cytochromep-450 3a (cyp3a) homologs are variably expressed in rat and human liver (4, 9), kidney (1, 5), adrenal gland (8), and anterior pituitary gland (15). Although the physiological consequences of CYP3A enzyme activity have not been defined, some observations support a role in blood pressure (BP) control. Renal CYP3A content and catalytic activity in the spontaneously hypertensive rat are higher than in the normotensive Wistar-Kyoto rat (1, 16). Treatment of the spontaneously hypertensive rat with the CYP3A-selective inhibitor troleandomycin lowers BP (16). Several human diseases associated with elevated BP, including Cushing's syndrome and toxemia of pregnancy (1), are characterized by increased urinary excretion of 6β-hydroxycortisol, a marker of in vivo CYP3A activity (16).

A role for CYP3A enzymes in BP control is scientifically plausible. 6β-Hydroxysteroid products of CYP3A-catalyzed reactions stimulate sodium transport across A6 toad kidney cells and sodium retention in rats (1, 3, 11). CYP3A-mediated 6β-hydroxylation is a major metabolic pathway for glucocorticoids and mineralocorticoids, including aldosterone, and previous evidence suggests CYP3A modulation of A6 cell sodium transport through a mineralocorticoid receptor mediated mechanism (13). Consistent with data from rat kidney (1), immunostaining for CYP3A highlights human distal tubules and collecting ducts (14), the principal sites at which mineralocorticoids stimulate sodium reabsorption. Renal CYP3A-produced 6β-hydroxysteroids have been postulated to function as mineralocorticoids near their site of synthesis (1), and CYP3A activity, along with that of 11β-hydroxysteroid dehydrogenase type 2, has been proposed to regulate glucocorticoid occupancy of mineralocorticoid receptors (13).

The lack of previous reports associating human renal CYP3A and BP may partly reflect the previous unavailability of noninvasive methods to predict renal CYP3A activity. It is now established that, among CYP3A isoforms, CYP3A5 expression prevails in human kidney (5). Recent reports demonstrate that CYP3A5 expression in liver (9) and intestine (10) is polymorphic and largely determined by a single-nucleotide polymorphism (A6986G) that distinguishes the CYP3A5*1 (“expressor”) allele from the *3 (a “reduced-expressor”) allele. We demonstrate that CYP3A5 genotype similarly predicts renal CYP3A5 expression. We also report preliminary evidence of CYP3A5 genotype prediction of BP and creatinine clearance (CrCl) among a group of healthy young adults.

METHODS

We assayed CYP3A catalytic activity and CYP3A5 immunoreactivity in renal tissue samples from 21 organ donors: 17 were Caucasian, 1 was Hispanic, and 3 were of unknown ethnicity. All methods were described previously (10). CYP3A5 genotypes were determined by direct genomic DNA sequencing. Microsomes were isolated by differential centrifugation. CYP3A5 content in 50 μg of microsomes was measured by Western blot with a CYP3A5-specific antibody (BD Gentest, Woburn, MA); undetectable bands were assigned a value of 0.25 pmol/mg, the limit of quantitation (LOQ; Fig. 1A). Midazolam (MDZ) 1′-hydroxylation was measured after a 15-min incubation of 200 μg protein with 8 μM MDZ; microsomes with undetectable activity were given the LOQ value of 0.04 pmol · min^-1 · mg^-1.

We genotyped 89 unrelated volunteers for the CYP3A5 A6986G single-nucleotide polymorphism using previously described methods (9); all subjects were self-identified African-Americans. Among the subset of this group consisting of 25 healthy individuals (age range: 18-52 yr) who volunteered to be screened for a pharmacogenetic study, we tested the ancillary hypothesis of CYP3A5 genotype association with BP and CrCl. Nurses blind to subject genotypes and medical histories performed and recorded all clinical measurements. Studies were approved by the University of North Carolina Committee on the Protection of Human Research Subjects.

RESULTS

The average CYP3A5 content (1.98 pmol/mg, n = 5) of microsomes from *1/*3 kidneys exceeded the average of those from *3/*3 kidneys (n = 16, Wilcoxon P = 0.0001), which was at or below the LOQ (Fig. 1A). None of the kidneys was *1/*1. The closely related CYP3A4 was not detected with a CYP3A4-specific antibody in any of the microsomal preparations. Mean microsomal CYP3A activity, reflected by MDZ 1′-hydroxylation, was 18-fold higher in *1/*3 kidney microsomes (8.04 pmol · min^-1 · mg^-1) than in those from *3/*3 kidneys (0.43 pmol · min^-1 · mg^-1, Wilcoxon P = 0.0137; Fig. 1B). Analysis of DNA from the *1/*3 outlier, indicated by arrows in Fig. 1, revealed an inactivating frame-shift mutation (27131-27132insT) that produces an inactivating allele known as *7 (6). Removal of this outlier slightly lowered P values.

CYP3A5*1 allele frequency among 89 African-Americans was 0.7. Among the 25-individual subset (Table 1), CYP3A5 genotype associated with seated systolic BP (SBP), mean arterial pressure, the product of SBP and heart rate (HR) (SBP * HR, an indicator of left-ventricular oxygen consumption), and Cockcroft-Gault CrCl. The *1/*1 group averaged the highest value for each measure. Average *1/*1 SBP exceeded that of the *3/*3 group by 19.3 mmHg (Fig. 2), and a gene-dose effect was apparent. CYP3A5 genotype associated significantly with combined BP strata (P = 0.0048; Table 1) from the Sixth Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High BP {N = optimal or normal [SBP <130 and diastolic BP (DBP) <85 mmHg]; H = high-normal or higher (SBP ≥130 or DBP ≥85 mmHg)} (7).

Sex-specific analysis detected CYP3A5 genotype associations with SBP, mean arterial pressure, and estimated CrCl among females (Table 1); similar trends were noted among the smaller sample of males, but these did not reach statistical significance. In multiple regression analyses, sex and genotype accounted for 70% of the variability in SBP. Compared with non-*1/*1 females, *1/*1 males showed a >20-fold increased risk of high normal or higher BP (P = 0.0002).

Subject age, averaging ∼25 yr, did not vary between genotype groups. Age and body mass index effects on BP were nonsignificant. Genotype association with body mass index, DBP, pulse pressure, and HR did not reach statistical significance.

DISCUSSION

CYP3A5*1 allele associates with CYP3A5 expression in human kidney, as previously reported for liver and intestine. However, in the liver and intestine, there is also substantial CYP3A4 expression (9, 10). As a result, hepatic and intestinal aggregate CYP3A activities associate only weakly with the CYP3A5*1 allele. In contrast, CYP3A4 was not detected in renal microsomes, and mean CYP3A activity differed markedly between CYP3A5 genotype groups.

Our preliminary finding of CYP3A5 genotype association with resting BP among healthy adults may be consistent with a role for CYP3A enzymes in BP control. Young adults with supernormal BP have an increased long-term risk of death due to cardiovascular and coronary heart disease (12); the identification of common genetic polymorphisms relevant to BP control is thus an important line of investigation. We are presently testing our hypothesis of a CYP3A5 genotype-BP correlation in a larger sample of adults. If shown true, our conjecture might portend a role for CYP3A5 inhibition in the treatment of some forms of hypertension.

The CYP3A5*1 allele frequency among our cohort of African-Americans agrees roughly with a previous report (6). This frequency exceeds those among all other ethnic populations studied to date (6, 9). Interethnic differences in the prevalence of sodium sensitivity (17) parallel those of hypertension (2), with African-Americans having the highest global prevalence of each. A possible link between CYP3A5 activity and the high prevalence of sodium-sensitive hypertension among African-Americans may merit further study.

DISCLOSURES

This study was supported by the Doris Duke Medical Research Program and National Institutes of Health Grants RR-00046, GM-37149, GM-63666, and GM-60346.

FOOTNOTES

• The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.