Renin angiotensin system gene polymorphisms and cerebral blood flow regulation: The MOBILIZE Boston study

Abstract

Background and purpose

Our objective was to investigate the associations between polymorphisms in representative genes of the renin angiotensin system with measures of cerebral blood flow regulation in older adults.

Methods

Participants in this analysis were white subjects (n=335) in the MOBILIZE Boston study, an observational study of community-dwelling elders who underwent transcranial Doppler while sitting and standing and during hypercapnea and hypocapnea. Autoregulation phenotype was the change in cerebrovascular resistance from sit to stand. Vasoreactivity (VR) phenotype was the slope of the change in cerebrovascular conductance vs change in end-tidal CO₂. Total of 33 tagged single nucleotide polymorphisms (SNP) were selected in the angiotensinogen gene (AGT), the angiotensin converting enzyme (ACE) gene and the angiotensin receptor gene (AGTR). Regression analyses adjusted for age, gender, body mass index, mean arterial blood pressure, stroke and use of antihypertensives were conducted for each SNP and outcome. Bonferroni corrections were used to adjust p-values for multiple testing.

Results

In the AGT gene, only the rs699 SNP was associated with VR after Bonferroni correction (p=0.00028). Homozygous carriers of the CC genotype of this SNP had lower VR compared to the CT or TT genotypes. There were no significant associations with autoregulation measures. None of the SNP’s in the other genes was associated with our phenotypes.

Conclusion

This analysis suggests that the AGT gene may be involved in vasoreactivity independent of blood pressure. Larger studies are needed to confirm the role of this gene in cerebrovascular health and aging.

Introduction

Both hypertension and aging lead to alterations in cerebral blood flow regulation ¹. These alterations may be associated with stroke² and cognitive impairment ³. Recent evidence suggest that angiotensin II plays an integral role in the regulation of cerebral blood flow.⁴ The mechanisms are not fully understood but are likely to be mediated by the endothelial response to angiotensin II in the brain.⁵ In particular, angiotensin II affects the endothelial production of nitric oxide, and inhibiting the angiotensin receptors is associated with improved cerebral autoregulation⁶, reduction in cerebral blood flow decline after middle cerebral artery occlusion⁷, and normalization of endothelial nitric oxide production⁸. Endothelial nitric oxide production is a key factor in the cerebrovascular response to carbon dioxide (CO₂)⁹. The latter can be clinically assessed by transcranial Doppler technology (TCD) and is a measure of cerebrovascular health and function. Further, renin angiotensin system gene polymorphisms are associated with angiotensin II levels and measures of systemic blood pressure regulation ¹⁰ and baroreflex sensitivity.¹¹ Therefore, we hypothesized that polymorphisms in genes of the renin angiotensin system are also associated with autoregulation and the cerebrovascular response to changes in CO₂.

Multiple genes in this system have been linked with various aging and vascular phenotypes. Of those, the angiotensinogen gene (AGT), the angiotensin converting enzyme (ACE) gene and the angiotensin receptor gene (AGTR) are the most widely studied. AGT controls angiotensin II activity by promoting the transcription of the angiotensinogen protein. ACE controls angiotensin II activity by promoting the transcription of the ACE protein, the enzyme that metabolizes angiotensin I to angiotensin II. AGTR controls the transcription of the receptor that binds angiotensin II to produce its vascular effects. Although these genes have been linked to various vascular phenotypes, no study has investigated the association of these genes to cerebrovascular function and regulation. Since blood flow regulation and vascular reactivity are heritable traits (h² is at least 60%), a genetic association study is justifiable¹².

Autoregulation and cerebral vasoreactivity to carbon dioxide (VR) are rarely collected in large cohort studies. The MOBILIZE Boston Study (which stands for “Maintenance of Balance, Independent Living, Intellect, and Zest in the Elderly of BOSTON”) is a population-based prospective observational study with genetic data that is measuring cerebrovascular function using TCD in elderly individuals.

Therefore our objective was to investigate the associations between polymorphisms in the three representative genes of the renin angiotensin system and measures of cerebral blood flow regulation in the MOBILIZE Boston study.

Methods

Study design

MOBILIZE Boston is a population-based prospective observational study funded through a National Institute on Aging program project grant. The details on the design and recruitment are described elsewhere^{13, 14}. The Institutional Review Board at Hebrew SeniorLife approved this study and each participant provided written informed consent.

Participants

The recruitment process included a door-to-door population-based recruitment of a probability sample. ^{13, 14} Eligibility criteria included age 70 years or older, ability to speak and understand English, and plans to be living in the recruitment area for at least 2 years. Exclusion criteria included cognitive impairment defined as a Mini-Mental-State-Examination (MMSE) score less than 18¹⁵, hearing or visual impairment that interfered with communication, having a terminal illness and inability to walk 20 feet without assistance. Of the 4319 individuals who were age-eligible, 1616 agreed to be screened, 765 were eligible and enrolled and 686 (545 white) agreed to have DNA collected and were genotyped ¹³. Participants’ assessments included anthropometric and blood pressure measurement, health habits, medical history (self-reported stroke, diabetes, heart disease, congestive heart failure, hypertension, other medical diagnoses), medication inventory, and functional and cognitive evaluations (MMSE, Trail Making Test) and cholesterol measurements¹⁴.

TCD Procedure and measures of cerebral blood flow regulation

Subjects were instrumented for heart rate (HR, ECG) and beat-to-beat arterial pressure monitoring (blood pressure, Finapres, Ohmeda Monitoring Systems, Englewood, CO) as previously described.¹⁴ End-tidal CO₂ was measured using a Vacumed CO₂ Analyzer (Ventura, CA) attached to a nasal cannula. TCD ultrasonography (MultiDop X4, DWL-Transcranial Doppler Systems Inc., Sterling, VA) was used to measure middle cerebral artery (MCA) mean blood flow velocity (BFV) at rest and in response to: 1) changes in end-tidal CO₂ and 2) blood pressure during a sit-to-stand protocol, as previously described.¹⁶ The MCA signal was identified according to the criteria of Aaslid et al ¹⁷ and recorded at a depth of 50 to 60 mm. A Mueller-Moll probe fixation device was used to stabilize the Doppler probe at the temporal bone window for the duration of the study. The envelope of the velocity waveform, derived from a fast-Fourier analysis of the Doppler frequency signal, was digitized at 500 Hz, displayed simultaneously with the blood pressure, ECG, and end-tidal CO₂ signals, and stored for later off-line analysis. TCD procedures were conducted by one dedicated TCD technician for the MOBILIZE Boston. The correlations between 2 TCD measurements on 21 subjects collected 6 months apart were excellent with R² Correlation Coefficients of 0.79 and 0.83, and the intra-class Correlation Coefficients of 0.92 and 0.95, respectively.

CO₂ Breathing Protocol to measure vasoreactivity: BFV in the MCA was measured continuously while subjects inspired a gas mixture of 8% CO₂, 21% O₂, and balance nitrogen for 2 minutes and then mildly hyperventilated to an end-tidal CO₂ of approximately 25 mmHg for 2 minutes. CO₂ itself may also affect blood pressure and the degree of change is dependent on baseline blood flow velocity¹⁸. Therefore, we used cerebrovascular conductance (cerebral blood flow/mean arterial blood pressure) and the percent change from baseline as our measure for VR.^{19, 20} VR hence was calculated as the slope of the percent change in cerebrovascular conductance vs. the change in end tidal carbon dioxide²⁰. This measure is more reflective of change in the vascular response to end tidal CO₂²¹.

Sit-to-Stand Protocol to measure cerebral autoregulation

The active sit-to-stand procedure is previously described in detail ²². After instrumentation, subjects sat in a straight-backed chair with their legs elevated at 90 degrees in front of them on a stool. For each of 2 active stands, subjects rested in the sitting position for 5 minutes, then stood upright for 1 minute. The initiation of standing was timed from the moment both feet touched the floor. Data were collected continuously during the final 1-minute of sitting and 1 minute of standing. To ensure that a sufficient stimulus was applied, a blood pressure drop of at least 10 mmHg during the stand event was required when calculating the outcome measures. Cerebrovascular resistance (CVR) was calculated from the ratio of blood pressure/BFV. Autoregulation was quantified by the difference between CVR sitting and standing (del CVR=CVRstand - CVR_sit).

Gene and SNP selection

The following genes in the renin angiotensin system were included: Angiotensinogen (AGT), Angiotensin Converting Enzyme 1 (ACE), and Angiotensin II Receptor 1(AGTR1). To capture the variation in these candidate genes, we selected tagged single nucleotide polymorphisms (SNP) within each gene . These Tag SNP captured most of the genetic information in a region through linkage disequilibrium (LD). Non-redundant Tag SNPs were selected for pairwise correlation (r²) ≥ 0.80 of HapMap II in the Caucasian population using Haploview. We selected Tag SNP that covered the entire gene region as well as its 10 kb 5′ upstream and 10 kb 3′ downstream regions. We added empirical SNP in the target genes that have been previously reported in the literature to be related to other vascular phenotypes. We identified 33 tag and empirical SNP in the selected genes based on our criteria. Table 1 provides the list of genes and selected SNP, p-values for the tests for Hardy Weinberg equilibrium, and the minor allele frequency in the MOBILIZE Boston study.

Table1.

Selected Renin Angiotensin System genes/SNP, Minor Allele Frequencies and their HWE in the white participants of MOBO

Gene_SNP	HWE	MAF	Gene_SNP	HWE	MAF
ACE_rs4293	0.7919	0.24	AGT_rs4762	0.324	0.01
ACE_rs4295	0.7782	0.18	AGT_rs11122576	0.1811	0.11
ACE_rs4309	0.5607	0.14	AGT_rs11122577	0.539	0.02
ACE_rs4311	0.2304	0.26	AGT_rs2004776	0.447	0.04
ACE_rs4329	0.9069	0.18	AGT_rs3889728	0.6523	0.04
ACE_rs4341	0.8569	0.18	AGT_rs4028824	0.3227	0.004
ACE_rs4357	0.8622	0.01	AGTR1_rs4681157	0.9429	0.18
ACE_rs4461142	0.3841	0.23	AGTR1_rs2933250	0.173	0.05
ACE_rs4459610	0.492	0.16	AGTR1_rs2638363	0.6138	0.05
ACE_rs8066276	0.3407	0.13	AGTR1_rs10935724	0.8723	0.10
ACE_rs12451328	0.7243	0.18	AGTR1_rs4681444	0.4035	0.04
AGT_rs7079	0.653	0.15	AGTR1_rs1492100	0.0764	0.05
AGT_rs2478523	0.7171	0.14	AGTR1_rs1492099	0.9518	0.03
AGT_rs2478545	0.127	0.05	AGTR1_rs389566	0.6865	0.10
AGT_rs2478544	0.2938	0.03	AGTR1_rs385338	0.7642	0.04
AGT_rs6687360	0.1706	0.10	AGTR1_rs5182	0.8081	0.18
AGT_rs699	0.1368	0.16	AGTR1_rs5186	0.2399	0.08

Genotyping

Genotyping was conducted at the Harvard Medical School-Partners Healthcare Center for Genetics and Genomics using the Sequenom iPLEX SNP genotyping.. Multiplex PCR assays are designed using Sequenom SpectroDESIGNER software (version 3.0.0.3) by inputting sequence containing the SNP site and 100 base pairs of flanking sequence on either side of the SNP. Quality control was conducted using a subsample (5%) of duplicate genotyping to identify any discordance in the results.

Statistical analyses

Hardy-Weinberg equilibrium (HWE) was examined for each SNP using the Fisher exact test. We used multiple regression analyses to test the association between cerebral blood flow phenotypes and each SNP in the selected genes. The additive genetic model was used for the main analyses. Covariate selection was based on clinical and prior evidence for potential confounding in the gene-cerebral blood flow relation. In addition to demographics and BMI, stroke affects the TCD measures and may be associated with renin angiotensin gene polymorphisms.²³ Both genes and TCD measures are associated with blood pressure and hence this was included in the model. ²⁴ Antihypertensives affect blood pressure and may interact with renin angiotensin genes in their blood pressure effect. Due to our sample size, we limited covariates to age, gender, mean blood pressure during the TCD procedure, body mass index, prior stroke, and use of anti-hypertensive medications.

To adjust for the single-point significance level for multiple testing with correct type-I error rate, we used a very conservative Bonferroni correction method. We investigated 2 phenotypes (VR and del-CVR) and 33 SNP. Therefore, we considered a p-value of less than 0.05/66 =0.00075 to be “analysis-wide” significant. We also calculated empirical p-values based on global random 10,000 permutations tests as described previously.²⁵ All reported p values are from two-sided tests. In order to minimize population admixture, we limited our analysis to white participants. Finally, in the SNP that we identify an association with our outcome, we explored the possibility that it will interact with angiotensin converting enzyme inhibitors (ACEINH). We provide these results in the online supplement.

Results

Study sample and characteristics

Of the 545 enrolled white participants, all had successful genotyping data. Of those 210 (39%), did not have cerebral blood flow data secondary to inability to insonate the middle cerebral artery or poor data quality. Therefore, this analysis was completed on 335 white participants (age: 77.8±0.3 years, BMI: 26.8±0.2, women: 54%, stroke: 9%, receiving antihypertensives: 62%, blood pressure: 127.4±0.8/69.6±0.4 mm Hg). Those without a Doppler insonation window were older (79.3±0.4, p=0.03), more likely to be women (75%, p<0.001), and had higher systolic blood pressure (132.8±1.7 p=0.001). There were no differences in the genotype distributions between subjects with and without TCD data of all the SNP’s except for ACE_rs12451328 (p=0.0026). (Online Table A)There was also no association between demographic or vascular factors and the autoregulation phenotype. In contrast, lower VR was associated with older age, female sex, higher mean arterial blood pressure, smoking and higher cholesterol levels. (Online Table B)

Del-CVR

Only 3 polymorphisms in the AGT gene and one in the ACE gene were associated with the autoregulation phenotype after adjusting for covariates. However, after adjusting for multiple testing none of these associations were significant. Table 2

Table 2.

Significant associations between polymorphisms in the ACE, AGT, and EDN1 genes and autoregulation of cerebral blood flow or vasoreactivity in the white participants of Mobilize Boston

			P-value corrected for multiple
Phenotype	SNP	P-value	testing by permutation
DEL-CVR	AGT_rs7079	0.0065	0.45
	ACE_rs4341	0.0115	0.57
	AGT_rs2478523	0.0149	0.73
	AGT_rs699	0.0217	0.83
VR	AGT_rs699*	0.000277	0.03
	AGT_rs2478544	0.00142	0.13
	AGT_rs2478545	0.00175	0.16
	AGT_rs2478523	0.0037	0.31

All analyses adjusted for age, gender, BMI, stroke, antihypertensive use and mean blood pressure during TCD procedures. DEL-CVR: Change in cerebrovascular resistance from sit to stand. VR: cerebrovascular reactivity

^*In the multivariate model for AGT_rs699, p value for age=0.0002, gender= 0.2852, BMI=0.1348, stroke=0.0387, Antihypertensives=0.0033, and mean blood pressure during TCD procedure=0.0076

VR

Four polymorphisms in the AGT gene were associated with the cerebral vasoreactivity phenotypes. Only one SNP, however, remained significant after Bonferroni correction (p=0.00027) and permutation adjustment (p=0.03). Table 2 shows the nominal and adjusted p-values for these associations. As shown in Figure 1, carriers of the CC genotype (112(33%))of rs699 had lower VR compared to either the CT or TT genotypes or the combined CT/TT genotype groups (223(67%)). Since the differences in VR maybe related to other factors, we compared clinical, biochemical and physiological factors in the CC genotype to the CT or TT genotype. As shown in Table 3, there were no significant differences in these factors between the 2 groups, suggesting that the association is unlikely to be related to non-genetic factors.

Figure 1.

Vasoreactivity in the genotypes of the AGT_rs699 SNP: (A) additive model with three genotypes. (B) Dominant model with the 2 genotypes.

Table 3.

Demographic, clinical, and neuropsychological, and biochemical comparison between 2 groups (CC vs CT and TT) of the RS699 SNP in the AGT gene. Numbers are means ± standard errors or N (%).

Variable	CC genotype (n=112)	CT or TT genotype (n=223)	p-value
Age, years	77.5±0.5	78.1±0.4	0.34
Gender, women	68 (61%)	114 (51%)	0.1
BMI, Kg/m2	27.5±0.5	26.6±0.3	0.11
General health, very good or excellent	105 (93%)	200 (90%)	0.26
Alcohol, none	80 (87%)	157 (89%)	0.58
Smokers, current	1 (1%)	11 (6%)	0.12
Blood pressure and blood flow
SBP, mm Hg	129±1	126±1	0.09
DBP, mm Hg	70±1	69±1	0.33
PP, mm Hg	59±1	57±1	0.24
BFV, cm/sec	41.2±0.9	40.2±0.7	0.39
Medical HX
Stroke	12 (11%)	18 (8%)	0.45
Hypertension	67 (60%)	125 (71%)	0.6
Anti-hypertensive use	67 (60%)	136 (61%)	0.8
Heart disease	26 (23%)	57 (26%)	0.59
Congestive heart failure	5 (4%)	9 (4%)	0.87
Diabetes	14 (12%)	26 (12%)	0.85
Cognitive function
Mini Mental Status Exam	27.7±0.2	27.8±	0.77
Trail Making Test, part B, seconds	133.6±7.9	125.3±5.5	0.38
Laboratory
Low density lipoprotein, mg/dl	108±3	108±2	0.96
Cholesterol, mg/dl	187±3	186±3	0.84

BMI: Body Mass Index, SBP: systolic blood pressure, DBP: diastolic blood pressure, PP: pulse pressure, BFV: blood flow velocity in the middle cerebral artery.

None of the other SNP in the AGT, ACE or AGTR genes was significantly associated with VR or del-CVR. The full list of SNP-phenotype associations is provided as an online supplement (Table C). Further, none of the haplotypes were associated with either phenotype. Finally, ACEINH were not associated with VR and did not interact with the rs699 SNP in the AGT gene. (online supplement Table D and E).

Discussion

This study provides preliminary evidence that a polymorphism in the AGT gene is independently associated with cerebral vasoreactivity in white elderly individuals. Homozygous carriers of the CC genotype of the rs699 SNP have lower cerebral CO₂ vasoreactivity compared to the other genotypes.

To our knowledge, this is the first study to provide evidence that renin angiotensin system genes are also involved in cerebral vasoreactivity. Prior evidence suggests a genetic role of this system in brain health and diseases such as stroke, depression and cognitive impairment ^{26, 27}. This study adds evidence that this system may also be involved in VR which is linked with aging outcomes such as stroke² and dementia²⁸.

CO₂-dependent vasoreactivity is mediated in part by the endothelium and is related to changes in nitric oxide.^{29, 30} Changes in end-tidal CO₂ are associated with fast changes in pH which modulate the effect of nitric oxide synthetase leading to changes in nitric oxide production. ³¹ In addition, ATP-dependent K+ channel activation may mediate CO₂- induced NO activity in the pial arterioles.³² Angiotensin II modulates nitric oxide production ³³ and affects K+ channels.³⁴ Our finding that a polymorphism in the AGT gene is associated with vasoreactivity, lends further support to the role of angiotensin II in the endothelium and cerebrovascular response to CO₂, possibly by affecting nitric oxide production and potassium channels. This role is further supported by the fact that in animal models, hypercapnia has been associated with increased angiotensin II levels.³⁵

Within the AGT gene, only one SNP was associated with VR, rs699 (also termed M235T)³⁶. This polymorphism has been previously reported to be associated with angiotensin II levels³⁷. Carriers of the CC genotypes have lower angiotensin II levels, which is likely to be associated with lower type 2 angiotensin II receptor activation and decreased endothelial-mediated vasodilatation. The fact that the heterozygous, CT genotype, did not have an intermediate level between CC and TT genotypes suggests that the genetic model for VR is a dominant one, where having T/C or T/T is associated with higher VR.

The lack of interaction between ACEINH and this SNP in AGT gene is likely related to our small sample size especially those on ACEINH (n=87), In addition, the genotype that is likely associated with low angiotensin II levels is also associated with lower VR. Therefore, it is unlikely to find in a cross sectional study a relation between drugs that lower Angiotensin II levels and VR. To test this hypothesis, a longitudinal study design is needed where vasoreactivity is measured before and after exposure to ACEINH. Such a design will allow us to test if the genotype with lower level or higher level of RAS is more likely to demonstrate and ACEINH by gene interaction.

We did not identify a genetic association with autoregulation in our sample. The robustness of autoregulation in the face of aging and vascular changes³⁸ may explain the lack of a significant association between our genetic polymorphisms and autoregulation phenotype. This may also be due to the relatively small sample size of our study.

One important limitation of this study is the relatively small sample size. Collecting cerebral blood flow data is limited by the time, cost, and ability to insonate the middle cerebral artery. We were not able to obtain cerebral blood flow data in 39% of those enrolled in the MOBILIZE Boston study. This TCD failure rate is lower than prior TCD population studies.² Although we found that those without TCD had higher vascular risk, there was no association with the AGT_rs4699 SNP suggesting that the potential bias from TCD measurement failure is likely to have a small impact on our results. Although a larger study is necessary to confirm and replicate our findings, such a study will be extremely expensive and resource intensive. Our study is also cross-sectional, raising the possibility that there is selective drop out of those with impaired cerebral blood flow prior to reaching older age. Finally, population admixture is a concern in any genetic association study, especially in vascular phenotypes where racial differences are critical. We limited our analysis to white participants to decrease the effect of admixture on our findings. However, this limits the generalizability of our findings to non-whites. We also did not investigate other genes in the renin angiotensin system pathway, such as the renin gene that may have an impact on cerebrovascular function.

Conclusion

This study suggests that a polymorphism in the AGT gene known to be involved in blood pressure control is also associated with cerebral vasoreactivity. This association is independent of blood pressure and stroke. Although it maybe difficult to achieve, larger studies are needed to confirm the role of renin angiotensin system genes in abnormal cerebral blood flow regulation and its clinical consequences.