Introduction: The role of ARRB2 in cardiovascular disease has recently gained increasing attention. However, the association between ARRB2 polymorphisms and heart failure (HF) has not yet been investigated. Methods: A total of 2,386 hospitalized patients with chronic HF were enrolled as the first cohort and followed up for a mean period of 20.2 months. Meanwhile, ethnically and geographically matched 3,000 individuals without evidence of HF were included as healthy controls. We genotyped the common variant in ARRB2 gene to identify the association between variant and HF. A replicated independent cohort enrolling 837 patients with chronic HF was applied to validate the observed association. A series of function analyses were conducted to illuminate the underlying mechanism. Results: We identified a common variant rs75428611 associated with the prognosis of HF in two-stage population: adjusted p = 0.001, hazard ratio (HR) = 1.31 (1.11–1.54) in additive model and adjusted p = 0.001, HR = 1.39 (1.14–1.69) in dominant model in first-stage population; adjusted p = 0.04, HR = 1.41 (1.02–1.95) in additive model and adjusted p = 0.03, HR = 1.51 (1.03–2.20) in dominant model in replicated stage. However, rs75428611 did not significantly associate with the risk of HF. Functional analysis indicated that rs75428611‐G allele increased the promoter activity and the mRNA expression level of ARRB2 by facilitating transcription factor SRF binding but not the A allele. Conclusions: Our findings demonstrated that rs75428611 in promoter of ARRB2 was associated with the risk of HF mortality. It is a promising potential treatment target for HF.

Heart failure (HF) is the serious end-stage clinical syndrome caused by various factors including coronary heart disease, hypertension, valvular heart disease, and idiopathic dilated cardiomyopathy [1]. Although significant advances have been made in the treatment of chronic HF, the mortality and rehospitalization rates remain high [2]. Aging population and rising prevalence of poorly controlled risk factors (i.e., hypertension, diabetes, and obesity) have increased the incidence of HF, which brought great financial burden to society [3]. Better understanding of the genetic basis of HF may improve the prevention, diagnosis, and treatment of HF [4].

ARRB2 was ubiquitously expressed intracellular protein and reported involved in the regulation of many important physiological function [5‒10]. As an endocytic adapter, ARRB2 could mediate the desensitization, internalization, and trafficking of G-protein-coupled receptors (GPCRs) [11]. In addition, substantial evidence has revealed that ARRB2 could act as scaffold proteins linking activated GPCRs to downstream signaling events in a G-protein-independent manner, namely biased GPCR signaling [12, 13]. Over the past decade, it has been well accepted that ARRB2 is cardioprotective in many cardiovascular diseases [11, 14‒16]. Kim et al. [17] demonstrated that ARRB2-biased AT1R stimulation could protect the heart from ischemia-reperfusion injury or mechanical stretch. ARRB2-mediated β1-adrenergic receptor transactivation of the EGFR has also been demonstrated as cardioprotective [16]. Besides, cardiac β-arrestin2 (ARRB2) signaling was indispensable for carvedilol-mediated improvement of cardiac remodeling induced by high-fructose/high-fat diet in mice [14]. While the results from Wang et al. were in opposition to the previous notion that Arrb2 is required for cardioprotection induced by biased ligand and the activation of some other receptors, they found that the expression of arrb2 was upregulated in cardiac I/R injury, and ARRB2 induced cardiac ischemia-reperfusion injury via inhibiting GPCR-independent cell survival signaling [7]. These opposite conclusions revealed the two-faced function of ARRB2 in the heart, which could be attributed to GPCR-dependent and -independent ARRB2 signaling pathways and different upstream stimuli. Based on the double role of ARRB2 in cardiovascular disease, we speculated that there may exist functional variants in ARRB2 gene that modify the prognosis of HF.

In this study, we identified 4 common variants in the promoter and exon of ARRB2 referring to the Chinese data of the 1000 Genomes. A total of 3,223 HF (two-stage population) patients and 3,000 healthy controls were genotyped to investigate the association between variants and the occurrence and prognosis of HF. A series of functional assays were conducted to illuminate the underlying mechanism.

Study Population

This study conformed to the principles of the Declaration of Helsinki and was approved by Review Board of Suining Central Hospital; approval number is 20200009. All patients have signed the informed consents. In the first stage, we recruited a total of 2,386 HF patients between March 2011 and June 2016 in Cardiology Division of Suining Central Hospital in Sichuan. Ethnically and geographically matched 3,000 individuals without evidence of HF, normal ventricular function assessed by echocardiography, and no evidence of coronary artery disease as determined either by a negative treadmill exercise test or by maximal coronary stenoses of 20% on coronary angiography were included as healthy controls. In the replicated stage, 837 HF patients were recruited from January 2016 to October 2018. Patients diagnosed with chronic HF were based on medical history, physical examination, and relevant investigations. The detailed clinical characteristics of individuals are listed in Table 1. Details on inclusion and exclusion criteria of HF and definition of risk factors have been described in online supplementary file 1 (for all online suppl. material, see https://doi.org/10.1159/000530827).

Table 1.

Baseline characteristics of the study population

CharacteristicsHealthy populationHF population
controls (n = 3,000)first stage (n = 2,386)replicated stage (n = 837)
Men, % 46 67 65 
Age, years 59.93±10.0 59.3±14.5* 59.4±14.4 
Glucose, mmol/L 5.01±0.52 6.67±3.35* 6.75±3.40 
TC, mmol/L 4.93±0.99 3.98±1.77* 3.91±1.32 
TG, mmol/L 1.46±0.98 1.42±1.14 1.37±0.93 
HDL, mmol/L 1.44±0.35 1.00±0.0.52* 0.98±0.0.42 
LDL-C, mmol/L 2.79±0.81 2.39±0.92* 2.37±0.96 
Hypertension, % 67 76 79 
Diabetes, % 32 30 
Hyperlipidemia,% 20 25 20 
Smoking status, % 39 35 
β-blocker use, % 92 93 
CharacteristicsHealthy populationHF population
controls (n = 3,000)first stage (n = 2,386)replicated stage (n = 837)
Men, % 46 67 65 
Age, years 59.93±10.0 59.3±14.5* 59.4±14.4 
Glucose, mmol/L 5.01±0.52 6.67±3.35* 6.75±3.40 
TC, mmol/L 4.93±0.99 3.98±1.77* 3.91±1.32 
TG, mmol/L 1.46±0.98 1.42±1.14 1.37±0.93 
HDL, mmol/L 1.44±0.35 1.00±0.0.52* 0.98±0.0.42 
LDL-C, mmol/L 2.79±0.81 2.39±0.92* 2.37±0.96 
Hypertension, % 67 76 79 
Diabetes, % 32 30 
Hyperlipidemia,% 20 25 20 
Smoking status, % 39 35 
β-blocker use, % 92 93 

Data are expressed as means ± SD or percentages.

TC, total cholesterol; TG, triglyceride; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol.

*p < 0.05 first stage versus controls.

End Point Assessment

The two-stage HF population was followed up for an average period of 20.2 and 18.3 months, respectively. A standard HF questionnaire was conducted during regular outpatient clinics or by telephone contact. The end points were defined as cardiovascular deaths or cardiac transplantation.

DNA Extraction, SNP Selection, and Genotyping

We extracted genomic DNA from peripheral blood leukocytes using Tiangen commercially available kit (Tiangen, Beijing, China) according to the protocol. The DNA was stored at −80°C for further use. Referring to the Chinese data of the 1000 Genomes, we selected common genetic variants with minor allele frequency (MAF) >0.05 in the functional region of ARRB2 for further genotyping. The probe for genotyping came from ABI with the following assay ID: rs1045280 (C___8718195_20, 4351379). The common variant in ARRB2 was genotyped using the TaqMan assay on the TaqMan 7900HT Sequence Detection System (Applied Biosystems, Foster City, CA, USA) with the following conditions: 10 min at 95°C (enzyme activation) followed by 45 cycles at 95°C for 15 s and 60°C for 1 min (annealing/extension). An end point read was performed for allelic discrimination after amplification.

Plasmids Construction, Cell Culture, Transient Transfection, and Luciferase Activity Assays

To determine the causal variant, we constructed the reporter plasmids including their wild and mutant type using PGL3‐basic. Meanwhile, the coding region of ARRB2 gene and the predicted transcriptional factors SRF, HAND2, HOXA5, and ZNF263 were cloned into pcDNA3.1, using human cDNA. Related primer and restriction enzyme cutting sites are showed below: forward primer with MluI 5′ cgA​CGC​GTG​AAA​GGA​ACT​CCC​ACA​GTG​CAG​T 3′, reverse primer with HindIII 5′ CCC​AAG​CTT​GCC​TTC​CAG​GGG​ATC​TAA​GTC 3′ for reporter plasmids construction, forward primer with HindIII 5′ CCC​AAG​CTT​ATG​ATC​ACC​AGT​GAG​ACC​GGC 3′, reverse primer with EcoRI 5′ CCG​GAA​TTC​TCA​TTC​ACT​CTT​GGT​GCT​GTG​G 3′ for SRF expression vector construction, forward primer with HindIII 5′ CCC​AAG​CTT​ATG​AGT​CTG​GTA​GGT​GGT​TTT​CC 3′, reverse primer with EcoRI 5′ CCG​GAA​TTC​TCA​CTG​CTT​GAG​CTC​CAG​GG 3′ for HAND2 expression vector construction, forward primer with HindIII 5′ CCC​AAG​CTT​ATG​AGC​TCT​TAT​TTT​GTA​AAC​TCA​TTT​T 3′, reverse primer with EcoRI 5′ CCGGAATTC CTC​AGA​TAC​TCA​GGG​ACG​GAA​GG 3′ for HOXA5 expression vector construction, forward primer with KpnI 5′ GGG​GTA​CCA​GGC​GCT​CTG​GAG​ACC​TGA​C 3′, reverse primer with NotI 5′ TTG​CGG​CCG​CTC​ACA​CTG​TGT​GAG​TTC​TCT​GAT​GAC 3′ for ZNF263 expression vector construction, and forward primer with HindIII 5′ CCC​AAG​CTT​ATG​GGG​GAG​AAA​CCC​GGG 3′, reverse primer with EcoRI 5′ CCG​GAA​TTC​CTA​GCA​GAG​TTG​ATC​ATC​ATA​GTC​GTC 3′ for ARRB2 expression vector construction. Cell culture and transient transfection procedures have been described in online supplementary file 1. Cells were harvested 48 h after transfection using the Passive Lysis Buffer (SIRIUS, Pforzheim, Germany). The data of luciferase expression levels were adjusted with reference to Renilla luciferase activity and relative to the average values of mutant type for corresponding variants. Each reporter was performed six independent experiments to avoid potential experimental errors.

Western Blotting for ARRB2

AC16 cells were grown in 6‐well plates and transfected with ARRB2 constructs or empty vector pcDNA3.1 using Lipofectamine™ 2000 transfection reagent (Invitrogen) according to the manufacturer’s instructions. Cells were harvested 48 h after transfection and lysed with lysis solution (50 mmol/L Tris‐Cl, pH 8.0; 150 mmol/L NaCl; 0.02% sodium azide; 0.1% SDS; 1 μg/mL aprotinin; 1% Nonidet P‐40; and 0.5% sodium deoxycholate) containing protease inhibitors (100 μg/mL phenylmethylsulfonyl fluoride, 2 μg/mL aprotinin, 2 μg/mL leupeptin). After centrifuging at 12,000 g for 20 min at 4°C, supernatant was collected, and protein concentrations were measured using the BCA protein assay reagent kit (Boster, China). Lysates were resolved by 10% SDS‐PAGE and transferred to polyvinylidene difluoride membranes. After blocking with 5% nonfat milk, blots were probed with ARRB2 antibody (Cell Signaling Technology, #3857) and incubated with a peroxidase‐conjugated secondary antibody. Bands were visualized by enhanced chemiluminescence reagents (Pierce Chemical, Rockford, IL, USA) and quantified by densitometry.

Transcription Assays of the ARRB2 Gene

Total of 204 samples of peripheral blood lymphocytes from participants undergoing coronary angiography were collected for mRNA assessment. Details about RNA isolation, mRNA transcription, and PCR performed in this study were described in online supplementary materials. ARRB2 and ACTB mRNA were measured using absolute quantification methods with each sample in triplicate. Related primer sequences are listed in online supplementary Table S1. Detailed characteristics of individuals are shown in online supplementary Table S2. Expression of ARRB2 relative to ACTB was compared among individuals with GG genotype, GA genotype, and AA genotype. This study protocol was reviewed and approved by the Review Board of Suining Central Hospital; approval number is 20200009. All patients have signed the informed consents.

Chromatin Immunoprecipitation Assay

Chromatin immunoprecipitation assays were carried out using commercially available assay kits. Detailed descriptions are available in online supplementary materials.

Statistical Analysis

Statistical analyses were performed with SPSS version 13.0 (SPSS, Inc., Chicago, IL, USA) for Windows (Microsoft Corp., Redmond, WA, USA). The polymorphisms were tested for Hardy‐Weinberg equilibrium among HF patients and the controls using χ2 test. Linkage disequilibrium was calculated using Haploview version 4.1. Multivariate logistic regression analyses based on different genetic models were used to test the association between variant and HF. In addition, we conducted Cox proportional hazards regression model to assess the association of variants with prognosis of HF, with or without adjustment for traditional risk factors including sex, age, hypertension, diabetes, hyperlipidemia, smoking state, and β‐blocker use. Data are expressed as means ± standard deviation (SD). Two independent samples were compared using Student’s t test. p < 0.05 was considered statistically significant.

Clinical Characteristics of Participants

As shown in Table 1, there were a total of 3,000 healthy controls and 3,223 HF patients (2,386 in the first stage, 837 in replicated stage) enrolled in our study. The proportion of diabetes, hypertension, dyslipidemia, and smoking state are significantly higher in HF population than in controls. While the level of total cholesterol and LDL in HF patients seems slightly lower than in controls, which can be attributed to lipid-lowering therapy.

Correlation between Variant in ARRB2 Gene and Risk of HF

Referring to the Chinese data of the 1000 Genomes, we identified 4 common variants with MAF >0.05 in the promoter and coding region of ARRB2 (Table 2). Notably, these 4 variants were in strong linkage disequilibrium with each other (r2 > 0.9). Finally, we selected rs1045280 as the tagged SNP for further genotyping. Hardy-Weinberg equilibrium analysis was conducted using the χ2 test with one degree of freedom, and the results showed that all these SNPs were in Hardy-Weinberg equilibrium (p > 0.05).

Table 2.

Characteristics of ARRB2 variants referring to the Chinese data of the 1000 Genomes

Gene positionadbSNP IDbGene regionMaj>mincMAF
chr17:4612866 rs8071086 Promoter A>G 0.166  
chr17:4612874 rs8069382 Promoter G>T 0.166  
chr17:4613232 rs75428611 Promoter G>A 0.163  
chr17:4622638 rs1045280 Synonymous T>C 0.166  
Gene positionadbSNP IDbGene regionMaj>mincMAF
chr17:4612866 rs8071086 Promoter A>G 0.166  
chr17:4612874 rs8069382 Promoter G>T 0.166  
chr17:4613232 rs75428611 Promoter G>A 0.163  
chr17:4622638 rs1045280 Synonymous T>C 0.166  

MAF, minor allele frequency; SNPs, single nucleotide polymorphisms.

aBase pair position is based on NCBI GRCH37.

bPolymorphism are numbered relative to transcription start site.

cWith major allele given first, followed by minor allele.

We investigated the association of rs1045280 with risk of HF using multivariate logistic regression analysis. As showed in Table 3, no significant difference was observed in genotype of rs1045280 between cases and controls in additive, dominant, and recessive models, which indicated that this SNP is not associated with the risk of HF.

Table 3.

Association between ARRB2 variant and HF

SNP rs IDFunctionPopulationMAFGenotypeModelCrude ORs (95% CI)Adjusted p valueAdjusted ORs (95% CI)
    TT TC CC Additive 1.00 (0.90–1.11) 0.63 0.97 (0.87–1.09) 
rs1045280 Synonymous Control 0.16 2119 791 90 Dominant 1.00 (0.89–1.13) 0.45 0.95 (0.83–1.08) 
T>C HF 0.16 1686 628 72 Recessive 0.99 (0.73–1.36) 0.59 0.91 (0.64–1.28) 
SNP rs IDFunctionPopulationMAFGenotypeModelCrude ORs (95% CI)Adjusted p valueAdjusted ORs (95% CI)
    TT TC CC Additive 1.00 (0.90–1.11) 0.63 0.97 (0.87–1.09) 
rs1045280 Synonymous Control 0.16 2119 791 90 Dominant 1.00 (0.89–1.13) 0.45 0.95 (0.83–1.08) 
T>C HF 0.16 1686 628 72 Recessive 0.99 (0.73–1.36) 0.59 0.91 (0.64–1.28) 

OR, odds ratio.

ORs and 95% CIs were obtained by logistic regression, with and without adjustment for sex, age, hypertension, diabetes, hyperlipidemia, and smoking status.

Association of rs1045280 in ARRB2 Gene with the Prognosis of HF

In the first stage, we performed genotyping of 2,386 HF patients with an average follow-up time of 20.2 months, during which 419 cardiovascular deaths or cardiac transplantation occurred. Cox proportional hazards regression model was conducted to assess the effect of rs1045280 on the prognosis of HF. The results revealed that rs1045280 was significantly associated with the mortality risk of HF in additive and dominant model both with or without adjustment for conventional risk factors (including sex, age, hypertension, diabetes, hyperlipidemia, smoking state) and β‐blocker use (shown in Fig. 1a–c; Table 4). The cardiovascular death or cardiac transplantation had occurred in 263 patients (15.6%) in TT genotype group, 137 patients (21.8%) in TC genotype group, and 19 patients (26.4%) in CC genotype group. Cox proportional hazards model analysis showed that the C allele of rs1045280 is significantly associated with an increased risk of cardiovascular death and cardiac transplantation (p = 0.00008, hazard ratio [HR] = 1.39, 95% confidence interval [CI] = 1.18–1.63). The statistical significance in multivariate analysis remained after adjustments for sex, age, hypertension, diabetes, hyperlipidemia, smoking state, and β‐blocker use (adjusted p = 0.001, HR = 1.31, 95% CI = 1.11–1.54).

Fig. 1.

Effects of rs1045280 on the prognosis of HF patients. Cox proportional hazards model analysis showed the association of rs1045280 genotypes with cardiovascular deaths or cardiac transplantation in the first-stage population (a–c), replicated-stage population (d–f), and combined population (g–i). a, d, g In additive model, rs1045280 is associated with the prognosis of HF. b, e, h In dominant model, patients carrying rs1045280-TT genotypes showed reduced risk of cardiovascular deaths or cardiac transplantation when compared with patients carrying TC or CC genotype. c, f, i In recessive model, the association between rs1045280 and prognosis of HF showed no statistical significance in the first-stage and replicated-stage populations (c, f); the combined p value is statistically significant but disappear after adjustment for conventional risk factors (i).

Fig. 1.

Effects of rs1045280 on the prognosis of HF patients. Cox proportional hazards model analysis showed the association of rs1045280 genotypes with cardiovascular deaths or cardiac transplantation in the first-stage population (a–c), replicated-stage population (d–f), and combined population (g–i). a, d, g In additive model, rs1045280 is associated with the prognosis of HF. b, e, h In dominant model, patients carrying rs1045280-TT genotypes showed reduced risk of cardiovascular deaths or cardiac transplantation when compared with patients carrying TC or CC genotype. c, f, i In recessive model, the association between rs1045280 and prognosis of HF showed no statistical significance in the first-stage and replicated-stage populations (c, f); the combined p value is statistically significant but disappear after adjustment for conventional risk factors (i).

Close modal
Table 4.

Association of rs1045280 with HF prognosis in two-stage population

Additive modelDominant modelRecessive model
p valueAdjusted p valueHR (95% CI)p valueAdjusted p valueHR (95% CI)p valueAdjusted p valueHR (95% CI)
First stage 0.00008 0.001 1.31 (1.11–1.54) 0.0001 0.001 1.39 (1.14–1.69) 0.063 0.15 1.4 (0.88–2.22) 
Replicated stage 0.011 0.04 1.41 (1.02–1.95) 0.01 0.03 1.51 (1.03–2.20) 0.36 0.5 1.42 (0.51–3.93) 
Combined 0.000002 0.00006 1.35 (1.17–1.56) 0.000003 0.00006 1.43 (1.20–1.71) 0.036 0.08 1.45 (0.96–2.21) 
Additive modelDominant modelRecessive model
p valueAdjusted p valueHR (95% CI)p valueAdjusted p valueHR (95% CI)p valueAdjusted p valueHR (95% CI)
First stage 0.00008 0.001 1.31 (1.11–1.54) 0.0001 0.001 1.39 (1.14–1.69) 0.063 0.15 1.4 (0.88–2.22) 
Replicated stage 0.011 0.04 1.41 (1.02–1.95) 0.01 0.03 1.51 (1.03–2.20) 0.36 0.5 1.42 (0.51–3.93) 
Combined 0.000002 0.00006 1.35 (1.17–1.56) 0.000003 0.00006 1.43 (1.20–1.71) 0.036 0.08 1.45 (0.96–2.21) 

HR and 95% CI were obtained using Cox regression, with and without adjustment for sex, age, hypertension, diabetes, hyperlipidemia, and smoking status.

SNP, single‐nucleotide polymorphism.

To further substantiate our findings from the first stage, we subsequently carried out an independent replication study by genotyping another cohort of 837 chronic HF patients to validate the observed association. The results showed that patients carrying TC+CC genotype displayed increased risk of cardiovascular death and cardiac transplantation when compared with patients with TT genotype (adjusted p = 0.03, HR = 1.51, 95% CI = 1.03–2.20) (shown in Fig. 1d–f; Table 4), which is consistent with the results of the first stage. The cardiovascular death or cardiac transplantation had occurred in 75 patients (15.6%) in TT genotype group, 48 patients (21.8%) in TC genotype group, and 5 patients (26.4%) in CC genotype group. The statistical significance remained after adjustment for conventional risk factors. In the combined results from two-stage population, rs1045280-C allele is significantly associated with an increased risk of cardiovascular death and cardiac transplantation (adjusted p = 0.00006, HR = 1.35, 95% CI = 1.17–1.56) (shown in Fig. 1g–i; Table 4).

Finally, we conducted stratification analysis to separately evaluate the association of rs1045280 with cardiovascular deaths or cardiac transplantation. In the first stage, a total of 419 cardiovascular deaths or cardiac transplantation occurred, including 385 cardiovascular deaths and 34 cardiac transplantations. In the replicated stage, a total of 128 cardiovascular deaths or cardiac transplantation occurred including 119 cardiovascular deaths and 9 cardiac transplantations. As shown in Table 5, rs1045280 was significantly associated with the risk of cardiovascular deaths in additive and dominant models, while the association disappeared with the risk of cardiac transplantation for two-stage population.

Table 5.

Association of rs1045280 with each outcome measure in two-stage population

First stageReplicated stage
cardiovascular deathscardiac transplantationcardiovascular deathscardiac transplantation
p valueHR (95% CI)p valueHR (95% CI)p valueHR (95% CI)p valueHR (95% CI)
Additive model 0.0001 1.41 (1.19–1.67) 0.69 1.14 (0.61–2.10) 0.009 1.5 (1.11–2.04) 0.95 1.05 (0.3–3.67) 
Dominant model 0.0001 1.52 (1.23–1.85) 0.93 0.97 (0.46–2.03) 0.009 1.61 (1.12–2.33) 0.82 0.85 (0.21–3.42) 
Recessive model 0.1 1.49 (0.92–2.44) 0.31 0.48 (0.12–2.0) 0.29 0.62 (0.25–1.51) 0.76 21 (6 × 10−8–8 × 1010
First stageReplicated stage
cardiovascular deathscardiac transplantationcardiovascular deathscardiac transplantation
p valueHR (95% CI)p valueHR (95% CI)p valueHR (95% CI)p valueHR (95% CI)
Additive model 0.0001 1.41 (1.19–1.67) 0.69 1.14 (0.61–2.10) 0.009 1.5 (1.11–2.04) 0.95 1.05 (0.3–3.67) 
Dominant model 0.0001 1.52 (1.23–1.85) 0.93 0.97 (0.46–2.03) 0.009 1.61 (1.12–2.33) 0.82 0.85 (0.21–3.42) 
Recessive model 0.1 1.49 (0.92–2.44) 0.31 0.48 (0.12–2.0) 0.29 0.62 (0.25–1.51) 0.76 21 (6 × 10−8–8 × 1010

HR and 95% CI were obtained using Cox regression.

Functional Analysis of the Causal SNP

Considering that fact that rs8071086, rs8069382, rs75428611, and rs1045280 were in linkage disequilibrium with each other, we performed fluorescent reporter gene assay and Western blot to identify the causal variant in AC16 cell. As shown in Figure 2a, the luciferase activity of rs75428611-G was significantly higher than that of rs75428611-A allele. However, we did not observe any effect on luciferase activity in rs8071086 and rs8069382 luciferase assays in AC16 cells. Similarly, we assessed the impact of rs1045280 on ARRB2 gene expression derived from full-length cDNA expression plasmids engineered to harbor either wild-type or mutant-type allele. No difference in protein expression level was observed for wild-type and mutant-type constructs of rs1045280 after transfected into AC16 (shown in Fig. 2b). Besides, we also assessed the functional possibility of 9 variants in the intron region, all of which were in strong linkage disequilibrium with rs75428611 (online suppl. Table S3). However, we did not observe any effect on luciferase activity in rs2036656, rs16954146, rs11324202, rs35592645, rs3786047, rs11868227, rs35376679, rs2271167, and rs3841623 luciferase assays in AC16 cells (shown in Fig. 3).

Fig. 2.

Functional analysis. a Firefly luciferase assays were performed in AC16, and rs75428611-G allele displayed higher luciferase activity than A allele; no effect on luciferase activity in rs8071086 and rs8069382 luciferase assays was observed. b No difference in protein expression level was observed for wild-type and mutant-type constructs of rs1045280. c SRF could significantly increase the transcriptional activity of rs75428611-G allele but not -A allele. d Compared with AG and AA genotypes, samples carrying GG genotype showed higher mRNA level of ARRB2; the difference between AG and AA is also statistically significant. *p < 0.05, NS, not significant.

Fig. 2.

Functional analysis. a Firefly luciferase assays were performed in AC16, and rs75428611-G allele displayed higher luciferase activity than A allele; no effect on luciferase activity in rs8071086 and rs8069382 luciferase assays was observed. b No difference in protein expression level was observed for wild-type and mutant-type constructs of rs1045280. c SRF could significantly increase the transcriptional activity of rs75428611-G allele but not -A allele. d Compared with AG and AA genotypes, samples carrying GG genotype showed higher mRNA level of ARRB2; the difference between AG and AA is also statistically significant. *p < 0.05, NS, not significant.

Close modal
Fig. 3.

Functional analysis of 9 variants in strong LD with rs75428611. The luciferase activity between wild-type and mutant-type of these 9 variants showed no difference. NS, not significant.

Fig. 3.

Functional analysis of 9 variants in strong LD with rs75428611. The luciferase activity between wild-type and mutant-type of these 9 variants showed no difference. NS, not significant.

Close modal

Subsequently, we conducted the bioinformatics analysis with Jaspar (http://jaspar.genereg.net/) and found that rs75428611-A allele destroyed that binding sites of SRF, HAND2, HOXA5, and ZNF263 in the promoter of ARRB2. We then experimentally investigated the function of rs75428611 on these transcriptional factor-mediated gene transcription using luciferase activity assays. As shown in Figure 2c, transfection of SRF expression plasmids increased the transcription activity for rs75428611-G but not -A ARRB2 promoter construct. This indicated that the rs75428611-A allele could disrupt the binding site of SRF in the ARRB2 promoter and reduced the transcriptional activity of the gene. However, HAND2, HOXA5, and ZNF263 transfection did not differentially affect the transcription activity of rs75428611-G and A ARRB2 promoter construct (online suppl. Fig. 1).

Meanwhile, we compared ARRB2 mRNA relative expression in lymphocytes including 204 samples (141 GG genotype, 57 AG genotype, and 6 AA genotype for rs75428611) and found that samples with GG genotype displayed significantly higher ARRB2 mRNA level than samples with AG and AA genotype (shown in Fig. 2d). The difference between AG and AA genotypes also showed statistical difference.

In order to further assess the functional relevance of rs75428611, we performed DNA binding analysis in AC16 cells. Chromatin immunoprecipitation assay was conducted and the results showed that SRF binds to the region encompassing rs75428611 (shown in Fig. 4).

Fig. 4.

Chromatin immunoprecipitation with antibody against Flag‐SRF in AC16 with transfection with Flag‐SRF construct. Quantitative PCR was used to measure the immunoprecipitation of DNA sequence surrounding rs75428611, normalized to background (control condition with IgG, no antibody).

Fig. 4.

Chromatin immunoprecipitation with antibody against Flag‐SRF in AC16 with transfection with Flag‐SRF construct. Quantitative PCR was used to measure the immunoprecipitation of DNA sequence surrounding rs75428611, normalized to background (control condition with IgG, no antibody).

Close modal

HF is a complex clinical syndrome with high hospitalization and mortality risk [18, 19]. In this study, we identified a genetic variant in the promoter of ARRB2 associated with the prognosis of HF. In the first-stage population, rs75428611-A allele displayed increased risk of cardiovascular death and cardiac transplantation (adjusted p = 0.001, HR = 1.31, 95% CI = 1.11–1.54) when compared with rs75428611-G allele. Importantly, we identified the consistent results in the replicated-stage population (adjusted p = 0.04, HR = 1.41, 95% CI = 1.02–1.95). Subsequently, we combined two HF populations for prognosis analysis, and the results indicated that rs75428611-A allele is associated with poorer prognosis of HF in total population (adjusted p = 0.00006, HR = 1.35, 95% CI = 1.17–1.56).

ARRB2 was universally expressed and originally identified for its negative regulator role of G-protein signaling via sterically blocking GPCR coupling to G-proteins in a process called “desensitization” [20, 21]. In addition, ARRB2 is also implicated in the recruitment of GPCRs into intracellular compartments, a process broadly known as receptor internalization [22]. Importantly, ARRB2 has been identified to function as scaffold proteins linking GPCR activation to several downstream effectors in G-protein-independent signaling [11]. Under the condition of HF, ARRB2 could relieve the deleterious response to excessive sympathetic stimulation through attenuating β-adrenergic and angiotensin II receptor signaling while simultaneously preserving cardiac structure and function by transactivating cardioprotective signaling cascades in response to injury [23]. For example, Noma et al. [16] have demonstrated the protective role of ARRB2-mediated β1-adrenergic receptor transactivation of the EGFR [24]. ARRB2 was identified as indispensable for EGFR transactivation by urotensin II receptor and conferred cardioprotection [15]. Notably, ARRB2 delays inflammatory responses by interfering with macrophage recruitment to the infarcted area, which could protect mice from MI-induced myocardial damage and reduce the mortality [25]. However, no study has investigated the association of variants in ARRB2 with the risk or prognosis of HF. In the present study, we conducted association analysis and found that patients carrying rs75428611-G displayed better prognosis than those carrying rs75428611-A allele. Functional assays revealed that the reporter activity of rs75428611-G was significantly higher than that of A allele. In addition, the mRNA level of ARRB2 with rs75428611-G was also higher than A allele using human lymphocytes. Our results demonstrated the fact that ARRB2 expression level was positively correlated to prognosis of HF patients, which favors the protective role of ARRB2 in HF and is in line with previous reports [23, 24].

However, recent study conducted by Wang et al. [7] has found that ARRB2 could mediate cardiac ischemia-reperfusion injury by inhibiting GPCR-independent cell survival signaling. They demonstrated that ARRB2 induced cardiomyocyte death by negatively blocking activation of PI3K-Akt-GSK3β cell survival signaling pathway. This contradiction reflected different, or even opposite function of GPCR-dependent and -independent ARRB2 signaling pathways. In fact, HF is characterized by sympathetic hyperactivity, which includes hyperactivity of β-adrenergic and angiotensin II receptor signaling and is demonstrated to be a “lethality factor” [26, 27]. ARRB2 is capable of limiting the deleterious response to excessive sympathetic stimulation through desensitization, internalization, and degradation of GPCRs. In addition, function as a scaffold protein, ARRB2 could link activated GPCRs to downstream signaling events in a G-protein-independent manner, which favor protective for HF. In brief, ARRB2 is likely to play a protective role in the condition of HF, which is also demonstrated by our study.

However, there exist some limitations to our study. First, we only focused on functional variant in ARRB2 gene. Other functional variants in nearby genes, which were in strict linkage disequilibrium with rs75428611 may also participate in the prognosis of HF, which needs further investigation in other studies. Additionally, rs75428611 is almost exclusive to East Asian population (MAF = 0.17), while the three other SNPs are common in other 1000 Genome populations. So the association of these variants with the prognosis of HF may be different in the perspective of replications and/or trans-ethnic studies, which need attention. Finally, we demonstrated SRF as the causal transcriptional factor for rs75428611-mediated differential expression of NID1 gene. In fact, SRF cooperates with various cofactors to manage the specification of diverse gene expression [28]. In the heart, SRF interacts with cofactors including NKX2.5 and GATA4 to regulate cardiac-specified gene activity [29]. In this study, we just focused on SRF and ignored the involvement of its cofactors. GATA4, NKX2.5, and other cofactors might also participate in SRF-mediated expression of NID gene, which need further investigation in the future.

In conclusion, our results demonstrated that rs75428611 in promoter region of ARRB2 was associated with the prognosis of HF. The rs75428611‐A allele may destroy the binding site of transcriptional factor SRF, thus reducing the transcription of ARRB2 and further leading to poorer prognosis of HF. Further prevention and treatment strategies through targeting ARRB2 are an attractive way to improve the prognosis of HF.

The authors would like to thank the staff of the Division of Cardiology and follow‐up nurses for their help in recruitment and follow‐up of HF patients.

This study protocol was reviewed and approved by Review Board of Suining Central Hospital; approval number is 20200009. This study conformed to the principles outlined in the Declaration of Helsinki and was approved by Review Board of Suining Central Hospital. All patients have signed the informed consents.

The authors have no conflicts of interest to declare.

The authors received no funding for this study.

Hongqiang Ren and Li Zhao developed the study concept and design, interpreted the data, and drafted the manuscript. Yijun Liu performed the research. Zhen Tan performed the data analysis. Guiquan Luo and Mei Zhang helped with specimen collection. Shuang Li and Tingwei Tang supervised the design of the study and revised the manuscript.

Data are not publicly available due to ethical reasons. Further inquiries can be directed to the corresponding author.

1.
Sun
Q
,
Jiang
S
,
Wang
X
,
Zhang
J
,
Li
Y
,
Tian
J
.
A prediction model for major adverse cardiovascular events in patients with heart failure based on high-throughput echocardiographic data
.
Front Cardiovasc Med
.
2022
;
9
:
1022658
.
2.
Harjula
A
,
Baldwin
JC
,
Starnes
VA
,
Stinson
EB
,
Oyer
PE
,
Jamieson
SW
.
Proper donor selection for heart-lung transplantation
.
J Thorac Cardiovasc Surg
.
1987
;
94
(
6
):
874
80
.
3.
Rossignol
P
,
Hernandez
AF
,
Solomon
SD
,
Zannad
F
.
Heart failure drug treatment
.
Lancet
.
2019
;
393
(
10175
):
1034
44
.
4.
Feldman
AM
,
Kontos
CD
,
McClung
JM
,
Gerhard
GS
,
Khalili
K
,
Cheung
JY
.
Precision medicine for heart failure: lessons from oncology
.
Circ Heart Fail
.
2017
;
10
(
6
):
e004202
.
5.
Coureuil
M
,
Lécuyer
H
,
Scott
MGH
,
Boularan
C
,
Enslen
H
,
Soyer
M
.
Meningococcus Hijacks a β2-adrenoceptor/β-Arrestin pathway to cross brain microvasculature endothelium
.
Cell
.
2010
;
143
(
7
):
1149
60
.
6.
Luan
B
,
Zhao
J
,
Wu
H
,
Duan
B
,
Shu
G
,
Wang
X
.
Deficiency of a beta-arrestin-2 signal complex contributes to insulin resistance
.
Nature
.
2009
;
457
(
7233
):
1146
9
.
7.
Wang
Y
,
Jin
L
,
Song
Y
,
Zhang
M
,
Shan
D
,
Liu
Y
.
β-arrestin 2 mediates cardiac ischemia-reperfusion injury via inhibiting GPCR-independent cell survival signalling
.
Cardiovasc Res
.
2017
;
113
(
13
):
1615
26
.
8.
Zhu
L
,
Rossi
M
,
Cui
Y
,
Lee
RJ
,
Sakamoto
W
,
Perry
NA
.
Hepatic β-arrestin 2 is essential for maintaining euglycemia
.
J Clin Invest
.
2017
;
127
(
8
):
2941
5
.
9.
Zhu
L
,
Almaça
J
,
Dadi
PK
,
Hong
H
,
Sakamoto
W
,
Rossi
M
.
β-arrestin-2 is an essential regulator of pancreatic β-cell function under physiological and pathophysiological conditions
.
Nat Commun
.
2017
;
8
:
14295
.
10.
Ravier
MA
,
Leduc
M
,
Richard
J
,
Linck
N
,
Varrault
A
,
Pirot
N
.
β-Arrestin2 plays a key role in the modulation of the pancreatic beta cell mass in mice
.
Diabetologia
.
2014
;
57
(
3
):
532
41
.
11.
Noor
N
,
Patel
CB
,
Rockman
HA
.
Β-arrestin: a signaling molecule and potential therapeutic target for heart failure
.
J Mol Cell Cardiol
.
2011
;
51
(
4
):
534
41
.
12.
Monasky
MM
,
Taglieri
DM
,
Henze
M
,
Warren
CM
,
Utter
MS
,
Soergel
DG
.
The β-arrestin-biased ligand TRV120023 inhibits angiotensin II-induced cardiac hypertrophy while preserving enhanced myofilament response to calcium
.
Am J Physiol Heart Circ Physiol
.
2013
305
6
H856
66
.
13.
Boerrigter
G
,
Soergel
DG
,
Violin
JD
,
Lark
MW
,
Burnett
JC
.
TRV120027, a novel β-arrestin biased ligand at the angiotensin II type I receptor, unloads the heart and maintains renal function when added to furosemide in experimental heart failure
.
Circ Heart Fail
.
2012
;
5
(
5
):
627
34
.
14.
Ibrahim
WS
,
Ibrahim
IAAE-H
,
Mahmoud
MF
,
Mahmoud
AAA
.
Carvedilol diminishes cardiac remodeling induced by high-fructose/high-fat Diet in mice via enhancing cardiac β-arrestin2 signaling
.
J Cardiovasc Pharmacol Ther
.
2020
;
25
(
4
):
354
63
.
15.
Esposito
G
,
Perrino
C
,
Cannavo
A
,
Schiattarella
GG
,
Borgia
F
,
Sannino
A
.
EGFR trans-activation by urotensin II receptor is mediated by β-arrestin recruitment and confers cardioprotection in pressure overload-induced cardiac hypertrophy
.
Basic Res Cardiol
.
2011
;
106
(
4
):
577
89
.
16.
Noma
T
,
Lemaire
A
,
Naga Prasad
SV
,
Barki-Harrington
L
,
Tilley
DG
,
Chen
J
.
Beta-arrestin-mediated beta1-adrenergic receptor transactivation of the EGFR confers cardioprotection
.
J Clin Invest
.
2007
;
117
(
9
):
2445
58
.
17.
Kim
KS
,
Abraham
D
,
Williams
B
,
Violin
JD
,
Mao
L
,
Rockman
HA
.
β-Arrestin-biased AT1R stimulation promotes cell survival during acute cardiac injury
.
Am J Physiol Heart Circ Physiol
.
2012
303
8
H1001
10
.
18.
Smadja
DM
,
Melero-Martin
JM
,
Eikenboom
J
,
Bowman
M
,
Sabatier
F
,
Randi
AM
.
Standardization of methods to quantify and culture endothelial colony-forming cells derived from peripheral blood: position paper from the International Society on Thrombosis and Haemostasis SSC
.
J Thromb Haemost
.
2019
;
17
(
7
):
1190
4
.
19.
Li
Y
,
Wang
H
,
Luo
Y
.
Improving fairness in the prediction of heart failure length of stay and mortality by integrating social determinants of health
.
Circ Heart Fail
.
2022
;
15
(
11
):
e009473
.
20.
Lefkowitz
RJ
,
Shenoy
SK
.
Transduction of receptor signals by beta-arrestins
.
Science
.
2005
;
308
(
5721
):
512
7
.
21.
Lohse
MJ
,
Benovic
JL
,
Codina
J
,
Caron
MG
,
Lefkowitz
RJ
.
beta-Arrestin: a protein that regulates beta-adrenergic receptor function
.
Science
.
1990
;
248
(
4962
):
1547
50
.
22.
Moore
CAC
,
Milano
SK
,
Benovic
JL
.
Regulation of receptor trafficking by GRKs and arrestins
.
Annu Rev Physiol
.
2007
;
69
:
451
82
.
23.
Boussi
L
,
Frishman
WH
.
β-Arrestin as a therapeutic target in heart failure
.
Cardiol Rev
.
2021
;
29
(
5
):
223
9
.
24.
Mangmool
S
,
Parichatikanond
W
,
Kurose
H
.
Therapeutic targets for treatment of heart failure: focus on GRKs and β-arrestins affecting βAR signaling
.
Front Pharmacol
.
2018
;
9
:
1336
.
25.
Watari
K
,
Nakaya
M
,
Nishida
M
,
Kim
KM
,
Kurose
H
.
β-arrestin2 in infiltrated macrophages inhibits excessive inflammation after myocardial infarction
.
PLoS One
.
2013
;
8
(
7
):
e68351
.
26.
Gronda
E
,
Dusi
V
,
D’Elia
E
,
Iacoviello
M
,
Benvenuto
E
,
Vanoli
E
.
Sympathetic activation in heart failure
.
Eur Heart J Suppl
.
2022
24
Suppl E
E4
11
.
27.
Koba
S
.
Angiotensin II, oxidative stress, and sympathetic nervous system hyperactivity in heart failure
.
Yonago Acta Med
.
2018
;
61
(
2
):
103
9
.
28.
Zheng
G
,
He
Z
,
Lu
Y
,
Zhu
Q
,
Jiang
Y
,
Chen
D
.
SRF-derived miR210 and miR30c both repress beating cardiomyocyte formation in the differentiation system of embryoid body
.
Biochem Biophys Res Commun
.
2022
;
626
:
58
65
.
29.
Xiao
S
,
Liang
R
,
Lucero
E
,
McConnell
BK
,
Chen
Z
,
Chang
J
.
STEMIN and YAP5SA synthetic modified mRNAs regenerate and repair infarcted mouse hearts
.
J Cardiovasc Aging
.
2022
;
2
(
3
):
31
.