Identification of Potential Therapeutic targets for Stroke and Its Subtypes by Integrating Proteomes and Genetics from Human Plasma

Hanchen Liu, Xiaoxi Zhang, Hongyu Ma, Thanh N. Nguyen, Weilong Hua, Shaojun Mo, Qinghai Huang, Jianmin Liu, Yu Zhou, Pengfei Yang
doi: https://doi.org/10.1101/2023.08.29.23294808

ABSTRACT

Background Previous genome-wide association studies (GWAS) have identified several risk genes for stroke; however, it remains unclear how they confer risk for the disease. We conducted an integrative analysis to identify candidate genes for stroke and stroke subtypes by integrating blood-derived multi-omics data with genetic data.

Method We systematically integrated the latest stroke GWAS database (73,652 patients and 1,234,808 controls) with human plasma proteomes (N=7,213) and performed proteome-wide association studies (PWAS), Mendelian randomization (MR), Bayesian colocalization analysis, and transcriptome-wide association study (TWAS) to prioritize genes that associate the risk of stroke and its subtypes with their expression and protein abundance in plasma. Cell-type specificity and functional enrichment analysis using single-cell RNA sequencing (scRNA-seq) and Gene Ontology (GO) databases were then performed to select target genes. A two-step MR analysis was followed to explore the potential mechanisms.

Results We found that the protein abundance of seven genes (MMP12, F11, SH3BGRL3, ENGASE, SCARA5, SWAP70, and SPATA20) in the plasma was associated with stroke and its subtypes, with six genes (MMP12, F11, SH3BGRL3, SCARA5, SWAP70, and SPATA20) causally related with stroke and its subtypes (P < 0.05/proteins identified for PWAS; P < 0.05/8 for MR; posterior probability of hypothesis 4 ≥ 75 % for Bayesian colocalization). The effect of F11, SH3BGRL, SPATA20, and SWAP70 on each subtype was mediated by Factor XI inhibitors (FXI), atrial fibrillation, T2D, and SBP respectively (p<0.05). We also found that SCARA5 and SWAP70 were related to stroke and ischemic stroke at the transcriptome level.

Conclusions Our present proteomic findings have identified new causal genes in the pathogenesis of stroke, which may offer potential future therapeutic targets for stroke prevention.

INTRODUCTION

Stroke is the leading cause of death and long-term disability worldwide, which imposes a great burden on global health.1 Despite significant advancements in the past decades, critical gaps persist in the management of stroke and its subtypes(i.e. large artery atherosclerosis, cardioembolic stroke, or small artery occlusion).2 As such, better understanding the mechanism of stroke occurrence, and identifying novel therapeutic targets are needed to tackle this challenge.

In recent years, large-scale genome-wide association studies (GWAS) have greatly aided the discovery of genetic loci linked to stroke and its subtypes.3 However, despite this progress in understanding the genetic risk of stroke, genes undergo regulation at the post-transcriptional, translational, and post-translational levels, predisposition to stroke involves a complex, pleiotropic, and polygenic genetic architecture. Deciphering the underlying mechanism for these genetic effects on stroke is always challenging, which hinders the translation of identified genetic findings into drug development.

Proteins are the final products of gene expression and the main functional components of cells and biological processes.4,5 Directly investigating proteins that are linked to stroke would be of great help in identifying potential therapeutic targets regarding stroke management, especially when this link is supported by the genetic association.6 In the past few years, several genetic articles concerned the role of circulating proteins in stroke, including a proteome-wide association study and one Mendelian randomization (MR) analysis.7,8 However, it is likely that the identification of protein quantitative trait locus (pQTL) in their study was incomplete, both due to inadequate sample size or/and lack of comprehensive protein measurements, and also multi-omics verification was lacking. With higher-quality and larger sample sizes of GWAS and, especially, human plasma proteomics data, becoming available, it would be promising to perform a new analysis using the newest data and tools to identify novel therapeutic targets for stroke.

Accordingly, we conducted an integrative analysis to identify candidate genes for stroke and stroke subtypes by integrating blood-derived multi-omics data with genetic data.

METHODS

Human plasma proteomic and transcriptomic data

The plasma proteomic data were derived from a comprehensive set of analyses of cis-genetic regulation of the plasma proteome in large European cohorts of the Atherosclerosis Risk in Communities (ARIC) study, comprising cleaned plasma protein data of 7,213 European American participants.9 Relative concentrations of plasma proteins or protein complexes were measured by the Slow-Of rate Modified Aptamer (SOMAmer) via a proteomic profiling platform.10,11 4657 SOMAmers were analyzed, which tagged proteins or protein complexes encoded by 4,435 genes. Finally, a total of 2,004 significant SOMAmers were identified.

The whole-blood eQTL data was derived from the Young Finns Study (YFS), in which the expression of peripheral blood genes (n=1264) was comprehensively analyzed.12

GWAS data of stroke and its subtypes

Genetic association estimates for stroke were obtained from the GIGASTROKE consortium, the largest published stroke GWAS meta-analysis to date.13 Populations were restricted to European participants. Data for 1,308,064 European descent individuals (73,652 any stroke cases and 1,234,808 controls) were used as a stroke GWAS summary dataset. The primary outcomes for main analysis were the occurrence of any stroke (including both ischemic and hemorrhagic stroke; AS; 73,652 cases), any ischemic stroke (AIS; 62,100 cases), or three aetiologic ischemic stroke subtypes: large artery stroke (LAS; 6,399 cases), cardioembolic stroke (CES; 10,804 cases) and small vessel stroke (SVS; 6,811 cases).

GWAS data on risk factors for stroke

The secondary outcomes we considered were stroke risk factors, which were selected from a literature review.14 Five risk factors were considered for the two-sample MR analyses and we sought well-powered publicly available GWAS summary statistics for these risk factors, including systolic blood pressure (SBP),15 atrial fibrillation (AF),16 type 2 diabetes (T2D),17 body mass index (BMI),18 and coagulation factor XI (FXI)19. The full details of the data sources and sample size for these GWASs were provided in supplementary materials (Table S1).

The RNA-Sequencing Data Availability of stroke

Single-cell RNA sequencing (scRNA-seq) data were used to identify the cell-type specificity and functional pathways of target genes.20 The scRNA-seq data were downloaded from the GEO dataset (https://www.ncbi.nlm.nih.gov/geo/) with accession number GSE174574. This study provided a scRNA-seq landscape of 17 cell populations of mice brain cortex with cell-type specific gene expression profiles based on the middle cerebral artery occlusion (MCAO) model.20 Functional pathways of target genes were enriched by Gene ontology (GO) analysis. Cell type-specific expression of the risk genes was displayed with a dot plot according to the cell classification from Zhang et al.20 The scRNA-seq analysis was performed using R packages of Seurat and ggsci. The enrichment score of risk genes from PWAS was calculated by R package AUCell.

Statistical analysis

Proteome-wide and transcriptome-wide association studies

The overall design of this study is shown in Figure 1. Proteome-wide association studies (PWAS) and transcriptome-wide association studies (TWAS) were carried out using FUSION.21 A linkage disequilibrium (LD) reference panel was used to minimize the influence of LD on the estimated test statistics.21 We used FUSION to compute the effect of single nucleotide polymorphisms (SNPs) on protein abundance for proteins with significant heritability (heritability P < 0.01). Two predictive models, enet and top1 were adopted in the analysis.9 Protein weights from the most predictive model were selected,9 and the expression weights were derived from transcriptomic data generated from YFS. Subsequently, we used FUSION to combine the genetic effect of stroke and stroke subtypes (stroke and its subtypes GWAS z-score) with the protein or expression weights by calculating the linear sum of z-score × weight for the independent SNPs at the locus to perform the PWAS or TWAS. The P values were adjusted for false discovery rate (FDR) using Benjamini-Hochberg (BH) method and the statistical significance was set at P<0.05.

Figure 1.
Figure 1. Overview of this study

First, we used protein quantitative trait locus (pQTL) datasets obtained from plasma and findings from stroke and its subtypes GWAS to perform a PWAS analysis. Second, we performed cell enrichment and functional enrichment analysis of the identified risk genes using single-cell transcriptomes and Gene Ontology (GO) databases. Third, we used independent Mendelian randomization (MR) analysis to verify PWAS-significant genes. Fourth, we used a colocalization to integrate GWAS data and plasma pQTL using a Bayesian colocalization analysis to explore whether two associated signals are consistent with shared causal variant(s). Fifth, we included five common risk factors for stroke in two-step MR analyses to explore the clinical mechanisms of risk genes. Finally, we explored the significant genes driving GWAS signals at the transcriptional level by leveraging gene expression data. ARIC, Atherosclerosis Risk in Communities; GWAS, genome-wide association studies; PWAS, proteome-wide association studies; AS: any stroke; AIS: any ischemic stroke; LAS: large artery stroke; CES: cardioembolic stroke; SVS: small vessel stroke; scRNA-seq, single-cell RNA sequencing; pQTL: protein quantitative trait locus; MR, Mendelian randomization; SBP, systolic blood pressure; AF, atrial fibrillation; T2D, type 2 diabetes; BMI, body mass index; FXI, coagulation factor XI; GTEx, Genotype-Tissue Expression; TWAS, transcriptome-wide association study.

Mendelian Randomization analysis

We used two-sample MR to verify whether stroke and its subtypes PWAS-significant genes (from the FUSION approach) were associated with stroke and its subtypes via their cis-regulated plasma protein abundance and estimate the role of protein expressed by these genes in stroke subtypes and its risk factors. The SNPs included in the study robustly and independently (r2 < 0.1; clumping window, 10 000 kb) predicted exposures at a genome-wide level (P < 5 × 10-8). We harmonized the SNP alleles across studies and removed palindromic SNPs with ambiguous allele frequencies (0.42–0.58).

Principal analyses were conducted using the inverse-variance weighted (IVW) MR method. This method assumes that all genetic variants are valid instrumental variables and produces the most accurate estimates.22 Wald’s ratio method was applied when there was only one SNP available for the target exposure.23 The weighted median and MR-Egger regression were adopted as sensitivity analyses. The weighted median approaches give more weight to the instrumental variables that are more precise, and the estimate is consistent even when up to 50 % of the information comes from invalid or weak instruments.24 The MR-Egger could detect and adjust for directional pleiotropy, albeit with low precision.22 We used Cochran’s Q statistic in the IVW model to assess the heterogeneity between variant-specific estimates. The MR Pleiotropy Residual Sum and Outlier (PRESSO) approaches were used to identify possible outliers. Finally, a leave-one-SNP-out analysis was performed in which SNPs were systematically removed to assess whether a single SNP drove the results. We set the P value for statistical significance at P<0.007, after correcting for multiple exposures using the Bonferroni method (α=0.05/7 proteins). Statistical analyses were conducted in R (version 4.2.2) and MR analyses were conducted using “TwoSampleMR”.

Bayesian colocalization analysis

We conducted colocalization analysis for each of the genes associated with one or more of the stroke outcomes to investigate whether the protein levels of a gene and stroke outcome genetic associations are due to the same causal variants. We estimated the posterior probability (PP) of multiple traits sharing the same causal SNP simultaneously using a multi-trait colocalization method.25 We used the default colocalization priors of P1 = 10−4, P2 = 10−4, and P12 = 10−5, where P1 is the probability that a given variant is associated with stroke or its subtypes, P2 is the probability that a given variant is a significant pQTL, and P12 is the probability that a given variant is both a stroke or its subtypes result and a pQTL. colocalization uses computed approximation Bayes factors and summary association data to generate a posterior probability for the following 5 hypotheses: H0, No association with either GWAS or pQTL; H1, Association with GWAS, not with pQTL; H2, Association with pQTL, not with GWAS; H3, Association with GWAS and pQTL, two independent SNPs; and H4, Association with GWAS and pQTL, one shared SNP. We mainly focused on the last hypothesis H4 and posterior probability (PP) was used to quantify support for H4 (denoted as PPH4). We defined strong evidence of colocalization at PPH4 ≥ 0.75.26

Mediation analysis

For protein levels of a gene that causally associate with both stroke and risk factors, we conducted a mediation analysis to quantify the effects of protein levels of a gene on stroke outcomes via risk factors. The “total” effect of exposure on outcome includes both the “direct” effect and any “indirect” effect via one or more mediators. In this study, the total effect is captured by a standard univariable MR analysis—the primary MR. To decompose direct and indirect effects, we used results from two-step MR and chose the Product method to estimate the beta of indirect effect and the Delta method to estimate the standard error (SE) and confidence interval (CI).27

Functional enrichment analysis and Cell-type specificity analysis

Gene ontology (GO) analysis was performed to do functional enrichment analysis. Subsequently, using mouse single-cell RNA-seq data of the CoW profiled, we investigated the cell type-specific expression of the risk genes. The cell suspensions underwent scRNAseq using standard 10X Chromium Single Cell Chemistry V3. Raw sequencing data were processed using Cellranger to produce gene-level counts for each cell in each sample. The CellRanger mkfastq command was used to generate Fastq files. Data were mapped to a prebuild mouse reference genome. All subsequent analysis was performed using R packages of Seurat and ggsci. The enrichment score of risk genes from PWAS was calculated by R package AUCell. The detailed parameters of scRNA analysis could refer to previous works.20

RESULTS

PWAS of stroke and its subtypes

The PWAS identified 7 genes (MMP12, F11, ENGASE, SH3BGRL3, SPATA20, SWAP70, SCARA5) whose cis-regulated plasma protein levels were associated with stroke and its subtypes at a false discovery rate (FDR) of P<0.05 (Figure 2 and Table S2). The protein abundances of MMP12 were associated with AS, AIS, and LAS (AS: Z-score: -5.881, PWAS FDR P = 2.73×10-6; AIS: Z-score: -6.149, PWAS FDP P= 5.18×10-7; LAS: Z-score: -5.674, PWAS FDR P= 7.41×10-6), and the protein abundances of F11 was associated with AS, AIS, and CES (AS: Z-score: 5.972, PWAS FDR P = 3.12×10-6; AIS: Z-score: 6.718, PWAS FDR P = 2.45×10-8; CES: Z-score: 5.683, PWAS FDR P = 1.39×10-5). The protein abundances of ENGASE and SH3BGRL3 were associated with AS and AIS (ENGASE: AS: Z-score= 5.055, PWAS FDR P = 1.91×10-4; AIS: Z-score= 4.805, PWAS FDR P = 6.87×10-4), (SH3BGRL3: AS: Z-score= -4.284, PWAS FDR P = 6.14×10-3; AIS: Z-score= -4.322, PWAS FDR P = 5.12×10-3). The protein abundances of SCARA5, SWAP70 and SPATA20 were associated with AS, AIS and SVS respectively (SCARA5: Z-score: -3.712, PWAS FDR P = 4.58×10-2; SWAP70: Z-score: -3.955, PWAS FDR P = 2.03×10-2; SPATA20: Z-score: 4.55, PWAS FDR P = 5.21×10-3). (Table S1)

Figure 2.
Figure 2. The Manhattan plot for the PWAS of stroke and stroke subtypes.

The Manhattan plot shows the genes identified using PWAS for AS, AIS, LAS, CES, and SVS, respectively. Each point represents a single association test between a gene and stroke and its subtypes ordered by genomic position on the x axis and the association strength on the y axis as the −log10(P) of a z-score test. The discovery PWAS identified 7 genes whose cis-regulated plasma protein abundance was associated with stroke subtypes at an FDR of P < 0.05. The red horizontal line reflects the significant threshold of the FDR P < 0.05.

Functional enrichment analysis and Cell-type specificity analysis

We investigated whether the risk genes identified by PWAS were enriched in some particular functions and some particular brain cell types. Go analyses suggest that MMP12 and F11 were significantly related to serine-type endopeptidase activity, as well as serine-type peptidase activity, wound healing, and serine hydrolase activity et.al. MMP12 and SWAP70 were significantly related to negative regulation of cell adhesion and positive regulation of response to external stimulus. MMP12 and ENGASE were significantly related to the glycoprotein metabolic process. SWAP70 and SH3BGRL3 were significantly related to the cell leading edge (Figure 3A). Single-cell RNA-seq data showed different expression patterns of the risk genes in the brain. MMP12 was mainly expressed in microglia. F11 was mainly expressed in fibroblast. SCARA5 was mainly expressed in perivascular fibroblast-like cells and SWAP70 was mainly expressed in endothelial. SH3BGRL3 was abundantly expressed in most cell types of the brain. The expression of SPATA20 was not detected in this system (Figure 3B and Table S3).

Figure 3.
Figure 3. The functional enrichment analysis and Cell-type specificity analysis among candidate genes.

A: functional enrichment analysis; B: cell-type specificity analysis, Single-cell-type expression of the potentially stroke-risk genes. Bar graph of single-cell-type enrichment for risk genes in stroke and its subtypes from the discovery PWAS. The diagram depicts expression of every gene (y axis) for each cell (x axis).

Genetically determined PWAS-significant genes and risk of stroke and its subtypes

Cis-regulated plasma protein levels of seven PWAS-significant genes were tested for causal relationships with stroke and its subtypes by using MR analysis (P<0.007 = 0.05/7). As shown in Figure 4, we found that the concentration of MMP12 was inversely associated with AS, AIS, and LAS risk (AS: OR [95% CI]: 0.93 [0.91, 0.95]; AIS: OR [95% CI]: 0.92 [0.91, 0.94]; LAS: OR [95% CI]: 0.82 [0.75, 0.91]), while F11 was positively associated with AS, AIS, and CES risk (AS: OR [95% CI]: 1.07 [1.05, 1.10]; AIS: OR [95% CI]: 1.09 [1.06, 1.13]; CES: OR [95% CI]: 1.23 [1.15, 1.31]). The concentration of ENGASE was positively associated with AS and AIS (AS: OR [95% CI]: 1.05 [1.03, 1.06]; AIS: OR [95% CI]: 1.04 [1.03, 1.06]), but the concentration of SH3BGRL3 was inversely associated with AS, AIS (AS: OR [95% CI]: 0.96 [0.94, 0.98]; AIS: OR [95% CI]: 0.95 [0.93, 0.97]). SCARA5 and SWAP70 were associated with a lower risk of AS (OR [95% CI]: 0.92 [0.87, 0.98]) and AIS (OR [95% CI]: 0.96 [0.94, 0.99]) while SPATA20 was associated with a higher risk of SVS (OR [95% CI]: 1.16 [1.09, 1.23]). All sensitivity analyses supported the main results of the MR Analysis (Table S4). The results of MR further confirmed the association between seven genes and their corresponding stroke subtypes.

Figure 4.
Figure 4. Effects of stroke and its subtypes PWAS-significant genes on stroke subtypes.

The squares are the causal estimates on the OR scale, and the whiskers represent the 95% confidence intervals for these ORs. SNPs: number of SNPs used for the estimation of the causal effects in this plot. P-value were determined from the inverse-variance-weighted two-sample MR method. P-het, P value in the Q statistic for heterogeneity; P-pleio, P value in the Egger intercept.

Colocalization between stroke risk genes and pQTLs in the plasma

We examined the PP for a shared causal variant between a pQTL and stroke for the seven genes which met the corrected P value threshold in previous PWAS and MR analysis (Table S10). The colocalization suggested the genetic variants associated with MMP12 (pQTL) were due to the same genetic variants underlying the association with AS, AIS, and LAS (PPH4 ≥ 0.75). Similarly, F11 pQTLs colocalized with AS, AIS, and CES genetic associations; SH3BGRL3 pQTLs colocalized with AS and AIS; SCARA5, SWAP70, and SPATA20 pQTLs colocalized with AS, AIS, and SVS respectively. It suggests that six of seven proteins play an important role in the pathophysiology of stroke subtypes.

Significance of the protein findings

To determine the importance of the six of seven (ENGASE was excluded because its results of Bayesian colocalization analysis were negative) potentially causal genes identified from the PWAS, MR, and Bayesian colocalization analyses, we obtained the lowest P values for the SNPs within 1 Mb of each of these seven genes using the summary statistics from the most extensive stroke GWAS (N=446, 696).28 The most significant P values were less than 5×10-8 in two genes (MMP12 and F11), while the P values of SNPs in the remaining 4 genes ranged from 1.32×10-5 to 2.36×10-6 (Table S5). The findings suggest that specific plasma proteins likely contribute to the pathogenesis of stroke subtypes.

Mediation Effect of six genes on stroke subtypes via risk factors

To investigate the indirect effect of cis-regulated plasma protein levels of six genes on stroke subtypes via risk factors, we carried out a mediation analysis using the effect estimates from two-step MR and the total effect from primary MR. We calculated the causal relationship between five risk factors and stroke using MR Analysis (Table S6).

This analysis was restricted to four genes, F11, SWAP70, SH3BGRL3, and SPATA20, that showed evidence of an effect in both MRs with risk factors and stroke outcomes (Table S6-S7). The mediation effect of F11 via FXI is the highest (63.8%). The indirect effect of SPATA20 on the risk of SVS via T2D contributes to one-fourth of the total effect (25.2%). Similarly, the proportion of the mediation effect of SWAP70 and SH3BGRL3 on AIS and AS via SBP and AF is one-tenth, respectively (SWAP70: 14.1%; SH3BGRL3: 12.2%). (Figure 5).

Figure 5.
Figure 5. Mediation effects of genes on stroke via risk factors.

Mediation analyses to quantify the effects of four genes on stroke subtypes via risk factors. A. F11 effect on CES mediated by FXI. B. SH3BGRL3 effect on AS mediated by AF. C. SPATA20 effect on SVS mediated by T2D. D. SWAP70 effect on AIS mediated by SBP. FXI, coagulation factor XI; CES, cardioembolic stroke; SBP, systolic blood pressure; AF, atrial fibrillation; T2D, type 2 diabetes; SVS, small vessel stroke; AS, any stroke; AIS, any ischemic stroke.

Examination of the potential stroke-causal proteins at the mRNA level

We combined the stroke subtypes GWAS data with human plasma transcriptomes to conduct stroke and its subtypes TWAS using FUSION. We found that the cis-regulated plasma mRNA expression of the forty-one genes was associated with stroke and its subtypes (FDR P < 0.05) (Table S8). Interestingly, we found two genes identified in TWAS, i.e. SWAP70 and SPATA20, were also significant in the AIS and SVS PWAS respectively, suggesting joint evidence from PWAS and TWAS for its role in stroke etiology.

DISCUSSION

In the present study, we employed an integrated analysis based on GWAS, PWAS, TWAS, MR, and Bayesian colocalization to investigate the causal relationship between proteins in the plasma and stroke and its subtypes. We identified 7 potential risk genes (MMP12, F11, SH3BGRL3, ENGASE, SCARA5, SWAP70, and SPATA20) of stroke with altered protein abundances in the plasma. We enriched all seven genes into corresponding functional pathways and cell types. Six (MMP12, F11, SH3BGRL3, SCARA5, SWAP70, and SPATA20) of the above genes were replicated in the independent MR and Bayesian colocalization analysis validation analyses of stroke and its subtypes, providing a higher confidence level. Furthermore, SWAP70 and SPATA20 were supported at the transcriptional level. Also, through a mediation analysis, we found the effect of F11 on CES, SH3BGRL3 on AS and AIS, SPATA20 on SVS, and SWAP70 on AIS were partly mediated by FXI, AF, T2D, and SBP respectively, demonstrating a critical role of these risk factors in the pathogenesis of stroke in consistence with previous epidemiological studies.29–33

Of the 7 genes studied, three (MMP12, F11, and SCARA5) have been implicated in previous articles.7,8 MMP12 is a member of the matrix metalloproteinase family that contributes to vascular remodeling. To date, the relationship between its circulating level and stroke is unclear and contradicted in previous studies. In one population-based study containing 2983 participants, higher MMP12 level was independently associated with ischemic stroke risk (HR=1.30, 95% CI 1.16-1.45, P=4.55×10-06).34 In contrast, our study is consistent with another genetic study which showed an inverse relationship between MMP12 circulating level and risk of ischemic stroke.35 FXI is mediated by the F11 gene and has become increasingly interesting for its role in the pathogenesis of thrombosis. A GWAS analysis containing 371,695 participants found a lower FXI level was associated with reduced risks of venous thrombosis (OR=0.1, 95%CI 0.07-0.14; P=3×10-43) and ischemic stroke (OR=0.47, 0.36-0.61; P=2×10-8).36 MR analysis of our study also revealed a causal relationship between F11 and CES, and 63.8% of the effect is mediated by FXI. Currently, FXI inhibitors have been developed as potential stroke targets, and their role is evaluated by some ongoing trials.37–39 SCARA5 is a scavenger receptor that exports ferritin-bound iron from circulation to parenchymal tissues, including the heart and brain.40 Previous MR studies revealed SCARA5 genetically lower serum iron which further decreases CES risk.8,41. Using multi-omics analysis, we confirmed this result, SCARA5 was associated with a lower risk of AS (OR [95% CI]: 0.92 [0.87, 0.98]).

To our best knowledge, the other four genes including ENGASE, SPATA20, SWAP70, and SH3BGRL3 have not been extensively studied in stroke research. And we were able to identify a shared causal variant between a pQTL and stroke for the four genes except for ENGASE. And we identified SPATA20 as associated with the SVS subtype. Moreover, mediation analysis of the effects of these three genes on stroke subtypes indicated potential mechanisms. For example, the indirect effect of SPATA20 on the risk of SVS via T2D contributes to one-fourth of the total effect (25.2%). and the proportion of the mediation effect of SWAP70 and SH3BGRL3 on AIS and AS via SBP and AF is one-tenth, respectively (SWAP70: 14.1%; SH3BGRL3: 12.2%). However, the mechanisms of the effects of these proteins are not clear yet. ENGASE is an enzyme involved in the processing of free oligosaccharides in the cytosol and is associated with glycosylation. Studies have shown protein glycosylation may affect the occurrence of AIS by regulating the progression of atherosclerosis and AF.42 SWAP70 is a guanine nucleotide exchange factor that participates in the regulation of many cellular processes, its role in many diseases is not clear yet, however, studies showed that SWAP70 is a protective molecule that can suppress the progression of nonalcoholic fatty liver disease by inhibiting hepatic steatosis and inflammation,43 and Pathological Cardiac Hypertrophy, et al.44 Both of which were associated with stroke occurrence. SWAP70 also organizes the actin cytoskeleton, which is crucial for phagocytosis.45 The role of SWAP70 in immune function may be related to its protective effect against AIS, and need further study in the future. SH3BGRL3 may play a role in maintaining cerebrovascular integrity and maintaining the function of cerebral vascular endothelial cells by activating STAT3 signaling, thereby preventing AS and AIS.46,47 Little is known about SPATA 20, it is predicted to be located in the extracellular region, and involved in the carbohydrate metabolic process and cell differentiation. Whatever, further studies are needed to confirm the role of these genes and proteins in stroke.

The strengths of our study include the utilization of the largest and most comprehensive human proteome and summary statistics from the most recent GWAS and protein quantitative trait locus (pQTL) datasets to date, and integrative analysis of multi-omics data including GWAS, PWAS, and TWAS. The stroke GWAS summary datasets we used included data for 1,308,064 European descent individuals, nearly three times the size of previous studies (MRGASTROKE n=446, 696). 7,8 The plasma proteomic data we used included a total of 7, 213 European Americans, and the number of significant SOMAmers in this study is almost three times that of plasma pQTL studies conducted in the past in the European ancestry sample.48 We further identified risk genes through cell enrichment and functional enrichment analysis using single-cell transcriptomes and GO databases. Afterward, we used independent MR analysis to verify PWAS-significant genes. Then we used a colocalization to integrate GWAS data and plasma pQTL using a Bayesian colocalization analysis to explore whether two associated signals are consistent with shared causal variant(s). Finally, we included five common risk factors of stroke in two-step MR analyses to explore the clinical mechanisms of risk genes and explored the significant genes driving GWAS signals at the transcriptional level by leveraging gene expression data.

Our study also has some limitations. First, although the plasma proteomic data we used included SOMAmers for ∼ 5,000 proteins or protein complexes, it does not provide coverage for the entire plasma proteome.9 Also, the data did not explore the effects of uncommon and rare variants, as well as complex trans-associations, analysis with substantial discovery on an even larger sample size is likely to need. Second, it is insufficient to elucidate the numerous stroke PWAS-identified loci from genetic and transcriptional levels. Functional genomic approaches are needed to identify the complex molecular mechanisms of stroke. Third, our study mainly focused on European subjects, and caution should be taken when generalizing our results to other ethnicities.

CONCLUSION

In conclusion, we found causal evidence supporting three classic genes (MMP12, F11, and SCARA5) and three novel genes (SH3BGRL, SWAP70, and SPATA20) associated with ischemic stroke and their subtypes. Our findings provide genetic evidence underlying ischemic stroke and subtypes, allowing novel therapeutic targets to be further identified.

Data Availability

All data in the manuscript is truly usable

Sources of funding

This study was partially supported by grants from the National Natural Science Foundation of China (No. 81400979 and 81870931), the SanHang Program of the Naval Medical University, and the ‘Climbing’ program of Changhai Hospital.

Disclosure

The authors report no disclosures relevant to the manuscript.

Supplemental Material

Tables S1-S10

Footnotes

  • Abbreviations: GWAS: large-scale genome-wide association studies; MR, Mendelian randomization; pQTL: protein quantitative trait locus; ARIC: Atherosclerosis Risk in Communities; SOMAmer: Slow-Of rate Modified Aptamer; YFS: Young Finns Study; AS: any stroke; AIS: any ischemic stroke; LAS: large artery stroke; CES: cardioembolic stroke; SVS: small vessel stroke; SBP: systolic blood pressure; AF: atrial fibrillation; T2D: type 2 diabetes; BMI: body mass index; FXI: coagulation factor XI; scRNA-seq: single-cell RNA sequencing; MCAO: middle cerebral artery occlusion; LD: linkage disequilibrium; PWAS: Proteome-wide association studies; TWAS: transcriptome-wide association studies; SNPs: single nucleotide polymorphisms; FDR: false discovery rate; BH: Benjamini-Hochberg; IVW: inverse-variance weighted; PRESSO: Pleiotropy Residual Sum and Outlier; PP: posterior probability; SE: standard error; CI: confidence interval; GO: Gene ontology.

Reference

  1. 1.↵
    Jang H, Park S, Kim MS, Yon DK, Lee SW, Koyanagi A, Kostev K, Shin JI, Smith L. Global, regional and national burden of alopecia areata and its associated diseases, 1990-2019: A systematic analysis of the Global Burden of Disease Study 2019. European Journal of Clinical Investigation. 2023. doi: 10.1111/eci.13958
    OpenUrlCrossRef
  2. 2.↵
    Wu SM, Wu B, Liu M, Chen ZM, Wang WZ, Anderson CS, Sandercock P, Wang YJ, Huang YN, Cui LY, et al. Stroke in China: advances and challenges in epidemiology, prevention, and management. Lancet Neurology. 2019;18:394–405. doi: 10.1016/s1474-4422(18)30500-3
    OpenUrlCrossRefPubMed
  3. 3.↵
    Tam V, Patel N, Turcotte M, Bosse Y, Pare G, Meyre D. Benefits and limitations of genome-wide association studies. Nature Reviews Genetics. 2019;20:467–484. doi: 10.1038/s41576-019-0127-1
    OpenUrlCrossRefPubMed
  4. 4.↵
    Vogel C, Marcotte EM. Insights into the regulation of protein abundance from proteomic and transcriptomic analyses. Nature Reviews Genetics. 2012;13:227–232. doi: 10.1038/nrg3185
    OpenUrlCrossRefPubMed
  5. 5.↵
    Rolland DCM, Basrur V, Jeon YK, McNeil-Schwalm C, Fermin D, Conlon KP, Zhou YQ, Ng SY, Tsou CC, Brown NA, et al. Functional proteogenomics reveals biomarkers and therapeutic targets in lymphomas. Proceedings of the National Academy of Sciences of the United States of America. 2017;114:6581–6586. doi: 10.1073/pnas.1701263114
    OpenUrlAbstract/FREE Full Text
  6. 6.↵
    Nelson MR, Tipney H, Painter JL, Shen JD, Nicoletti P, Shen YF, Floratos A, Sham PC, Li MJ, Wang JW, et al. The support of human genetic evidence for approved drug indications. Nature Genetics. 2015;47:856-+. doi: 10.1038/ng.3314
    OpenUrlCrossRefPubMed
  7. 7.↵
    Wu BS, Chen SF, Huang SY, Ou YN, Deng YT, Chen SD, Dong Q, Yu JT. Identifying causal genes for stroke via integrating the proteome and transcriptome from brain and blood. Journal of Translational Medicine. 2022;20. doi: 10.1186/s12967-022-03377-9
    OpenUrlCrossRef
  8. 8.↵
    Chong M, Sjaarda J, Pigeyre M, Mohammadi-Shemirani P, Lali R, Shoamanesh A, Gerstein HC, Pare G. Novel Drug Targets for Ischemic Stroke Identified Through Mendelian Randomization Analysis of the Blood Proteome. Circulation. 2019;140:819–830. doi: 10.1161/circulationaha.119.040180
    OpenUrlCrossRefPubMed
  9. 9.↵
    Zhang JN, Dutta D, Kottgen A, Tin A, Schlosser P, Grams ME, Harvey B, Yu B, Boerwinkle E, Coresh J, et al. Plasma proteome analyses in individuals of European and African ancestry identify cis-pQTLs and models for proteome-wide association studies. Nature Genetics. 2022;54:593-+. doi: 10.1038/s41588-022-01051-w
    OpenUrlCrossRef
  10. 10.↵
    Gold L, Ayers D, Bertino J, Bock C, Bock A, Brody EN, Carter J, Dalby AB, Eaton BE, Fitzwater T, et al. Aptamer-Based Multiplexed Proteomic Technology for Biomarker Discovery. Plos One. 2010;5. doi: 10.1371/journal.pone.0015004
    OpenUrlCrossRefPubMed
  11. 11.↵
    Williams SA, Kivimaki M, Langenberg C, Hingorani AD, Casas JP, Bouchard C, Jonasson C, Sarzynski MA, Shipley MJ, Alexander L, et al. Plasma protein patterns as comprehensive indicators of health. Nature Medicine. 2019;25:1851-+. doi: 10.1038/s41591-019-0665-2
    OpenUrlCrossRefPubMed
  12. 12.↵
    Raitakari OT, Juonala M, Ronnemaa T, Keltikangas-Jarvinen L, Rasanen L, Pietikainen M, Hutri-Kahonen N, Taittonen L, Jokinen E, Marniemi J, et al. Cohort Profile: The Cardiovascular Risk in Young Finns Study. International Journal of Epidemiology. 2008;37:1220–1226. doi: 10.1093/ije/dym225
    OpenUrlCrossRefPubMedWeb of Science
  13. 13.↵
    Mishra A, Malik R, Hachiya T, Jurgenson T, Namba S, Posner DC, Kamanu FK, Koido M, Le Grand Q, Shi MY, et al. Stroke genetics informs drug discovery and risk prediction across ancestries. Nature. 2022;611:115-+. doi: 10.1038/s41586-022-05165-3
    OpenUrlCrossRef
  14. 14.↵
    Feigin VL, Stark BA, Johnson CO, Roth GA, Bisignano C, Abady GG, Abbasifard M, Abbasi-Kangevari M, Abd-Allah F, Abedi V, et al. Global, regional, and national burden of stroke and its risk factors, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet Neurology. 2021;20:795–820. doi: 10.1016/s1474-4422(21)00252-0
    OpenUrlCrossRefPubMed
  15. 15.↵
    Evangelou E, Warren HR, Mosen-Ansorena D, Mifsu B, Pazoki R, Gao H, Ntritsos G, Dimou N, Cabrer CP, Karaman I, et al. Genetic analysis of over 1 million people identifies 535 new loci associated with blood pressure traits. Nature Genetics. 2018;50:1412-+. doi: 10.1038/s41588-018-0205-x
    OpenUrlCrossRefPubMed
  16. 16.↵
    Nielsen JB, Thorolfsdottir RB, Fritsche LG, Zhou W, Skov MW, Graham SE, Herron TJ, McCarthy S, Schmidt EM, Sveinbjornsson G, et al. Biobank-driven genomic discovery yields new insight into atrial fibrillation biology. Nature Genetics. 2018;50:1234-+. doi: 10.1038/s41588-018-0171-3
    OpenUrlCrossRefPubMed
  17. 17.↵
    Xue A, Wu Y, Zhu ZH, Zhang FT, Kemper KE, Zheng ZL, Yengo L, Lloyd-Jones LR, Sidorenko J, Wu YD, et al. Genome-wide association analyses identify 143 risk variants and putative regulatory mechanisms for type 2 diabetes. Nature Communications. 2018;9. doi: 10.1038/s41467-018-04951-w
    OpenUrlCrossRefPubMed
  18. 18.↵
    Yengo L, Sidorenko J, Kemper KE, Zheng ZL, Wood AR, Weedon MN, Frayling TM, Hirschhorn J, Yang J, Visscher PM, et al. Meta-analysis of genome-wide association studies for height and body mass index in similar to 700 000 individuals of European ancestry. Human Molecular Genetics. 2018;27:3641–3649. doi: 10.1093/hmg/ddy271
    OpenUrlCrossRefPubMed
  19. 19.↵
    Suhre K, Arnold M, Bhagwat AM, Cotton RJ, Engelke R, Raffler J, Sarwath H, Thareja G, Wahl A, DeLisle RK, et al. Connecting genetic risk to disease end points through the human blood plasma proteome. Nature Communications. 2017;8. doi: 10.1038/ncomms14357
    OpenUrlCrossRefPubMed
  20. 20.↵
    Zheng K, Lin LM, Jiang W, Chen L, Zhang XY, Zhang Q, Ren Y, Hao JW. Single-cell RNA-seq reveals the transcriptional landscape in ischemic stroke. Journal of Cerebral Blood Flow and Metabolism. 2022;42:56–73. doi: 10.1177/0271678x211026770
    OpenUrlCrossRef
  21. 21.↵
    Gusev A, Ko A, Shi H, Bhatia G, Chung W, Penninx B, Jansen R, de Geus EJC, Boomsma DI, Wright FA, et al. Integrative approaches for large-scale transcriptome-wide association studies. Nature Genetics. 2016;48:245–252. doi: 10.1038/ng.3506
    OpenUrlCrossRefPubMed
  22. 22.↵
    Burgess S, Butterworth A, Thompson SG. Mendelian Randomization Analysis With Multiple Genetic Variants Using Summarized Data. Genetic Epidemiology. 2013;37:658–665. doi: 10.1002/gepi.21758
    OpenUrlCrossRefPubMed
  23. 23.↵
    Davey Smith G, Hemani G. Mendelian randomization: genetic anchors for causal inference in epidemiological studies. Human Molecular Genetics. 2014;23:R89–R98. doi: 10.1093/hmg/ddu328
    OpenUrlCrossRefPubMedWeb of Science
  24. 24.↵
    Davies NM, Holmes MV, Smith GD. Reading Mendelian randomisation studies: a guide, glossary, and checklist for clinicians. Bmj-British Medical Journal. 2018;362. doi: 10.1136/bmj.k601
    OpenUrlFREE Full Text
  25. 25.↵
    Giambartolomei C, Vukcevic D, Schadt EE, Franke L, Hingorani AD, Wallace C, Plagnol V. Bayesian Test for Colocalisation between Pairs of Genetic Association Studies Using Summary Statistics. Plos Genetics. 2014;10. doi: 10.1371/journal.pgen.1004383
    OpenUrlCrossRefPubMed
  26. 26.↵
    Kia DA, Zhang D, Guelfi S, Manzoni C, Hubbard L, Reynolds RH, Botia J, Ryten M, Ferrari R, Lewis PA, et al. Identification of Candidate Parkinson Disease Genes by Integrating Genome-Wide Association Study, Expression, and Epigenetic Data Sets. Jama Neurology. 2021;78:464–472. doi: 10.1001/jamaneurol.2020.5257
    OpenUrlCrossRef
  27. 27.↵
    Carter AR, Sanderson E, Hammerton G, Richmond RC, Smith GD, Heron J, Taylor AE, Davies NM, Howe LD. Mendelian randomisation for mediation analysis: current methods and challenges for implementation. European Journal of Epidemiology. 2021;36:465–478. doi: 10.1007/s10654-021-00757-1
    OpenUrlCrossRefPubMed
  28. 28.↵
    Malik R, Chauhan G, Traylor M, Sargurupremraj M, Okada Y, Mishra A, Rutten-Jacobs L, Giese AK, van der Laan SW, Gretarsdottir S, et al. Multiancestry genome-wide association study of 520,000 subjects identifies 32 loci associated with stroke and stroke subtypes. Nature Genetics. 2018;50:524-+. doi: 10.1038/s41588-018-0058-3
    OpenUrlCrossRefPubMed
  29. 29.↵
    Kivimäki M, Kuosma E, Ferrie JE, Luukkonen R, Nyberg ST, Alfredsson L, Batty GD, Brunner EJ, Fransson E, Goldberg M, et al. Overweight, obesity, and risk of cardiometabolic multimorbidity: pooled analysis of individual-level data for 120 813 adults from 16 cohort studies from the USA and Europe. The Lancet Public health. 2017;2:e277–e285. doi: 10.1016/s2468-2667(17)30074-9
    OpenUrlCrossRef
  30. 30.
    Kwan TW, Wong SS, Hong Y, Kanaya AM, Khan SS, Hayman LL, Shah SH, Welty FK, Deedwania PC, Khaliq A, et al. Epidemiology of Diabetes and Atherosclerotic Cardiovascular Disease Among Asian American Adults: Implications, Management, and Future Directions: A Scientific Statement From the American Heart Association. Circulation. 2023. doi: 10.1161/cir.0000000000001145
    OpenUrlCrossRef
  31. 31.
    Siegerink B, Govers-Riemslag JW, Rosendaal FR, Ten Cate H, Algra A. Intrinsic coagulation activation and the risk of arterial thrombosis in young women: results from the Risk of Arterial Thrombosis in relation to Oral contraceptives (RATIO) case-control study. Circulation. 2010;122:1854–1861. doi: 10.1161/circulationaha.110.943738
    OpenUrlAbstract/FREE Full Text
  32. 32.
    Lin Q, Ye T, Ye P, Borghi C, Cro S, Damasceno A, Khan N, Nilsson PM, Prabhakaran D, Ramirez A, et al. Hypertension in stroke survivors and associations with national premature stroke mortality: data for 2 · 5 million participants from multinational screening campaigns. The Lancet Global health. 2022;10:e1141–e1149. doi: 10.1016/s2214-109x(22)00238-8
    OpenUrlCrossRef
  33. 33.↵
    Zhou B, Perel P, Mensah GA, Ezzati M. Global epidemiology, health burden and effective interventions for elevated blood pressure and hypertension. Nature reviews Cardiology. 2021;18:785–802. doi: 10.1038/s41569-021-00559-8
    OpenUrlCrossRefPubMed
  34. 34.↵
    Kalani R, Bartz TM, Psaty BM, Elkind MSV, Floyd JS, Gerszten RE, Shojaie A, Heckbert SR, Bis JC, Austin TR, et al. Plasma Proteomic Associations With Incident Ischemic Stroke in Older Adults: The Cardiovascular Health Study. Neurology. 2023. doi: 10.1212/wnl.0000000000207242
    OpenUrlCrossRef
  35. 35.↵
    Misra S, Talwar P, Kumar A, Kumar P, Sagar R, Vibha D, Pandit AK, Gulati A, Kushwaha S, Prasad K. Association between matrix metalloproteinase family gene polymorphisms and risk of ischemic stroke: A systematic review and meta-analysis of 29 studies. Gene. 2018;672:180–194. doi: 10.1016/j.gene.2018.06.027
    OpenUrlCrossRef
  36. 36.↵
    Georgi B, Mielke J, Chaffin M, Khera AV, Gelis L, Mundl H, van Giezen JJJ, Ellinor P, Kathiresan S, Ziegelbauer K, et al. Leveraging Human Genetics to Estimate Clinical Risk Reductions Achievable by Inhibiting Factor XI. Stroke. 2019;50:3004–3012. doi: 10.1161/strokeaha.119.026545
    OpenUrlCrossRef
  37. 37.↵
    Greco A, Laudani C, Spagnolo M, Agnello F, Faro DC, Finocchiaro S, Legnazzi M, Mauro MS, Mazzone PM, Occhipinti G, et al. Pharmacology and Clinical Development of Factor XI Inhibitors. Circulation. 2023;147:897–913. doi: 10.1161/circulationaha.122.062353
    OpenUrlCrossRef
  38. 38.
    Kluge KE, Seljeflot I, Arnesen H, Jensen T, Halvorsen S, Helseth R. Coagulation factors XI and XII as possible targets for anticoagulant therapy. Thrombosis research. 2022;214:53–62. doi: 10.1016/j.thromres.2022.04.013
    OpenUrlCrossRef
  39. 39.↵
    Sharma M, Molina CA, Toyoda K, Bereczki D, Kasner SE, Lutsep HL, Tsivgoulis G, Ntaios G, Czlonkowska A, Shuaib A, et al. Rationale and design of the AXIOMATIC-SSP phase II trial: Antithrombotic treatment with factor XIa inhibition to Optimize Management of Acute Thromboembolic events for Secondary Stroke Prevention. Journal of stroke and cerebrovascular diseases : the official journal of National Stroke Association. 2022;31:106742. doi: 10.1016/j.jstrokecerebrovasdis.2022.106742
    OpenUrlCrossRef
  40. 40.↵
    Li JY, Paragas N, Ned RM, Qiu A, Viltard M, Leete T, Drexler IR, Chen X, Sanna-Cherchi S, Mohammed F, et al. SCARA5 is a ferritin receptor mediating non-transferrin iron delivery. Developmental cell. 2009;16:35–46. doi: 10.1016/j.devcel.2008.12.002
    OpenUrlCrossRefPubMed
  41. 41.↵
    Gill D, Monori G, Tzoulaki I, Dehghan A. Iron Status and Risk of Stroke. Stroke. 2018;49:2815–2821. doi: 10.1161/strokeaha.118.022701
    OpenUrlCrossRefPubMed
  42. 42.↵
    Li JZ, Qiu YM, Zhang CL, Wang HL, Bi RT, Wei YH, Li YA, Hu B. The role of protein glycosylation in the occurrence and outcome of acute ischemic stroke. Pharmacological Research. 2023;191. doi: 10.1016/j.phrs.2023.106726
    OpenUrlCrossRef
  43. 43.↵
    Qian QF, Li Y, Fu JJ, Leng DW, Dong Z, Shi JJ, Shi HJ, Cao DW, Cheng X, Hu YF, et al. Switch-associated protein 70 protects against nonalcoholic fatty liver disease through suppression of TAK1. Hepatology. 2022;75:1507–1522. doi: 10.1002/hep.32213
    OpenUrlCrossRef
  44. 44.↵
    Qian QF, Hu FJ, Yu WJ, Leng DW, Li Y, Shi HJ, Deng DW, Ding KH, Liang C, Liu JP. SWAP70 Overexpression Protects Against Pathological Cardiac Hypertrophy in a TAK1-Dependent Manner. Journal of the American Heart Association. 2023;12. doi: 10.1161/jaha.122.028628
    OpenUrlCrossRef
  45. 45.↵
    Baranov MV, Revelo NH, Dingjan I, Maraspini R, ter Beest M, Honigmann A, van den Bogaart G. SWAP70 Organizes the Actin Cytoskeleton and Is Essential for Phagocytosis. Cell Reports. 2016;17:1518–1531. doi: 10.1016/j.celrep.2016.10.021
    OpenUrlCrossRefPubMed
  46. 46.↵
    Davis CM, Lyon-Scott K, Varlamov EV, Zhang WH, Alkayed NJ. Role of Endothelial STAT3 in Cerebrovascular Function and Protection from Ischemic Brain Injury. International Journal of Molecular Sciences. 2022;23. doi: 10.3390/ijms232012167
    OpenUrlCrossRef
  47. 47.↵
    Nie Z, Cheng DT, Pan CL, Wei ZM, Wang CY, Wang CY. SH3BGRL3, transcribed by STAT3, facilitates glioblastoma tumorigenesis by activating STAT3 signaling. Biochemical and Biophysical Research Communications. 2021;556:114–120. doi: 10.1016/j.bbrc.2021.03.165
    OpenUrlCrossRef
  48. 48.↵
    Sun BB, Maranville JC, Peters JE, Stacey D, Staley JR, Blackshaw J, Burgess S, Jiang T, Paige E, Surendran P, et al. Genomic atlas of the human plasma proteome. Nature. 2018;558:73-+. doi: 10.1038/s41586-018-0175-2
    OpenUrlCrossRefPubMed
Back to top
Posted August 31, 2023.
Download PDF
Data/Code
Email
Identification of Potential Therapeutic targets for Stroke and Its Subtypes by Integrating Proteomes and Genetics from Human Plasma
Hanchen Liu, Xiaoxi Zhang, Hongyu Ma, Thanh N. Nguyen, Weilong Hua, Shaojun Mo, Qinghai Huang, Jianmin Liu, Yu Zhou, Pengfei Yang
medRxiv 2023.08.29.23294808; doi: https://doi.org/10.1101/2023.08.29.23294808
Twitter logo Facebook logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area