G

P

TS

1

2

CQ1

G3

4

a5b6c7

8

a9

10

A11

R12

R13

A14

A

15

K16

C17

218

G19

M20

C21

T22

123

24

b25

S26

r27

h28

r29

a30

c31

o32

Cf

1h

ARTICLE IN PRESS Model

RO 5983 1–6

Prostaglandins & other Lipid Mediators xxx (2012) xxx– xxx

Contents lists available at SciVerse ScienceDirect

Prostaglandins and Other Lipid Mediators

he functional variant V433M of the CYP4F2 and the metabolic syndrome inwedes

ristiano Favaa,b,∗, Martina Montagnanaa,c,1, Elisa Danesea,b,c,1, Marketa Sjögrena, Peter Almgrena,ian Cesare Guidic, Bo Hedblada, Gunnar Engströma, Pietro Minuzb, Olle Melandera

Department of Clinical Sciences, Lund University, Skane University Hospital, Malmö, SwedenDepartment of Medicine, ItalyDepartment of Life and Reproduction Sciences, University Hospital of Verona, Italy

r t i c l e i n f o

rticle history:eceived 10 January 2012eceived in revised form 20 February 2012ccepted 22 March 2012vailable online xxx

eywords:ytochrome P4500-HETEeneticsetabolic syndrome

YP4F2riglycerides

a b s t r a c t

Background and aim: The genetic basis of Metabolic syndrome (MetS) is largely unknown but a link withsalt sensitivity is recognized. The cytochrome P450 isoform 4F2 (CYP4F2) is involved in renal productionof 20-hydroxyeicosatethraenoic acid (20-HETE), a natriuretic substance associated with salt sensitivity.The same enzyme is implicated in -hydroxylation of very long and medium chain fatty acids in theliver suggesting its possible influence on gluco-metabolic components of MetS. The aim of the presentstudy was to evaluate the effect of CYP4F2 V433M, a functional polymorphism previously associated withhypertension via renal salt reabsorption, on the individual components of MetS and MetS itself.Methods: The polymorphism was genotyped in the cardiovascular cohort of the Malmö Diet and Cancer(MDC-CVA) study and successively in the Malmö Preventive Project (MPP) cohort. Different definitionsof the MetS were applied.Results: In the MDC-CVA, male, but not female, CYP4F2 M433 carriers had significantly higher levelsof waist, triglycerides, BP and a composite sum of MetS phenotypes (MetS score) beside lower HDL-cholesterol respect to V-homozygotes. MetS, as defined in the ATPIII and the AHA/NHLBI definitions, was

more prevalent in M-carriers with respect to V-homozygotes. In the MPP cohort, significant associationwas detectable only for triglycerides at baseline and for Diastolic BP at reinvestigation in male M-carriers.Conclusion: The initial positive association of the CYP4F2 V433M polymorphism with components of MetSand MetS itself, found in MDC-CVA, was partially denied in another large cohort. The first associationeither could be due to a false positive result or alternatively, different genetic background or populationstratification could have hidden the effect of the polymorphism in the replication cohort.

33

34

35

36

37

38

39

40

. Introduction

20-HETE is a vasoconstrictor and natriuretic substance, derivedy metabolism of arachidonic acid by cytochrome p450 (CYP).ince the 1980s its role in hypertension development in differentats and mice strains has been established [1] and more recentlyas been proposed also in humans [2]. CYP4F2 is the enzymeesponsible for most of 20-HETE production at renal level whereas

Please cite this article in press as: Fava C, et al. The functional variant V433MOther Lipid Mediat (2012), http://dx.doi.org/10.1016/j.prostaglandins.2012

common, functional polymorphism, CYP4F2 V433M, was asso-iated with hypertension development and other cardiovascularutcomes in different studies [3–5]. The functional significance of

∗ Corresponding author at: Department of Medicine, Division of Internal Medicine, Piazzale LA Scuro 10, 37134 Verona, Italy. Tel.: +39 45 8124414;ax: +39 45 8027465.

E-mail addresses: cristiano.fava@med.lu.se, cristiano.fava@univr.it (C. Fava).1 Contributed equally to this work.

41

42

43

44

45

46

098-8823/$ – see front matter © 2012 Published by Elsevier Inc.ttp://dx.doi.org/10.1016/j.prostaglandins.2012.03.001

© 2012 Published by Elsevier Inc.

these mutations on the ability of the CYP4F2 enzyme to metabo-

lize arachidonic acid was previously demonstrated in recombinant

baculoviruses, where the presence of the M433 allele, decreased

20-HETE production to more than 50% of control [6].

Other authors found that a haplotype construct of the CYP4F2 is

involved in hypertension development and cardiovascular disease

[7,8].

A link between salt sensitivity and other gluco-metabolic traits

has long been recognized and is partially demonstrated by the clus-

tering of all these features in the metabolic syndrome (MetS). The

genetic basis of this clustering is largely unknown and most of the

studies converge about the relation between insulin resistance and

salt sensitivity [9].

Since the CYP4F2 is expressed also at liver level and is implicated

of the CYP4F2 and the metabolic syndrome in Swedes. Prostaglandins.03.001

in -hydroxylation of very long and medium chain fatty acids, a 47

process that is increased in different metabolic diseases and differ- 48

ently influenced by starvation, excess lipid and alcohol intake [10], 49

we were interested in exploring a possible role of the functional 50

ING Model

P

2 her Lip

V51

b52

53

o54

p55

U56

M57

v58

259

60

l61

T62

l63

264

Q365

c66

w67

c68

e69

t70

(71

f72

t73

I74

t75

a76

y77

(78

o79

c80

281

82

183

e84

(85

i86

a87

l88

a89

i90

o91

w92

p93

o94

i95

t96

D97

i98

299

100

i101

o102

(103

b104

s105

I106

107

w108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

ARTICLERO 5983 1–6

C. Fava et al. / Prostaglandins & ot

433M polymorphism in different features of the MetS beyondlood pressure (BP).

Thus, the aim of the present study was to test the associationf the CYP4F2 V433M polymorphisms with MetS individual com-onents and MetS as defined by different Scientific Societies [11].sing standardized residuals from all the single components ofetS we constructed also a MetS score to have a single continuous

ariable to be analyzed for MetS [12].

. Materials and methods

All study participants had given written informed consent. Theocal ethics committee of Southern Sweden approved the studies.he procedures were in accordance with the institutional guide-ines.

.1. Malmö Diet and Cancer study (MDC)

The study population is derived from the Malmö Diet and Can-er study (MDC) The BP along with other cardiovascular risk factorsere measured in a sub-sample of MDC referred to as the MDC-

ardiovascular arm (MDC-CVA) (n = 6103) [13,14]. Successfullyxtracted genomic DNA was available from 5763 MDC-CVA par-icipants. Fasting plasma samples were available in 5,533 subjectssee supplementary material for further details). Complete dataor the components of the metabolic syndrome [National Choles-erol Education Program/Adult Treatment Panel III (NCEP/ATP III),nternational Diabetes Federation (IDF), American Heart Associa-ion/National Heart, Lung, Blood Institute (AHA/NHLBI) definitions]nd CYP4F2 genotype was available in 5383 individuals. In the anal-ses of the European Group for the Study of Insulin ResistanceEGIR) definition, 547 patients (with diabetes and/or with no datan fasting insulin concentration in plasma) were excluded from theohort of 5341 subjects leaving 4836 subjects for analysis.

.2. Malmö Preventive Project (MPP)

In the MPP, 33,346 Swedish participants (22,444 men and0,902 women, mean age 49 years, from the city of Malmö in south-rn Sweden participated in health screening during 1974–1992attendance rate 71%) [15]. All individuals underwent a phys-cal examination and measurements of fasting blood glucosend triglycerides concentrations were performed. Information onifestyle factors and medical history was obtained from a self-dministered questionnaire. Of individuals participating in thenitial screening, 4931 died and 551 were lost from follow-up forther reasons. Twenty-five thousand of the eligible individualsere invited to a re-screening visit during 2002–2006, including ahysical examination and fasting blood samples for measurementsf glucose, triglycerides and HDL-cholesterol concentrations. Of thenvited individuals, 18,240 participated in the re-screening; 705 ofhem were excluded from the present study because of the lack ofNA or crucial clinical information. Thus, 17,535 participants, were

ncluded in the follow-up study.

.3. Phenotyping

Waist circumference was measured with the patient stand-ng, at the umbilicus level. The BMI was calculated as the ratiof the weight in kilograms to the square of the height in meterskg/m2). BP was measured by specially trained nurses in the rightrachial artery with some differences between the studies (see

Please cite this article in press as: Fava C, et al. The functional variant V433MOther Lipid Mediat (2012), http://dx.doi.org/10.1016/j.prostaglandins.2012

upplementary methods). The systolic BP was defined by ‘phase’ and the diastolic BP defined by ‘phase V’ Korotkoff sounds.

When continuous traits where analyzed, a duplicate analysisas done: either excluding people under chronic therapy: that is

PRESSid Mediators xxx (2012) xxx– xxx

antihypertensive treatments (n = 876 in MDC-CVA; n = 5915 in MPP

at reinvestigation; glucose lowering agents (n = 73 in MDC-CVA;

n = 1262 in MPP at reinvestigation; and lipid lowering drugs (n = 120

in MDC-CVA; n = 168 in MPP at reinvestigation; or adding 1 SD to

the standardized age/sex adjusted residuals (see below).

2.4. Laboratory analysis

After an overnight fast, blood samples were drawn for the deter-

mination of serum lipids, serum insulin and whole blood glucose.

Samples were analyzed by standard methods at the Department of

Clinical Chemistry, Malmö University Hospital. Details of the mea-

surement of serum lipids, serum insulin and whole blood glucose

have been published elsewhere [16].

When serum triglyceride levels were above 4.00 mmol/l, LDL

cholesterol was not determined.

2.5. Metabolic syndrome definitions

MetS was defined according to NCEP/ATP III criteria, according tothe IDF definition, according to the EGIR definition (subjects with-

out diabetes only) and according to the latest AHA/NHLBI definition(see the Supplementary methods for details).

For the EGIR definition it is necessary to detect insulin resistance.

We defined insulin resistance as HOMA values >75th percentile

(i.e. >1.813 mmol mU L−2 for females and >2.133 mmol mU L−2 for

males, based on the distribution of all subjects without diabetes

included in the MDC-CVA).

In the MPP at baseline, as waist and HDL-cholesterol measure-

ments were not available for most participants, we were unable to

construct the definitions of the metabolic syndrome according to

any of the common definitions.

2.6. MetS score

In the MDC-CVA and in the MPP at reinvestigation we con-

structed a MetS score to analyze MetS as a quantitative trait. First,

we adjusted each of the variables included in the MetS definitions,

for gender and age. The residuals from these linear regression anal-yses were then standardized by subtracting the age and gender

specific mean and dividing by the age and gender-specific stan-

dard deviation. Since the resultant residuals were approximately

normally distributed, the standardization process yielded trait val-

ues that followed a N ∼ (0, 1) distribution. Since many people, were

under antihypertensive, or antilipemic or antidiabetic medications

we replicated the analysis either without considering the medi-

cations or adding 1SD of the trait in question, if the patient was

treated. Outliers for each trait were handled as follows: subjects

with values > ±3 SD were set as ±3 SD.

To generate the MetS score for each of the subjects’

available observations, we combined the standardized resid-

uals for each component trait as follows: [MetS score = mean

BP + triglycerides + glucose + waist − HDL]; the HDL was subtracted

from the score because it is protective against cardiovascular

diseases, and tends to be inversely correlated with the other com-

ponent traits; thus, lower HDL values correspond to a higher MetS

score.

2.7. Genotyping

of the CYP4F2 and the metabolic syndrome in Swedes. Prostaglandins.03.001

Genotypes of the CYP4F2 V433M polymorphism (dbSNP 161

accession number, rs2108622) were determined by end-point flu- 162

orescent measurements (see Supplementary Material for details). 163

ARTICLE IN PRESSG Model

PRO 5983 1–6

C. Fava et al. / Prostaglandins & other Lipid Mediators xxx (2012) xxx– xxx 3

Table 1Anthropometric and metabolic features of the investigated subjects in the MDC-CVA and MPP (baseline and reinvestigation).

All subjectsWith at least one valid genotype Numb. MDC-CVAAll subjects Numb. MPPAt baseline Numb. MPPAt follow-up

Gender, male (%) 5154 42.2 17,669 63.4 17,924 63.4Age, years 5154 57.67 ± 5.95 17,669 45.09 ± 7.40 17,924 68.23 ± 5.77Systolic blood pressure, mmHg 5154 141.57 ± 19.04 17,669 126.77 ± 14.10 17,924 145.01 ± 19.97Diastolic blood pressure, mmHg 5154 87.00 ± 9.35 17,669 85.27 ± 8.69 17,924 83.56 ± 10.56Body mass index, kg/m2 † 5154 25.82 ± 3.93 17,669 24.29 ± 3.37 17,924 27.16 ± 4.15Waist, cm † 5154 83.92 ± 12.86 606 79.64 ± 10.26 17,924 94.8 ± 12.3Glucose, mmol/L || 5154 5.18 ± 1.38 17,669 4.91 ± 0.75 17,924 5.84 ± 1.43Cholesterol, mmol/L || 5150 6.17 ± 1.08 17,669 5.61 ± 1.05 17,924 5.60 ± 1.09Triglycerides, mmol/L || 5154 1.37 ± 0.74 17,669 1.28 ± 0.80 17,924 1.27 ± 0.80HDL-cholesterol, mmol/L || 5154 1.38 ± 0.37 0 – 17,924 1.41 ± 0.42Obesity, % † 5154 13.2 17,669 5.6 17,924 21.7Hypertension (ATPIII definition), % 5154 76.7 17,669 58.9 17,924 84.6Diabetes, % ‡ 5148 8.5 17,669 3.2 17,924 13.2MetS, ATPIII definition 5154 22.1 0 – 17,924 29.3MetS, IDF definition 5154 29.5 0 – 17,924 45,3MetS, EGIR definition 4612 20.6 0 – 0 –MetS, AHA/NHLBI definition 5154 37.2 0 – 17,924 47,8

M oject;T for tH

2164

165

e166

s167

t168

f169

a170

m171

s172

s173

a174

175

t176

t177

s178

3179

180

f181

9182

h183

M184

g185

h186

187

e188

189

190

T191

S192

3193

194

o195

b196

m197

198

t199

M200

g201

t202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

DC CVA, Malmö Diet and Cancer Cardiovascular Arm; MPP, Malmö Preventive Prreatment Panel III; IDF, International Diabetes Federation; EGIR, European Groupeart, Lung and Blood Institute (AHA/NHLBI).

.8. Statistics

Continuous variables are presented as the mean ± SD. All data,xcept for the power analysis, were analyzed with SPSS statisticaloftware (version 18.0; SPSS Inc. Chicago, Illinois, USA). Chi-squareest (Pearson) was used to compare group frequencies and to testor deviations from Hardy-Weinberg equilibrium. Analysis of vari-nce followed by Tukey’s test and t-test was used to compare groupeans of continuous variables. Multiple logistic and linear regres-

ion analyses were used in the multivariate models with MetStatus and/or the MetS score as dependent variables and genotype,ge, gender as independent variables.

For variables with skewed distributions such as glucose andriglycerides, log-normalized values were used in the analysis. Allests were two-sided and P values less than 0.05 were consideredtatistically significant.

. Results

Genotyping success rate in the subsamples with complete dataor the components of the MetS was 95.7% (n = 5154) in MDC-CVA,9.1% (18,084) in MPP. In MDC-CVA we had 2825 (54.8%) V433Vomozygotes, 1981 (38.4%) V433M heterozygotes and 348 (6.8%)433M homozygotes; in MPP we had 9814 (54.3%) V433V homozy-

otes, 7057 (39.0%) V433M heterozygotes and 1213 (6.7%) M433Momozygotes.

Genotype distributions did not deviate from Hardy–Weinbergquilibrium in all the samples (2 = 0.01, P = 0.99 in the MDC-CVA,2 = 0.59, P = 0.74 in the MPP).

Clinical characteristics of all participants are presented inable 1. Detailed analyses about statistical power are reported inupplementary material.

.1. MDC-CVA

On the basis of the findings in our previous study [15] we focusedur attention especially on the Autosomal Dominant model, whichetter fits our data. Results about autosomal recessive and additiveodels are also reported in Supplementary Tables S1–S4.As shown in Table 2, a statistically significant difference in

Please cite this article in press as: Fava C, et al. The functional variant V433MOther Lipid Mediat (2012), http://dx.doi.org/10.1016/j.prostaglandins.2012

riglycerides and HDL-cholesterol level as well as the compositeetS score was observed in carriers of different CYP4F2 V433M

enotypes. When we stratified the population by gender, we foundhat, the results about HDL-cholesterol and triglycerides remained

MetS, metabolic syndrome; NCEP, National Cholesterol Education Program/Adulthe Study of Insulin Resistance; AHA/NHLBI, American Heart Association/National

significant in male but not in female M-carriers (Table 2); indeed the

differences in waist and BP became statistically significant (Table 2).

The MetS score remained significantly higher only in male carriers

of the CYP4F2 M-allele. Repetitions of all these analyses with-

out adjusting the crude values by antihypertensive, antilipemic or

antidiabetic medications gave similar results (data not shown).

Analysis of the dichotomous trait MetS as defined in the

ATPIII, IDF, EGIR and AHA/NHLBI definitions showed an increase

in the prevalence of the syndrome (ATPIII and AHA/NHLBI def-

initions) in carriers of the M-allele with a trend using the IDF

definition and no difference using the EGIR definition in which

diabetic subjects were excluded (Table 3). When excluding dia-

betic subjects also from the ATPIII and AHA/NHLBI definitions,

the positive association remained statistically significant (ATPIII:

OR[95%CI] 1.189 [1.040–1.359] and AHA/NHLBI: O.R. [95%CI] 1.137

[1.005–1.288]). When stratified by gender the increase in the

prevalence of the syndrome in carriers of the M-allele was still

present and of the same magnitude in males but not in females

(Table 3).

3.2. MPP

In the MPP all MetS individual components were not statistically

different in carriers of different CYP4F2 genotype with the excep-

tion of a mild difference in triglycerides at baseline, which resulted

higher in carriers of the CYP4F2 M-allele (Table 2). After stratifica-

tion for gender, triglycerides at baseline as well as diastolic BP at

reinvestigation resulted significantly higher in male carriers of the

CYP4F2 M-allele (Table 2). Repetitions of all these analyses with-

out adjusting the crude values by antihypertensive, antilipemic or

antidiabetic medications gave similar results (data not shown).

In Table 3 analyses of different definitions of MetS and CYP4F2

genotype are presented showing that only using the AHA/NHLBI

definition a positive trend toward significance was evident in car-

riers of the CYP4F2 M-allele. When excluding diabetic subjects,

carriers of the M-allele showed a significantly higher OR for MetS

prevalence only when the AHA/NHLBI definition was used (O.R.

[95%CI]: 1.068 [1.001–1.140] p = 0.046).

Stratifying the results by gender, we found only a non significant

trend toward higher risk of MetS as defined according to the ATPIII

of the CYP4F2 and the metabolic syndrome in Swedes. Prostaglandins.03.001

in male carriers of the CYP4F2 M-allele. 241

When 1293 male subjects (11.3%) who also participated in the 242

MDC-CVA were excluded from the MPP, the association between 243

the M-allele, diastolic BP ( ± SE: 0.041 ± 0.023; p = 0.066) and 244

ARTICLE IN PRESSG Model

PRO 5983 1–6

4 C. Fava et al. / Prostaglandins & other Lipid Mediators xxx (2012) xxx– xxx

Table 2Beta coefficient and standard error (on parenthesis) for the CYP4F2 V433M polymorphism tested by linear regression under the autosomal dominant genetic model in thewhole population and after stratification for gender.

MetS individual components Whole population Males Females

ˇ-coeff. (SEM) p-value ˇ-coeff. (SEM) p-value ˇ-coeff. (SEM) p-value

Malmö Diet and CancerSystolic BP, mmHg 0.049 (0.032) 0.12 0.145 (0.05) 0.004 −0.021 (0.042) 0.62Diastolic BP, mmHg 0.046 (0.032) 0.15 0.127 (0.052) 0.015 −0.009 (0.04) 0.82Waist, cm 0.045 (0.027) 0.10 0.094 (0.043) 0.029 0.010 (0.037) 0.78Triglycerides, mmol/L 0.073 (0.029) 0.01 0.111 (0.046) 0.017 −0.06 (0.036) 0.09Glucose, mmol/L −0.005 (0.024) 0.83 −0.014 (0.030) 0.64 −0.005 (0.024) 0.83HDL-cholesterol, mmol/L −0.082 (0.028) 0.004 −0.127 (0.043) 0.003 −0.05 (0.037) 0.18MetS score 0.272 (0.098) 0.006 0.521 (0.154) 0.001 0.091 (0.127) 0.47

Malmö Preventive Project (baseline)Systolic BP, mmHg 0.008 (0.016) 0.61 0.004 (0.018) 0.82 0.015 (030) 0.62Diastolic BP, mmHg 0.021 (0.016) 0.19 0.025 (0.020) 0.20 0.013 (0.027) 0.63BMI 0.003 (0.014) 0.82 0.000 (0.017) 0.99 0.008 (0.026) 0.75Triglycerides, mmol/L 0.029 (0.015) 0.04 0.044 (0.019) 0.023 0.004 (0.023) 0.87Glucose, mmol/L −0.017 (0.014) 0.21 −0.016 (0.017) 0.33 −0.019 (0.023) 0.42

Malmö Preventive Project (reinvestigation)Systolic BP, mmHg 0.016 (0.017) 0.34 0.026 (0.021) 0.22 −0.001 (0.030) 0.97Diastolic BP, mmHg 0.029 (0.018) 0.11 0.044 (0.021) 0.04 0.001 (0.028) 0.98Waist, cm −0.002 (0.015) 0.92 0.005 (0.018) 0.77 −0.014 (0.027) 0.59Triglycerides, mmol/L 0.0.22 (0.015) 0.13 0.028 (0.019) 0.14 0.012 (0.023) 0.61Glucose, mmol/L 0.006 (0.017) 0.75 0.028 (0.023) 0.22 −0.033 (0.026) 0.21HDL-cholesterol, mmol/L −0.005 (0.015) 0.76 −0.012 (0.017) 0.49 0.008 (0.028) 0.76MetS score 0.035 (0.057) 0.55 0.113 (0.070) 0.11 −0.103 (0.097) 0.29

B All ths

t245

s246

4247

248

m249

f250

d251

r252

w253

u254

c255

w256

257

a258

p259

fi260

t261

c262

m263

p264

p265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

TOm

MS

P, blood pressure; HDL, high density lipoproteins; MetS, metabolic syndrome. N.B.ubject treated by specific drugs (see Section 2 for detailed description).

riglycerides ( ± SE: 0.039 ± 0.020; p = 0.054) remained borderlineignificant.

. Discussion

This study provides evidence that a common functional poly-orphism of the CYP4F2 V433M, might be associated with several

eatures of the MetS and MetS itself, at least in males However, aefinitive conclusion about the effect of this polymorphism shouldemain cautious: in fact, after a clear result in the MDC-CVAhere male M-carriers were associated with all the MetS individ-al components with the exception of glucose, the result was lessonvincing in the MPP where only diastolic BP and triglyceridesere higher in male carriers of the CYP4F2 M-allele.

This CYP4F2 V433M SNP is already well known for its associ-tion with BP/hypertension and stroke but several considerationsrompted us to test its effect on the other features of the MetS:rst, the CYP4F2 enzyme is expressed also in the liver where par-icipates in the -hydroxylation of medium and very long/long

Please cite this article in press as: Fava C, et al. The functional variant V433MOther Lipid Mediat (2012), http://dx.doi.org/10.1016/j.prostaglandins.2012

hain fatty acids (FA), saturated and unsaturated FA which areetabolized by oxidation to shorter chain FA incorporated into

hospholipids, triglycerides and cholesterol esters. Normally thisathway is of minor importance (5–15%) in the metabolism of FA

able 3dds ratio and 95% CI (on parenthesis) for Metabolic syndrome conferred by the CYP4Fodel in the whole population and after stratification for gender.

Whole population p-value

Malmö Diet and CancerMetS, ATPIII definition 1.193 (1.056–1.349) 0.005

MetS, IDF definition 1.111 (0.984–1.253) 0.09

MetS, EGIR definition 1.046 (0.871–1.256) 0.63

MetS, AHA/NHLBI definition 1.135 (1.011–1.275) 0.03

Malmö Preventive Project (reinvestigation)MetS, ATPIII definition 1.046 (0.980–1.115) 0.18

MetS, IDF definition 1.040 (0.981–1.104) 0.19

MetS, AHA/NHLBI definition 1.056 (0.995–1.121) 0.07

etS, metabolic syndrome; NCEP, National Cholesterol Education Program/Adult Treatmtudy of Insulin Resistance; AHA/NHLBI, American Heart Association/National Heart, Lun

e standardized residuals were adjusted for age and sex and 1 SD was added to each

but can increase markedly during starvation, ethanol intake and

in different metabolic disorders [10]. Moreover in animals CYP4F

can be induced by insulin and high fat diet can differently regulate

the gene expression of CYPs since it increases CYP4A but suppressCYP4F, a mechanism that has been implicated in liver protection

from steatosis and steatohepatitis [10]. Second, it is established the

link between MetS and a low grade inflammatory state: CYP4F2

can metabolize both pro and anti-inflammatory leukotrienes along

with other pro-inflammatory eicosanoids including 5-HETE, 12-

HETE and 8-HETE [6].

Third, the fact that in our previous study the effect of this poly-

morphism on the incidence of stroke resulted independent from

hypertension raised the suspicion that its action could be mediated

by another, still unknown, metabolic risk factor [4].

However, the fact that the CYP4F2 V433M is associated not only

with MetS as defined by the ATPIII and AHA/NHLBI but also with all

the features of the MetS except from glucose, at least in the MDC-

CVA, was mostly unexpected. To our knowledge there are only a

few SNPs simultaneously associated with 3 or more features of the

of the CYP4F2 and the metabolic syndrome in Swedes. Prostaglandins.03.001

MetS in the same population [17]. On the other hand, the results 285

for HDL-cholesterol and waist were no longer present in the MPP 286

where only Diastolic BP and triglycerides were slightly higher in 287

male carriers of the CYP4F2 M-allele. 288

2 V433M polymorphism tested by logistic regression according to the AD genetic

Males p-value Females p-value

1.285 (1.075–1.535) 0.006 1.117 (0.944–1.323) 0.201.143 (0.958–1.365) 0.14 1.051 (0.889–1.244) 0.561.132 (0.87–1.474) 0.36 0.972 (0.753–1.255) 0.831.205 (1.015–1.431) 0.03 1.080 (0.923–1.265) 0.34

1.074 (0.991–1.164) 0.08 1.037 (0.935–1.151) 0.491.031 (0.958–1.111) 0.41 1.064 (0.963–1.175) 0.221.060 (0.985–1.142) 0.12 1.058 (0.958–1.169) 0.27

ent Panel III; IDF, International Diabetes Federation; EGIR, European Group for theg and Blood Institute (AHA/NHLBI). Adjustment for age, sex.

ING Model

P

her Lip

289

M290

s291

h292

g293

294

s295

c296

p297

m298

(299

[300

301

t302

o303

a304

i305

s306

c307

t308

r309

v310

311

l312

p313

o314

e315

e316

f317

d318

p319

i320

f321

a322

K323

m324

325

l326

l327

c328

L329

(330

s331

332

s333

r334

i335

l336

s337

s338

e339

b340

y341

r342

343

l344

s345

i346

b347

p348

a349

d350

y351

352

l353

a354

355

356

357

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

402

403

404

405

406

407

408

ARTICLERO 5983 1–6

C. Fava et al. / Prostaglandins & ot

The results about BP confirm our previous findings in theDC-CVA and also in the MPP was evident only in males further

upporting the idea that this Cytochrome could be regulated by sexormones as pointed out in animal models [18,19] and in otherenetic epidemiological studies [20].

Recent GWAS on MetS itself and individual components haveeldom pointed out loci close to the CYP4F2 which maps onhromosome 19p13.11. The SNP rs16996148 and rs17216525,reviously associated with triglycerides level are in the same chro-osomal region but 5,000,000 bp away and with no apparent LD

according to the Hapmap project) with respect to the rs210862221–23].

The studies that till now explored the link between metabolicraits of the MetS and hypertension focused especially on the rolef insulin and insulin resistance acting as salt sensitizers. In fact thentinatriuretic effect of insulin is exacerbated in states of metabolicnsulin resistance, representing a major mechanism for blood pres-ure elevation in this setting [24]. On the other hand hypertensionan cause pruning of the microvascular tree with capillary rarefac-ion provoking a restriction in maximum tissue perfusion and mayesult in impaired glucose extraction by skeletal muscle creating aicious circle [25].

Our data did not show any effect of the SNP on glucose, insulinevels or HOMA index (in the MDC-CVA ± SE: 0.001 ± 0.007;

= 0.89; and ± SE: −0.001 ± 0.008; p = 0.93) as a crude indexf insulin resistance. Thus, our data suggest a different way toxplore the link between CYP4F2 and both BP and MetS: the CYP4F2nzyme, at renal tubular level increase the reabsorption of saltavoring hypertension, whereas at hepatic level could metabolizesifferent lipid compounds, including triglycerides influencing theirlasma level. Moreover, a recent work emphasizes the functional

mportance of the hepatic CYP4F2 V433M polymorphism since war-arin dosage had to be augmented in people carriers of the 433Mllele because they have decreased capacity to metabolize vitamin

[26]. Unfortunately, we could not dose 20-HETE or other lipidediators derived by CYP4F2 in our populations.CYP4F2 is known to metabolize the leukotriens such as the

eukotriene B4 (LTB4) and the V433M could be expected to modu-ate inflammatory responses but in their in vitro work, Stec andoauthor did not find this polymorphism differently metabolizeTB4 [6]. It could also be that other SNPs in linkage disequilibriumLD) with the CYP4F2 V433M are implicated even if functional datauggest it is the major candidate [6].

Most of the positive associations we found in MDC-CVA wereuccessively partially denied in another large urban-based survey,ising the suspicion of a false positive results in the first cohort. Its important to emphasize some differences between these popu-ation: in particular the mean age was different between all thesetudies and this effect can influence all the analyzed parameters,econdly being MetS one of the most dangerous risk factor for CVvents, it is worth taking into account that 551 people died in MPPefore the reinvestigation and did not give their DNA for the anal-sis: thus, it is possible that this bias could have altered also ouresults.

We have to acknowledge some other limitations of our popu-ation studies. More than 2000 subjects (nearly 15%) of the MPPample were previously investigated in the MDC-CVA, thus includ-ng these subjects in the final analysis could result in a “duplication”ias, while deleting them could result in an “inappropriate sam-ling” bias. Probably due to the loss of statistical power the resultsbout DBP and triglycerides were only borderline significant wheneleting these subjects but seems consistent between the two anal-

Please cite this article in press as: Fava C, et al. The functional variant V433MOther Lipid Mediat (2012), http://dx.doi.org/10.1016/j.prostaglandins.2012

ses.Different ways exist to overcome the bias derived from ana-

yzing continuous values of subjects treated by antihypertensive,ntidiabetic and antilipemic drugs [27]. In our previous studies we

PRESSid Mediators xxx (2012) xxx– xxx 5

adopted different BP adjustment, on the basis of a method pro-

posed by Harrap and colleagues with mostly similar results. In this

study we chose to standardize all continuous trait and to adjust

for specific medications just adding 1 SD to the measured values.

This method could be seen as an oversimplification but we know

that by excluding these subjects we risk to lose important data

about the most informative subjects, and by including them with-

out adjusting their values we do not take into account the effect of

drugs.

In any case the results were mostly similar also when the treated

subjects were excluded from the analysis or the continuous value is

not adjusted for specific medication apart from BP that was no more

significant if treated subjects were excluded (data not shown).

Moreover, MetS is a heterogeneous condition characterized

by different metabolic and haemodynamic disturbances pooled

together with the presumption to share insulin resistance as theunifying factor; however the final prove it is a definite disease andnot the sum of different conditions is lacking. Nevertheless, what-ever the utilized definition a strong association of the MetS with

cardiovascular mortality [28] and morbidity [16,29,30] has beenfound. The use of all the major current definitions was an attempt

to circumvent any bias derived from misclassification.

In conclusion, in a large urban based survey a common V433M

polymorphism of the CYP4F2 increased the risk of MetS individual

components, a composite MetS score and MetS defined according to

the ATPIII and AHA/NHLBI definitions. In a large confirmatory study

most of these findings were denied but a positive result remained

at least for triglycerides and BP. Both false positive and false neg-

ative results could explain these contradictory results. Moreover

MetS, is a syndrome difficult to define, as witnessed by the con-

tinuous attempts of definition by different Scientific Societies, and

this could add to the complication. Our study underlines how it is

difficult to get undoubted results on genetic findings even if large

samples are utilized. Further studies in different populations are

warranted.

Disclosures statement

GE is employed as senior epidemiological scientist at

AstraZeneca R&D.

Other authors have nothing to disclose.

Funding

This study was supported by grants from the Swedish Medi-

cal Research Council, the Swedish Heart and Lung Foundation, the

Medical Faculty of Lund University, Malmö University Hospital, the

Albert Påhlsson Research Foundation, the Crafoord Foundation, the

Ernhold Lundströms Research Foundation, the Region Skane, Hulda

and Conrad Mossfelt Foundation and King Gustaf V and Queen

Victoria Foundation. The Knut and Alice Wallenberg Foundation

provided economic support of the SWEGENE DNA extraction facil-

ity.

Appendix A. Supplementary data

Supplementary data associated with this article can be

found, in the online version, at http://dx.doi.org/10.1016

/j.prostaglandins.2012.03.001.

References

of the CYP4F2 and the metabolic syndrome in Swedes. Prostaglandins.03.001

[1] McGiff JC, Quilley J. 20-HETE and the kidney: resolution of old problems and 409

new beginnings. Am J Physiol 1999;277:R607–23. 410

[2] Minuz P, Jiang H, Fava C, et al. Altered release of cytochrome p450 metabolites 411

of arachidonic acid in renovascular disease. Hypertension 2008;51:1379–85. 412

ING Model

P

6 her Lip

413

414

415

416

417

418

419

420

421

422

423

424

425

426

427

428

429

430

431

432

433

[434

435

436

[437

438

439

440

441

442

[443

444

445

[446

447

[448

449

450

451

[452

453

[ 454

455

456

[ 457

458

459

[ 460

461

462

[ 463

464

465

466

[ 467

468

469

470

[ 471

472

[ 473

474

[ 475

476

477

[ 478

479

480

[ 481

482

[ 483

484

485

[ 486

487

[ 488

489

490

ARTICLERO 5983 1–6

C. Fava et al. / Prostaglandins & ot

[3] Ward NC, Tsai I, Barden A, et al. A single nucleotide polymorphism in the CYP4F2but not CYP4A11 gene is associated with increased 20-HETE excretion andblood pressure. Hypertension 2008;51:1393–8.

[4] Fava C, Montagnana M, Almgren P, et al. The V433M variant of the CYP4F2is associated with ischemic stroke in male Swedes beyond its effect on bloodpressure. Hypertension 2008;52:373–80.

[5] Ding H, Cui G, Zhang L, et al. Association of common variants of CYP4A11 andCYP4F2 with stroke in the Han Chinese population. Pharmacogenet Genomics2010;20:187–94.

[6] Stec DE, Roman RJ, Flasch A, Rieder MJ. Functional polymorphism inhuman CYP4F2 decreases 20-HETE production. Physiol Genomics 2007;30:74–81.

[7] Niu W, Gao P, Zhu D. Haplotype-based case-control study of the humanCYP4F2 gene and essential hypertension in Japanese subjects. Hypertens Res2009;32:529.

[8] Fu Z, Nakayama T, Sato N, et al. Haplotype-based case-control study of thehuman CYP4F2 gene and essential hypertension in Japanese subjects. Hyper-tens Res 2008;31:1719–26.

[9] Melander O, Groop L, Hulthén UL. Effect of salt on insulin sensitivitydiffers according to gender and degree of salt sensitivity. Hypertension2000;35:827–31.

10] Zhang X, Chen L, Hardwick JP. Promoter activity and regulation of the CYP4F2leukotriene B(4) omega-hydroxylase gene by peroxisomal proliferators andretinoic acid in HepG2 cells. Arch Biochem Biophys 2000;378:364–76.

11] Alberti KG, Eckel RH, Grundy SM, et al. Harmonizing the metabolic syndrome: ajoint interim statement of the International Diabetes Federation Task Force onEpidemiology and Prevention; National Heart, Lung, and Blood Institute; Amer-ican Heart Association; World Heart Federation; International AtherosclerosisSociety; and International Association for the Study of Obesity. Circulation2009;120:1640–5.

12] Fava C, Montagnana M, Guerriero M, et al. Chromosome 2q12, theADRA2B I/D polymorphism and metabolic syndrome. J Hypertens 2009;27:1794–803.

13] Berglund G, Elmstahl S, Janzon L, Larsson SA. The Malmo Diet and Cancer studydesign and feasibility. J Intern Med 1993;233:45–51.

14] Hedblad B, Nilsson P, Janzon L, Berglund G. Relation between insulin resis-tance and carotid intima-media thickness and stenosis in non-diabetic

Please cite this article in press as: Fava C, et al. The functional variant V433MOther Lipid Mediat (2012), http://dx.doi.org/10.1016/j.prostaglandins.2012

subjects results from a cross-sectional study in Malmo, Sweden. Diabet Med2000;17:299–307.

15] Berglund G, Eriksson KF, Israelsson B, et al. Cardiovascular risk groups and mor-tality in an urban Swedish male population: the Malmo Preventive Project. JIntern Med 1996;239:489–97.

[

[

PRESSid Mediators xxx (2012) xxx– xxx

16] Nilsson PM, Engstrom G, Hedblad B. The metabolic syndrome and incidence

of cardiovascular disease in non-diabetic subjects—a population-based study

comparing three different definitions. Diabet Med 2007;24:464–72.

17] Monda KL, North KE, Hunt SC, Rao DC, Province MA, Kraja AT. The genetics

of obesity and the metabolic syndrome. Endocr Metab Immune Disord Drug

Targets 2010;10:86–108.

18] Singh H, Cheng J, Deng H, et al. Vascular cytochrome P450 4A expression and 20-

hydroxyeicosatetraenoic acid synthesis contribute to endothelial dysfunctionin androgen-induced hypertension. Hypertension 2007;50:123–9.

19] Nakagawa K, Marji JS, Schwartzman ML, Waterman MR, Capdevila JH.

Androgen-mediated induction of the kidney arachidonate hydroxylases is asso-

ciated with the development of hypertension. Am J Physiol Regul Integr CompPhysiol 2003;284:R1055–62.

20] Fava C, Ricci M, Melander O, Minuz P. Hypertension, cardiovascular risk and

polymorphisms in genes controlling the cytochrome P450 pathway of arachi-

donic acid: a sex-specific relation? Prostaglandins Other Lipid Mediat 2011.November 30 [Epub ahead of print].

21] Willer CJ, Sanna S, Jackson AU, et al. Newly identified loci that influence lipid

concentrations and risk of coronary artery disease. Nat Genet 2008;40:161–9.

22] Kathiresan S, Willer CJ, Peloso GM, et al. Common variants at 30 loci contribute

to polygenic dyslipidemia. Nat Genet 2009;41:56–65.

23] Kathiresan S, Melander O, Guiducci C, et al. Six new loci associated with blood

low-density lipoprotein cholesterol, high-density lipoprotein cholesterol ortriglycerides in humans. Nat Genet 2008;40:189–97.

24] Sarafidis PA, Bakris GL. The antinatriuretic effect of insulin: an unappreciated

mechanism for hypertension associated with insulin resistance. Am J Nephrol

2007;27:44–54.

25] Mayet J, Hughes A. Cardiac and vascular pathophysiology in hypertension.

Heart 2003;89:1104–9.

26] McDonald MG, Rieder MJ, Nakano M, Hsia CK, Rettie AE. CYP4F2 is a vitamin

K1 oxidase: an explanation for altered warfarin dose in carriers of the V433Mvariant. Mol Pharmacol 2009;75:1337–46.

27] Cui J, Hopper J, Harrap S. Antihypertensive treatments obscure familial contri-

butions to blood pressure variation. Hypertension 2003;41:207–10.

28] Ford ES. The metabolic syndrome and mortality from cardiovascular disease

and all-causes: findings from the National Health and Nutrition Examination

Survey II Mortality Study. Atherosclerosis 2004;173:309–14.

of the CYP4F2 and the metabolic syndrome in Swedes. Prostaglandins.03.001

29] Noto D, Barbagallo CM, Cefalu AB, et al. The metabolic syndrome predicts car- 491

diovascular events in subjects with normal fasting glucose: results of a 15 years 492

follow-up in a Mediterranean population. Atherosclerosis 2008;197:147–53. 493

30] Dekker JM, Girman C, Rhodes T, et al. Metabolic syndrome and 10-year cardio- 494

vascular disease risk in the Hoorn Study. Circulation 2005;112:666–73. 495