Differential Secretome Profiling of Human Osteoarthritic Synoviocytes Treated with Biotechnological Unsulfated and Marine Sulfated Chondroitins

Abstract

Symptomatic slow-acting drugs (SYSADOA) are increasingly used as effective therapies for osteoarthritis, representing an attractive alternative to analgesics or non-steroidal anti-inflammatory drugs to relieve disease symptoms. Pharmaceutical preparations of chondroitin sulfate, derived from animal sources, alone or in combination with glucosamine sulfate, are widely recognized for their beneficial effect on osteoarthritis treatment. A growing interest has also been devoted to understanding the molecular mechanisms modulated by SYSADOA using -omic strategies, most of which rely on chondrocytes as a model system. In this work, by using an integrated strategy based on unbiased proteomics and targeted cytokine profiling by a multiplexed protein array, we identified differences in the secretomes of human osteoarthritic synoviocytes in response to biotechnological unsulfated, and marine sulfated chondroitins treatments. The combined strategy allowed the identification of candidate proteins showing both common and distinct regulation responses to the two treatments of chondroitins. These molecules, mainly belonging to ECM proteins, enzymes, enzymatic inhibitors and cytokines, are potentially correlated to treatment outcomes. Overall, the present results provide an integrated overview of protein changes in human osteoarthritic synoviocytes secretome associated to different chondroitin treatments, thus improving current knowledge of the biochemical effects driven by these drugs potentially involved in pathways associated to osteoarthritis pathogenesis.

Keywords: biotechnological unsulfated chondroitin; marine chondroitin; mass spectrometry.; osteoarthritis; secretome.

Figure 1

Flowchart of the workflow applied for secretome analysis of human OA synoviocytes treated with a marine CS and an unsulfated biotechnological (BC) chondroitin: (a) Workflow used for the high-resolution LC-MS/MS analyses of CS-/BC-treated synoviocytes secretomes. (b) Schematic analysis pipeline showing main sequential steps for LC-MS/MS data processing and bioinformatic analyses performed on filtered proteomic data.

Figure 2

Overview of secretory pathway prediction results for differentially expressed proteins in the OA synoviocytes secretome following BC and CS treatments: (a) Pie chart representing the secretory pathway prediction for differentially expressed proteins by SignalP (classical secretion), SecretomeP (non-classical secretion) and TMHMM (transmembrane) servers. (b) Bar chart of the significantly enriched cellular components GO terms obtained for differentially expressed proteins in CS-/BC-treated synoviocytes secretomes. The number of proteins belonging to the enriched gene ontology term and the p-values are reported on the graph. (c) Venn diagram showing the overlapping of proteins secreted via classical/non-classical secretion pathways and by exosomes. (d) Details of genes within each subset of the Venn diagram reported in (c).

Figure 3

Dot plot of the top 10 over-represented biological processes of differentially expressed proteins in secretomes of CS-/BC-treated vs. pCTR synoviocytes. The dot size is proportional to the number of differentially expressed proteins associated with the process and the dot colour gradients indicated the significance of the enrichment (-log10(FDR corrected p-values)).

Figure 4

Heatmap representing the log2 fold-change values of differentially expressed proteins in CS-/BC-treated vs. pCTR synoviocytes secretomes. Down-regulated and up-regulated proteins are coloured in green and red, respectively. Missing expression values for undetected proteins in one out of the two conditions are reported in grey.

Figure 5

Bipartite view of the networks of enriched “molecular function” GO categories for the differentially expressed proteins in secretomes of (a) CS-treated vs. pCTR and (b) BC-treated vs. pCTR synoviocytes. The size of the node corresponds to the number of proteins included in the list and mapped on the specific term. The colour scale of the nodes is reported according to the p-value of the enriched category. The smaller nodes, corresponding to individual proteins/genes, are depicted in red (up-regulation) and green (down-regulation) based on their log 2 fold-change values. No change nodes are in light blue.

Figure 6

(a) Cytoscape network of enriched “biological processes” GO categories for proteins differentially expressed in both CS- and BC-treated with respect to pCTR synoviocytes secretomes identified by high-resolution LC-MS/MS. Up- and down-regulated proteins are reported as red and green nodes, respectively. Node size and edge thickness are related to number of peptides identified by MS and interaction confidence scores, respectively. Network clustering was performed according to the top enriched terms from the Cytoscape stringApp functional enrichment analysis (FDR-corrected p-value < 0.05) after redundancy filtering; (b) Validation of FBN1 up-regulation by Western blot analysis in CS- and BC-treated OA synoviocytes. β-Actin was used as the loading control. Relative changes of treated samples versus pCTR are given as mean + SD of three independent analyses, * p < 0.05.

Figure 7

Expression levels of significantly differentially modulated cytokines in CS-/BC-treated with respect to pCTR synoviocytes secretomes. Measured concentrations are referred to CM collected from 10 × 10⁴ cells for all conditions. CM were simultaneously screened for determining the cytokines concentration by interpolation on standard curves. All measurements were performed in triplicate. Data are reported as means ± SD. (* p ≤ 0.05; ** p ≤ 0.01; *** p ≤ 0.001; **** p ≤ 0.0001).