In this analysis we have examined relations between two non-synonymous KCNMB1 polymorphisms, which were previously shown to be associated with reductions in blood pressure and cardiovascular outcomes, and components of the arterial waveform. E65K was associated with lower augmented pressure and central pulse pressure, without effects on aortic PWV or muscular artery diameter or flow. The association was not affected by age, gender, and anti-hypertensive therapy. We found no association between tonometry measures and V110L.
was first reported to have a protective effect on diastolic hypertension in a Spanish cohort (n=3876) [4
]. Subsequently, the INVEST trial found that K
-carriers achieved blood pressure control more rapidly and were less likely to require multiple medications compared to non-carriers [7
]. Most recently we have shown that E65K
-carriers have lower systolic and diastolic pressure in the setting of anti-hypertensive therapy in two independent cohorts [6
]. Physiology studies in β1-deficient mice, which are hypertensive, demonstrate profound abnormalities of blood vessel relaxation. While altered potassium handling with acute volume expansion has also been described in these mice, the mean arterial blood pressure in β1-deficient mice is not salt-sensitive [21
], suggesting that abnormalities of systemic arterial relaxation rather than salt handling have the greatest effect on blood pressure in these animals. Consistent with a direct effect on arterial stiffness in humans we now report associations with lower augmented pressure and central pulse pressure in E65K
-carriers. These associations were not explained by age, sex, anti-hypertensive therapy or atherosclerotic disease risk, suggesting that the genetic effect may be related directly to alterations in vascular structure or function. By comparison, the V110L
polymorphism, which has not been shown to affect BKCa-channel function, was not associated with tonometry measures in our cohort. While our studies support an important role for E65K
and arterial stiffness, the potential for altered salt and mineralocorticoid effects mediated by E65K
warrants further study.
Our findings add to a growing body of literature suggesting that E65K
has protective effects on the vasculature. Fernandez-Fernandez, et al.[5
], demonstrated that the E65K
gain-of-function mutation may result in membrane hyperpolarization, which they predicted would cause vascular smooth muscle cell relaxation and blood vessel dilation. To our knowledge our study is the first to test for this predicted effect using direct measures of vascular function. It was difficult to predict whether E65K
would cause a generalized effect on all blood vessels because the BKCa channel is expressed in a range of arteries, including the aorta, large muscular arteries and small resistance arterioles [22
]. In this analysis, K
-allele carriers had a reduction in augmented pressure, suggesting a reduction in reflected wave amplitude. The reflected pressure wave is formed as forward traveling pressure waves reflect from sites of impedance mismatch. Mismatch occurs when larger, more compliant (low impedance) arterial segments connect with smaller, stiffer (high impedance) arteries and resistance vessels [16
]. Diminished augmented pressure in K
-carriers suggest that E65K
reduces arterial impedance mismatch somewhere in the arterial tree, perhaps by altering muscular artery or resistance vessel diameter or elastance. The latter physical attributes are the principal determinants of local vascular impedance of an arterial segment. Therefore, our results support a role for E65K
in the modulation of wave reflection in humans.
The amplitude of the reflected wave is determined by both the size of the initial forward traveling wave and the proportion of that wave that is reflected, which can be estimated by using augmentation index. Since neither an index of relative wave reflection (the augmentation index) nor forward wave amplitude had clear association with E65K, it is difficult to determine whether lower augmented pressure and central pulse pressure was attributable to reduced relative wave reflection or combined reductions in relative wave reflection and forward wave amplitude. However, our subgroup analysis in subjects not receiving antihypertensive therapy found reduced augmentation index in K-carriers, suggesting an effect via the reflected wave. Additional studies that measure both pressure and flow, so that true forward and reflected wave amplitude can be evaluated, may clarify this remaining question.
We found no difference in carotid-femoral PWV, suggesting that the mutant channel does not affect aortic wall stiffness, which is closely related to carotid-femoral PWV. Because there were no significant differences in brachial artery diameters at rest in K-carriers, we conclude that the mutant BKCa channels did not have a general effect on the diameter of the brachial artery and presumably other medium-sized muscular arteries. We did not observe an association with mean arterial pressure, suggesting that if there was a reduction in global peripheral resistance, it was offset by an increase in cardiac output, resulting in no change in mean arterial pressure. Further studies that include measures of cardiac output and total peripheral resistance will be required to clarify the relation between E65K and global resistance vessel function.
Measures of brachial artery diameter and flow were potentially confounded by the use of anti-hypertensive agents from a variety of medication classes in our sample. In a subgroup analysis of subjects not receiving anti-hypertensive agents, we found a greater increase in brachial artery size and flow volume in the setting of hyperemia in K
-carriers, suggesting that E65K
may have an effect on muscular arteries. The effect of E65K
on brachial artery measures was identified only in the setting of hyperemia, suggesting that K
-carriers may have increased sensitivity to endothelium-dependent or other mediators of smooth muscle cell relaxation. Indeed, nitric oxide has been shown to positively regulate smooth muscle cell calcium “sparks” that trigger KCNMB1
-dependent BK channel opening in pressurized rat cerebral arteries[25
]. Our finding that flow-mediated brachial artery dilation was greater in K
-carriers, if confirmed, would support a mechanistic link between E65K
and nitric oxide.
Because our results are consistent with a prior linkage study [12
] and findings in the Spanish cohort and the INVEST trial [4
], we consider a false positive association to be less likely. We sought to reduce the effects of multiple testing by restricting the initial analyses to the linear analysis of several measures of arterial stiffness with two non-synonymous polymorphisms, after taking into account known covariates. We consider our exploration for modifiers of E65K
as secondary analyses and restricted these analyses to pre-specified interaction terms in order to minimize the risk of false-positive associations.
Strengths and limitations
The routine ascertainment of tonometric measures, genotypes, and covariate data in a community-based sample are strengths of the study. Our estimates of forward and reflected wave amplitude were based on assessment of pressure only. In order to ascertain the true forward and reflected wave amplitude, both pressure and flow are required. We acknowledge that we did not account for multiple testing and that our findings will need to be replicated in an external cohort. In addition, the generalizability of our findings to individuals who are younger, or are not white or of European descent will need to be tested. Because the genotyping was conducted in the standard Framingham unrelated DNA plate set, many of the individuals who were phenotyped were not genotyped.