Journal Article

Glioblastoma stem cells express non‐canonical proteins and exclusive mesenchymal‐like or non‐mesenchymal‐like protein signatures

Pernemalm, Maria, Babačić, Haris, Hellström, Mats, Lehtiö, Janne, Michienzi, Alessandro, Cardinali, Deborah, Ciafrè, Silvia Anna, Trevisi, Gianluca, Uhrbom, Lene, Pellegatta, Serena, Mangiola, Annunziato, Maturi, Nagaprathyusha, Umer, Husen M., Galardi, Silvia

2023

Overview

Glioblastoma (GBM) cancer stem cells (GSCs) contribute to GBM's origin, recurrence, and resistance to treatment. However, the understanding of how mRNA expression patterns of GBM subtypes are reflected at global proteome level in GSCs is limited. To characterize protein expression in GSCs, we performed in‐depth proteogenomic analysis of patient‐derived GSCs by RNA‐sequencing and mass‐spectrometry. We quantified > 10 000 proteins in two independent GSC panels and propose a GSC‐associated proteomic signature characterizing two distinct phenotypic conditions; one defined by proteins upregulated in proneural and classical GSCs (GPC‐like), and another by proteins upregulated in mesenchymal GSCs (GM‐like). The GM‐like protein set in GBM tissue was associated with necrosis, recurrence, and worse overall survival. Through proteogenomics, we discovered 252 non‐canonical peptides in the GSCs, i.e., protein sequences that are variant or derive from genome regions previously considered non‐protein‐coding, including variants of the heterogeneous ribonucleoproteins implicated in RNA splicing. In summary, GSCs express two protein sets that have an inverse association with clinical outcomes in GBM. The discovery of non‐canonical protein sequences questions existing gene models and pinpoints new protein targets for research in GBM. Cancer stem cells (GSCs) drive malignancy in glioblastoma. However, their molecular phenotype is not well understood. Here, we report proteomic profiling of GSCs and protein sets that separate two GSC types, which are differentially associated with overall survival in glioblastoma. Through proteogenomics, we detect matching mRNA and protein sequences mapping to gene regions previously considered non‐coding for proteins, including variants of HNRNPs.

• Publisher: John Wiley & Sons, Inc,John Wiley and Sons Inc,Wiley
• Subject: Brain cancer / Brain Neoplasms - metabolism / Cell culture / Cell Line, Tumor / Chromatography / Consortia / Epidermal growth factor / Ethics / Experiments / Gene expression / Genomes / Genomics / Glioblastoma / Glioblastoma - genetics / Glioblastoma - metabolism / Glioblastoma multiforme / glioblastoma stem cells / Humans / Mass spectrometry / mesenchymal / Mesenchyme / Neoplastic Stem Cells - metabolism / proneural / Proteins / proteogenomics / Proteomes / Proteomics / Ribonucleoproteins / RNA / RNA sequencing / Scientific imaging / signature / Stem cell research / Stem cells / Tumors