JASN Express. Published on April 2, 2008 as doi: 10.1681/ASN.2008010063

CLINICAL RESEARCH www.jasn.org

Association of a CYP4A11 Variant and Blood Pressure in Black Men

ʈ James V. Gainer,*† Michael S. Lipkowitz,‡ Chang Yu,§ Michael R. Waterman, Elliott P. Dawson,¶ Jorge H. Capdevila,** Nancy J. Brown,* and the AASK Study Group

Divisions of *Clinical Pharmacology and **Nephrology, Department of Medicine, and Departments of §Biostatistics ʈ and Biochemistry, Vanderbilt University Medical Center, and †Veterans Administration Hospital, Nashville, Tennessee; and ¶BioVentures, Inc., and the ‡Division of Nephrology, Department of Medicine, Mount Sinai School of Medicine, New York, New York

ABSTRACT CYP4A11 monooxygenase oxidizes endogenous arachidonic acid to 20-hydroxyeicosa- tetraenoic acid, a renal vasoconstrictor and natriuretic. Cyp4a deficiency causes hypertension in male mice, and a loss-of-function variant (T8590C) of CYP4A11 is associated with hypertension in white individuals. Hypertension and hypertensive renal disease are more common among black than white individuals, but the relationship between genetic variation at CYP4A11 and hypertension in black individuals is not known. This study tested the hypothesis that the CYP4A11 T8590C polymorphism is associated with higher BP or clinical outcomes in 732 black Americans with hypertensive renal disease participating in the African American Study of Kidney Disease (AASK). Men with the 8590CC genotype had significantly higher systolic BP (CC 156.5 Ϯ 22.6 versus 148.4 Ϯ 24.3 mmHg in CT and TT combined; P ϭ 0.04) and pulse pressure (P ϭ 0.04) at baseline; this association was not observed among women. In addition, this genotype was associated with higher systolic and diastolic BP at 36-mo follow-up among those randomly assigned to the lower BP arm of the AASK. Among all participants (or men but not women) with proteinuria, the 8590CC genotype was associated with an increased cumulative incidence of ESRD or death, controlling for randomization and clinical characteristics. In summary, the CYP4A11 8590CC genotype is associated with increased BP in black men with hypertensive nephrosclerosis and is associated with adverse clinical outcomes in those with baseline proteinuria. These data support a role for renal monooxygenases and 20-hydroxyeicosatetraenoic acid in the regulation of BP and renal function in men.

J Am Soc Nephrol ●●: –, 2008. doi: 10.1681/ASN.2008010063

Hypertension is a significant independent risk for ase, expressed in the human kidney, catalyzes the the development and progression of chronic kidney conversion of endogenous arachidonic acid to 20- disease.1 Studies in men and women have demon- hydroxyeicosatetraenoic acid (20-HETE). The 20- strated a strong, graded relationship between in- HETE metabolite regulates salt and water ho- creasing hypertension severity and declining renal meostasis at multiple sites within the kidney.12 function.2 Differences in the incidence and progres- sion of chronic kidney disease among ethnic groups Received January 17, 2008. Accepted February 26, 2008. have been reported,3–5 and black Americans are Published online ahead of print. Publication date available at four times as likely to develop ESRD as white Amer- www.jasn.org. icans.6 Animal7–9 and human studies10,11 support a Correspondence: Dr. Nancy J. Brown, 550 Robinson Research causal role for genetic factors, but the genetic deter- Building, Vanderbilt University Medical Center, Nashville, TN minants of hypertension-associated renal disease in 37232-6602. Phone: 615-343-8701; Fax: 615-343-2551; E-mail: humans are largely unknown. [email protected] The CYP4A11 arachidonic acid monooxygen- Copyright ᮊ 2008 by the American Society of Nephrology

J Am Soc Nephrol ●●: –, 2008 ISSN : 1046-6673/●●00- 1 CLINICAL RESEARCH www.jasn.org

Experimental models of hypertension indicate that 20-HETE can act in either a pro- or antihypertensive manner, depending on its site-specific expression in the renal vasculature or tubule, respectively.13–16 In previous work, we showed that knockout of Cyp4a14, a murine homologue of CYP4A11, caused gender- specific and androgen-sensitive hypertension.17 Several studies supported the involvement of 20-HETE, CYP4A, and omega hydroxylase activity in animal models of renal injury.18–20 In humans, we21 previously reported an association of a functional CYP4A11 variant, T8590C, characterized by re- duced 20-HETE synthase activity, with hypertension in two independent white American cohorts. Mayer et al.22 also demon- strated an association of the T8590C variant with hypertension and systolic BP (SBP) in a German-based white population co- hort. In our original study, we did not detect an association of CYP4A11 T8590C genotype with hypertension in an black Amer- ican cohort, but the study was limited by a small sample size. The primary goal of this study was to determine whether genetic variation in CYP4A11 (i.e., T8590C variant) influences hemodynamic indices or the progression of renal disease in hypertensive black Americans and whether gender-dependent genotype effects are detectable. To this end, we genotyped 732 participants of the African American Study of Kidney Disease (AASK), a trial designed to evaluate the effect of two different BP goals and three different treatment regimens on progres- sion of hypertensive kidney disease. We assessed the effect of the CYP4A11 T8590C variant on BP and on clinical end points.

Figure 1. Relationship between CYP4A11 T8590C genotype RESULTS and prerandomization SBP, DBP, and pulse pressure (PP) in men and women. *P Ͻ 0.05 versus CT and TT groups. Baseline Characteristics previously in black individuals (Table 1).21 There were no sig- CYP4A11 T8590C genotype and allele frequencies were in nificant differences among genotype groups in age, gender dis- Hardy-Weinberg equilibrium and similar to those reported tribution, body mass index (BMI), baseline renal function, his-

Table 1. Patient characteristicsa P (65 ؍ CC (n (322 ؍ CT (n (345 ؍ Characteristic TT (n Age (yr) 53.6 Ϯ 10.9 53.9 Ϯ 10.2 53.7 Ϯ 10.9 0.90 Women (%) 40.0 41.0 35.4 0.70 BMI (kg/m2) 30.9 Ϯ 6.3 31.4 Ϯ 6.9 29.6 Ϯ 6.3 0.12 GFR (ml/min per 1.73 m2) 47.6 Ϯ 13.5 47.7 Ϯ 13.9 46.1 Ϯ 13.5 0.66 Serum creatinine (mg/dl)b 2.02 Ϯ 0.73 1.98 Ϯ 0.69 2.05 Ϯ 0.79 0.60 UP/Cr† 0.32 Ϯ 0.53 0.31 Ϯ 0.49 0.27 Ϯ 0.48 0.74 History of heart disease (%) 48.4 50.3 50.8 0.87 Years of hypertension 13.5 Ϯ 10.3 13.9 Ϯ 9.3 15.4 Ϯ 11.3 0.38 Antihypertensive meds 2.5 Ϯ 1.2 2.5 Ϯ 1.1 2.7 Ϯ 1.1 0.23 Antihypertensive use (%) angiotensin-converting enzyme inhibitor 44.6 38.9 35.6 0.21 ␤ blocker 27.4 30.6 40.7 0.11 calcium channel blocker 65.8 60.5 71.2 0.18 dihydropyridine calcium channel blocker 48.2 43.6 54.2 0.24 diuretic 62.2 69.4 72.9 0.08 Randomization to ramipril:metoprolol:amlodipine (%) 43.8:38.3:18.0 39.4:41.0:19.6 27.7:43.1:29.2 0.10 Randomization to low BP group (%) 50.4 53.7 46.2 0.46 aData are means Ϯ SD unless otherwise indicated. UP/Cr, urine to creatinine ratio at baseline. bTo convert creatinine mg/dl to ␮ mol/L, multiply by 88.4. To convert urine protein g/d to mg/d, multiply by 1000.

2 Journal of the American Society of Nephrology J Am Soc Nephrol ●●: –, 2008 www.jasn.org CLINICAL RESEARCH

Table 2. CYP4A11 genotype and BP during follow-upa Genotype 3Mo 36Mo Parameter TT CT CC P TT CT CC P (57 ؍ n) (292 ؍ n) (314 ؍ n) (65 ؍ n) (322 ؍ n) (345 ؍ n) Usual BP medications, n men 1.7 Ϯ 1.3b 2.1 Ϯ 1.5 1.8 Ϯ 1.2 0.18 2.1 Ϯ 1.5b 2.3 Ϯ 1.7b 2.0 Ϯ 1.2b 0.46 women 1.5 Ϯ 1.3 1.3 Ϯ 1.1b,c 1.8 Ϯ 1.5 0.32 2.0 Ϯ 1.5 2.0 Ϯ 1.5b 2.4 Ϯ 1.7 0.72 all 1.6 Ϯ 1.3b 1.8 Ϯ 1.4b 1.8 Ϯ 1.3 0.48 2.1 Ϯ 1.5b 2.2 Ϯ 1.6b 2.1 Ϯ 1.4 0.74 SBP (mmHg) men 139.1 Ϯ 16.5b 143.4 Ϯ 16.1b 146.3 Ϯ 12.0b 0.07 140.2 Ϯ 13.7b 142.3 Ϯ 17.1b 143.3 Ϯ 9.0 0.58 women 139.4 Ϯ 17.7b 145.7 Ϯ 19.0 145.3 Ϯ 14.7b 0.12 141.7 Ϯ 15.9b 143.9 Ϯ 20.6b 135.2 Ϯ 12.5 0.36 all 139.3 Ϯ 17.0b 144.3 Ϯ 17.3b,d 145.9 Ϯ 12.8b,d 0.01 140.9 Ϯ 14.7b 142.9 Ϯ 18.5b 140.5 Ϯ 10.9 0.54 DBP (mmHg) men 87.7 Ϯ 9.8b 89.4 Ϯ 10.9b 88.7 Ϯ 11.5b 0.53 86.8 Ϯ 10.5b 87.2 Ϯ 10.5b 87.1 Ϯ 9.5 0.97 women 87.3 Ϯ 11.2b 84.8 Ϯ 11.0† 90.4 Ϯ 10.8b 0.21 84.8 Ϯ 9.3b 82.1 Ϯ 11.6c 87.2 Ϯ 8.8b 0.23 all 87.5 Ϯ 10.4b 87.7 Ϯ 11.2b 89.3 Ϯ 11.1b 0.66 85.9 Ϯ 10.0b 85.2 Ϯ 11.2b 87.1 Ϯ 9.1 0.66 Lower BP medications, n men 2.5 Ϯ 1.4 2.3 Ϯ 1.1 2.2 Ϯ 1.3 0.55 2.9 Ϯ 1.5 3.1 Ϯ 1.5 2.9 Ϯ 1.2 0.57 women 1.7 Ϯ 1.3c 1.8 Ϯ 1.3c 1.5 Ϯ 0.9 0.71 2.4 Ϯ 1.6 2.7 Ϯ 1.5 2.0 Ϯ 1.1 0.31 all 2.2 Ϯ 1.4 2.1 Ϯ 1.2 1.9 Ϯ 1.2 0.52 2.7 Ϯ 1.5 2.9 Ϯ 1.5 2.6 Ϯ 1.2 0.36 SBP (mmHg) men 128.2 Ϯ 17.2 126.8 Ϯ 19.9 131.7 Ϯ 20.2 0.55 124.6 Ϯ 17.0 123.3 Ϯ 17.0 139.1 Ϯ 24.0d,e 0.004 women 132.2 Ϯ 23.6 138.9 Ϯ 24.2c 125.6 Ϯ 15.2 0.10 126.4 Ϯ 15.2 135.5 Ϯ 22.8c,d 124.1 Ϯ 19.3 0.03 all 129.6 Ϯ 19.7 132.0 Ϯ 22.6 129.5 Ϯ 18.5 0.52 125.2 Ϯ 16.3 128.5 Ϯ 20.5 133.5 Ϯ 23.2 0.08 DBP (mmHg) men 80.2 Ϯ 12.7 80.1 Ϯ 12.9 80.0 Ϯ 14.9 1.00 76.8 Ϯ 10.6 75.9 Ϯ 12.2 83.8 Ϯ 16.4e 0.048 women 81.3 Ϯ 14.0 82.8 Ϯ 14.2 78.9 Ϯ 9.2 0.61 76.4 Ϯ 10.4 78.6 Ϯ 11.5 77.8 Ϯ 9.8 0.57 all 80.5 Ϯ 13.1 81.3 Ϯ 13.5 79.6 Ϯ 12.9 0.76 76.7 Ϯ 10.5 77.1 Ϯ 11.9 81.6 Ϯ 14.4 0.13 aData are means Ϯ SD. bP Ͻ 0.05 versus lower BP group. cP Ͻ 0.05 versus men. dP Ͻ 0.05 versus 8590TT. eP Ͻ 0.05 versus 8590CT. tory of heart disease, or antihypertensive medication use. higher in men with the 8590CC genotype compared with men CYP4A11 T8590C genotype tended to be associated with di- with CT and TT genotypes (Figure 1). uretic use. In men, BMI correlated with diastolic BP (DBP; r ϭ There was no difference in antihypertensive medication use 0.197, P Ͻ 0.001) but not SBP. There was no relationship be- among CYP4A11 genotypes at 3 or 36 mo (Table 2). At 3 mo tween BMI and SBP or DBP in women. after randomization, there was a significant relationship be- Randomization to treatment arm or to level of BP control tween CYP4A11 T8590C genotype and SBP in men and women was similar among T8590C genotype groups (Table 1). Pa- combined who were randomly assigned to usual BP control. In tients in the 8590CC genotype were more likely to have been patients who were randomly assigned to lower BP, SBP and randomly assigned to amlodipine (29.2 versus 18.7%; P ϭ DBP were significantly higher at 36 mo in men with the 0.04). This resulted from an imbalanced randomization to am- CYP4A11 8590CC genotype compared with men with CT and lodipine of patients without baseline proteinuria (32.7% of CC TT genotypes, whereas SBP was highest in women with the versus 18.2% of CT and TT; P ϭ 0.02); randomization to amlo- 8590CT genotype. dipine was similar in the CC versus the CT and TT groups among patients with baseline proteinuria (18.8 versus 20.2%; P ϭ 0.93). Clinical End Points Overall, there was no effect of CYP4A11 T8590C genotype on Blood Pressure clinical end points (Table 3). This was true when analyzed sep- There was no effect of CYP4A11 T8590C genotype on baseline arately in patients with and without baseline proteinuria. BP in the combined group. Analyzed separately in men and There was a NS trend toward a decreased incidence of GFR women, however, there was a significant effect of T8590C ge- events in patients with the 8590CC genotype and without base- notype BP in men, such that SBP and pulse pressure were line proteinuria.

J Am Soc Nephrol ●●: –, 2008 CYP4A11 Variant in AASK 3 CLINICAL RESEARCH www.jasn.org

Table 3. Number and rate of clinical eventsa (Parameter TT CT CC P P (CC versus CT ؉ TT GFR total 60 (17.4) 59 (18.3) 6 (9.2) 0.20 0.08 UP/Cr Ͼ0.22b 34 (32.4) 36 (34.6) 5 (31.3) 0.94 0.84 ESRD total 42 (12.2) 47 (14.6) 8 (12.3) 0.64 0.81 UP/Cr Ͼ0.22b 33 (31.4) 36 (34.6) 7 (43.8) 0.63 0.39 Death total 4 (1.2) 3 (0.9) 2 (3.1) 0.35 0.19 UP/Cr Ͼ0.22b 1 (1.0) 0 (0.0) 1 (6.3) 0.14 0.14 ESRD or death total 46 (13.3) 50 (15.5) 10 (15.4) 0.71 0.83 UP/Cr Ͼ0.22b 34 (32.4) 36 (34.6) 8 (50.0) 0.40 0.19 GFR, ESRD or death total 81 (23.5) 79 (24.5) 12 (18.5) 0.57 0.32 UP/Cr Ͼ0.22b 49 (46.7) 50 (48.1) 9 (56.3) 0.79 0.50 aData are n (%). bA baseline ratio of 0.22 corresponds approximately to proteinuria of 300 mg/d. For the end point death, the Fisher exact test was used because expected cell counts were Ͻ5.

Figure 2 shows the cumulative incidence of ESRD and death regression, which included drug treatment group, urinary pro- by CYP4A11 T8590C genotype group for all patients and for tein-to-creatinine ratio, and the presence or absence of electro- those with baseline proteinuria. There was no effect of geno- cardiographic evidence of myocardial infarction, the P value type on the cumulative incidence of ESRD or death for all for CYP4A11 T8590C was 0.06. Among patients with baseline patients combined (P ϭ 0.61). In a Cox proportional hazard proteinuria, CYP4A11 8590CC genotype was associated with an increased risk for ESRD and death over time (hazard ratio 3.16; P ϭ 0.004), as were drug treatment group (hazard ratio 2.44; P ϭ 0.04), urinary protein-to-creatinine ratio (P Ͻ 0.001), and the presence or absence of electrocardiographic evidence of previous myocardial infarction (P ϭ 0.02). Base- line mean arterial pressure (MAP), BMI, BP group, and age were NS and were excluded from the model. When analyzed separately for men and women with proteinuria, the effect of CYP4A11 8590CC genotype on the cumulative incidence of ESRD or death was statistically significant for men (P ϭ 0.003) but not for women (P ϭ 0.20) by Cox regression analysis. When MAP at 36 mo, rather than BP group, was included in the model, CYP4A11 8590CC genotype (P ϭ 0.007), baseline MAP (P ϭ 0.005), drug treatment group (0.02), urinary pro- tein-to-creatinine ratio (P ϭ 0.001), and the presence or ab- sence of electrocardiographic evidence of myocardial infarc- tion (P ϭ 0.04) significantly predicted progression to ESRD or death for men with proteinuria. MAP at 36 mo was NS. There was no effect of T8590C genotype on the cumulative incidence of the combined end point of GFR event, ESRD, or death either for all patients or for those with proteinuria (data not shown).

DISCUSSION

CYP4A11 catalyzes the metabolism of arachidonic acid to 20- Figure 2. Product limit estimate of ESRD and death for all HETE in humans. A functional CYP4A11 C variant at nucleo- patients (above) and for those with a baseline urinary protein-to- tide 8590, characterized by decreased 20-HETE synthase activ- creatinine ratio (UP/Cr) Ͼ0.22 (below). ity, has been associated with hypertension in three white

4 Journal of the American Society of Nephrology J Am Soc Nephrol ●●: –, 2008 www.jasn.org CLINICAL RESEARCH populations: The Framingham Offspring Cohort, a Tennessee of 14,000 cases of seven common diseases identified 24 inde- Cohort, and the MONitoring trends and determinants In Car- pendent association signals at P Ͻ 5 ϫ 10Ϫ7 with six common diovascular disease (MONICA) Augsburg echocardiographic diseases but found no SNP associated with hypertension at that substudy.21,22 Here, we report that the CYP4A11 8590CC ge- level of significance.27 The authors considered the possibilities notype is associated with increased BP and with increased pro- that hypertension has fewer common risk alleles of larger effect gression to ESRD or death in male hypertensive black Ameri- size or that the study failed to detect susceptibility variants with cans with mild to moderate renal disease and proteinuria. large effect size because they were poorly tagged. In addition, In our previous report of an association of the CYP4A11 C hypertension may be more susceptible to the dilution effect allele with hypertension in the Framingham Offspring study resulting from the inclusion of individuals with hypertension cohort and in a Tennessee cohort, we modeled the C allele as a among control subjects. The effects of multiple susceptibility dominant allele, primarily because of the low frequency of the alleles with small effect size and of confounding by misclassifi- 8590CC genotype in these populations. In this study, the in- cation can be minimized by studying homogeneous, precisely creased frequency of the 8590CC genotype in the AASK cohort phenotyped populations such as the AASK cohort. together with the size of the cohort allowed us to explore other More dramatic than the effect of CYP4A11 T8590C variant models of association with BP. The association of the CYP4A11 on BP was the effect of the variant on the cumulative incidence 8590CC genotype with baseline SBP in men and with 36-mo of ESRD or death. As reported previously for the effect of SBP and DBP in men randomly assigned to low-salt intake is ramipril versus amlodipine on decline in GFR,28 the association remarkably similar to that observed among white European of ESRD or death with CYP4A11 genotype was observed only participants in the MONICA Augsburd echocardiographic in patients who had elevated urinary protein excretion at base- substudy.22 In that study, the investigators reported a recessive line. It is known that proteinuria is associated with endothelial effect of the C allele on SBP in men and on SBP and DBP in dysfunction and predicts risk for death as a result of cardiovas- men and women combined. As in the MONICA study, the cular disease.29 The association of the CYP4A11 8590CC geno- degree of elevation of SBP at baseline in men with the 8590CC type with ESRD and death could not be attributed to differ- genotype (8 mmHg) and of SBP (16 mmHg) and DBP (7.5 ences in baseline serum creatinine, GFR, or randomization to mmHg) at 36 mo in men with the 8590CC genotype in the level of BP control or treatment among genotype groups, be- lower BP group is clinically significant in AASK. cause these all were similar in patients with baseline protein- The finding that the loss-of-function CYP4A11 8590C vari- uria. Interestingly, genotype did not affect the cumulative in- ant associates with increased SBP supports the concept that the cidence of GFR events or the combined end point of GFR effect of 20-HETE on tubular sodium transport predominates event, ESRD, or death. On the basis of the magnitude of BP over its vasoconstrictor effect in determining sodium excretion elevation associated with the 8590CC genotype, it is attractive and BP in humans. Tubular CYP4A11 deficiency is associated to hypothesize that relative hypertension in patients of the with salt-sensitive hypertension in rodent models.23 Humans 8590CC genotype group contributed to the increased inci- with salt-sensitive hypertension demonstrate decreased uri- dence of ESRD or death. Although, as reported previously in nary 20-HETE excretion in response to furosemide.24 Al- AASK,30 randomization to usual or lower BP control did not though participants in AASK were not specifically character- affect the incidence of ESRD or death, baseline BP did predict ized according to the salt sensitivity of their BP, patients who progression to ESRD or death. Taken together with the finding carried a C allele tended to be taking a diuretic before random- that CYP4A11 8590CC genotype predicted progression to ization, suggesting the possibility that they were more likely to ESRD or death in men but not in women, the data suggest that have salt-sensitive or volume-dependent hypertension. the effect of the genotype on BP contributed to the effect of We used a single-nucleotide polymorphism (SNP) associa- genotype on these outcomes in men. tion analysis to determine the relationship between the The CYP4A11 8590CC genotype was associated with in- CYP4A11 gene and BP in the AASK population. Although it is creased baseline SBP in men but not in women. This interac- possible that haplotype analysis can provide increased power tive effect of CYP4A11 genotype and gender is consistent with to detect associations, this is not universal and depends on data from mice genetically deficient in Cyp4a14, the murine patterns of linkage disequilibrium among markers in the study homologue of CYP4A11.17 Disruption of the Cyp4a14 gene population in general and between the studied SNP and func- causes a significantly greater increase in systolic and diastolic tional variants.25 Fu et al.26 detected an association between the BP in male compared with female mice. In addition, castra- CYP4A11 gene and essential hypertension in Japanese men us- tion-normalized BP and androgen replacement restore hyper- ing haplotype analysis; however, the significance of the associ- tension in male Cyp4a14 (Ϫ/Ϫ) mice, consistent with andro- ation depended on the inclusion of the CYP4A11 T8590C ge- gen-dependent regulation of CYP4A arachidonic acid notype, suggesting that the haplotype analysis afforded no monooxygenase. In the MONICA Augsburg echocardio- advantage over the SNP analysis. graphic substudy, homozygosity for CYP4A11 8590C allele was Genome-wide association (GWA) studies provide a partic- associated with increased SBP in men but not in women.22 ularly powerful new approach to identify involved in Likewise, Fu et al.26 reported that the CYP4A11 T8590C poly- complex diseases such as hypertension. One such GWA study morphism is associated with essential hypertension in Japa-

J Am Soc Nephrol ●●: –, 2008 CYP4A11 Gene Variant in AASK 5 CLINICAL RESEARCH www.jasn.org nese men but not in Japanese women. In short, this study is the protein-to-creatinine ratio of 0.22 (a value corresponding approxi- third to report a gender-specific association of the CYP4A11 mately to a threshold for clinically significant proteinuria of 300 T8590C genotype with BP. mg/d) was used for subgroup analysis on the basis of the use of this In summary, we report the association of homozygosity for cutoff in the original comparison of ramipril versus amlodipine after a functional variant of the CYP4A11 with increased baseline BP an interim analysis of AASK detected significant interactions between in black American men with hypertensive nephrosclerosis, as outcome variables and baseline proteinuria.28 Cumulative incidence well as with an increased cumulative incidence of ESRD or rates for clinical end points were estimated using the product-limit death among those with baseline proteinuria. These data con- method. Cox proportional hazard regression analysis was used to de- firm an association between BP and the CYP4A11 gene previously termine the effect of CYP4A11 genotype and other clinical factors on reported in white cohorts and, thus, provide strong evidence of a the cumulative incidence of ESRD or death. Variables were excluded role for the renal monooxygenases and 20-HETE in the regulation using a backward conditional method. Analyses were performed us- of BP in humans. If confirmed prospectively in other populations, ing SPSS 15.0 for Windows (SPSS, Chicago, IL). This study was an then the data suggest that CYP4A11 genotype may be used to ancillary study to AASK, and, as such, the analyses were not per- stratify patients with hypertensive renal disease and proteinuria formed by the AASK Data Coordinating Center. according to their risk for ESRD and death.

ACKNOWLEDGMENTS CONCISE METHODS The rationale for the AASK and details of the study design have been This work was funded by National Institutes of Health grants reported previously.30–32 In brief, participants were self-identified black DK038226, DK048689, DK057867, HL065193, HL079184, Americans who had hypertension, were aged 18 to 70 yr, and had a GFR HL004221, RR000071, and RR000095. 2 between 20 and 65 ml/min per 1.73 m and no other identified causes of We thank Dr. Scott M. Williams for review of the manuscript. renal insufficiency. Exclusion criteria were (1) diastolic BP Ͻ95 mmHg; (2) known history of diabetes; (3) urinary protein-to-creatinine ratio Ͼ 2.5; (4) accelerated or malignant hypertension within 6 mo; (5) accel- DISCLOSURES erated, malignant, or secondary hypertension; (6) evidence of non–hy- E.P.D. is employed by BioVentures, Inc., a company that provides products pertension-related causes of renal disease; (7) serious systemic disease and technical support for genotyping. including congestive heart failure; or (8) specific indication for or contra- indication to a study drug or study procedure. The protocol was ap- proved by the institutional review board at each center, and all partici- REFERENCES pants gave written informed consent.30 The AASK used a 3 ϫ 2 factorial design in which participants were 1. Bakris GL, Williams M, Dworkin L, Elliott WJ, Epstein M, Toto R, Tuttle randomly assigned to a usual MAP goal of 102 to 107 mmHg or to a K, Douglas J, Hsueh W, Sowers J: Preserving renal function in adults Յ low MAP goal of 92 mmHg and to treatment with one of three with hypertension and diabetes: A consensus approach. National antihypertensive study drugs: A sustained-release ␤ blocker (meto- Kidney Foundation Hypertension and Diabetes Executive Committees prolol), an angiotensin-converting enzyme inhibitor (ramipril), or a Working Group. Am J Kidney Dis 36: 646–661, 2000 dihydropyridine calcium channel blocker (amlodipine). The primary 2. Haroun MK, Jaar BG, Hoffman SC, Comstock GW, Klag MJ, Coresh J: Risk factors for chronic kidney disease: A prospective study of 23,534 composite clinical end point was prespecified as the time from ran- men and women in Washington County, Maryland. J Am Soc Nephrol domization to any of the following: (1) GFR event, defined as a con- 14: 2934–2941, 2003 firmed Ն50% or 25-ml/min per 1.73 m2 decline in GFR; (2) ESRD, 3. Jones CA, McQuillan GM, Kusek JW, Eberhardt MS, Herman WH, defined as the need for replacement therapy; or (3) death.30,33,34 Dur- Coresh J, Salive M, Jones CP, Agodoa LY: Serum creatinine levels in ing the trial, the data safety and monitoring board recommended the the US population: Third National Health and Nutrition Examination Survey. Am J Kidney Dis 32: 992–999, 1998 termination of the amlodipine arm on the basis of a significant treat- 4. Coresh J, Astor BC, Greene T, Eknoyan G, Levey AS: Prevalence of 30 ment effect in individuals with baseline proteinuria. chronic kidney disease and decreased kidney function in the adult US population: Third National Health and Nutrition Examination Survey. Genotyping Am J Kidney Dis 41: 1–12, 2003 Each of 732 participants from whom DNA was available were geno- 5. Norris KC, Agodoa LY: Unraveling the racial disparities associated with kidney disease. Kidney Int 68: 914–924, 2005 typed at the CYP4A11 T8590C (also referred to as F434S; rs1126742) 6. US Renal Data System: USRDS 2005 Annual Data Report: Atlas of 21 locus using methods as described previously. End-Stage Renal Disease in the United States, Bethesda, National Institutes of Health, National Institute of Diabetes and Digestive and Statistical Analysis Kidney Diseases, 2005 Data are presented as means Ϯ SD unless otherwise noted. Continu- 7. Brown DM, Provoost AP, Daly MJ, Lander ES, Jacob HJ: Renal disease susceptibility and hypertension are under independent genetic con- ous variables were compared using one-way ANOVA or an unpaired trol in the fawn-hooded rat. Nat Genet 12: 44–51, 1996 ␹2 t test. Categorical variables were compared using testing or the 8. Shiozawa M, Provoost AP, van Dokkum RP, Majewski RR, Jacob HJ: Fisher exact test when the expected cell count was Ͻ5. A urinary Evidence of gene-gene interactions in the genetic susceptibility to

6 Journal of the American Society of Nephrology J Am Soc Nephrol ●●: –, 2008 www.jasn.org CLINICAL RESEARCH

renal impairment after unilateral nephrectomy. J Am Soc Nephrol 11: 24. Laffer CL, Laniado-Schwartzman M, Wang MH, Nasjletti A, Elijovich F: 2068–2078, 2000 Differential regulation of natriuresis by 20-hydroxyeicosatetraenoic 9. Eikmans M, Aben JA, Koop K, Baelde HJ, de HE, Bruijn JA: Genetic acid in human salt-sensitive versus salt-resistant hypertension. Circu- factors in progressive renal disease: The good ones, the bad ones and lation 107: 574–578, 2003 the ugly ducklings. Nephrol Dial Transplant 21: 257–260, 2006 25. Morris RW, Kaplan NL: On the advantage of haplotype analysis in the 10. DeWan AT, Arnett DK, Atwood LD, Province MA, Lewis CE, Hunt SC, presence of multiple disease susceptibility alleles. Genet Epidemiol Eckfeldt J: A genome scan for renal function among hypertensives: 23: 221–233, 2002 The HyperGEN study. Am J Hum Genet 68: 136–144, 2001 26. Fu Z, Nakayama T, Sato N, Izumi Y, Kasamaki Y, Shindo A, Ohta M, 11. Satko SG, Freedman BI, Moossavi S: Genetic factors in end-stage Soma M, Aoi N, Sato M, Ozawa Y, Ma Y: A haplotype of the CYP4A11 renal disease. Kidney Int Suppl S46–S49, 2005 gene associated with essential hypertension in Japanese men. J Hy- 12. McGiff JC, Quilley J: 20-HETE and the kidney: Resolution of old pertens 26: 453–461, 2008 problems and new beginnings. Am J Physiol 277: R607–R623, 1999 27. Genome-wide association study of 14,000 cases of seven common 13. Carroll MA, McGiff JC: A new class of lipid mediators: Cytochrome diseases and 3,000 shared controls. Nature 447: 661–678, 2007 P450 arachidonate metabolites. Thorax 55[Suppl 2]: S13–S16, 2000 28. Agodoa LY, Appel L, Bakris GL, Beck G, Bourgoignie J, Briggs JP, 14. Capdevila JH, Falck JR: The CYP P450 arachidonic acid monooxygen- Charleston J, Cheek D, Cleveland W, Douglas JG, Douglas M, Dowie ases: From cell signaling to blood pressure regulation. Biochem Bio- D, Faulkner M, Gabriel A, Gassman J, Greene T, Hall Y, Hebert L, phys Res Commun 285: 571–576, 2001 Hiremath L, Jamerson K, Johnson CJ, Kopple J, Kusek J, Lash J, Lea 15. McGiff JC, Quilley J: 20-Hydroxyeicosatetraenoic acid and epoxyei- J, Lewis JB, Lipkowitz M, Massry S, Middleton J, Miller ER, III, Norris K, cosatrienoic acids and blood pressure. Curr Opin Nephrol Hypertens O’Connor D, Ojo A, Phillips RA, Pogue V, Rahman M, Randall OS, 10: 231–237, 2001 Rostand S, Schulman G, Smith W, Thornley-Brown D, Tisher CC, Toto 16. Roman RJ: P-450 metabolites of arachidonic acid in the control of RD, Wright JT Jr, Xu S: Effect of ramipril vs amlodipine on renal cardiovascular function. Physiol Rev 82: 131–185, 2002 outcomes in hypertensive nephrosclerosis: A randomized controlled 17. Holla VR, Adas F, Imig JD, Zhao X, Price E Jr, Olsen N, Kovacs WJ, trial. JAMA 285: 2719–2728, 2001 Magnuson MA, Keeney DS, Breyer MD, Falck JR, Waterman MR, 29. Gerstein HC, Mann JF, Yi Q, Zinman B, Dinneen SF, Hoogwerf B, Halle Capdevila JH: Alterations in the regulation of androgen-sensitive Cyp JP, Young J, Rashkow A, Joyce C, Nawaz S, Yusuf S: Albuminuria and 4a monooxygenases cause hypertension. Proc Natl Acad Sci U S A 98: risk of cardiovascular events, death, and heart failure in diabetic and 5211–5216, 2001 nondiabetic individuals. JAMA 286: 421–426, 2001 18. Frisbee JC, Falck JR, Lombard JH: Contribution of cytochrome P-450 30. Wright JT Jr, Bakris G, Greene T, Agodoa LY, Appel LJ, Charleston J, omega-hydroxylase to altered arteriolar reactivity with high-salt diet Cheek D, Douglas-Baltimore JG, Gassman J, Glassock R, Hebert L, and hypertension. Am J Physiol Heart Circ Physiol 278: H1517–H1526, Jamerson K, Lewis J, Phillips RA, Toto RD, Middleton JP, Rostand SG: 2000 Effect of blood pressure lowering and antihypertensive drug class on 19. Muller DN, Theuer J, Shagdarsuren E, Kaergel E, Honeck H, Park JK, progression of hypertensive kidney disease: Results from the AASK Markovic M, Barbosa-Sicard E, Dechend R, Wellner M, Kirsch T, Fie- trial. JAMA 288: 2421–2431, 2002 beler A, Rothe M, Haller H, Luft FC, Schunck WH: A peroxisome proliferator-activated receptor-alpha activator induces renal CYP2C23 31. Gassman JJ, Greene T, Wright JT Jr, Agodoa L, Bakris G, Beck GJ, activity and protects from angiotensin II-induced renal injury. Am J Douglas J, Jamerson K, Lewis J, Kutner M, Randall OS, Wang SR: Pathol 164: 521–532, 2004 Design and statistical aspects of the African American Study of Kidney 20. Seki T, Ishimoto T, Sakurai T, Yasuda Y, Taniguchi K, Doi M, Sato M, Disease and Hypertension (AASK). J Am Soc Nephrol 14[Suppl 2]: Roman RJ, Miyata N: Increased excretion of urinary 20-HETE in rats S154–S165, 2003 with cyclosporine-induced nephrotoxicity. J Pharmacol Sci 97: 132– 32. Appel LJ, Middleton J, Miller ER, III, Lipkowitz M, Norris K, Agodoa LY, 137, 2005 Bakris G, Douglas JG, Charleston J, Gassman J, Greene T, Jamerson 21. Gainer JV, Bellamine A, Dawson EP, Womble KE, Grant SW, Wang Y, K, Kusek JW, Lewis JA, Phillips RA, Rostand SG, Wright JT: The Cupples LA, Guo CY, Demissie S, O’Donnell CJ, Brown NJ, Waterman rationale and design of the AASK cohort study. J Am Soc Nephrol MR, Capdevila JH: Functional variant of CYP4A11 20-hydroxyeicosa- 14[Suppl 2]: S166–S172, 2003 tetraenoic acid synthase is associated with essential hypertension. 33. Lewis J, Agodoa L, Cheek D, Greene T, Middleton J, O’Connor D, Ojo Circulation 111: 63–69, 2005 A, Phillips R, Sika M, Wright J Jr: Comparison of cross-sectional renal 22. Mayer B, Lieb W, Gotz A, Konig IR, Aherrahrou Z, Thiemig A, Holmer function measurements in African Americans with hypertensive ne- S, Hengstenberg C, Doering A, Loewel H, Hense HW, Schunkert H, phrosclerosis and of primary formulas to estimate glomerular filtration Erdmann J: Association of the T8590C polymorphism of CYP4A11 rate. Am J Kidney Dis 38: 744–753, 2001 with hypertension in the MONICA Augsburg echocardiographic sub- 34. Lewis J, Greene T, Appel L, Contreras G, Douglas J, Lash J, Toto R, study. Hypertension 46: 766–771, 2005 Van LF, Wang X, Wright JT Jr: A comparison of iothalamate-GFR and 23. Moreno C, Maier KG, Hoagland KM, Yu M, Roman RJ: Abnormal serum creatinine-based outcomes: Acceleration in the rate of GFR pressure-natriuresis in hypertension: Role of metab- decline in the African American Study of Kidney Disease and Hyper- olites of arachidonic acid. Am J Hypertens 14: 90S–97S, 2001 tension. J Am Soc Nephrol 15: 3175–3183, 2004

J Am Soc Nephrol ●●: –, 2008 CYP4A11 Gene Variant in AASK 7