Identification of Central Nervous System Oncologic Disease Biomarkers in EVs from Cerebrospinal Fluid (CSF) of Pediatric Patients: A Pilot Neuro-Proteomic Study

Abstract

Cerebrospinal fluid (CSF) is a biochemical–clinical window into the brain. Unfortunately, its wide dynamic range, low protein concentration, and small sample quantity significantly limit the possibility of using it routinely. Extraventricular drainage (EVD) of CSF allows us to solve quantitative problems and to study the biological role of extracellular vesicles (EVs). In this study, we implemented bioinformatic analysis of our previous data of EVD of CSF and its EVs obtained from congenital hydrocephalus with the aim of identifying a comprehensive list of potential tumor and non-tumor biomarkers of central nervous system diseases. Among all proteins identified, those enriched in EVs are associated with synapses, synaptosomes, and nervous system diseases including gliomas, embryonal tumors, and epilepsy. Among these EV-enriched proteins, given the broad consensus present in the recent scientific literature, we validated syntaxin-binding protein 1 (STXBP1) as a marker of malignancy in EVD of CSF and its EVs from patients with pilocytic astrocytoma and medulloblastoma. Our results show that STXBP1 is negatively enriched in EVs compared to non-tumor diseases and its downregulation correlates with adverse outcomes. Further experiments are needed to validate this and other EV markers in the blood of pediatric patients for translational medicine applications.

Proteomic Profiling of Cerebrospinal Fluid and Its Extracellular Vesicles from Extraventricular Drainage in Pediatric Pilocytic Astrocytoma, towards Precision Oncology

“…The expression profile of these proteins, after Z-score normalization, was visualized in the heatmap shown in Figure S3 . Notably, 136 of these overlap with proteins identified as promising biomarkers of brain tumors in our previous work [ 28 ] ( Table S2 ). Next, a T-test was applied to identify the proteins that distinguish between PA and HC, or PA and MB, or PA and non-PA patients for each fraction.…”

“…The CPXM2 protein was overexpressed in PA patients in all the comparisons, whereas AQP4 showed the opposite protein profile ( Figure 2 , Figures S4 and S5 ). Notably, of these 50 highlighted proteins, 12 (24%) overlap with core panel proteins identified as the most promising biomarkers of brain tumors in our previous work, including AQP4 [ 28 ] ( Table S2 ).…”

“…The role of EVs in intercellular communication and the transfer of bioactive molecules positions them as promising candidates for biomarker discovery in childhood PA. Recent proteomic data from the control EVD CSF showed that EVs are more informative than Tot CSF, as they express specific brain disease markers and therapeutic targets that cannot be detected in the Tot CSF, due to its considerable dynamic range, whereas the enriched EV fractions display a markedly lower variability [ 28 ]. Consistently, here 1447 (47.1%), 2105 (65.6%), and 2648 (86.2%) proteins were found expressed in the total, SEV, and LEV fractions, respectively.…”

“…The selection of the optimal sample size is a critical parameter in the experimental design of biomarker discovery studies. The number of controls and patients were determined a priori, considering the biological variability of cerebrospinal fluid (CSF) extraventricular drainage (EVD) samples obtained in our previous studies [ 28 ]. The design included 80% of power and alpha error ≤ 0.05 after correction for multiple interactions (Benjamini–Hochberg) using the equations and procedures suggested by Dell RB et al [ 29 ] and Forshed J [ 30 ].…”

“…Statistical analysis was performed as previously described [ 28 ], with some modifications. Briefly, after log2 conversion, identified proteins were filtered for 70% presence in at least one fraction.…”