In a recent study published in the journal Stroke, researchers identify genetic and molecular risk factors for subsequent cardiovascular outcomes after incident stroke in an effort to identify potential therapeutic targets to improve patient prognoses.
Study: Protein Identification for Stroke Progression via Mendelian Randomization in Million Veteran Program and UK Biobank. Image Credit: crystal light / Shutterstock.com
Identifying the causes of stroke
Stroke is a major global health issue that causes significant disability and mortality, particularly arterial ischemic stroke (AIS). AIS, which is a type of stroke caused by blocked blood flow to the brain, is responsible for up to 85% of stroke cases. AIS arises due to cerebral blood vessel blockage, with modifiable risk factors including hypertension, diabetes, dyslipidemia, atrial fibrillation, obesity, and lifestyle behaviors.
Although genome-wide association studies (GWAS) often focus on incident strokes, studying subsequent events can provide new insights into stroke progression. Further research is crucial to identify genetic and molecular risk factors, establish novel therapeutic targets, and improve prognosis after an initial stroke.
About the study
Data for the current study were obtained from the United Kingdom Biobank (UKB), which includes over 500,000 participants between 40 and 69 years of age, and the Million Veteran Program (MVP), which comprises over 850,000 participants, 8% of whom are women with an average age of 61.9 years. Incident stroke was defined using hospital-linked data for AIS or transient ischemic attack.
Data were standardized to correct collider bias for subsequent stroke, and Slope-Hunter was used. GWAS results for subsequent strokes were compared to incident strokes using replication performance methods.
Multiancestry meta-analyses were conducted, including European-only and across all ancestries from MVP and UKB. Tissue expression analysis was performed using annotation of GWAS and functional mapping.
Mendelian randomization (MR) was used to identify causal relationships between protein abundance, subsequent AIS, and major adverse cardiovascular events (MACE) using protein quantitative trait loci (pQTL) data from the UKB Pharma Proteomics Project. MR estimates were calculated using the Wald ratio method.
Colocalization analysis determined shared genetic factors between traits using the coloc package. Collider bias was evaluated by checking single-nucleotide polymorphism (SNP) associations with stroke incidence and running MR on uncorrected and Slope-Hunter adjusted GWAS results. Significant SNPs and proteins identified from the MR analysis were cross-referenced with known druggable targets.
Study findings
After exclusions based on ancestry and relatedness, 93,422 individuals with incident stroke from the UKB and MVP were included in the study. Taken together, the study cohort comprised 51,929 cases of subsequent major adverse cardiovascular events (MACE) and 45,120 cases of subsequent AIS. Stroke cases were more likely to be older, male, smoke, have hypertension or type 2 diabetes, as well as use antihypertensive and lipid-lowering medications as compared to those without AIS.
GWAS revealed no significant associations in the multi-ancestry meta-analysis for subsequent AIS or MACE. However, two significant genetic variants were identified in specific ancestry analyses, including rs76472767 near the ring finger protein 220 (RNF220) gene in African ancestry for subsequent MACE and rs13294166 near the long intergenic non-protein coding ribonucleic acid (RNA) 1492 (LINC01492) gene in African ancestry for subsequent AIS. These variants were not associated with incident AIS, thus suggesting that the Slope-Hunter correction for collider bias may not have been necessary.
Replication analysis indicated that genetic factors for incident stroke did not fully replicate in subsequent strokes. Of the 91 SNPs associated with incident stroke, 77 replicated in the incidence GWAS, whereas 33 replicated in the subsequent MACE GWAS. This observation suggests distinct genetic etiologies for incident stroke and subsequent MACE.
MR against pQTL data identified six proteins with putative causal effects on incident AIS. These proteins included cystatin E/M (CST6), fibroblast growth factor 5 (FGF5), G-protein-coupled receptor kinase 5 (GPRK5), furin, paired basic amino acid cleaving enzyme (FURIN), matrix metalloproteinase 12 (MMP12), and scavenger receptor class A member 5 (SCARA5). However, none of these proteins were associated with any effects on subsequent MACE.
C-C motif chemokine ligand 27 (CCL27) and tumor necrosis factor receptor superfamily member 14 (TNFRSF14) were associated with causal effects on subsequent MACE. Whereas higher levels of CCL27 were protective, higher levels of TNFRSF14 increased the risk of this condition. Colocalization analysis provided moderate and strong evidence for CCL27 and TNFRSF14, respectively.
Verification using independent pQTL data sets confirmed MR results for five of the nine significant proteins. MR results for CCL27 and TNFRSF14 did not vary significantly across different ancestries; however, the sample sizes for Hispanic and African subgroups were small.
Comparison against potential druggable targets revealed no clinical trials for TNFRSF14 and CCL27. Angiopoietin 1 (ANGPT1), FGF5, furin, MMP12, and tissue factor pathway inhibitor (TFPI), all of which were related to ischemic stroke, were identified as putatively causal in MR for incident stroke.
Conclusions
The current study suggests that CCL27 and TNFRSF14 likely affect stroke progression. Whereas TNFRSF14 influences immune cell survival and plaque destabilization, CCL27 maintains immune homeostasis.
Genetic variants associated with subsequent MACE and AIS provide new insights into stroke progression, which are distinct from incident stroke factors. Some of the different proteins identified as therapeutic targets for incident AIS, including ANGPT1, MMP12, FGF5, furin, and TFPI, are not associated with subsequent MACE, thus suggesting different targets for incident and subsequent strokes.
Journal reference:
- Elmore, A. R., Adhikari, N., Hartley, A. E., et al. (2024). Protein Identification for Stroke Progression via Mendelian Randomization in Million Veteran Program and UK Biobank. Stroke. doi:10.1161/STROKEAHA.124.047103