Analysis and comparison of cerebrospinal fluid and serum proteomics in schizophrenia, multiple sclerosis, and clinical controls

Introduction

Multiple sclerosis (MS) is considered an autoimmune disease of the central nervous system, whereas schizophrenia (SCZ) is a psychiatric disorder whose underlying pathophysiological mechanisms remain poorly understood. Emerging evidence suggests that immune processes, inflammation, demyelination, and synaptic disturbances - key features of MS pathogenesis - may also contribute to the pathophysiology of SCZ. Notably, treatment-resistant schizophrenia (TRS) has been associated with increased immune-related processes and synaptic dysfunction compared with non-treatment-resistant SCZ, suggesting that these mechanisms may be particularly relevant in this subgroup[1].To investigate shared and distinct proteomic patterns, we aim to identify an MS-associated protein signature compared with clinical controls (CC) and subsequently compare this signature with proteomic alterations observed in SCZ overall, as well as in TRS and non-TRS subgroups. We hypothesize that the MS-specific protein signature partially overlaps with SCZ-associated proteomic alterations, with a potentially stronger overlap in TRS, whereas non-overlapping components may reflect disease-specific proteomic patterns distinguishing MS from SCZ and its subgroups.

Methods

This cross-sectional study analyzes 172 cerebrospinal fluid (CSF) and 172 serum samples collected between 2001 and 2025 during routine clinical practice at three clinical centers. The CSF cohort consists of 52 MS, 39 CC, and 81 SCZ samples, whereas the serum cohort comprises 40 MS, 39 CC, and 93 SCZ samples. Among the SCZ samples included in this study, 18 of the 81 CSF samples and 18 of the 93 serum samples serve as bridging samples to a previously generated Olink Explore HT SCZ dataset, enabling dataset integration and increasing the overall sample size.

Proteomic profiling was performed using Olink Explore HT, a highly sensitive multiplex proximity extension assay capable of quantifying up to 5,416 proteins from small sample volumes.

Planned analyses include univariate differential expression analysis, gene set enrichment analysis, and support vector machine modelling.

Results/Planned analyses

Univariate differential expression analysis will first be used to identify disease-associated protein signatures for MS compared with clinical controls and for SCZ compared with clinical controls. In addition, a direct comparison between MS and SCZ overall, as well as TRS and non-TRS subgroups, will be performed to determine disease-specific protein patterns and to identify overlapping and non-overlapping proteins. Next, gene set enrichment analysis will be applied to characterize biological pathways associated with overlapping and disease-specific protein patterns, with a particular focus on immune, inflammatory, myelin-related, and synaptic processes. Finally, support vector machine analyses will be used exploratively to assess whether the identified protein signatures can distinguish between MS, SCZ overall, TRS/non-TRS subgroups, and CC.

Conclusions

As this project is currently ongoing, results are not yet available at this stage. However, the analyses will be completed by the time of the ECNP Congress, allowing the results to be presented as part of the poster. This study may provide insight into shared and distinct proteomic patterns in MS and SCZ, with particular relevance to immune-related alterations in TRS.

References

Fischer, L.E., Gottschalk, M.G., Reitz, L.U.G., Moussiopoulou, J., Koutsouleris, N., Yakimov, V., Boudriot, E., Seeburger, S., Jiang, T., Weiske, J., Petrera, A., Maccarone, G., CDP Working Group, Andressoo, J.-O., Schmitt, A., Keeser, D., Hasan, A., Uhr, M., Falkai, P., Papiol, S., Wagner, E., Raabe, F.J., 2025. Cerebrospinal fluid proteomics reveals synaptic and immune dysregulation in treatment-resistant schizophrenia. medRxiv, preprint, http://dx.doi.org/10.1101/2025.05.14.25327624