Studies of aldosterone renin ratio and genetic variation at the CYP11B1/CYP11B2 locus in human essential hypertension.
PhD thesis, University of Glasgow.
Full text available as:
Essential hypertension is seen as a contemporary public health challenge not only because of its high prevalence and variable treatment response but also because it represents a major risk factor for cardiovascular disease. Both genetic and environmental factors contribute to the regulation of blood pressure, thus making the study of hypertension difficult and complex. Over recent years, with the advent of new molecular technologies, there has been an increasing interest in its genetic component.
Alterations in the rate and pattern of adrenal steroid biosynthesis can contribute to blood pressure changes and its heritable component. In humans, mutations in the genes encoding aldosterone synthase (CYP11B2) and 11β-hydroxylase (CYP11B1), responsible for the final stages of aldosterone and cortisol production respectively, lead to rare monogenic disorders. Both, glucocorticoid remediable aldosteronism and 11β-hydroxylase deficiency are associated with mineralocorticoid hypertension. More subtle but much more common genetic variations, such as single nucleotide polymorphisms (SNPs), are associated with more common intermediate phenotypes of essential hypertension such as the aldosterone to renin ratio (ARR). Up to 15 % of hypertensive patients may have an altered aldosterone biosynthesis as indicated by an elevated ARR. However, many studies using the ARR have widely variable results, as there is no standardization of the assays used and because many factors may influence the measurement of renin and aldosterone levels.
Previous studies have reported associations between a variation in the promoter region (-344T/C) and intron 2 (intron conversion) in the CYP11B2 gene with inappropriate aldosterone production and hypertension. However, the findings have not been consistent. Interestingly, associations have also been reported between these polymorphisms and raised basal and ACTH-stimulated levels of the 11-deoxysteroids (DOC and 11-deoxycortisol), indicating a reduced 11β-hydroxylase efficiency, encoded by CYP11B1. This is not entirely surprising considering that the polymorphisms within both genes are known to be in linkage disequilibrium. Recent findings indicate that two polymorphisms in the promoter region of CYP11B1 (-1889G/T and –1859A/G) are associated with altered efficiency of 11β-hydroxylase. Using this evidence, a hypothesis linking reduced 11β-hydroxylation efficiency to an elevated ARR and hypertension, involving a small but chronic increase in adrenocorticotrophic hormone (ACTH) stimulation of the adrenal gland, has been proposed to explain the undetermined interaction between the CYP11B genes. This hypothesis requires further testing. Thus, identification of the specific SNPs or group of SNPs along the CYP11B1/CYP11B2 locus responsible for the observed elevation in ARR and hypertension requires further investigation. In this thesis, the factors that influence the measurement of the ARR are examined and the relationship between ARR, blood pressure and genetic variation at the CYP11B1/CYP11B2 locus is studied in several large population samples.
Since minor changes in renin measurement may lead to significant changes in the ARR, an evaluation of methods to measure it in population studies was of fundamental importance. In Chapter 3, the performance of a novel automated plasma renin concentration (PRC) assay was evaluated in normotensive subjects and compared with the routinely used plasma renin activity (PRA) assay. The analytical and functional sensitivity, precision and stability of the new PRC assay were excellent. Overall, the correlation between the PRC and PRA assays was good. The only limitation of the PRC assay was a small decrease in sensitivity at lower concentrations, possibly due to cross-reactivity with prorenin. Although this limitation might be a drawback for its use in clinical diagnosis, for large scale population studies, practical issues as well as performance were relevant. The PRC assay offers several advantages over the traditional PRA assay; it allows efficient processing of more samples at a lower cost in less time. The assay has high precision, detecting small differences with greater certainty. In addition, the storage and handling of a radioactive reagent are eliminated. Thus, the new PRC assay was preferred as the method of choice for the family study in Chapter 4.
The study in Chapter 4 examined the significance of a raised ARR in normotensive and hypertensive subjects and the phenotypic and familial factors affecting it. The distribution of ARR and its heritability was estimated in 1172 individuals from 248 Caucasian families. The associations between ARR and blood pressure with polymorphic variations at the CYP11B1/CYP11B2 locus were also tested. Unadjusted and adjusted ARR values, for age, gender, body mass index, ACE inhibitors and beta-blocker use, were continuously distributed in normal and hypertensive cohorts; there was no evidence of a cut-off that would identify a separate population with primary aldosteronism. The median ARR was 4.19 ng/L per mIU/L (range 0.04-253.16). The majority of subjects (> 80%) had an ARR < 10ng/L per mIU/L; ARR levels were higher in females, and associated with age, body mass index and potassium. Antihypertensive agents had significant predictable effects on the ARR. Renin was negatively and ARR positively associated with ambulatory blood pressure readings (p<0.001) in subjects not taking antihypertensives. The heritability of the ARR was 38.1% (p<<10-8), and the heritability of renin and aldosterone was 27.4% and 28.7% respectively with p<<10-8. Plasma aldosterone, but not ARR, was influenced by intron 2 conversion genotype in the CYP11B2 gene (beta = -0.07, p=0.04).
There is substantial genetic determination of the ARR and its components. The effects of gender and other confounders indicate that care must be taken when interpreting the ARR as a screening test for primary aldosteronism. The ratio is not a marker of a distinct pathological abnormality but subjects with a high ratio tend to have higher blood pressure readings, possibly reflecting the long term influence of aldosterone on cardiovascular homeostasis.
The study in Chapter 5 sought to identify the most informative SNPs in the CYP11B1/CYP11B2 locus in Caucasians. There is tight linkage disequilibrium across the entire locus, facilitating the selection of the SNPs. For the study, the selection of the first four SNPs chosen for genotyping was based on previous indications of their association with hypertension. These SNPs were the -344T/C and the intron 2 conversion in the CYP11B2 and the -1889G/T and -1859A/G in the promoter region of CYP11B1. Using the HapMap data for Caucasians together with information from previous genotyping of this locus in 26 normotensive subjects from the MONItoring of trends and determinants in CArdiovascular disease Study (MONICA), a further eighteen SNPs were chosen for genotyping. A 90% coverage of this locus was achieved. The eighteen SNPs were located in the exon 3 and intron 6 of the CYP11B2, exon 1, intron 2, intron 5 and the 3’UTR of CYP11B1, and the intergenic region. The 22 SNPs thus selected were genotyped in 79 unrelated individuals from the British Genetics of Hypertension Study (BRIGHT). After evaluating the minor allele frequencies, Hardy-Weinberg equilibrium and percentage of genotyping, 15 polymorphisms were chosen for haplotype tagSNP analysis. The final set of 7 SNPs, which according to this tagSNP analysis maximize haplotype diversity in Caucasians and minimize genotyping redundancy, comprised: 3 SNPs in CYP11B1 (rs5301, rs6410, rs4313136 (-1889G/T)), 3 SNPs in CYP11B2 (rs4546, the intron conversion (IC) and rs1799998 (-344T/C)) and one in the intergenic region (rs4736354). This set of SNPs was used in Chapter 7 in a large case-control study to test for association with hypertension.
The study in Chapter 6 analyzed the allele frequency and linkage disequilibrium differences in the CYP11B1/CYP11B2 locus in 35 founders of families with African ancestry, 149 founders of families with Caucasian ancestry and the 79 Caucasian unrelated individuals used in Chapter 5. In addition, the genotyping results from the 35 Afro-Caribbean founders, the 79 Caucasians and the results available from eleven populations in HapMap for the CYP11B1/CYP11B2 locus were combined to generate phylogenetic trees.
From the 15 SNPs selected in the pilot study in Chapter 5, 12 SNPs were chosen. Considerable allelic variation was observed in these 12 SNPs genotyped in the Afro-Caribbean and the 79 subjects from the BRIGHT with Caucasian ancestry. Most significant differences were in the following SNP: rs4313136 (-1889G/T) and rs6471580, p<0.05; rs4736354, rs4546 and rs1799998 (-344T/C), p.<0.01 and rs5283 and intron conversion, p<0.0001. These differences are located at the 5’ end of the genes and the intergenic sections possibly because SNPs in regions crucial for the synthesis would result in deleterious effects in the enzymes. The comparison of allele frequencies of common SNPs between the Afro-Caribbean and Caucasian founders from the family study showed the same trend as with the subjects from the BRIGHT. The variation in linkage disequilibrium (LD) length was also compared between these groups. Subjects with African ancestry had shorter LD patterns in the CYP11B1/CYP11B2 locus when compared with individuals of Caucasian ancestry. The high levels of LD across these regions in Caucasians, while advantageous in designing efficient association studies, do not allow identification of causal SNPs. African ancestry is more conducive to the identification of such variants. It is clearly not recommended that individuals from different ethnic origin are combined in an association study as there is an increased risk of type 1 errors with population stratification.
The distribution of the Afro-Caribbean subjects and the 79 Caucasians from the BRIGHT in the branches of the phylogenetic trees was different, suggesting different patterns of evolution.
In Chapter 7, association with hypertension and the CYP11B1/CYP11B2 locus in a Caucasian case-control population was tested and replicated in an independent case-control population. In the discovery phase, the seven SNPs selected in Chapter 5 and rs4471016 (-1859A/G) were genotyped in 3340 unrelated individuals of the BRIGHT Case Control cohort. The SNPs significantly associated with hypertension (rs6410, rs4471016 (-1859A/G), rs43131369 (-1889G/T), rs6471581, rs4546, intron conversion and rs1799998(-344T/C)) were tested for replication in 2929 unrelated individuals of the Swedish study. Only the intron conversion showed significant association with hypertension in the single SNP analysis. Subjects with the conversion allele in the intron 2 of CYP11B2 had a higher risk of hypertension (OR=1.19 [1.10-1.29], p value=1.06 x 10-5). In the haplotype association analysis, three haplotypes showed significant association with hypertension in the BRIGHT Study and were replicated in the Swedish Study. The first haplotype was T/A/G comprising rs6410, rs4471016 and 4313136 in the CYP11B1 gene and this had an OR=3.14 [2.27-4.32], p=2.99x10-12. The second haplotype was A/G/G comprising SNPs from both genes (rs4471016, rs4313136 and rs6414) and this had an OR=2.64 [1.87-3.74], p=3.90x10-8. The third haplotype was A/T/Conv comprising rs6414, rs4546 and the intron conversion in the CYP11B2 and this had an OR=4.64 [2.63-8.18], p=1.11x10-7. These three haplotypes were present in approximately 2% of the populations studied. Thus, the regions most strongly associated with hypertension in Caucasians were narrowed to a region between intron 2 and intron 3 in the CYP11B2, and to another region between the promoter and exon 1 in the CYP11B1 gene. However, functional studies are required to elucidate the mechanism by which these genetic variations lead to alterations in aldosterone and cortisol production and subsequent hypertension.
In summary, the results show that the PRC assay is an entirely acceptable option for measuring renin levels in population studies. The ARR is continuously distributed, influenced by genetic factors and is not a marker of a distinct pathological abnormality. There is significant disparity in allele frequencies and linkage disequilibrium structure at the CYP11B locus between ethnic groups. These differences should allow for the design of efficient association studies and the identification of causal variants. Using sophisticated statistical analyses, the studies have allowed a more precise identification of the key regions in this locus associated with hypertension in Caucasians. Although the identification of causal variants remains to be elucidated, the findings of the studies presented in this thesis are of considerable interest. This is the first time a positive association of a locus that is known to be involved in a rodent model of hypertension, rare autosomal human disorders, and functional changes that lead to an altered biochemical intermediate phenotype has been confirmed. Additionally, the evidence in this thesis shows that differences in genotype at the CYP11B1/CYP11B2 locus might play an important role in modulating the adverse effects of lifelong exposure to inappropriate aldosterone levels. In the future, screening susceptible subjects by combining the use of the ARR and genotyping at the CYP11B1/CYP11B2 locus will enable early lifestyle interventions or a personalised therapy to slow the progression towards aldosterone modulated hypertension and cardiovascular disease.
Actions (login required)