study PCSK1 in three complex disorders simultaneously. A and other endocrinopathies in a pediatric cohort. Gastroenterology 2013; 145(1): number of limitations of our study need to be considered. Our 138–148. tagging SNPs did not cover the whole gene and negative findings 8 Yourshaw M, Solorzano-Vargas RS, Pickett LA, Lindberg I, Wang J, Cortina G et al. cannot rule out effects of SNPs outside the haploblock. Moreover, Exome sequencing finds a novel PCSK1 mutation in a child with generalized rare human mutations have been described to cause severe malabsorptive diarrhea and diabetes insipidus. J Pediatr Gastroenterol Nutr 2013; obesity and abnormal glucose homeostasis, but our study 57(6): 759–767. populations are too small to detect an effect of such rare variants. 9 Benzinou M, Creemers JW, Choquet H, Lobbens S, Dina C, Durand E et al. Com- We had sufficient power to detect effects of the studied variants mon nonsynonymous variants in PCSK1 confer risk of obesity. Nat Genet 2008; 40(8): 943–945. on the studied continuous traits. Also for obesity and hyperten- 10 Chang YC, Chiu YF, Shih KC, Lin MW, Sheu WH, Donlon T et al. Common PCSK1 sion, our studies offered enough power to detect reasonable haplotypes are associated with obesity in the Chinese population. Obesity (Silver effect sizes for the genetic variants that we studied. However, for Spring, MD) 2010; 18(7): 1404–1409. T2D power was limited. 11 Choquet H, Kasberger J, Hamidovic A, Jorgenson E. Contribution of common In conclusion, we found a relationship of a genetic variant PCSK1 genetic variants to obesity in 8359 subjects from multi-ethnic American (rs436321) in PCSK1 with blood pressure levels and hypertension. population. PLoS ONE 2013; 8(2): e57857. We replicated the association of a PCSK1 polymorphism (rs6232) 12 Qi Q, Li H, Loos RJ, Liu C, Hu FB, Wu H et al. Association of PCSK1 rs6234 with with BMI and obesity. No associations of PCSK1 polymorphisms obesity and related traits in a Chinese Han population. PLoS ONE 2010; 5(5): e10590. with fasting insulin levels and risk of T2D were found. Our findings 13 Rouskas K, Kouvatsi A, Paletas K, Papazoglou D, Tsapas A, Lobbens S et al. Common variants in FTO, MC4R, TMEM18, PRL, AIF1, and PCSK1 show evidence of suggest that genetic variation in PCSK1 may contribute to risk of, association with adult obesity in the Greek population. Obesity (Silver Spring. MD) at least, some of these interrelated disorders. 2012; 20(2): 389–395. 14 Villalobos-Comparan M, Villamil-Ramirez H, Villarreal-Molina T, Larrieta-Carrasco E, Leon-Mimila P, Romero-Hidalgo S et al. PCSK1 rs6232 is associated with childhood and adult class III obesity in the Mexican population. PLoS ONE 2012; 7(6): e39037. 15 Wen W, Cho YS, Zheng W, Dorajoo R, Kato N, Qi L et al. Meta-analysis identifies What is known about this topic common variants associated with body mass index in east Asians. Nat Genet 2012; PCSK1 activates precursors POMC, proinsulin and prorenin. 44(3): 307–311. Mutations in PCSK1 lead to obesity and impaired glucose 16 Kilpelainen TO, Bingham SA, Khaw KT, Wareham NJ, Loos RJ. Association of var- homeostasis. iants in the PCSK1 gene with obesity in the EPIC-Norfolk study. Hum Mol Genet Recently, an association was found between a PCSK1 polymorphism 2009; 18(18): 3496–3501. and essential hypertension in a Chinese Han population. 17 Renstrom F, Payne F, Nordstrom A, Brito EC, Rolandsson O, Hallmans G et al. Replication and extension of genome-wide association study results for obesity in What this study adds 4923 adults from northern Sweden. Hum Mol Genet 2009; 18(8): 1489–1496. In Caucasians, the rs436321 variant in PCSK1 was significantly 18 Willer CJ, Speliotes EK, Loos RJ, Li S, Lindgren CM, Heid IM et al. Six new loci associated with blood pressure and hypertension in Caucasians. associated with body mass index highlight a neuronal influence on body weight The association of genetic variation in PCSK1 polymorphism with BMI regulation. Nat Genet 2009; 41(1): 25–34. and obesity was replicated. 19 Martin LJ, Comuzzie AG, Dupont S, Vionnet N, Dina C, Gallina S et al. A quanti- tative trait locus influencing type 2 diabetes susceptibility maps to a region on 5q in an extended French family. Diabetes 2002; 51(12): 3568–3572. 20 Dusatkova L, Zamrazilova H, Sedlackova B, Vcelak J, Hlavaty P, Aldhoon Hainerova I et al. Association of obesity susceptibility gene variants with metabolic CONFLICT OF INTEREST syndrome and related traits in 1443 czech adolescents. Folia Biol (Praha) 2013; The authors declare no conflict of interest. 59(3): 123–133.

& 2015 Macmillan Publishers Limited Journal of Human Hypertension (2015) 82 – 86 PCSK1 influences blood pressure and BMI QGuet al 86 21 Kalidas K, Dow E, Saker PJ, Wareham N, Halsall D, Jackson RS et al. Prohormone diabetes mellitus provisional report of a WHO consultation. Diabet Med 1998; convertase 1 in obesity, gestational diabetes mellitus, and NIDDM: no evidence 15(7): 539–553. for a major susceptibility role. Diabetes 1998; 47(2): 287–289. 28 Gaudermann WJ. Quanto 1.1: A computer program for power and sample 22 Reudelhuber TL, Ramla D, Chiu L, Mercure C, Seidah NG. Proteolytic processing size calculations for genetic epidemiology studies. Available at: http://hydra. of human prorenin in renal and non-renal tissues. Kidney Int 1994; 46(6): usc.edu/gxe [website] 2012 [cited 2014]. Last accessed on 1 February 2014. 1522–1524. 29 Frank GR, Fox J, Candela N, Jovanovic Z, Bochukova E, Levine J et al. Severe 23 Hsiao TJ, Hwang Y, Chang HM, Lin E. Association of the rs6235 variant in the obesity and diabetes insipidus in a patient with PCSK1 deficiency. Mol Genet proprotein convertase subtilisin/kexin type 1 (PCSK1) gene with obesity and Metab 2013; 110(1–2): 191–194. related traits in a Taiwanese population. Gene 2014; 533(1): 32–37. 30 Zhu X, Zhou A, Dey A, Norrbom C, Carroll R, Zhang C et al. Disruption of PC1/3 24 Li XM, Ling Y, Lu DR, Lu ZQ, Liu Y, Chen HY et al. The obesity-related poly- expression in mice causes dwarfism and multiple neuroendocrine peptide pro- morphism PCSK1 rs6235 is associated with essential hypertension in the Han cessing defects. Proc Natl Acad Sci USA 2002; 99(16): 10293–10298. Chinese population. Hypertens Res 2012; 35(10): 994–999. 31 Jansen E, Ayoubi TA, Meulemans SM, Van de Ven WJ. Neuroendocrine-specific 25 Hofman A, Darwish Murad S, van Duijn CM, Franco OH, Goedegebure A, Ikram MA expression of the human prohormone convertase 1 gene. Hormonal regulation of et al. The Rotterdam Study: 2014 objectives and design update. Eur J Epidemiol transcription through distinct cAMP response elements. J Biol Chem 1995; 2013; 28(11): 889–926. 270(25): 15391–15397. 26 Report of the Expert Committee on the Diagnosis and Classification of Diabetes 32 Strawbridge RJ, Dupuis J, Prokopenko I, Barker A, Ahlqvist E, Rybin D et al. Gen- Mellitus. Diabetes Care 1997; 20(7): 1183–1197. ome-wide association identifies nine common variants associated with fasting 27 Alberti KG, Zimmet PZ. Definition, diagnosis and classification of diabetes proinsulin levels and provides new insights into the pathophysiology of type 2 mellitus and its complications. Part 1: diagnosis and classification of diabetes. Diabetes 2011; 60(10): 2624–2634.

Supplementary Information accompanies this paper on the Journal of Human Hypertension website (http://www.nature.com/jhh)

Journal of Human Hypertension (2015) 82 – 86 & 2015 Macmillan Publishers Limited