Proteomic analysis of genetically stratified grade-I meningioma

Abstract

Introduction: Meningioma is the most common primary intracranial tumour. Although often benign in nature, at least 20% have high rates of recurrence. Surgical resection and radiotherapy are the main therapeutic approach, yet tumour location can hamper complete resection and chemotherapies are ineffective. Therefore, more accurate prognostic biomarkers are needed to improve the clinical management of meningioma. The mutational profile of mutant meningioma is well defined, including in addition to most common NF2 loss, mutations of TRAF7, KLF4, and AKT1. The aim of this study is to identify novel biomarkers and therapeutic targets of genetically stratified meningioma by characterising the proteomic landscape.

Materials and methods: Frozen meningioma tissue samples were stratified according to mutational background: AKT1E17K/TRAF7, KLF4K409Q/TRAF7 and NF2-/-. For global proteomics, proteins were fractionated by SDS-PAGE followed by in-gel tryptic digestion and sample preparation for LCMS analysis. Raw mass spectrometry data files were analysed using MaxQuant (1.6.2.10) and Perseus software (1.6.1.3). Quantitative phosphoproteomics was performed using TMT 10plex labelling approach followed by kinase prediction using KinswingR. GO enrichment analyses were performed using DAVID against all human proteins. Proteomics data were also further analysed for pathway analysis by the Ingenuity® Pathway Analysis (IPA®) software. Followed by mutational subtype specific potential protein and kinase candidates were validated via Western blot and immunohistochemistry techniques.

Results: In total, 4162 proteins were quantified across all meningioma mutational subtypes and normal meninges (n=31). Hierarchical clustering analysis showed distinct proteomic profiles of meningioma mutational subtypes and corresponding clusters of differentially expressed proteins. Comparative analysis showed only 10 proteins were commonly significantly upregulated (log2 fold-change≥1; p<0.05) among all mutational subtypes vs. normal meninges which indicates mutational subtypes are very different from each other at their protein expression level. Mutational subtypes specific analysis identified 156 proteins to be commonly significantly upregulated (log2 fold-change≥1; p<0.05) in AKT1E17K/TRAF7, 14 proteins in KLF4K409Q/TRAF7, and 7 proteins in NF2-/- mutant meningioma subtypes compared to normal meninges. IPA and gene ontology analysis showed that AKT1E17K/TRAF7mutant subtype-specific upregulated proteins were highly enriched in oxidative phosphorylation and metabolic pathways, while KLF4K409Q/TRAF7 mutated meningioma-specific upregulated proteins were associated with extracellular matrix disassembly and focal adhesion GO terms. AKT1E17K/TRAF7 mutated meningioma-specific top candidate protein expressions including CLIC3, CRABP2, GMDS, and Pyruvate carboxylase were validated via WB and Simple wes and an increased expression of KLF4K409Q/TRAF7 specific protein Endoglin confirmed via both WB and IHC. NF2-/- mutated meningioma-specific one protein, Annexin-3 showed significant expression compared to other meningioma subtypes and NMT, and knocking down of this protein showed reduced Benmen-1 cell proliferation. Lastly, analysis of 6600 phosphosites (n=8) predicted regulatory kinases including PRKCA and EGFR, and the presence of an active form of PRKCA in benign meningioma was confirmed via both WB and IHC.

Conclusion: A proteomic approach is an effective tool to identify distinct protein expression in genetically distinct meningioma subgroups and has led to the identification of differentially expressed proteins and activated pathways in mutant subtypes. Further validation and functional verification of potential candidates will allow us to identify potential drug targets/biomarkers for benign meningiomas.