Fig 2: In silico prediction identifies TNC as a potentially novel GzmB substrate.Schematic diagram of predicted granzyme B (GzmB) cleavage sites on human TNC (hTNC). Letters represent the single–amino acid code of the tetrapeptide preceding as well as the dipeptide following the expected cleavage site. Arrows represent the location of the expected cleavage site (after aspartic acid, D) and correspond to the amino acid location referenced between brackets.

Fig 3: GzmB cleaves TNC substratum to impair fibroblast spreading and viability.(A) Schematic representation of the experimental procedure used to investigate how GzmB digestion of a TNC substratum impacts fibroblast behaviors. (B) Western blot analysis of the presence of soluble TNC fragments after 24-hour digestion of the immobilized TNC substratum as indicated in A. (C) Phase contrast microscopy performed on primary human dermal fibroblasts cultured for 72 hours onto noncoated (NC) wells, wells coated with nondigested (ND) rhTNC, or wells coated with rhTNC and digested with 50 nM rhGzmB as indicated in A. Bars, 400 μm. (D) Percentage of cell viability obtained by MTS assay performed on primary human dermal fibroblasts cultured for 72 hours onto NC wells, wells coated with ND rhTNC, or wells coated with rhTNC and digested with 50 nM rhGzmB as indicated in A. Results were normalized compared to the NC condition and are represented as means ± SD from 6 independent experiments. Statistical analyses were performed using nonparametric ANOVA (Kruskal-Wallis rank-sum test) followed by Dunn’s multiple comparisons test with Bonferroni’s P value adjustment. ***=P < 0.001. Images shown are representative of at least 3 independently repeated experiments.

Fig 4: TNC fragments detected in patients with RA align with in silico prediction and in vitro digestion by GzmB.GzmB (A) and TNC (B) quantification in healthy controls (HC, n = 5), rheumatoid arthritis (RA, n = 12), osteoarthritis (OA, n = 7), and inflammatory arthritis (IA, n = 6) patient SF represented as means ± SD. Statistical analyses were performed using nonparametric ANOVA (Kruskal-Wallis rank-sum test) followed by Dunn’s multiple comparisons test with Bonferroni’s P value adjustment. (C) Correlation (Spearman rank-order) between TNC and GzmB levels in HC, RA, OA, and IA patients’ SF (n = 30). (D) Western blot analysis demonstrating the presence of 2 TNC fragments (70 kDa and 30 kDa) in the SF of 12 patients with RA. *=P < 0.05, **=P < 0.01.

Fig 5: GzmB proteolysis of TNC releases its C-terminal FBG-like domain in vitro.GzmB-generated peptides from TNC were identified from the mass spectrometry data (available in Table 2) and mapped on the TNC protein. Intensity of the TNC peptide in the control, 24 hours’ digestion, or 48 hours’ digestion sample was plotted as line graphs.