Genetics, genomics and other molecular approaches

Salt sensitivity has long been suspected to be involved in not only the development of arterial hypertension but also to trigger endothelial dysfunction [1] and alter vessel wall properties of large arteries without any effect on or beyond blood pressure elevation [2]. It is, therefore, comprehensible that UK/US/Canada/Europe and WHO recommendations were launched to generally reduce salt intake to less than 6 and 5 g/day, respectively. In the US, it is even recommended to reduce daily salt intake to less than 4 g in individuals above 40 years of age, Blacks, and those with hypertension. He et al.[3] were able to show that a modest reduction in salt intake caused significant decrease in blood pressure (and urinary albumin and pulse wave velocity, but perhaps blood pressure-associated) in different ethnic groups consisting of Whites, Blacks and Asians. It has been suggested that very-low-salt diets may have unintended adverse consequences [4], especially among patients with a more prominent counter-regulatory response of the renin–angiotensin system. In an observational follow-up study of the Trials Of Hypertension Prevention, Cook et al.[5] showed that a 25–35% dietary salt reduction in individuals with prehypertension resulted in a 25–30% lower risk of cardiovascular outcomes in the long term (10–15 years post trial). With respect to more general health benefit approaches, using a coronary heart disease (CHD) policy model, salt intake reduction of about 3 g/day has been projected to substantially reduce the development of CHD, stroke, myocardial infarction and deaths from any cause in the US [6]. With these epidemiological studies in mind, possible interindividual differences in the clinical response to salt reduction (separately in salt-sensitive and nonsensitive individuals), short-term and long-term, may be due to modulating genetic factors. At this stage, the review by Kelly and He [7] in this issue of the Journal of Hypertension comes into play. The authors are members of the Genetic Epidemiology Network of Salt-Sensitivity (GenSalt) Study, a study devoted to specifically analyze mechanisms of salt sensitivity. In the current review, the authors mainly refer to results from salt-sensitivity-related monogenic blood pressure disorders, whole genome linkage analyses, genetic association studies and Genome-Wide Association (GWA) studies. Genetic linkage analyses for complex diseases – robust, often successful, but necessitating a large number of families [8] – have already reached their zenith some, perhaps, 10–15 years ago. Genetic association studies – less robust, but more powerful with less numbers of individuals [8] – have encountered their difficulties mostly due to lack of replication and inconsistent results, differences in sample sizes, phenotype definitions or measures, ethnic background, and the like. Even if genetic association studies, that is the candidate gene approach, have reached their limits, the principal idea that one may deduce appropriate candidate genes from monogenic or better oligogenic studies on blood pressure regulation was initially very promising. Apart from results from transgenic or knockout animal studies, candidate genes for association studies were deduced from human monogenic or oligogenic analyses. An example: when functional variants, for example, gain-of-function mutations that increase the constitutive activation of the amiloride-sensitive distal renal epithelial sodium channel (βENaC) in Liddle's syndrome [9], have a substantiated role in phenotype expression, that is elevated blood pressure (‘quasi’ one-to-one genotype–phenotype relationship), common gene polymorphisms [mostly single-nucleotide polymorphisms (SNPs)] at that very locus may also be associated with blood pressure phenotypes in the general population, even with smaller effect sizes [as also suggested for HSD11B2 (see references in the review) and CYP11B2] [10,11]. To only refer to the above example, the βENaC T594M mutation has been suggested to be associated with hypertension in Blacks [12,13], with inconsistent results [14], and was not identified in other ethnic groups [15,16], and an in-vitro functionality has not or not convincingly been demonstrated [16,17]. Even if common SNPs in the ENaC genes have been suggested to contribute to blood pressure response to dietary sodium intake [18], their molecular functionality remains to be demonstrated and the association data await replication. Apart from the above-mentioned lack of replication in different populations with different sample sizes and nonidentical phenotype measures, the effect sizes in association studies were often too small to draw any definite conclusions. It is, thus, comprehensible that the original enthusiasm for the candidate gene approach somehow declined. In a more audacious attempt to further identify yet unknown pathways, nonhypothesis-driven approaches have become available (as of 02/21/12, the catalog includes 1175 GWA study publications: http://www.genome.gov/page.cfm?pageid=26525384&clearquery=1#result_table) and meanwhile powerful enough to literally outpace other approaches. The increasing number of GWA replications clearly increased the confidence in this approach, but also in the results as such. However, a small bitter pill to swallow is that replicated genomic variants (almost always SNPs) are often not located within defined gene regions or very far from it (e.g. a 58-kb interval [19], which necessitates extensive resequencing to identify the ‘functional’ variant) or the related gene regions have not been sufficiently explored with respect to function. Another important issue is the possibility to identify yet unknown pathophysiological factors, which we were able to demonstrate, for example, for potential pro-inflammatory factors in venous thrombosis [20,21], a disease condition which was traditionally attributed to risk factors defined by the Virchow's trias (we will soon publish our results from molecular functional analyses). For the genomics of salt sensitivity, the aspect of environmental modulation of the phenotype, that is the response to salt intake, may be of crucial importance. In this respect, Citterio et al.[22], using a genome-wide genotyping array, have identified a cluster of variations associated with changes in both diastolic and systolic blood pressure in treatment-naïve hypertensive patients after salt load. The associated SNPs mapped to protein kinase, cGMP-dependent, regulatory, type I, solute carrier family 24 (sodium/potassium/calcium exchanger), member 3 and solute carrier family 8 (sodium-calcium exchanger), member 1, none of which overlapped with any gene annotated in this review by Kelly and He [7]. On the contrary, our recent GWA on blood pressure phenotypes, the largest one to date, conducted by the International Consortium for Blood Pressure, replicated loci within or near genes related to renal sodium handling as quoted by Kelly and He in this issue, namely natriuretic peptide A-natriuretic peptide B (NPPA-NPPB), CYP17A1 and natriuretic peptide receptor C (NPR3) [23], all of which have previously been implicated in blood pressure regulation. In this respect, NPR3-C5orf25 on chromosome 5 codes for the natriuretic peptide clearance receptor (NPR-C), NPR3 knockout mice exhibiting lower blood pressure [24]. CYP17A1 mutations cause congenital adrenal hyperplasia, a condition which is also caused by mutations in other genes (for references, see Kelly and He in this issue), even if to different phenotypic degrees. A common intronic variant at the CYP17A1 locus has also been associated with systolic blood pressure in a GWA by Levy et al.[25]; mutations in this gene have been described in patients with 17α-hydroxylase deficiency characterized by apparent mineralocorticoid excess, salt retention, hypokalemia, and hypertension. The NPPA and NPPB genes at the methylenetetrahydrofolate reductase-NPPB locus encode precursors for atrial and B-type natriuretic peptides and have recently been associated with plasma atrial natriuretic peptide/brain natriuretic peptide levels and blood pressure in a large GWA [26] and with blood pressure and gene expression (eSNP) in the Global BPgen consortium and the Cohorts for Heart and Aging Research in Genomic Epidemiology analysis, respectively [25]. Coming back to the specific nature of salt sensitivity, but also to the fact that different loci contribute to the phenotype under study, future GWA strategies [27] should account for gene–gene [28] and gene–environment [29] interactions. When we have solved the issue of a ‘clear-cut’ association with the trait, functional genomics with all its technical and molecular tools will more readily identify the molecular mechanism underlying the association with the trait in question. A convincing functional mechanism for regulatory sequences would be that only one specific allele (and not the other) binds to a transcription factor complex [30], or that a specific molecular haplotype (on one single strand!) binds other transcription factor complexes than specific alleles alone [31], or that different allelic constellation cross-interact between a distal promoter region and a proximal 5’-untranslated region resulting in different transcription factor modules and hence gene transcriptional regulation [32]. Functional genomics increasingly use quantitative trait measures with respect to transcript expression [33,34]. Conceivably, these studies may identify differential gene expression patterns that underlie environmental challenge, that is salt load in appropriate tissues and cells under short and long-term conditions. Genomics is getting more molecular functional and more attractive for researchers who will extend their interest in basic research. Despite all these efforts in identifying, for example, genotype-salt-sensitivity–phenotype associations, it is, however, surprising that pharmacogenetic studies, using, for example, diuretics, still failed to report clinically significant genotype-driven differential effects of this drug class on blood pressure, even if recommendations to tackle this issue are made [35,36]. What can we tell the practitioner? Even without any genetic/genomic knowledge, they are still doing well in recommending ‘modest’ dietary salt reduction (≤5–6 g/day) in patients with essential hypertension but also in the general population. Prescribing diuretics – with strong blood pressure-lowering effects as mono but also in combination therapy – only makes sense in combination with this dietary recommendation. ACKNOWLEDGEMENTS Conflicts of interest S.-M. Brand, formerly S.-M. Brand-Herrmann/S.-M. Herrmann, was supported by a grant from the EU-Project ICT in the FP7-ICT-2007-2, project number 224635, VPH2 – Virtual Pathological Heart of the Virtual Physiological Human.