| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Editorial |
Departments of Pediatrics (P.C.W.) and Obstetrics/Gynecology (W.E.R.), University of Texas Southwestern Medical Center, Dallas, Texas 75390-9063
Address all correspondence and requests for reprints to: Perrin C. White, M.D., Department of Pediatrics, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, Texas 75390-9063. E-mail: perrin.white{at}utsouthwestern.edu.
The final steps in the synthesis of cortisol and aldosterone (conversion from 11-deoxycortisol and deoxycortisone, respectively) are catalyzed by 95% identical cytochrome P450 isozymes, CYP11B1 and CYP11B2 (1). CYP11B1 is expressed in the adrenal zona fasiculata and is regulated by ACTH, whereas CYP11B2 is expressed in the zona glomerulosa and is regulated by angiotensin II and potassium, with ACTH having mostly a short-term effect on expression. The corresponding genes are located approximately 40 kb apart on chromosome 8q24.
There are three known Mendelian diseases arising from mutations in these genes. Inactivating mutations in CYP11B1 cause a form of virilizing congenital adrenal hyperplasia (2), which is distinguished clinically from the more common 21-hydroxylase deficiency by the lack of salt-wasting signs; to the contrary, many patients with 11-hydroxylase deficiency develop hypertension, presumably due to elevated levels of deoxycorticosterone, a known mineralocorticoid. These patients typically have suppressed secretion of renin and aldosterone consistent with the presence of a mineralocorticoid other than aldosterone.
Conversely, inactivating mutations in CYP11B2 cause aldosterone synthase deficiency, which is characterized by salt-wasting (mainly in childhood) with normal cortisol synthesis (3). These patients typically have elevated renin and very low aldosterone but elevated deoxycorticosterone levels.
Finally, recombinations (unequal crossing-over) between CYP11B1 and CYP11B2 can cause glucocorticoid suppressible hyperaldosteronism (glucocorticoid remediable aldosteronism), in which the transcriptional regulatory region of CYP11B1 is juxtaposed to coding sequences from CYP11B2 (4, 5, 6). This allows a chimeric enzyme with the activity of CYP11B2 to be expressed at high levels in the zona fasciculata under the control of ACTH, leading to aldosterone excess and hypertension.
Although all three of these diseases are rare, it is reasonable to hypothesize that formes frustes might be more common. For example, a mild nonclassic form of 21-hydroxylase deficiency (CYP21 mutations) occurs far more frequently (
1 in 500 in most populations) than the severe classic form (1 in 16,000) and may be an important cause of androgen excess in women (7). However, a survey of women in a reproductive endocrinology clinic failed to detect any CYP11B1 mutations that might cause nonclassic 11-hydroxylase deficiency (8). CYP11B2 mutations that reduce activity to as little as 0.6% of normal seem to have no effects on aldosterone secretion in vivo, so that a nonclassic form of aldosterone synthase deficiency also appears unlikely (3, 9).
Chimeric CYP11B1-CYP11B2 enzymes have similar enzymatic activity if the cross-over breakpoint is between exons 1 and 4 but are unable to synthesize aldosterone if the breakpoint is after exon 5 (6). Indeed, all patients with glucocorticoid suppressible hyperaldosteronism have their breakpoints before exon 5 (4, 6). There does not appear to be a region in which a cross-over will yield an intermediate level of aldosterone synthase activity.
Thus, other mutations that might affect CYP11B2 expression or activity have been sought with the expectation that this gene is a plausible candidate risk locus for hypertension or other cardiovascular disease. The first and most extensively studied is a C/T polymorphism in the 5' flanking region of CYP11B2 (–344C/T), which involves a binding site for the SF1 transcription factor (10).
The effects of this polymorphism are uncertain at all functional levels including gene expression, aldosterone secretion, blood pressure, and cardiovascular complications. Although the C allele binds SF1 approximately four times more strongly than does the T allele, reporter constructs containing the two alleles are expressed at equal levels in cultured human adrenocortical cells (11), and the entire SF1 binding site can be deleted without apparent effect in this system (12). A reporter construct carrying the C allele was slightly more responsive to angiotensin II than the T allele. Whereas SF1 is essential for adrenal development and is a potent stimulator of gene expression for most steroidogenic enzymes, it represses CYP11B2 expression in vitro, and so effects on CYP11B2 expression in vivo are difficult to predict (11).
With regard to aldosterone excretion and blood pressure, this polymorphism has been reported to have no effect (13, 14, 15, 16, 17, 18); the C allele has been associated with higher aldosterone excretion and/or blood pressure (19, 20); and the T allele has been associated with higher excretion and/or blood pressure (21, 22, 23, 24, 25, 26, 27, 28). The weight of evidence favors the T allele as being associated with higher aldosterone excretion and blood pressure.
The C allele has been associated with greater left ventricular size in some (13, 29) but not other (14) studies and with decreased baroreflex sensitivity (30). It has also been associated with exacerbating the adverse effects of smoking and decreased high-density lipoprotein cholesterol levels on cardiovascular risk in one study (31) but not others (16, 18, 32).
Unfortunately, such discrepant results are not unusual in association studies. There are several possible explanations. The simplest is that many studies use relatively small sample sizes rendering type 2 statistical errors (failing to reject a false null hypothesis, "false negative") more likely, and conversely publication bias favors the dissemination of positive results that may be due to type 1 errors (incorrectly rejecting a true null hypothesis, "false positive"). Another explanation is that some effects may occur only in the context of particular alleles of unlinked genes (i.e. epistasis) that may be more prevalent in certain study populations. A related but not identical explanation is that of population admixture. For example, if a particular allele is more common in Afro-Americans than in Caucasians [as is the case for the –344C allele of CYP11B2 (10)], and if Afro-Americans are more prone to hypertension for a variety of reasons, than in an ethnically mixed American population the frequencies of the –344C allele and hypertension might both be increased in proportion to the number of Afro-American individuals (or persons of mixed Afro-American and Caucasian ancestry). An association between –344C and hypertension would not have any functional implications under these circumstances.
Finally, there might be other polymorphisms in or near CYP11B2 that might be more important functionally. When particular alleles at genetically linked polymorphic loci are associated with each other, the phenomenon is termed "genetic linkage disequilbrium." This is usually quantitated as a variable D' that can have values between 0 and 1. A particular combination of alleles at two or more linked loci is a "haplotype."
Linkage disequilibrium between single stranded conformation polymorphisms in CYP11B1 and CYP11B2 was documented in 1992 when linkage of glucocorticoid-suppressible hyperaldosteronism to the combined locus was demonstrated (5). However, none of these polymorphisms were defined by sequence analysis. With regard to polymorphisms of defined sequence, the –344C/T polymorphism in CYP11B2 was found to be in linkage disequilibrium with a gene conversion in intron 2; the majority of CYP11B2 genes carrying the –344T allele had an intron 2 sequence corresponding in its entirety to the normal sequence of CYP11B1, whereas genes carrying the –344C allele rarely carried the gene conversion. Thus, three common haplotypes could be defined (–344C;no conversion, –344T;no conversion, –344T;conversion) (10). Subsequently, a conservative amino acid substitution (Lys173Arg, K173R, 2718A/G) was found to be in strong linkage disequilibrium with –344C/T (21, 33).
Because CYP11B1 and CYP11B2 were closely linked genetically, and polymorphisms within these genes were known to be in linkage disequilibrium, investigators previously tried to determine whether measures of CYP11B1 activity might be associated with alleles at these loci. Higher levels of both 11-deoxycorticosterone and 11-deoxycortisol—the main substrates for CYP11B1—after ACTH stimulation were indeed associated with the –344T allele of CYP11B2 in two studies, suggesting that CYP11B1 activity was lower when the –344T allele was present in CYP11B2 (34, 35). This was presumably due to an associated but undefined polymorphism in CYP11B1.
Two papers in this issue of the JCEM (36, 37) confirm and extend these findings. Both are very technical papers that would be most readily understood by geneticists with strong statistics backgrounds. The papers reach very similar conclusions, with the findings of Keavney et al. (36) being much more robust by virtue of studying a large number of families instead of a smaller number of unrelated individuals.
Both studies made use of polymorphisms, most of which were available in public databases, to document extensive linkage disequilbrium between polymorphisms in CYP11B1 and CYP11B2 over a genetic region spanning approximately 50 kb (Fig. 1
). Although terminology differed, five polymorphisms were used in both studies. Keavney et al. (36) were able to directly determine linkage disequilibrium and identify haplotypes by studying inheritance (segregation) of the various markers within individual families, whereas Ganapathipillai et al. (37) had to deduce linkage disequilibrium by studying occurrence of various combinations of alleles in unrelated individuals. With a few exceptions, the values for D' (indicated degree of linkage disequilibrium) agreed fairly well between studies. Both studies found a limited number of frequently occurring haplotypes. Ganapathipillai et al. (37) found four with frequency greater than 0.1, corresponding to the three defined by White and Slutsker (10) with the –344T;conversion haplotype split into two higher resolution haplotypes by alleles at a polymorphism in the 3' untranslated region of CYP11B2. Keavney et al. (36) found three such haplotypes with the –344C;no-conversion haplotype split by a polymorphism in the 3' portion of CYP11B2; on the other hand, the –344T;no-conversion haplotype occurred rarely in their population.
|
Keavney et al. (36) went on to demonstrate that allelism at the CYP11B1/B2 loci explained 5% of total variability in 11-deoxycortisol excretion vs. 30% for unlinked polygenes; the sum corresponds fairly well to the estimated 51% heritability overall.
Although it is presumed that a polymorphism affecting activity or expression of CYP11B1 is responsible for the observed associations, it is not obvious what that polymorphism is. Keavney et al. (36) found that a polymorphism in exon 1 (225A/G) had the strongest associations with 11-deoxycortisol excretion, but it does not change the encoded amino acid (CTA and CTG both encode leucine) and thus cannot directly affect enzymatic activity. Perhaps it affects CYP11B1 transcription or splicing of pre-mRNA. Additional experiments, such as transfecting minigenes into cultured adrenocortical cells, will be required to investigate such possibilities. Alternatively, other polymorphisms as yet unknown might be responsible for the findings.
Taken together with previous work, the two papers provide convincing evidence that 11-deoxycortisol secretion is influenced by allelism in or near CYP11B1. However, it cannot be demonstrated from the available data whether there is any functional relationship between the well-defined biochemical (or "intermediate") phenotype of increased 11-deoxycortisol excretion (presumably reflecting decreased 11-hydroxylase activity) and other phenotypes such as increased aldosterone:renin ratio or increased blood pressure. Keavney et al. (36) have suggested that decreased 11-hydroxylase activity might lead to chronically increased ACTH levels and that the resulting ACTH stimulation might increase aldosterone secretion independently of the renin-angiotensin system. This explanation seems implausible for several reasons. First, patients with classic 11-hydroxylase deficiency have suppressed rather than elevated aldosterone secretion, so it seems unlikely that a milder defect would have the opposite effect. Second, ACTH is a stimulus for aldosterone secretion mainly in the short term with long-term elevations in ACTH actually lowering aldosterone levels (40). Finally, other conditions in which ACTH might be expected to be similarly or more markedly elevated, such as nonclassic congenital adrenal hyperplasia due to 21-hydroxylase deficiency, are not known to be associated with increased aldosterone secretion.
Given the difficulty of studying complex traits in outbred human populations, it is gratifying to find a single locus that consistently influences a biochemical parameter. Determining what this means for the endocrinologist will require much additional study.
Received December 13, 2004.
Accepted December 15, 2004.
References
This article has been cited by other articles:
 |
|
 |
|
  Wenxia Chai, Y. M Hoedemaekers, R. H. van Schaik, M. van Fessem, I. M Garrelds, J. J Saris, D. Dooijes, F. J ten Cate, M. M. Kofflard, and A. J. Danser Cardiac aldosterone in subjects with hypertrophic cardiomyopathy Journal of Renin-Angiotensin-Aldosterone System, December 1, 2006; 7(4): 225 - 230. [Abstract] [PDF]
|
|
|
|
 |
|
 D. M. McNamara, S. W. Tam, M. L. Sabolinski, P. Tobelmann, K. Janosko, A. L. Taylor, J. N. Cohn, A. M. Feldman, and M. Worcel Aldosterone Synthase Promoter Polymorphism Predicts Outcome in African Americans With Heart Failure: Results From the A-HeFT Trial J. Am. Coll. Cardiol., September 19, 2006; 48(6): 1277 - 1282. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
