Promising therapeutic targets for ischemic stroke identified from plasma and cerebrospinal fluid proteomes: a multicenter Mendelian randomization study

Abstract

Background: Ischemic stroke (IS) is more common every year, the condition is serious, and have a poor prognosis. New, efficient, and safe therapeutic targets are desperately needed as early treatment especially prevention and reperfusion is the key to lowering the occurrence of poorer prognosis. Generally circulating proteins are attractive therapeutic targets, this study aims to identify potential pharmacological targets among plasma and cerebrospinal fluid (CSF) proteins for the prevention and treatment of IS using a multicenter Mendelian randomization (MR) approach.

Methods: First, the genetic instruments of 734 plasma and 151 CSF proteins were assessed for causative connections with IS from MEGASTROKE consortium by MR to identify prospective therapeutic targets. Then, for additional validation, plasma proteins from the deCODE consortium and the Fenland consortium, as well as IS GWAS data from the FinnGen cohort, the ISGC consortium and UK biobank, were employed. A thorough evaluation of the aforementioned possible pharmacological targets was carried out using meta-analysis. The robustness of MR results was then confirmed through sensitivity analysis using several techniques, such as bidirectional MR analysis, Steiger filtering, and Bayesian colocalization. Finally, methods like Protein-Protein Interaction (PPI) Networking were utilized to investigate the relationship between putative drug targets and therapeutic agents.

Results: The authors discovered three proteins that may function as promising therapeutic targets for IS and meet the Bonferroni correction ( P <0.05/885=5.65×10 -5 ). Prekallikrein (OR=0.41, 95% CI: 0.27-0.63, P =3.61×10 -5 ), a protein found in CSF, has a 10-fold protective impact in IS, while the plasma proteins SWAP70 (OR=0.85, 95% CI: 0.80-0.91, P =1.64×10 -6 ) and MMP-12 (OR=0.92, 95% CI: 0.89-0.95, P =4.49×10 -6 ) of each SD play a protective role in IS. Prekallikrein, MMP-12, SWAP70 was replicated in the FinnGen cohort and ISGC database. MMP-12 (OR=0.93, 95% CI: 0.91-0.94, P <0.001), SWAP70 (OR=0.92, 95% CI: 0.90-0.94, P <0.001), and prekallikrein (OR=0.53, 95% CI: 0.33-0.72, P <0.001) may all be viable targets for IS, according to the combined meta-analysis results. Additionally, no evidence of reverse causality was identified, and Bayesian colocalization revealed MMP-12 (PPH 4 =0.995), SWAP70 (PPH 4 =0.987), and prekallikrein (PPH 4 =0.894) shared the same variant with IS, supporting the robustness of the aforementioned causation. Prekallikrein and MMP-12 were associated with the target protein of the current treatment of IS. Among them, Lanadelumab, a new drug whose target protein is a prekallikrein, may be a promising new drug for the treatment of IS.

Conclusion: The prekallikrein, MMP-12, and SWAP70 are causally associated with the risk of IS. Moreover, MMP-12 and prekallikrein may be treated as promising therapeutic targets for medical intervention of IS.

Figure 1

Design of the study and its primary findings. CSF, cerebrospinal fluid; ISGC, International Stroke Genetics Consortium; PPI, Protein-Protein Interaction Networks.

Figure 2

Volcano plot of three proteins identified as potential targets in ischemic stroke from 734 plasma (A) and 151 CSF (B) proteins by Wald ratio or inverse variance weighted method. CSF, cerebrospinal fluid; PVE, proportion of variance explained.

Figure 3

Data from the Primary(MEGASTROKE consortium, A), ISGC consortium (B), FinnGen cohort (C) and UK biobank (D) externally validate the causal link between MMP-12, SWAP70, and prekallikrein and IS. CSF, cerebrospinal fluid. Primary(Exposure): 734 plasma proteins and 151 CSF proteins used in the initial phase. _Primary: IS GWAS from MEGASTROKE consortium. Pietzner et al.: Plasma proteins in the Fenland database. Ferkingstad et al.: Plasma proteins in the deCODE database. IS: ischemic stroke, Fenland: Plasma proteins in the Fenland database. deCODE: Plasma proteins in the deCODE database. ISGC: IS GWAS data from International Stroke Genetics Consortium, FinnGen: IS GWAS data from FinnGen consortium. UKB: IS GWAS data from UK Biobank. Arrows: 95% upper and lower CI outside the range of the horizontal axis.

Figure 4

Meta-analysis of the causal relationship between MMP-12 (A), SWAP70 (B), and prekallikrein (C) and IS risk in the initial phase and the replication phase (external validation). CSF, cerebrospinal fluid; IS, ischemic stroke. Primary_: 734 plasma proteins and 151 CSF proteins used in the initial phase, _Primary: IS GWAS data from the MEGASTROKE database, Fenland: Plasma proteins in the Fenland database, deCODE: Plasma proteins in the deCODE database, ISGC: IS GWAS data from the ISGC database, FinnGen: IS GWAS data from the FinnGen cohort, UKB: IS GWAS data from UK Biobank.

Figure 5

Regional association and matrix plots of colocalization analysis results. (A), (C), and (E) are the P-value distributions of the loci 100 KB above and below the pQTL for MMP-12, SWAP70, and Prekallikrein, respectively. (B), (D), and (F) are the P-value distributions of IS corresponding to the instrumental variables of MMP-12, SWAP70, and Prekallikrein. In (A–F), each point represents a SNP. Horizontal coordinates indicate the physical location on the chromosome, and the vertical axis indicates the -log P-value of GWAS of pQTL or IS.

Figure 6

Interactions between identified prospective medication targets and the present IS drug targets.