Fig 2: Ky3 reduces tissue expression of IL-6, OPG and TRAIL. Boxplots demonstrate a significant reduction in whole lung expression of OPG RNA (a), OPG Protein (b), TRAIL RNA (c), TRAIL protein (d) and IL-6 RNA (e) and plasma protein (f). Panel (g) shows representative photomicrographs of serial lung sections. Sections were stained for macrophages (F4/80), OPG, TRAIL, IL-6 and I?ßa. Box and Whisker plots represent the interquartile range (box) with the line representing the median and whisker the full range of the data, each animal is represented by a dot (Ctrl (white) n = 7, SuHx (blue) n = 5, IgG (grey) n = 7 & Ky3 (green) n = 12 animals per group). * p < 0.05, ** p < 0.01, compared to IgG treated rats using one-way ANOVA followed by Sidak’s multiple comparisons test. All images are presented at their original magnification ×400, scale bar represents 20 µm

Fig 3: OPG activates pro-proliferative signalling and a disease-relevant transcriptome. Panel (a) Signalling Pathway Impact Analysis (SPIA) with each pathway represented by one dot. The pathways to the right of the red diagonal line are significant after Bonferroni correction of the global p-values obtained using Fisher’s methods from the combination of pPERT and pNDE values, the pathways to the right of the blue line are significant after FDR correction. (b) shows a heat map of significant differentially regulated genes after gene enrichment against PAH-associated genes in OPG stimulated PASMCs, (c) TaqMan validation of gene expression microarray, TaqMan expression data normalised using ??CT with 18 s rRNA as the endogenous control gene. Panel (d) shows a heat map of cell cycle/CDK proteins significantly regulated by OPG at 10 and 60 min expressed as a ratio to unstimulated controls from the same time point from Kinex phospho-arrays identified, with (e) showing those specifically related to NF-?ß. f Western blot validation of Kinex array data in unstimulated (0.2% FCS, Un) or OPG-stimulated (50 ng ml-1) PASMCs at 10 min (10) and 60 min (60) with relative band densities of phospho-ERK1/2, phospho-HSP27, phospho-mTOR, phospho-CDK4 and total CDK5 are shown by the bar graphs and representative western blot images shown above the graph. Heat maps show Z-ratio gene or protein expression. Bars represent mean with error bars showing the standard error of the mean, n = 3 for pooled triplicate samples (a, b), n = 12 (c), n = 4 (d, e), n = 5 (f) from three donors of PASMCs, dots represent experimental repeats. Bars from unstimulated cells are white, OPG stimulated blue. *p < 0.05, ** p < 0.01, *** p < 0.001 compared OPG-stimulated to unstimulated PASMCs using one-way ANOVA followed by Bonferroni’s multiple comparisons post hoc test. When there were only two groups, unpaired t-tests were used

Fig 4: Effects of the removal of the dental epithelium on dental mesenchyme evaluated by in vitro and in vivo cultures. (A) H&E staining of tooth germs (a) and dental the mesenchyme (b) cultured for 7 days. RT-qPCR revealed that the level of DSPP (c) decreased in the dental mesenchyme, whereas that of OPG (d), OCN (e), RUNX2 (f) and ALP (g) in the dental mesenchyme was significantly higher than that in tooth germs. Each experiment was performed in triplicate. (B) In vivo subrenal capsule culture with gel sponges (schematic). Three E16.5 tooth germs and three samples of the dental mesenchyme were cultured for further study. (C) H&E staining (a and b) and Masson's trichrome staining (c and d) revealed that, while tooth germs could form normal tooth structures after 3 weeks of subrenal capsule culture, without the dental epithelium, the dental mesenchyme could only form bone-like tissue. (D) (a) Samples without primary antibody were used as negative control. Immunohistochemical staining for DSP (b), OPG (c) and OCN (d) revealed that, after 3 weeks in vivo culture, the DSP level was higher in tooth germ-derived tissue (b1), whereas levels for OPG and OCN were higher in dental mesenchyme-derived tissue (c2 and d2). (E) RT-qPCR revealed that, compared with the prenatal period, OPG expression in postnatal first-molar tooth germs significantly increased. Scale bars, 100 µm (A-a and b, C-a1-d1) and 50 µm (C-a2-d2 and D). Data are presented as the mean ± SD. ***P<0.001. The red arrows indicate parts of the tissue that were positive for immunohistochemical staining. DSPP, dentin sialophosphoprotein; OPG, osteoprotegerin; OCN, osteocalcin; RUNX2, runt-related transcription factor 2; ALP, alkaline phosphatase; DSP, dentin sialoprotein.

Fig 5: OPG-Fas interaction mediates the OPG-induced phenotypic response of PASMC. TaqMan expression of (a) VEGFA, (b) PDGFRA, (c) TNC, (d) Cav1 and (e) TRAIL in response to OPG in the presence (hash bars) or absence (Grey bars) of anti-Fas neutralising antibody (1500 ng ml-1). Panel (f) demonstrates OPG inhibition of FasL and TRAIL-induced apoptosis in HT1080 cells. g PASMC migration following 6 h stimulation with PDGF (20 ng ml-1), OPG (30 ng ml-1) or 0.2% FCS (serum-free media, SFM), in the presence or absence of Fas neutralising antibody. h Proliferation of PASMCs following stimulation with OPG for 72 h in the presence or absence of Fas neutralising antibody and/or TRAIL neutralising antibody (0.5 nM). Proliferation expressed as a percentage of proliferation to PDGF. Bars represent the mean with error bars showing the standard error of the mean. Dots represent experimental repeats, Panels (a–e) (n = 4), panel (f) (n = 3), panel (g) (n = 4), panel (h) (n = 4 for SFM, 10 for PDGF & OPG stimulations) * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 following one-way ANOVA with Bonferroni’s multiple comparisons post hoc test