GRK4 Genetics and Response to -Blocker

G protein–coupled receptors (GPCRs) such as dopamine receptor subtypes, angiotensin receptor subtypes, and β-adrenoceptors have important roles in intracellular signaling and are of considerable interest in cardiovascular regulation and disease.1,2 GPCRs can be phosphorylated by GPCR serine/threonine kinases (GRKs).1,2 This causes β-arrestin to bind to the GPCR and prevents coupling of G proteins. Such uncoupling results in endocytosis and down-regulation of the GPCR.1,2 Thus GRKs play a unique role in initiating the desensitization of GPCRs after their activation, so terminating the action of their ligand on cyclic adenosine monophosphate (cAMP) signaling within a target cell. The GRK family consists of seven members.1,2 These have been divided into three subgroups based on structure and function. Unlike most family members, whose expression is widespread, the family member GRK4 has an expression pattern that is restricted to the testis, myometrium, brain, and the renal proximal tubule.1,2 It is therefore likely to have a more specific function than the other members. GRK4 is known to suppress β-adrenoceptor density on the plasma membrane of cells and β-blockers suppress GRK4 activity.3 Several single-nucleotide polymorphisms (SNPs) of the GRK4 gene have been described. Three of them—R65L in exon 3, A142V in exon 5, and A486V in exon 14—are associated with impairment of D1 receptor mediated increase in cAMP.1 Since the natriuretic effect of dopamine is impaired in hypertension,2 could genetic variant(s) of GRK4 be contributing? GRK4 increases angiotensin II type 1 receptor expression and function.4 So, in various ways, dysregulation, resulting from constitutive (ligand-independent) activation of alleles of GRK4 that contain these genetic variants are of considerable interest in inherited aspects of essential hypertension, particularly in salt-sensitive hypertension. Indeed, the GRK4 locus on chromosome 4p16.3 is linked to essential hypertension and salt sensitivity.5 Genetic association studies have compared allele frequencies of the R65L, A142V, and A486V SNPs in the GRK4 gene in hypertensive and normotensive subjects. In an Anglo-Celtic Australian cohort with two hypertensive parents, and thus strong predisposition to hypertension, the V allele of A486V was associated with essential hypertension.6 This was also seen in an Italian cohort.7 Neither study found an association of the other two SNPs with hypertension. In the Australian study, the V allele of A486V was also associated with elevation in systolic pressure (P = 0.0017), and, in males only, the L65 and V142 alleles tracked with elevation in diastolic pressure.6 Alleles of these SNPs are in linkage disequilibrium, which means they tend to be coinherited, and haplotypes differed between the hypertensive and normotensive groups.6 In Chinese subjects, the A486V variant8 and the L65/V142/A486 haplotype9 have shown associations with hypertension. In black and white American twins, this haplotype was also associated with a 1.05-mm Hg steeper increase in systolic pressure per year of increase in age compared with the most common haplotype.10 Interestingly, mice transgenic for the V142 variant display impairment of the natriuretic action of the D1 receptor and as a result are hypertensive.1 In the present issue of American Journal of Hypertension, Bhatnagar and co-workers have examined whether there is an association between these three GRK4 variants and the response in mean arterial pressure (MAP) to the selective β1-adrenoceptor blocker metoprolol.11 The subjects were African Americans with early hypertensive nephrosclerosis. The target MAP was 107 mm Hg (i.e., a reduction to at least 140/90 mm Hg systolic/diastolic pressure) and the authors measured the time in days taken to reach this MAP for subjects of each genotype. They found that MAP in men with the A142 allele took longer to get down to 107 mm Hg or lower in response to metroprolol. With each additional A142 allele time to reach ≤107 mm Hg was slower by half than required for those with the V142/V142 genotype (P = 0.03). The authors also noted a possible interaction of the SNPs, with the most striking effect being seen for subjects who possessed both the A142 allele and the L65 allele (A142 + L65 haplotype). Indeed, those with the A142 allele responded more slowly only if they also had an L65 allele (P = 0.04). The A486V variant showed only a weak association with response to metproprolol (P = 0.04). When corrected for multiple comparisons, the statistical significance was ameliorated. Hazard ratios (HRs) indicated significance for the A142V variant (HR 1.54, P = 0.009) and for the R65/V142/A486 haplotype (HR 1.54, P = 0.03). Although 317 subjects were genotyped, which is a reasonable number for a preliminary survey of an effect such as this, when subdivided by gender the numbers were 197 men and 131 women. The latter, being lower in number, would thus have less power to show an existing relationship. Moreover, the significant findings were seen in men, with no genotypic difference observed for women. The lower n value could have contributed to the absence of an association with genotype in women, as the authors acknowledge. The number of V486 homozygotes was low (n = 6), making data for these rather tentative. Low n values applied in some of the haplotype analyses and the data obtained may be unreliable. Whereas the authors state that haplotype analyses are often an efficient approach to uncovering interactions, such analyses have also been criticized for generating spurious results, especially as n values diminish with sub-grouping of data. What this means is that the n values in some cells of the contingency table analysis could have been too low for a reliable result to be obtained. This is because a low n value can generate a spuriously high statistic and thus a P value that is significant, when a nonsignificant P value would arise had the overall n value of the study group been much higher. The authors note, moreover, problems with adequacy of power, the low n value being of particular concern in causing this. Thus, a bigger study would appear necessary. A “shot-gun” approach to genetic discovery has the potential to yield a significant new finding. But to be assured that such a finding was real would be assisted if any association observed was attained at a high level of statistical significance. Alas, the findings of Bhatnagar et al. include differences that are marginal at best, and sex-specific to boot. Nevertheless, there will be those who consider a marginal outcome, as seen in this study, as having merit regardless, since it could justify this or another group undertaking more extensive studies on the same SNPs in hypertensive African-American patients with kidney disease or in other patient cohorts. Furthermore, each population cohort will be different, so results could be valid in the group studied, but need not apply to another group. There is, moreover, genetic admixture among African Americans, as noted by the authors. To conclude, hypertensive Black American men with an A142 variant of GRK4, by possibly having higher β-adrenoceptor activity, appear less responsive to a β-blocker, in terms of the time required for the β-blocker to lower their blood pressure to normal. Although this article tests a hypothesis that is relevant intellectually and quite interesting, so making it novel, the results cannot be accepted at this stage, and would need to be confirmed by very much larger studies. Nevertheless, the findings offer food for thought in research on genetic factors that influence the response of hypertensive subjects to a β-blocker.