Fibromodulin selectively accelerates myofibroblast apoptosis in cutaneous wounds by enhancing interleukin 1β signaling

Abstract

Activated myofibroblasts deposit extracellular matrix material to facilitate rapid wound closure that can heal scarlessly during fetal development. However, adult myofibroblasts exhibit a relatively long life and persistent function, resulting in scarring. Thus, understanding how fetal and adult tissue regeneration differs may serve to identify factors that promote more optimal wound healing in adults with little or less scarring. We previously found that matricellular proteoglycan fibromodulin is one such factor promoting more optimal repair, but the underlying molecular and cellular mechanisms for these effects have not been fully elucidated. Here, we find that fibromodulin induces myofibroblast apoptosis after wound closure to reduce scarring in small and large animal models. Mechanistically, fibromodulin accelerates and prolongs the formation of the interleukin 1β-interleukin 1 receptor type 1-interleukin 1 receptor accessory protein ternary complex to increase the apoptosis of myofibroblasts and keloid- and hypertrophic scar-derived cells. As the persistence of myofibroblasts during tissue regeneration is a key cause of fibrosis in most organs, fibromodulin represents a promising, broad-spectrum anti-fibrotic therapeutic.

Fig. 1. Fibromodulin (FMOD) accelerates myofibroblast clearance in rat and pig models with reduced scar formation.

a Representative images of sections stained with α-smooth muscle actin (α-SMA) and cleaved caspase-3 by immunohistochemistry (IHC) staining from adult rat wounds at day 14 post-injury. Yellow lines outline the scar area. Dashed boxes in the lower magnification images represent the region of interest shown in the higher magnification images below. Scale bars, 100 μm. b Quantification of α-SMA⁺ myofibroblast density in adult rat wounds from (a). N = 12 (control) or 14 (FMOD)-treated rats, respectively. c Quantification of cleaved caspase-3⁺ cell density in adult rat wounds from (a). N = 12 rats. d Representative images of sections stained with hematoxylin & eosin (H&E) (first panel), picrosirius red (PSR) coupled with polarized light microscopy (PLM) (second panel), IHC for α-SMA (third panel) and cleaved caspase-3 (fourth panel) from excessive-mechanical-loading wounds of female adult red Duroc pigs at week 8 post-injury. Dermal scar areas are outlined in blue (for the H&E-stained image) or yellow (for PSR-PLM images), while white dotted lines in the PSR-PLM images outline the epidermal edge. The solid green-boxed regions (second panel) represent the region of α-SMA-stained images, while the dotted green-boxed regions (second panel) represent the region of cleaved caspase-3-stained images. Scale bars, 50 (black) or 500 μm (yellow). e Quantification of α-SMA⁺ myofibroblast density in control- vs. FMOD-treated adult pigs from d. f Quantification of cleaved caspase-3⁺ cell density in control- vs. FMOD-treated adult pigs from (d). N = 9 wounds from 3 pigs (e, f). All treatments were administrated at the time of surgery. The number of IHC-positively stained cells and nuclei across the entire wound area was counted under a microscope from two centrally bisected sections of each wound sample. The ratio of IHC-positively stained cells to the total number of cells (indicated by the number of nuclei) was then calculated to quantify the density of α−SMA⁺ (b, e) or cleaved caspase-3⁺ (c and f) cells. Data presented as mean ± standard deviation (s.d.) overlaying all the data points. P-values were determined by two-tailed unpaired t-tests (b–f). *P < 0.05; **P < 0.005. Source data are provided as a Source Data file.

Fig. 2. Fibromodulin (FMOD) induces BJ fibroblast-converted myofibroblast apoptosis.

All cells were subjected to serum starvation prior to treatment. a Representative images of human BJ fibroblasts stained with α-smooth muscle actin (α-SMA) by immunofluorescence staining before and after 5.0 ng/mL transforming growth factor (TGF)β1 treatment for 6 days. N = 3 biological replicates. b Representative flow cytometry plot of BJ fibroblasts before and after 5.0 ng/mL TGFβ1 treatment for 6 days, with mouse anti-α-SMA antibody [4A4] (GTX60466, GeneTex; 1: 250 dilution) and goat anti-mouse IgG(H + L) highly cross-adsorbed secondary antibody, Alexa Fluor^TM 405 (A-31553, Thermo Fisher Scientific; 2 μg/mL). c Quantification of α-SMA⁺ myofibroblast percentages in BJ fibroblasts from (b). N = 3 biological replicates. d Expression of ACTA2 (the gene encoding α-SMA) in BJ fibroblasts before and after fibroblast-myofibroblast conversion. Data were normalized to the ACTA2 level before TGFβ1-induced fibroblast-myofibroblast conversion. N = 3 biological replicates. e Representative images of BJ fibroblast-converted myofibroblasts (BJ-myofibroblasts) with α-SMA and TUNEL staining, accompanied by the respective flow cytometry plots stained with Annexin V-FITC and PI staining. N = 3 (for α-SMA and TUNEL staining) or 4 (for flow cytometry) biological replicates, respectively. f Quantification of apoptotic BJ-myofibroblasts by flow cytometry from (e). g Quantification of pyroptotic BJ-myofibroblasts by flow cytometry from (e). N = 4 biological replicates (f, g). Scale bars, 25 μm. Data presented as mean ± s.d. overlaying all the data points. P-values were determined by two-tailed unpaired t-tests (c–g). N.S., not significant, P > 0.05; *P < 0.05; **P < 0.005. Source data are provided as a Source Data file.

Fig. 3. Fibromodulin (FMOD) induces the apoptosis of the primary normal human dermal fibroblast (NHDF) FB-AA36-converted myofibroblasts.

All cells were subjected to serum starvation prior to treatment. a Representative images of the primary NHDF FB-AA36 (derived from a 36-year-old female African–American donor; Supplementary Table 1) stained with α-smooth muscle actin (α-SMA) by immunofluorescence staining before and after transforming growth factor (TGF)β1-induced fibroblast-myofibroblast conversion. b Expression of ACTA2 in FB-AA36 fibroblasts before and after fibroblast-myofibroblast conversion. Data were normalized to the ACTA2 level before TGFβ1-induced fibroblast-myofibroblast conversion. c Representative images of NHDF FB-AA36 with α-SMA and TUNEL staining, accompanied by the respective flow cytometry plots stained with Annexin V-FITC and PI staining. d Quantification of NHDF FB-AA36 apoptosis from flow cytometry in (c). e Representative images of FB-AA36 fibroblast-converted myofibroblasts with α-SMA and TUNEL staining, accompanied by the respective flow cytometry plots stained with Annexin V-FITC and PI staining. f Quantification of FB-AA36 fibroblast-converted myofibroblast apoptosis from flow cytometry in (e). Scale bars, 25 μm. Data presented as mean ± s.d. overlaying all the data points. N = 3 biological replicates; P-values were determined by two-tailed unpaired t-tests (b–f). N.S., not significant, P > 0.05; **P < 0.005. Source data are provided as a Source Data file.

Fig. 4. Fibromodulin (FMOD) induces the apoptosis and interleukin (IL)1β expression of the human keloid-derived KB-AA35 fibroblasts.

All cells were subjected to serum starvation prior to treatment. a Representative images of the human keloid-derived fibroblast KB-AA35 (derived from a 35-year-old female African–American donor; Supplementary Table 1) stained with α-smooth muscle actin (α-SMA) by immunofluorescence staining. Scale bar, 25 μm. b Expression of ACTA2 in FB-AA36 fibroblasts and KB-AA35 fibroblasts. Data were normalized to the ACTA2 level of FB-AA36 fibroblasts. c Quantification of KB-AA35 fibroblast apoptosis by flow cytometry with Annexin V-FITC and PI staining. d Expression of IL1B in KB-AA35 fibroblasts with or without FMOD treatment. Data were normalized to the IL1B level without FMOD treatment. e Active IL1β production in KB-AA35 fibroblasts with or without FMOD treatment. Data presented as mean ± s.d. overlaying all the data points. N = 3 biological replicates; P-values were determined by two-tailed unpaired t-tests (b–e). *P < 0.05; **P < 0.005. Source data are provided as a Source Data file.

Fig. 5. Fibromodulin (FMOD) induces apoptosis in dermal fibroblasts derived from the keloid and hypertrophic scar but not in the normal skin tissue of donor 1914138.

All cells were subjected to serum starvation prior to treatment. a Representative images of normal skin (from abdomen), keloid (from abdomen), and hypertrophic scar (from left shoulder) tissues collected from the donor with the National Disease Research Interchange (NDRI) reference ID 1914138 with α-smooth muscle actin (α-SMA) staining by immunohistochemistry (IHC; upper), and the representative images of dermal fibroblasts isolated from these tissues with α-SMA staining by immunofluorescence staining (lower). b Expression of ACTA2 in dermal fibroblasts derived from tissues described from (a). Data were normalized to the ACTA2 level of fibroblasts isolated from the normal skin. c Representative flow cytometry plots of fibroblasts derived from tissues described in (a) with Annexin V-FITC and PI staining. d Quantification of apoptotic dermal fibroblasts by flow cytometry from (c). e The influence of FMOD on cell apoptosis with and without IL1β was calculated from (d). Scale bars, 50 μm (black) or 25 μm (white). Data presented as mean ± s.d. overlaying all the data points. N = 3 biological replicates; P-values were determined by two-tailed unpaired t-tests (b–e). N.S., not significant, P > 0.05; *P < 0.05; **P < 0.005. Source data are provided as a Source Data file.

Fig. 6. Interleukin (IL)1B-knockout diminishes the pro-apoptosis effect of FMOD on BJ-myofibroblasts.

All cells were subjected to serum starvation prior to treatment. a Representative plots of wildtype (WT) and IL1B^-/- BJ-myofibroblasts stained with Annexin V-allophycocyanin (APC) and DAPI staining. b Quantification of BJ-myofibroblast apoptosis from (a). c Impacts of fibromodulin (FMOD) and IL1β on cell apoptosis determined from (b). Data presented as mean ± s.d. overlaying all the data points. N = 3 biological replicates; P-values were determined by two-tailed unpaired t-tests (b, c). N.S., not significant, P > 0.05; **P < 0.005. Source data are provided as a Source Data file.

Fig. 7. Knockdown of interleukin (IL)1B, interleukin 1 receptor type 1 (IL1R1), or interleukin-1 receptor accessory protein (IL1RAP) expression weakens the effects of fibromodulin (FMOD) on myofibroblast IL1β expression and apoptosis.

All cells were subjected to serum starvation prior to treatment. a ELISA assessment of active IL1β production from the IL1B-, IL1R1-, and IL1RAP-knockdown BJ-myofibroblasts, as well as the control cells transfected with non-targeting siRNA control, respectively. b Impacts of FMOD on active IL1β production determined from (a). c Representative flow cytometry plots of BJ-myofibroblasts described in (a). Cell apoptosis was evaluated with staining of Annexin V-FITC and PI staining. d Quantification of apoptotic BJ-myofibroblasts by flow cytometry from (c). e Impacts of FMOD on BJ-myofibroblast apoptosis determined from (d). f Quantification of pyroptotic BJ-myofibroblasts by flow cytometry from (c). g Impacts of FMOD on BJ-myofibroblast pyroptosis determined from (f). Data presented as mean ± s.d. overlaying all the data points. N = 3 biological replicates; P-values were determined by two-tailed unpaired t-tests (a–h). N.S., not significant, P > 0.05; *P < 0.05; **P < 0.005. Source data are provided as a Source Data file.

Fig. 8. Fibromodulin (FMOD) binds to the interleukin (IL)1β-interleukin 1 receptor type 1 (IL1R1)-interleukin-1 receptor accessory protein (IL1RAP) ternary complex.

a Image depicting the pull-down assay conducted with recombinant human His-tagged FMOD, which was incubated in vitro with recombinant IL1β. b Image depicting the pull-down assay conducted with recombinant human His-tagged FMOD, which was incubated in vitro with recombinant IL1R1. c Image depicting the pull-down assay conducted with recombinant human His-tagged FMOD, which was incubated in vitro with recombinant IL1RAP. d Image depicting the pull-down assay conducted with recombinant human His-tagged FMOD, which was incubated in vitro with IL1β-IL1R1-IL1RAP ternary complex (Com.) in an acellular system. e Image depicting the pull-down assay conducted with recombinant human His-tagged FMOD, which was incubated in vitro with the whole membrane protein (M.P.) extracted from BJ-myofibroblasts. Blot is representative of N = 2 biological replicates (a–e). f Surface plasmon resonance (SPR) spectrum characterizing the binding properties between IL1R1 and FMOD. g SPR spectrum characterizing the binding properties between IL1RAP and FMOD. h In silico analysis of protein-protein interactions between FMOD with IL1β-IL1R1 binary complex. i. In silico analysis of protein-protein interactions between FMOD and IL1β-IL1R1-IL1RAP ternary complex. j In silico analysis of binding energy among IL1β-IL1R1-IL1RAP ternary complex components. k In silico analysis of the average distance of the involved amino acid residues of IL1β-IL1R1-IL1RAP ternary complex. More details about the amino acid residue interactions are demonstrated in Supplementary Fig. 16. Source data are provided as a Source Data file.

Fig. 9. Fibromodulin (FMOD) promotes the direct binding of interleukin 1 receptor type 1 (IL1R1) and interleukin (IL)1β on the surface of BJ-myofibroblasts.

All cells were subjected to serum starvation prior to treatment. a Representative images of in situ IL1R1:IL1β-proximity ligation assay (PLA) staining on attached BJ-myofibroblasts. Phalloindin was used to stain F-actin, while DAPI was used for nuclei staining. b–g Quantification of IL1R1:IL1β-PLA⁺ signal per cell at 1- (b), 6- (c), 12- (d), 24- (e), 48- (f), and 72-h (g) from (a). h Representative flow cytometry plot and quantification of IL1R1:IL1β-PLA staining of BJ-myofibroblasts at 24 h post-treatment. Scale bar, 50 μm. Data presented as violin plots (b-g) or mean ± s.d. h overlaying all the data points. N = 10 (a–g) or 3 (h), respectively; P-values were determined by two-tailed Mann–Whitney U tests (b–g) or two-tailed unpaired t-tests (h), respectively. N.S., not significant, P > 0.05; *P < 0.05; **P < 0.005. Source data are provided as a Source Data file.

Fig. 10. Fibromodulin (FMOD) promotes the direct binding of interleukin 1 receptor type 1 (IL1R1) and interleukin-1 receptor accessory protein (IL1RAP) on the surface of BJ-myofibroblasts.

All cells were subjected to serum starvation prior to treatment. a Representative images of in situ IL1R1:IL1RAP-proximity ligation assay (PLA) staining on attached BJ-myofibroblasts. Phalloindin was used to stain F-actin, while DAPI was used for nuclei staining. b–g Quantification of IL1R1:IL1RAP-PLA⁺ signal per cell at 1- (b), 6- (c), 12- (d), 24- (e), 48- (f), and 72-h (g) after the treatment of FMOD and/or IL1β from (a). h Representative flow cytometry plot and quantification of IL1R1:IL1RAP-PLA staining of BJ-myofibroblasts at 24 h post-treatment. Scale bar, 50 μm. Data presented as violin plots (b–g) or mean ± s.d. (h) overlaying all the data points. N = 10 (a–g) or 3 (h), respectively; P-values were determined by two-tailed Mann-Whitney U tests (b–g) or two-tailed unpaired t-tests (h). N.S., not significant, P > 0.05; *P < 0.05; **P < 0.005. Source data are provided as a Source Data file.