A common variant of RIP3 promoter region is associated with poor prognosis in heart failure patients by influencing SOX17 binding

Abstract Receptor‐interacting protein kinase 3 (RIP3) is a key determinant of necroptosis and participates in ischaemia—and oxidative stress‐induced necroptosis, myocardial remodelling and heart failure (HF). In this study, we tested the hypothesis that common variants in RIP3 gene were associated with the risk and prognosis of HF in the Chinese Han population. By re‐sequencing and luciferase assays, we identified a common functional variant in the RIP3 promoter region. The rs3212247‐T allele suppressed RIP3 promoter activity by facilitating transcription factor SOX17 binding, but not the C allele. We further recruited 2961 control participants and 3194 HF patients who underwent a mean follow‐up of 19 months (6‐31 months) for this study. Rs3212247 and another missense variant rs3212254 were genotyped. Although rs3212247 did not significantly associate with increased risk of HF (odds ratio = 1.00, 95% CI = 0.92‐1.08, P = 0.91), it raised the risk for cardiovascular death and cardiac transplantation (hazard ratio = 1.47, 95% CI = 1.13‐1.91, P = 0.004). Moreover, participants carrying the rs3212247 CC genotype had higher plasma levels of RIP3 than those carrying the TT or TC genotype (p for trend = 0.02) in New York Heart Association class III HF group. No association was found between the RIP3 missense variant rs3212254 and risk or prognosis of HF after adjustment for traditional risk factors. In conclusion, genetic variant in RIP3 promoter region is associated with increased RIP3 transcription, thus contributed to the poor prognosis of HF patients. Clinical Trial Registration: https://www.clinicaltrials.gov/ct2/show/NCT03461107?term=03461107&cond=Heart+Failure&cntry=CN&rank=1. Unique identifier: NCT03461107.


| INTRODUC TI ON
Chronic heart failure (HF) is a complex clinical syndrome caused by various factors including coronary heart disease, hypertension, valvular heart disease and idiopathic dilated cardiomyopathy. 1 Although great advances have been achieved in medical and surgical therapy recent decades, the 5-and 10-year mortality rates of HF remain high. 2,3 As the ageing of population and HF risk factors increase, HF has become a serious challenge for the public health. 3,4 Better understanding of the genetic basis of HF may shed light on the prevention, diagnosis and treatment of HF. 5,6 Recently, genome wide association studies have successfully identified multiple genes associated with the risk of HF. [7][8][9][10] However, only a few studies focused on the association of genetic variation with its prognosis. 4,11 Up to now, genetic loci and related genes associated with mortality of HF have remained rarely discovered and need more investigation. Thus, we speculated that there may exist other functional variants which could modify the prognosis of HF.
Necroptosis is a form of necrosis strictly regulated by distinct signalling pathway. 12 Substantial evidence has demonstrated that RIPK1-RIPK3-MLKL axis represented the core components for tumour necrosis factor-induced necroptosis both in human and mouse, which could be abolished by Receptor-interacting protein kinase 3 (RIP3)-depletion. [12][13][14] Many pathological processes including malignant melanoma, 15 intestinal tumourigenesis 16 and Abdominal Aortic Aneurysms 17 also have the participation of RIP3. Importantly, RIP3 showed strong expression in hearts of murine and human, which implied the importance of this protein for the function of myocardium. [18][19][20] Recently, studies have shown that necroptosis played a vital role in ischaemia-and oxidative stress-induced myocardial remodelling and HF, which could be alleviated by RIP3-depletion. 18,21 Based on these findings, we hypothesized that RIP3 gene variation may affect its expression and modify the susceptibility and prognosis of HF.
In the present study, we re-sequenced the regulatory region of RIP3 gene including the promoter and 5'UTR region of 200 control participants to identify putative functional variants. The underlying mechanism was explored. Ultimately, a total of 3194 HF patients and 2961 control participants were genotyped to investigate association between the variants and the occurrence and prognosis of HF.
Here, we demonstrated that rs3212247-C allele in the promoter of RIP3 gene destroyed the binding of sox17, which could repress the expression of ripk3. Moreover, participants carrying the rs3212247 CC genotype had higher plasma levels of RIP3 than those carrying the TT or TC genotype (p for trend = 0.02) in New York Heart Association (NYHA) class III HF group and the plasma RIP3 levels of NYHA III and IV groups were significantly higher compared with NYHA II and control groups, which suggest positive correlation between ripk3 level and severity of HF. Importantly, population of HF carrying rs3212247-CC genotype showed poorer prognosis assessed by cardiovascular death combined with cardiac transplantation (hazard ratio = 1.47, 95% CI = 1.13-1.91, P = 0.004).

| Genetic variation screening
Genomic DNA was extracted from peripheral venous blood samples of 200 control participants. Common variants in the promoter and 5'UTR region of RIP3 were identified by Sanger sequencing.
Polymerase chain reaction products were amplified from regions 1.8 kb upstream to 0.2 kb downstream of transcription start site ( Figure 1A). Details regarding primers, re-sequencing procedures and data analysis are given in Table S1 and the Methods section in the supplemental materials.

| In silico analysis of RIP3 promoter region
A total of 14 common single nuclear polymorphism (SNPs; Minor allele frequency [MAF] > 0.01) in RIP3 were identified by Sanger sequencing. The respective regions flanking these variants were analysed for potential transcription factor binding sites using Jaspar (http://jaspar.gener eg.net/). [22]

| SNP selection and genotyping
Genomic DNA was extracted from peripheral leucocytes as previously reported. 23 ABI Primer Expression 3.0 software was used for design of probe and primer sequences, which were synthesized by Shanghai GeneCore Bio Technologies Co., Ltd, China. The common variants in RIP3 promoter and CDS regions were genotyped using the TaqMan assay on the TaqMan 7900HT Sequence Detection System (Applied Biosystems, Foster City, CA) with the following conditions: 10 minutes at 95°C (enzyme activation) followed by 45 cycles at 95°C for 15 seconds and 60°C for 1 minute (annealing/extension). An endpoint read was performed for allelic discrimination after amplification.
The details of the procedure for amplification and the quality control of genotyping were mentioned in our previous study. 24 Details regarding primers and probes can be found in Table S2.
Bands were visualized by enhanced chemiluminescence reagents (Pierce Chemical, Rockford, IL) and quantified by densitometry.

| Chromatin immunoprecipitation assay
Chromatin immunoprecipitation assays were carried out using commercially available assay kits. Detailed descriptions are available in supplemental materials. TA B L E 1 Baseline characteristics of the study population

| Electrophoretic mobility-shift assay
Detailed procedures performed in the study are described in the supplemental materials.

| Measurement of genotype-dependent plasma levels of RIP3
Sex, age and NYHA class matched 207 HF patients and 78 sex, agematched controls were selected to detect the plasma levels of RIP3.
The detailed clinical characteristics and biochemical profiles of the samples are shown in Table S3. Total plasma RIP3 levels were analysed using a Human RIP3 ELISA Kit (CUSABIO, Wuhan, China) according to the manufacturer's instructions.

| Statistical analyses
Statistical analyses were performed with spss version 13.0 (SPSS, Because the distribution of RIP3 plasma levels is skewed non-normally, log transformation was performed and the results fitted normal distribution. All biostatistics calculations were performed with Prism (GraphPad). Data are expressed as mean ± SEM of n experiments.
All probability values were two-sided and P < 0.05 was considered to be significant.

| Effects of polymorphisms on the transcriptional activity
Firstly, we performed functional analyses comparing the activities of the 'major' (Hap1) and 'minor' (Hap2) haplotype in AC16 cells. As shown in Figure 1B, firefly luciferase reporter gene expression of the minor haplotype increased significantly compared to the major haplotype. The results showed that reporter gene expression of the rs3212247-C allele was significantly increased compared with the rs3212247-T allele ( Figure 1C), which could be replicated in HEK293T cell ( Figure   S2A). However, we did not observe any effect on luciferase activity in rs3834521, rs3759625, rs3212249, rs3759630 and rs3212246 luciferase assays in AC16 cells ( Figure 1D-H), expect that the wild-type of haplotype, which contains rs3212250 and rs3212251, showed higher reporter gene expression than mutant type ( Figure 1I). These results indicate that rs3212247 in the promoter of RIP3 gene may affect the binding domain of an endogenous cardiac regulator factor and in turn affect the gene expression.

| SOX17 modulates RIP3 expression
Bioinformatics analysis indicated that rs3212247 located in the binding site of SOX17, PBX1, SRY, ZNF35, TEAD4, ZNF354C, SOX9, SOX10 and SOX21. In order to confirm the effects of all these transcription factors (TFs) on endogenous RIP3 expression, we sequenced AC16 and HEK293T cell lines and found them to be TT genotype ( Figure S3). Subsequently, AC16 and HEK293T cells were transfected with pcDNA3.1 and TFs respectively. Western blot results revealed that SOX17 down-regulated RIP3 expression in both AC16 ( Figure 2B) and HEK293T cells ( Figure S2B). No changes were observed in RIP3 expression for other eight TFs in AC16 cells ( Figure S4). We then experimentally investigated the effect of rs3212247 on SOX17-mediated gene expression using Luciferase Activity Assays. Rs3212247 promoter reporter plasmids were cotransfected with SOX17 into AC16 cells. As shown in Figure 2A, compared with PGL3 empty vector, transcription activity of rs3212247-T allele reduced after transfected with SOX17 expression plasmids, which was in line with previous report that SOX17 could act as an inhibitory transcription factors. 30 Meanwhile, rs3212247-T allele exhibited lower firefly luciferase activity compared with rs3212247-C allele after transfected with SOX17. Consistent result was observed in HEK293T ( Figure S2A).
This indicated that the rs3212247-C allele could disrupt the binding site of SOX17 in the RIP3 promoter and increased the transcriptional activity of the gene.

| In vivo and vitro analyses of SOX17 binding for the RIP3 rs3212247-T>C allele
In order to further assess the functional relevance of rs3212247, we performed DNA binding analysis in AC16 and HEK293 cells.
Chromatin Immunoprecipitation Assay was conducted and the results showed that SOX17 binding to the region encompassing rs3212247 was detected in both cell lines ( Figure 3).
Meanwhile, we carried out electrophoretic mobility-shift assays in AC16. As shown in Figure 4, SOX17-containing nuclear extracts prefer to bind to biotin-labelled rs3212247-T probe but not to rs3212247-C probe and additional unlabelled oligonucleotide partially competed with this binding. However, no supershift band was present.

| Association of rs3212247 with plasma RIP3 levels
We measured the plasma RIP3 concentrations of HF patients to further investigate the functional role of rs3212247. As shown in Figure 5, plasma concentrations of RIP3 were higher in individuals with the CC genotype than in those with TT or TC genotypes among the NYHA class III patients (p for trend = 0.02). However, we did not observe the same results in other groups including NYHA II, NYHA IV and control groups. Interestingly, the plasma RIP3 levels of NYHA III and IV groups were significantly higher compared with NYHA II and control groups, suggesting that higher RIP3 levels may aggravate HF.
Furthermore, we assessed the effect of β-blocker use on plasma RIP3 level. No difference was observed between HF patients with and without β-blocker use in NYHA II, NYHA III, NYHA IV and combined groups ( Figure S5).

| Association of rs3212247 with the risk and prognosis of HF
Since it affected the expression and plasma concentration of RIP3, we further tested the association of rs3212247 with risk and prognosis of HF. Considering that rs3212247 was in strong linkage disequilibrium with rs3212250 and rs3212251 (r 2 = 0.99), we chose rs3212250 combined with rs3212251 as a tagging SNP for genotyping because they were only 6-bp distance. As a result, no significant difference was observed in genotype of rs3212247 between cases and controls (Table 3), which indicated that this SNP is not associated with the risk of HF. Interestingly, rs3212247 was significantly associated with the prognosis of HF using genetic recessive model, both with or without adjustment for conventional risk factors (including sex, age, hypertension, diabetes, hyperlipidemia, smoking state) and β-blocker use ( Figure 6A).  (Tables 3 and 4).

| D ISCUSS I ON
Heart failure is a common chronic disorder accompanied by loss of terminally differentiated cardiomyocyte. 21 As a passive and unregulated death form, necrosis has long been considered as the main form of cell death in myocardial injury. 33,34 Emerging evidence has shown that RIP3 participated in necroptosis, which is a programmed necrosis regulated by RIP1-RIP3-dependent death signalling pathway. 21,35 In heart, RIP3 enrolled in I/R and MI-induced myocardial injury and remodelling, which could be ameliorated by RIP3 knockout. 18,21 Considering the important role of RIP3 in myocardium, we investigated the association of single-nucleotide polymorphisms in RIP3 gene with HF using case-control and case-only

CO N FLI C T O F I NTE R E S T
The authors declare no competing financial interests.