Comparative Evaluation of Adaptive Focused Acoustics‐ and SP3‐Based Workflows for Deep Proteomic Profiling of 30‐Year‐Old Formalin‐Fixed Paraffin‐Embedded Tissues

Formalin‐fixed paraffin‐embedded (FFPE) tissues represent a vast archive for translational research that offers access to specimens with clinical outcomes. However, formalin‐induced crosslinking hinders efficient protein extraction and limits proteomic applications. We systematically compared three protein extraction workflows, adaptive focused acoustics (AFA) with Covaris buffer (AFA/Covaris‐b), AFA with SDS buffer (AFA/SDS‐b), and SDS‐based lysis followed by SP3 clean‐up and digestion (SP3/SDS‐b), using hepatocellular carcinoma (HCC) tissues archived for over 30 years. Benchmarking was performed on two HCC cases using three serial tumor sections each, followed by protein quantification using data‐independent acquisition mass spectrometry on a ZenoTOF 7600+ instrument. All workflows identified 5206 proteins, with similar mean identifications (AFA/Covaris‐b: 4068; AFA/SDS‐b: 4134; SP3/SDS‐b: 4048), but distinct reproducibility (CVs: 11.9%, 12.2%, and 20.8%, respectively), leading to the exclusion of SP3. Method selection across eight HCC cases, including paired tumor and adjacent non‐tumorous tissues, revealed a total of 6018 proteins, with mean identifications of 3876 (AFA/Covaris‐b) and 3852 (AFA/SDS‐b). Both AFA workflows showed high reproducibility, with slightly lower variability for AFA/Covaris‐b (CVs: HCC: 13.5%, non‐tumorous liver tissue (NTL): 11.0%) compared with AFA/SDS‐b (CVs: HCC: 14.8%, NTL: 11.3%). While upregulated pathways were not consistently detected across workflows, downregulated pathways showed greater concordance between methods. These findings should be interpreted cautiously given the lack of appropriate temporal or fresh‐frozen controls. Collectively, AFA‐based workflows enable reproducible proteomic profiling of long‐archived FFPE tissues and provide a practical approach for methodological evaluation in retrospective proteomics studies.

Significance of the study

FFPE tissues are one of the most valuable clinically annotated biospecimen resources worldwide. However, their routine use in proteomics is hindered by extensive protein cross‐linking, particularly in long‐term archived blocks. Our study addresses this challenge by systematically evaluating and validating protein extraction strategies for FFPE hepatocellular carcinoma tissues stored for more than 30 years. By directly benchmarking AFA‐based and SP3 workflows, we show that AFA protocols, particularly those with the Covaris buffer, provide superior reproducibility while maintaining broad proteome coverage. Importantly, we showed that archival samples yielded more than 6,000 quantifiable proteins, with reproducible pathway‐level patterns observed across methods, although interpretation of biological signals remains limited by the absence of appropriate controls and potential extraction‐related biases. These findings suggest that long‐preserved FFPE specimens can serve as useful resources for modern data‐independent acquisition mass spectrometry. The ability to extract reproducible proteomic information from decades‐old FFPE blocks has important implications; it enables retrospective biomarker discovery, integration with genomic data, and validation of therapeutic targets in richly annotated patient cohorts. By providing a reproducible workflow, this study supports the expanded use of archived FFPE tissues and facilitates future efforts for large‐scale, longitudinal cancer proteomics.