Unraveling the relationships between alpha- and beta-adrenergic modulation and the risk of heart failure

Background The effects of α and ß adrenergic receptor modulation on the risk of developing heart failure (HF) remains uncertain due to a lack of randomized controlled trials. This study aimed to estimate the effects of α and ß adrenergic receptors modulation on the risk of HF and to provide proof of principle for genetic target validation studies in HF. Methods Genetic variants within the cis regions encoding the adrenergic receptors α1A, α2B, ß1, and ß2 associated with blood pressure in a 757,601-participant genome-wide association study (GWAS) were selected as instruments to perform a drug target Mendelian randomization study. Effects of these variants on HF risk were derived from the HERMES GWAS (542,362 controls; 40,805 HF cases). Results Lower α1A or ß1 activity was associated with reduced HF risk: odds ratio (OR) 0.83 (95% CI 0.74–0.93, P = 0.001) and 0.95 (95% CI 0.93–0.97, P = 8 × 10−6). Conversely, lower α2B activity was associated with increased HF risk: OR 1.09 (95% CI 1.05–1.12, P = 3 × 10−7). No evidence of an effect of lower ß2 activity on HF risk was found: OR 0.99 (95% CI 0.92–1.07, P = 0.95). Complementary analyses showed that these effects were consistent with those on left ventricular dimensions and acted independently of any potential effect on coronary artery disease. Conclusions This study provides genetic evidence that α1A or ß1 receptor inhibition will likely decrease HF risk, while lower α2B activity may increase this risk. Genetic variant analysis can assist with drug development for HF prevention.

Several ADRA1A (α1A), ADRA2B (α2B), and ADRB2 (β2) variants were associated with a decrease in the risk of diabetes and/or HBA1c, while several ADRB1 (β1) variants were associated with an increase in the risk of diabetes (Supplementary Table 3) One ADRA1A (α1A) and one ADRB1 (β1) variant were associated with a decrease in HDL cholesterol.Two ADRB1 (β1) variants were associated with an increase in LDL cholesterol, while an ADRB2 (β2) variant was associated with the reverse (increase in HDL and decrease in LDL).Finally, several ADRB1 (β1) variants were associated with an increase in triglycerides.

Supplementary MR analyses for α1A and α2B
As specified in the "Selection of genetic instruments" section of the main manuscript, since less than three SNPs weighted by diastolic BP could be selected for α1A and α2B, we identified two additional α1A SNPs using the same previous criteria but with an associated Pvalue ≤ 1×10 -4 ; and three additional α2B SNPs using the same previous criteria but with a clumping threshold r² < 0.6 to perform the MR analyses (Figure 1 and Supplementary Figure 1, Supplementary Table 3).
Supplementary MR analyses for α1A and α2B were performed with only the SNPs selected using the same criteria as for ß1 and ß2: association at genome-wide significance (P ≤ 5×10 -8 ) and clumping at a LD threshold of r 2 < 0.1 (one SNP for α1A and two SNPs for α2B).
They provided similar results for HF risk (Supplementary Figure 3), LV dimensions, CAD risk, and HF risk-adjusted for CAD risk as outcomes.
There were two exceptions, both associated with the lowering of α2B: its deleterious effect on LVEF while numerically and directionally similar was no more statistically significant (β = -0.07% 95% CI -0.16-0.01,P=0.10), as well as its protective effect on the CAD risk (OR 0.96 95% CI 0.88-1.04,P=0.28).

Additional MR analyses with systolic BP or heart rate as proxies
The analyses were repeated, but this time with the genetic instruments weighted by systolic BP as proxies for ß1 and ß2 activities and by heart rate as a proxy for α2B activity.The ORs derived from the corresponding MR estimate for each adrenergic receptor are given a 1 beat per min (bpm) decrease in heart rate.

Heart failure
The results were consistent with the main analysis.

LV dimensions
Similar results as the main ones were found for ß1 (Supplementary Figure 8 and Supplementary Table 8) and ß2 (Supplementary Figure 8 and Supplementary Table 8) activity modulation regarding LV volumes, LV mass, and LVEF as outcome.
While the size and direction of the effects of a lower α2B activity on LVESV, LVEDV, LV mass, or LVEF were similar to those found in the main analyses, they were not statistically significant (Supplementary Figure 9).

Coronary Artery Disease and mediation analysis
A lower ß1 activity was associated with lower CAD risk (Supplementary Figure 6, Supplementary Table 9).However, contrary to the main analyses a lower α2B activity was not associated with CAD risk (Supplementary Figure 5).MR mediation analyses using the HF GWAS adjusted for CAD found no statistically significant association between a lower ß1 activity and the risk of HF: OR 0.99 (95% CI 0.97-1.01,P=0.19) (Supplementary Figure 6 and Supplementary Table 10).The deleterious effect associated with a lower α2B activity on HF risk was similar when calculated using the HF GWAS adjusted for CAD or the whole HF GWAS: OR 1.14 (95% CI 1.01-1.29,P=0.03) (Supplementary Figure 5).

Supplementary Figures
Supplementary Figure 1: Principle of Mendelian Randomization (MR) as a tool for drug target validation.The classical MR approach relies on the selection of variants in the whole genome that are associated at genome wide significance with the exposure of interest.Drug target MR follows the same principle as the classical MR method but restricts the genetic variants selection to the region of the gene encoding the drug target of interest, here one of the sympathetic nervous system genes.For each of these genes, the corresponding genetic variants are expected to affect its activity and then the biomarker trait of interest: blood pressure (BP) or heart rate, in this study.These traits are used to weight genetic variants for the drug target genes since the adrenergic receptor activities cannot be directly measured.The MR estimates calculated using these variants predict whether there is an effect of the modifiable exposure, i.e., the drug target activity, on the outcome, HF risk, left ventricular (LV) dimensions, CAD risk or HF risk adjusted for CAD risk.Supplementary Figure 2: Scheme of the localization of the genetic variants used to build the genetic instrument.To build the genetic instrument, the gene encoding regions of the nine adrenergic receptors, ADRA1A (α1A), ADRA1B (α1B), ADRA1D (α1D), ADRA2A (α2A), ADRA2B (α2B), ADRA2C (α2C), ADRB1 (ß1), ADRB2 (ß2) and ADRB3 (ß3), as well as their promoter and cis-enhancer regions were first selected.As illustrated in this figure, the promoter and cis-enhancer regions were found either in the regions between the adrenergic receptor (AR) encoding region and the encoding region of the upstream or downstream genes, and/or within the encoding region of the upstream or downstream genes.The genetic variants used to build the genetic instrument were found within these regions.AR = adrenergic receptor.
Supplementary  The references number refer to the number of the manuscript in the supplemental references list Supplementary Table 2: Sympathetic nervous system genes Single Nucleotide Polymorphisms (SNPs) associated at genome wide significance with at least one of the three exposures.

GWAS
diastolic BP systolic BP heart rate Number of SNPs reported for four sympathetic nervous system genes associated at genome wide significance (P value ≤ 5×10 -8 ) with at least one of the three biologic traits (diastolic BP, systolic BP and heart rate), with a Minor Allele Frequency (MAF) > 0.01 and clumped to a linkage disequilibrium (LD) threshold of r 2 < 0.1.

GWAS: Genome Wide Association Studies
Supplementary Table 3: SNPs identified in the four adrenergic receptor genes ADRA1A, ADRA2B, ADRB1 and ADRB2 to perform MR analyses.* These SNPs were selected using a threshold P value ≤ 1×10 -4 and a threshold clumping r² < 0.1 ** These SNPs were selected using a threshold P value ≤ 5×10 -8 and a threshold clumping r² < 0.6 (1) Corresponding literature: Literature either citing the SNP or associated SNP (LD r² >0.1).The number refer to the number of the manuscript in the supplemental references list Supplementary Table 4 Individual effect estimates of each SNP selected for the MR analyses showing the effect of α1A (ADRA1A), α2B (ADRA2B), β1 (ADRB1) and β2 (ADRB2) activity on left ventricular dimensions (LVESV, LVEDV, LV mass and LVEF).The selected genetic instruments weighted by diastolic BP (1 mm Hg decrease) were used as proxy for the activity of α1A, α2B, β1 and β2.The effect size is reported in mL for left ventricular end-systolic volume (LVESV) and left ventricular end-diastolic volume (LVEDV), in g for left ventricular (LV) mass and in percentage for left ventricular ejection fraction (LVEF).

Figure 3 :
Number of SNPs obtained after performing the different steps of the MR study for the genes (a) ADRA1A (α1A), (b) ADRA2B (α2B), (c) ADRB1 (ß1) and (d) ADRB2 (ß2).The steps 1, 2, 3, and 4 refer to the steps described in Figure1.MR: Mendelian Randomization; SNP: Single Nucleotide PolymorphismSupplementary Figure4: Individual effect estimates of each SNP selected for the MR analyses showing the effect of α1A (a), β1 (b), α2B (c) and β2 (d) activity on heart failure risk.The selected genetic instruments weighted by diastolic BP (1 mm Hg decrease) were used as proxy for the activity of each adrenergic receptor.BP: Blood Pressure; MR: Mendelian Randomization; SNP: Single Nucleotide Polymorphism.Mendelian Randomization estimates showing the effect of ADRA1A and ADRA2B adrenergic receptors activity on Heart Failure risk using SNP selected with a threshold P value ≤ 5×10-8 and a threshold clumping r² < 0.1.The selected genetic instruments weighted by diastolic Blood Pressure (1 mm Hg decrease) were used as proxy for the activity of each receptor.protective effect deleterious effect Supplementary Figure 8: MR estimates showing the effect of β1 and β2 activity on left ventricular dimensions (LVESV, LVEDV, LV mass and LVEF).The selected genetic instruments weighted by systolic BP (1 mm Hg decrease) were used as proxy for the activity of β1 and β2.The effect size is reported in mL for left ventricular end-systolic volume (LVESV) and left ventricular end-diastolic volume (LVEDV), in g for left ventricular (LV) mass and in percentage for left ventricular ejection fraction (LVEF).a: MR results using LVESV as outcome.b: MR results using LVEDV as outcome.c: MR results using LV mass as outcome.d: MR results using LVEF as outcome.BP: Blood Pressure; MR: Mendelian Randomization.MR estimates showing the effect of α2B activity on left ventricular dimensions (LVESV, LVEDV, LV mass and LVEF).The selected genetic instrument weighted by heart rate (1 beat per minute decrease) was used as proxy for the activity of α2B.The MR estimates were calculated with the Wald estimator method using the SNP rs749459.The effect size is reported in mL for left ventricular endsystolic volume (LVESV) and left ventricular end-diastolic volume (LVEDV), in g for left ventricular (LV) mass and in percentage for left ventricular ejection fraction (LVEF).a: MR result using LVESV as outcome.b: MR result using LVEDV as outcome.c: MR result using LV mass as outcome.d: MR result using LVEF as outcome.MR: Mendelian Randomization; SNP: Single Nucleotide Polymorphism.
The SNPs reported in this table are the genetic variants that were used to perform the different MR analyses.Localization is reported for the corresponding gene indicated in the "gene" column.The Outcome GWAS column reports the MR analyses outcomes for which the SNPs were used: HF: heart failure, LVESV: left ventricular end-systolic volume, LVEDV: left ventricular end-diastolic volume, LV mass: left ventricular mass, LVEF: left ventricular ejection fraction, CAD: coronary artery disease, HFcondCAD: heart failure adjusted for coronary artery disease.