Fig 2: Rcn3 facilitates TGFβ1 signalling by transcriptionally maintaining the expression of TGFBR1. A qPCR analyses of TGFBR1 and TGFBR2 in NHLF with Rcn3 knockdown or overexpression upon TGFβ1 exposure for 12 h. The qPCR data were normalised to the GAPDH content and analysed by the 2−∆∆Ct method relative to the vehicle Ctl-siRNA group (n = 6 independent biological replicates). The immunoblot assay indicates that Rcn3 knockdown by siRNA significantly inhibits the expression of TGFBR1 with intact TGFBR2 expression in NHLF upon TGFβ1 exposure for 12 h (B) and in DHLF-IPF (C). The ratios to tubulin expression are presented in dot graph relative to Ctl-siRNA (n = 3 independent biological replicates). D The immunoblot of cytoplasmic and membrane protein fractions from NHLF with Rcn3-siRNA and Ctl-siRNA. The β-tubulin expression and Na–K ATPase are as loading controls for cytoplasmic and membranous proteins, respectively. Cytoplasmatic-TGFBR1/tubulin and membrane-TGFBR1/Na–K ATPase ratios are presented in dot graph relative to Ctl-siRNA (n = 6 independent biological replicates). E The overexpression of Rcn3 causes marked increases in TGFBR1, collagen I, and α-SMA in NHLF. The ratios to tubulin expression are presented in dot graph relative to lenti-Ctl (n = 5 independent biological replicates). F Knockdown of Rcn3 failed to change TGFBR1 protein degradation in NHLF, as indicated by cycloheximide chase assay. (n = 3 independent experiments). Data are presented as the mean ± SD, and differences between two groups were analysed by unpaired student’s t-tests or Mann–Whitney U-tests as appropriate. *P < 0.05 versus Ctl-siRNA or lenti-Ctl. NHLF normal human lung fibroblast, DHLF-IPF disease human lung fibroblasts-idiopathic pulmonary fibrosis, Ctrl-siRNA control-siRNA, lenti-Rcn3 lentivirus-Rcn3, lenti-Ctl lentivirus-control

Fig 3: Schematic summary of the proposed profibrotic mechanisms, a positive feedback of TGFβ1 signalling, by which Rcn3 contributes to the pathogenesis of pulmonary fibrosis. Upon the activation of pulmonary injury-repair, TGFβ1 enhances Rcn3 expression in interstitial lung fibroblast through an ER stress signalling dependent manner; those increased Rcn3, in turn, detains EZH2 in the cytoplasm, leading to the decrease of H3K27me3 in TGFBR1 promoter region and the following enhancements of TGFBR1 expression and TGFβ1 signalling

Fig 4: TGFβ1 exposure promoted Rcn3-EZH2 interaction which restrained nuclear EZH2 level and decreased EZH2/H3K27me3 enrichment at the TGFBR1 promoter region. A A schematic diagram of BioID labelling strategy (up-left panel). The BioID-Rcn3 fusion protein was constructed by inserting Rcn3 (without signal peptide) into the C-terminal of BioID2 and the signal peptide fragment of Rcn3 into of N-terminal of BioID2; BioID-Rcn3 and BioID-only (bearing Rcn3 signal peptide in the N-terminal) vectors were transfected into NHLF. Upon TGFβ1 exposure, selective biotinylation of proximal proteins is followed by stringent cell lysis for streptavidin-affinity purification and identification by LC–MS/MS. By eliminating the common proteins in BioID-Rcn3, BioID-only and Biotin controls, 12 proteins were identified as specific potential interaction proteins of Rcn3, including EZH2 (right panel). The amino acid sequence of the EZH2 peptide fragment identified by MS was shown in a table (lower panel). B Co-IP validated Rcn3-EZH2 interaction in NHLF co-transfected with Flag-Rcn3, Myc-EZH2 or Flag-Rcn3 + Myc-EZH2. C Biolayer interferometry (BLI) studies by using recombination human Rcn3 and EZH2 protein indicated the direct binding between Rcn3 and EZH2 and the binding with a clear concentration gradient with an affinity (KD) of 1.54 µM. BLI binding representative of at least 3 independent traces. D The immunoblot of EZH2 in nuclear protein fraction from NHLF. The ratios to histone 3 expression were presented as dot graph relative to controls (n = 4, independent biological replicates). E Immunofluorescence staining of Rcn3 (green) and EZH2 (red) on the NHLF cells in response to TGFβ1 exposure and NHLF with Rcn3 overexpression. The DAPI (blue) is used for nuclear staining. TGFβ1 treatment and Rcn3 overexpression promote Rcn3-EZH2 interaction in the cytoplasm (the yellow dots in the merge views). Scale bar: 10 μm. F The immunoblot of H3K27me3. The H3K27me3/Histone 3 ratio was present as a dot graph relative to Ctl-siRNA or Lenti-Ctl (n = 3, independent biological replicates). G, H chromatin immunoprecipitation (ChIP) assay was used to analyse H3K27me3 in TGFBR1 promoter in NHLF with Rcn3-siRNA. Schematics of TGFBR1 and related four sets of ChIP qRT-PCR primers (up panel). qPCR analysis of TGFBR1 promoter regions conducted with 4 sets of specific primers (R1-R4) after a ChIP assay using anti-H3K27me3 antibody or the IgG isotype control. The results were presented as fold enrichments relative to Ctl-siRNA (n = 4 independent biological replicates). H Schematic summary of the proposed profibrotic mechanisms, a positive-feedback loop TGFβ1-Rcn3-TGFBR1, by which Rcn3 contributes to the pathogenesis of pulmonary fibrosis. Data are presented as the mean ± SD with unpaired student’s t-tests or Mann–Whitney U-tests to compare two groups. *P < 0.05 versus Ctl-siRNA or lenti-Ctl. Co-IP co-immunoprecipitation, NHLF normal human lung fibroblast, Ctrl-siRNA control-siRNA, lenti-Rcn3 lentivirus-Rcn3, lenti-Ctl lentivirus-control

Fig 5: TGFβ1 exposure induced Rcn3 expression in ER Stress dependent manner and Rcn3 induction are critical for the activation of pulmonary fibroblast. The immunoblot analyses of the mouse (A) and human (B) lung fibroblast show the upregulations of Rcn3, α-SMA, collagen I, and GRP78 in response to TGFβ1 exposure at indicated concentrations. The ratios to tubulin expression are presented in dot graph relative to vehicle control (n = 3–5 independent biological replicates, #P < 0.05 versus vehicle). C The ER-stress inhibitor 4-PBA significantly blunts the TGFβ1-induced upregulation of GRP78, Rcn3 and TGFBR1. The ratios to tubulin expression are presented in dot graph relative to vehicle control (n = 5 independent biological replicates, #P < 0.05 versus vehicle). D The immunoblot analyses indicate that Rcn3 depression by siRNA markedly constricts TGFβ1-induced upregulations of α-SMA and collagen I. The ratios to tubulin expression are presented in dot graph relative to the vehicle-treated Ctl-siRNA group (n = 5 independent biological replicates). E qPCR analyses of Rcn3, Col1a1, αSMA, CCND1 and PCNA in human lung fibroblast exposed to TGFβ1 (5 ng/ml) for 24 h. The data were normalised to the GAPDH content and analysed by the 2−∆∆Ct method relative to the vehicle Ctl-siRNA group (n = 6 independent biological replicates per group). Data presented as mean ± SD; 2-way ANOVA (Tukey post hoc test) was performed. #p < 0.05 vs vehicle treatment at same siRNA group; D and E *p < 0.05 vs Ctl-siRNA at same treatment. CKO conditional knockout, BLM bleomycin, Ctl-siRNA control-siRNA, Col1a1: collagen I a1, CCND1 Cyclin D1, PCNA proliferating cell nuclear antigen