Mikolajewicz, Nicholas

Abstract Cerebrospinal fluid (CSF) is the biofluid most proximal to the central nervous system (CNS), yet its proteomic organization across neurological disease remains unresolved. We assembled a meta-atlas of 35 CSF proteomic datasets (2016–2026) comprising 10,560 quality-controlled samples spanning neurodegenerative, neuroinflammatory, neuro-oncological, infectious and structural-injury contexts. Meta-analysis of cross-cohort detection defined a 747-protein core mass-spectrometry-accessible CSF proteome that generalizes to held-out cohorts. We derived thirteen consensus meta-programs that transfer across independent cohorts and span the major neural, barrier, immune and plasma compartments. Unexpectedly, these programs are dominated by a single axis of variation opposing complement- and plasma-associated inflammation against neuronal-synaptic homeostasis. This injury–homeostasis axis is diagnosis-spanning and tracks clinical severity and disease activity across diverse neurological conditions. We show that a two-protein C5–NRCAM model can approximate the injury–homeostasis axis, bridging deep proteomic profiling toward bedside assessment.

Abstract Introduction Leptomeningeal disease (LMD) is the seeding of tumor cells in the leptomeninges and the cerebrospinal fluid (CSF) and carries a poor prognosis of 2-4 months. CSF sampling is required to study LMD, but genomic analysis may not be able to identify non-genomic and post-translational modifications. Proteomics can address this gap but is limited by reproducibility and disease heterogeneity. Integrating proteomic with genomic data may prioritize driver genes that promote LMD. Methods 23 patients with breast cancer-associated LMD with matched CSF proteomic and genomic data were studied. Variant allele frequencies (VAFs) for genes TP53, NFE2L2, ERBB2 and PIK3CA were derived, and proteomic profiling was performed on matched CSF. Spearman correlations between driver-gene VAF and protein expression were computed, restricting to genes with VAF>0 in ≥ 3 samples, and Benjamini-Hochberg correction was applied across gene-protein pairs. Proteins significantly associated with VAF were analyzed using over representation analysis (ORA) and gene segment enrichment analysis (GSEA) with KEGG, Gene Ontology, and MSigDB. Results TP53 VAF was associated with upregulation of proteins involved in glycolysis and cell proliferation, with enrichment of hallmark glycolysis (FDR=2.6x10-8), mTORC1 signalling (FDR=8.5x10-8), and reactive oxygen species pathways (FDR=2.7x10-3), and downregulation of complement (FDR=6.8 x 10-6) and coagulation cascades (FDR=0.011). NFE2L2 VAF was associated with epithelial-mesenchymal transition by both GSEA (FDR=2.3x10-9) and ORA (FDR=4.6x10-3), TGF-β signalling (FDR=0.015) and glycoprotein metabolism (FDR=1.6x10-7). ERBB2 VAF was associated with an upregulation of cell-cycle regulation (FDR=6.4x10-10), mTORC1 pathway (FDR=3.2x10-4), and a downregulation of proteins involved in cell-cell adhesion (FDR=4.0x10-9), axon guidance (GSEA FDR=0.025), and developmental growth (GOBP FDR=3.4x10-4). Conclusion Distinct CSF proteomic signatures were associated with specific driver mutations, implicating metabolic and oxidative stress pathways (TP53), stromal remodelling (NFE2L2), and disruption of CNS structural pathways (ERBB2). Validation in prospective LMD cohorts is needed to confirm these mutation-specific signatures and inform diagnostic and therapeutic strategies.