Proteogenomics in cerebrospinal fluid and plasma reveals new biological fingerprint of cerebral small vessel disease

Stéphanie Debette, Ilana Caro, Daniel Western, Shinichi Namba, Na Sun, Shuji Kawaguchi, Yunye He, Masashi Fujita, Gennady V. Roshchupkin, Tim D’Aoust, Marie‐Gabrielle Duperron, Muralidharan Sargurupremraj, Ami Tsuchida, Masaru Koido, Marziehsadat Ahmadi, Chengran Yang, Jigyasha Timsina, Laura Ibáñez, Koichi Matsuda, Yutaka Suzuki, Yoshiya Oda, Akinori Kanai, Pouria Jandaghi, Hans Markus Münter, Daniel Auld, Iana Astafeva, Raquel Puerta, Jerome I. Rotter, Bruce M. Psaty, Joshua C. Bis, W.t. Longstreth, Thierry Couffinhal, Pablo García‐González, Vanesa Pytel, Marta Marquié, Amanda Cano, Merçé Boada, Marc Joliot, Mark Lathrop, Quentin Le Grand, Lenore J. Launer, Joanna M. Wardlaw, Myriam Heiman, Agustı́n Ruiz, Paul M. Matthews, Sudha Seshadri, Myriam Fornage, Hieab H.h. Adams, Aniket Mishra, David‐Alexandre Trégouët, Yukinori Okada, Manolis Kellis, Philip L. De Jager, Christophe Tzourio, Yoichiro Kamatani, Fumihiko Matsuda, Carlos CruchagaResearch Square 2024 · DOI

Introduction

Cerebral small vessel disease (cSVD) represents a significant global health challenge, acting as a primary driver of both stroke and vascular dementia. Characterized by damage to the brain's microvasculature, cSVD often manifests on MRI as white matter hyperintensities (WMH) or enlarged perivascular spaces (PVS). Despite its prevalence and devastating clinical impact, the underlying molecular mechanisms remain poorly understood, and no specific mechanism-based treatments currently exist. This study aims to bridge the gap between genetic risk and clinical outcomes by identifying the specific proteins and biological pathways that drive cSVD progression.

Study Design and Methodology

To uncover the causal drivers of cSVD, researchers implemented a comprehensive proteogenomic framework. The core of this approach was Two-Sample Mendelian Randomization (2SMR), which utilizes genetic variants—specifically protein quantitative trait loci (pQTLs)—as instrumental variables to estimate the causal effect of protein levels on disease markers. This method helps circumvent common issues like reverse causation, where the disease state itself alters protein expression.

Figure 1: Overview of the proteogenomic study design integrating CSF and plasma pQTLs with cSVD GWAS data.

The study leveraged a unique resource of pQTLs from both cerebrospinal fluid (CSF, N=3,107) and blood plasma (N=35,559). By comparing these two fluids, the team could distinguish between systemic signals and those localized to the central nervous system. These pQTLs were integrated with large-scale genome-wide association studies (GWAS) for MRI markers of cSVD, including WMH (N=48,454) and PVS (N=40,095). Findings were further validated across different proteomic platforms (Somascan and Olink), diverse ancestries (European and East Asian), and across the human lifespan.

A New Biological Fingerprint for cSVD

The analysis identified a "biological fingerprint" consisting of 49 proteins causally associated with cSVD markers. A key finding was the "CSF advantage": the study identified five times more protein-cSVD associations in the CSF than in plasma, despite the CSF sample size being significantly smaller. This underscores the importance of analyzing the fluid that directly bathes the brain to capture neurological disease signals.

Of the 49 proteins, 16 were robustly associated with cSVD in both CSF and plasma, while 24 were identified exclusively in the CSF. Pathway enrichment analysis revealed that these proteins are predominantly involved in the extracellular matrix (the "matrisome" or cellular scaffolding) and immune response pathways. Integration with single-nucleus RNA sequencing (snRNAseq) data from post-mortem brain samples further localized these signals, showing that many cSVD-associated genes are highly expressed in microglia—the brain's resident immune cells.

Clinical Relevance and Early Detection

The clinical implications of these findings are profound. Approximately half of the identified cSVD-associated proteins were also linked to clinical risks of stroke, Alzheimer’s disease, or both. Furthermore, the study investigated the timing of these biological changes. Remarkably, immune-related proteins were associated with MRI markers of vessel damage in individuals as young as 20 years old. This suggests that the biological processes leading to late-life cSVD and dementia are active decades before clinical symptoms appear, opening a window for very early intervention.

Therapeutic Opportunities

One of the most promising aspects of this research is the identification of seven cSVD-associated proteins that are already targets for existing drugs. These include proteins like EPO, PCSK9, and TFPI. For example, PCSK9 inhibitors, which are currently used to manage cholesterol, showed a potential protective effect against certain cSVD markers. This discovery provides a shortlist of candidates for drug repurposing, potentially accelerating the development of treatments for a condition that currently has none.

Conclusion and Future Directions

This study provides the first comprehensive proteogenomic signature of cSVD, highlighting the critical roles of the extracellular matrix and microglial inflammation. While the use of different proteomics platforms presents some technical challenges due to protein isoforms, the identified proteins represent a goldmine for both biomarker development and therapeutic targeting. Future research will focus on validating these drug targets in clinical trials and refining the use of protein-based Genetic Risk Scores (GRS) to identify individuals at high risk for severe cSVD.

Summary updated

Abstract

Cerebral small vessel disease (cSVD) is a leading cause of stroke and dementia with no specific mechanism-based treatment. We used Mendelian randomization to combine a unique cerebrospinal fluid (CSF) and plasma pQTL resource with the latest European-ancestry GWAS of MRI-markers of cSVD (white matter hyperintensities, perivascular spaces). We describe a new biological fingerprint of 49 protein-cSVD associations, predominantly in the CSF. We implemented a multipronged follow-up, across fluids, platforms, and ancestries (Europeans and East-Asian), including testing associations of direct plasma protein measurements with MRI-cSVD. We highlight 16 proteins robustly associated in both CSF and plasma, with 24/4 proteins identified in CSF/plasma only. cSVD-proteins were enriched in extracellular matrix and immune response pathways, and in genes enriched in microglia and specific microglial states (integration with single-nucleus RNA sequencing). Immune-related proteins were associated with MRI-cSVD already at age twenty. Half of cSVD-proteins were associated with stroke, dementia, or both, and seven cSVD-proteins are targets for known drugs (used for other indications in directions compatible with beneficial therapeutic effects. This first cSVD proteogenomic signature opens new avenues for biomarker and therapeutic developments.
BibTeX
@inproceedings{debette2024proteogenomics,
  author = {Stéphanie Debette and Ilana Caro and Daniel Western and Shinichi Namba and Na Sun and Shuji Kawaguchi and Yunye He and Masashi Fujita and Gennady V. Roshchupkin and Tim D’Aoust and Marie‐Gabrielle Duperron and Muralidharan Sargurupremraj and Ami Tsuchida and Masaru Koido and Marziehsadat Ahmadi and Chengran Yang and Jigyasha Timsina and Laura Ibáñez and Koichi Matsuda and Yutaka Suzuki and Yoshiya Oda and Akinori Kanai and Pouria Jandaghi and Hans Markus Münter and Daniel Auld and Iana Astafeva and Raquel Puerta and Jerome I. Rotter and Bruce M. Psaty and Joshua C. Bis and W.T. Longstreth and Thierry Couffinhal and Pablo García‐González and Vanesa Pytel and Marta Marquié and Amanda Cano and Merçé Boada and Marc Joliot and Mark Lathrop and Quentin Le Grand and Lenore J. Launer and Joanna M. Wardlaw and Myriam Heiman and Agustı́n Ruiz and Paul M. Matthews and Sudha Seshadri and Myriam Fornage and Hieab H.H. Adams and Aniket Mishra and David‐Alexandre Trégouët and Yukinori Okada and Manolis Kellis and Philip L. De Jager and Christophe Tzourio and Yoichiro Kamatani and Fumihiko Matsuda and Carlos Cruchaga},
  title = {{Proteogenomics in cerebrospinal fluid and plasma reveals new biological fingerprint of cerebral small vessel disease}},
  booktitle = {Research Square},
  year = {2024},
  doi = {10.21203/rs.3.rs-4535534/v1},
}
Figure