Perilymph metabolomic and proteomic MALDI-ToF profiling with porous silicon chips: A proof-of-concept study

Introduction: Sensorineural hearing losses (SNHLs) are a significant public health issue, and the hear-ing loss field is desperately in need of effective therapy. Pathophysiological mechanisms are not yet clearly understood in the absence of validated methods to assess the inner ear content. Proteomic and metabolomic analysis of perilymph is opening new research perspectives for SNHLs. We aimed to demon-strate the feasibility of an innovative mass spectrometry (MS) strategy using porous silicon chips (PSCs) to investigate the low molecular weight (LMW) protein and metabolite content of human perilymph. Our second objective was to stratify perilymph samples according to their MS profiles and compare these results with clinical data. Material and methods: Perilymph samples obtained during cochlear implant surgery from patients with SNHLs were retrieved from a validated biobank. To focus on LMW entities, we used a PSC enrichment protocol before MALDI-ToF MS analysis. PSCs were used as a LMW molecular preanalytical stabilizer and amplifier. Patients' clinical data and SNHL characteristics were retrieved retrospectively from med-ical charts. Results: We successfully acquired and compared 59 exploitable MS profiles out of 71 perilymph samples. There was a good correlation between duplicates. Comparing both ears from the same patient, we found good reproducibility even when there was a one-year interval between samplings. We identified three distinct groups when comparing the samples' metabolomic profiles and four homogeneous groups com-paring their LMW proteome profiles. Clinical data analysis suggested that some groups shared clinical or preanalytical characteristics. Conclusion: This proof-of-concept study confirms that LMW proteome and metabolome content of per-ilymph can be analyzed with PSCs. Based on protein profiles, we managed to stratify perilymp samples according to their molecular composition. These results must be confirmed with a larger population, and sampling methods require improvement, but this approach seems promising. In the future, this approach may pave the way for companion test strategies to precisely diagnose and define potential molecular targets for audioprotective therapies.(c) 2022 Elsevier B.V. All rights reserved.