SILAC-Based Characterization of Plasma-Derived Extracellular Vesicles in Patients Undergoing Partial Hepatectomy

Abstract

Post-hepatectomy liver failure (PHLF) remains a significant risk for patients undergoing partial hepatectomy (PHx). Reliable prognostic markers and treatments to enhance liver regeneration are lacking. Plasma nanoparticles, including lipoproteins, exosomes, and extracellular vesicles (EVs), can reflect systemic and tissue-wide proteostasis and stress, potentially aiding liver regeneration. However, their role in PHLF is still unknown.

Methods: Our study included nine patients with hepatocellular carcinoma (HCC) undergoing PHx: three patients with PHLF, three patients undergoing the associating liver partition and portal vein ligation for staged hepatectomy (ALPPS) procedure, and three matched controls without complications after PHx. Patient plasma was collected before PHx as well as 1 and 5 days after. EVs were isolated by ultracentrifugation, and extracted proteins were subjected to quantitative mass spectrometry using a super-SILAC mix prepared from primary and cancer cell lines.

Results: We identified 2625 and quantified 2570 proteins in the EVs of PHx patients. Among these, 53 proteins were significantly upregulated and 32 were downregulated in patients with PHLF compared to those without PHLF. Furthermore, 110 proteins were upregulated and 78 were downregulated in PHLF patients compared to those undergoing ALPPS. The EV proteomic signature in PHLF indicates significant disruptions in protein translation, proteostasis, and intracellular vesicle biogenesis, as well as alterations in proteins involved in extracellular matrix (ECM) remodelling and the metabolic and cell cycle pathways, already present before PHx.

Conclusions: Longitudinal proteomic analysis of the EVs circulating in the plasma of human patients undergoing PHx uncovers proteomic signatures associated with PHLF, which reflect dying hepatocytes and endothelial cells and were already present before PHx.

Keywords: extracellular vesicles; liver regeneration; post-hepatectomy liver failure; proteomics.

Figure 1

Study design and workflow of sample processing for proteomic analysis of plasma EVs.

Figure 2

Qualitative proteome signatures of plasma EVs: (A) SDS-PAGE protein profiles, with numbers of identified proteins in LC-MS analysis denoted to individual samples (A = before PHx, PRE; B = 1 day after PHx; C = 5 days after PHx) and Venn diagrams showing common and unique proteins identified in respective sample groups (PHLF, noPHLF, ALPPS) after filtering for at least 2 valid values on basis of LFQ-L intensities. (B) GOCC categorization of all identified proteins. (C) Venn diagram of overlapping and “only” proteins identified in outcome groups PHLF, noPHLF, and ALPPS, irrespective of sampling time, after filtering for at least 3 valid values (LFQ-L intensities). Corresponding source data can be found in Supplementary Data S1. (D) Localization and functional categorization of proteins identified only in PHLF (225), or noPHLF and ALPPS (198), on basis of Uniprot-Keywords, KEGG, Reactome, and literature. * Category of cell death, encompassing apoptosis, necroptosis, and ferroptosis. (E) Top five enriched molecular signatures of proteins only identified in PHLF, noPHLF, or ALPPS, with −log10 p-Values on x-axis and numbers of proteins in each category denoted. Corresponding source data can be found in Supplementary Data S2.

Figure 3

Quantitative proteome signatures of plasma EVs: (A) Unsupervised PCA scaled to unit variance of proteins identified in all samples (296) with outcome groups indicated. (B) Analysis of workflow for statistical analysis of differentially expressed proteins (DEPs) on basis of SILAC ratios. (C,D) Volcano plots depicting DEPs considering conservative p-value cutoff (p < 0.05/−log10 p > 1.3) as significant. (E) Heatmap of commonly up- and downregulated proteins in pairwise comparisons of PHLF versus noPHLF and ALPPS. (F,G) Bubble plot depicting enriched pathway terms denoting PHLF EV proteins significantly (F) upregulated or (G) downregulated in comparison to both noPHLF and ALPPS. Source data can be found in Supplementary Data S3.

Figure 4

EV signatures in PHLF patients prior to surgery: (A) PCA of EV protein cargo before PHx in outcome groups PHLF, noPHLF, and ALPPS. (B,C) Heatmap summarizing enriched GO terms of up- or downregulated and “only” proteins in (B) PHLF versus noPHLF and (C) PHLF versus ALPPS. Source data can be found in Supplementary Data S4.