Propolis compound inhibits profibrotic TGF-β1/SMAD signalling in human fibroblasts

Abstract

Hypertrophic scarring of the skin is a cause of pain, disfigurement, and restricted mobility. Excessive TGF-β1 signalling leads to SMAD3 phosphorylation, which is implicated in hypertrophic scarring. In this study, we examined the mechanism of action of tomentosenol A, a small compound that we isolated from the propolis of the Australian stingless bee Tetragonula carbonaria. Cultured adult human dermal fibroblasts and HEK293 cells were stimulated with TGF-β1, with or without tomentosenol A, and were assessed for phosphorylation of SMADs 2/3 (Western blot, AlphaLISA assay), SMAD signalling (HEK293 cells expressing a SMAD3 reporter gene), and profibrotic gene transcription using RTqPCR for ACTA2 (smooth muscle α-actin), COL1A1 and COL3A (collagens), CCN2 (connective tissue growth factor) and FN1 (fibronectin). Protein expression was measured using ELISA (fibronectin) and visualised via confocal microscopy (smooth muscle α-actin). TGF-β1 increased SMAD3 phosphorylation by 44.3-fold above baseline levels, and this effect was inhibited by tomentosenol A in a concentration-dependent manner (IC₅₀, 99.0 nM). TGF-β1 stimulated SMAD3 reporter gene expression and upregulated ACTA2, COL1A1, COL3A1, FN1 and CCN2 transcription; fibronectin protein expression; and smooth muscle α-actin filament formation in fibroblasts. These responses were inhibited by 6.25 μM tomentosenol A. These findings indicate that tomentosenol A inhibits TGF-β1/SMAD3 signalling and downstream profibrotic gene transcription and protein expression. As this pathway is implicated in hypertrophic scarring of the skin, tomentosenol A can be developed as a novel therapy for the management of scars caused by deep dermal injuries that are associated with surgery, trauma and burns.

Introduction

Hypertrophic scarring (HS) is a cutaneous fibrotic disorder of dysregulated wound healing that is characterised by the aberrant deposition of extracellular matrix macromolecules and tissue contraction¹. HS affects more than 100 million people annually in the developed world and is associated with deep dermal wounds². The pathogenesis of scar tissue formation involves excessive deposition of collagens I and III, with an associated high collagen I/III ratio, disorganisation of eosinophilic hyalinized collagen bundles and aberrant crosslinking, a flattened epidermis lacking epidermal appendages and rete ridges and a thickened dermis³. Increased smooth muscle α-actin (α-SMA) expression is evident in transformed myofibroblasts and contributes to scar contraction³. HS develops in the month after injury and presents as raised, thickened and often red scars within the bounds of the wound³. HS can cause chronic pain, pruritis, contractures, disfigurement and tissue dysfunction and may have a negative psychosocial impact, particularly when it appears on visible parts of the body, such as the face or neck². HS tends to regress within 1 to 2 years; however, once mature, regeneration of normal tissue does not occur⁴.

Normal wound healing consists of highly ordered, interdependent processes involving haemostasis, inflammation, proliferation and resolution². The regulation of these processes occurs through autocrine and paracrine signalling initiated by a variety of cytokines, chemokines and growth factors, including transforming growth factor-beta 1 (TGF-β1)⁵. TGF-β1 has a central role in regulating angiogenesis, inflammation, proliferation, re-epithelialisation, granulation tissue formation, extracellular matrix deposition and cellular apoptosis following deep dermal injury. However, persistent signalling by TGF-β1 is implicated in the development of fibrotic disorders, including HS⁴.

Resident dermal and subcutaneous fibroblasts have an obligatory role in skin wound repair⁶. Myofibroblasts are characterised by elevated structural collagen expression and de novo expression of α-SMA, which is incorporated into stress fibres and is the primary cell type contributing to the reestablishment of the extracellular matrix after injury⁷. TGF-β1 is the key driver of myofibroblast trans-differentiation and fibrogenesis⁷. TGF-β1 binds to cell surface TGF-β type II receptors (TGF-βRII), inducing the phosphorylation of threonine and/or serine residues of the TGF-βRI cytoplasmic glycine/serine (GS) domain⁸. Activated TGF-βRI phosphorylates cytosolic small mothers against decapentaplegic 2/3 (SMAD2/3), which then forms heterotrimeric complexes with SMAD4 (Fig. 1A)⁸. The complex translocates to the nucleus and, in conjunction with cofactors such as cyclic adenosine monophosphate (cAMP) response element-binding protein (CREB)/p300, regulates the expression of target genes, including ACTA2 (α-SMA), COL1A1 and COL1A2 (collagen type 1 α1 chain and collagen type 1 α2 chain, respectively), COL3A1 (collagen type 3 α1 chain) and SMAD7^9,10. SMAD7 provides feedback inhibition of TGF-β/SMAD3 signalling¹¹.

[See PDF for image]

Fig. 1

The effects of tomentosenol A on canonical TGF-β pathway SMADs were compared to those of the TGF-βRI inhibitor, GW788388 and the SMAD3-specific inhibitor, SIS3 (A). Effects of tomentosenol A on phosphorylated SMAD2 in NAHDFs (B, D, Western blot). Effects of tomentosenol A on total SMAD2/3, phosphorylated SMAD3 and total SMAD4 levels in NAHDFs (C, Western blot). Effects of tomentosenol A on SMAD3 phosphorylation in NAHDFs (E, F; AlphaLISA immunoassay). The cells were incubated with media, a solvent control or TGF-β1 (10 ng mL⁻¹) with or without tomentosenol A 6.25 μM (B, C, F) or 0.1–6.25 μM (e), GW788388 (1 μM) (B, F) or SIS3 (10 μM) (F) for 1 h. For Western blotting (B, C, D), equal amounts of protein (or positive control cell lysate 4 µL) were electrophoresed on 10% polyacrylamide gels and transferred to nitrocellulose membranes. Membranes were cut horizontally between 37 and 50 kDa and hybridised overnight with rabbit anti-human phospho-SMAD2 (Ser 465/467) mAb, rabbit anti-human phospho-SMAD3 (Ser423 and 425) pAb, rabbit anti-human SMAD2/3 mAb, rabbit anti-human SMAD4 mAb or rabbit anti-human GAPDH pAb before incubation with IRDye 680LT goat anti-rabbit IgG. Blots were imaged using an Odyssey CLx scanner at 700 nm, and the fluorescence signal was analysed using Image Studio Software version 6.0. The data (D) represent the mean ± SEM of three independent experiments. The AlphaLISA signal (E, F) was measured via a multimode plate reader with an excitation wavelength of 680 nm and an emission wavelength of 615 nm. The data represent the mean ± SEM of three independent experiments carried out in triplicate. The data were analysed using GraphPad Prism (10.2.3), and statistical significance was determined via ANOVA with Tukey’s post hoc multiple comparisons of the signal intensity (615 nm wavelength) (*p < 0.05; **p < 0.005; ***p < 0.0005; ****p < 0.0001).

Propolis is a resinous substance produced by stingless bees from local flora and is incorporated into the hive by the bees¹². Propolis provides structural support and antimicrobial and antimycotic activity and protects the nest from predators¹². Our research identified potential anti-inflammatory, antioxidant and vasomodulatory properties of methanolic extracts of Australian stingless bee propolis^{13, 14–15}. Tomentosenol A, which is isolated from stingless bee propolis, inhibits the TGF-β1-stimulated migration and proliferation of cultured fibroblasts and inhibits TGF-β1-stimulated myofibroblast differentiation and soluble collagen production¹⁶. These findings raise the possibility that tomentosenol A could be developed as an antifibrotic agent with the potential to manage hypertrophic scarring of the skin¹⁶. The precise mechanism by which tomentosenol A mediates its effects is not yet known. In this study, we investigated the mechanism of action of tomentosenol A by using cultured, normal adult human dermal fibroblasts (NAHDFs), human hypertrophic scar fibroblasts (HSFs), and human embryonic kidney (HEK)-Blue TGF-β reporter cells, with a focus on the canonical TGF-β1/SMAD3 signalling pathway.

Results

Tomentosenol A impedes the transduction of TGF-β1 signalling by inhibiting the phosphorylation of SMAD3

The effect of tomentosenol A on SMAD protein levels and TGF-β1 stimulated SMAD phosphorylation (Fig. 1A) in NAHDFs was determined using Western blot and an AlphaLISA phospho-SMAD3 specific immunoassay. Western blot demonstrated that TGF-β1 (10 ng mL⁻¹) stimulated SMAD2 phosphorylation (P < 0.0001, N = 3). This effect was inhibited by the TGF-βRI antagonist, GW788388 (single observation; Fig. 1B) but not by tomentosenol A (6.25 μM) (Fig. 1D). There was no effect by tomentosenol A (6.25 μM) or TGF-β1 (10 ng mL⁻¹) on total SMAD2/3, or total SMAD4 protein levels (N = 3, P > 0.05, Fig. 1C). Tomentosenol A (6.25 μM) inhibited TGF-β1-stimulated SMAD3 phosphorylation (Fig. 1C). In the absence of TGF-β1, tomentosenol A had no effect on SMAD3 phosphorylation (Fig. 1C).

Tomentosenol A inhibited TGF-β1-stimulated SMAD3 phosphorylation in a concentration-dependent manner (IC₅₀, 99.0 nM; R², 96.0; Fig. 1E). Recombinant TGF-β1 (10 ng mL⁻¹) increased SMAD3 phosphorylation above that observed in media containing solvents (P < 0.0001; Fig. 1F). The stimulatory effect of TGF-β1 on SMAD3 phosphorylation was abolished by GW788388 (1 μM) and partially inhibited by the SMAD3 inhibitor, SIS3 (10 μM) (Fig. 1F). In the absence of TGF-β1, tomentosenol A (6.25 μM), GW788388 (1 μM) and SIS3 (10 μM) had no effect on SMAD3 phosphorylation (Supplementary Figure S1A). The combination of SIS3 (10 μM) and tomentosenol A (6.25 μM) abolished the response to TGF-β1 (Supplementary Figure S1B). No further inhibition of SMAD3 phosphorylation was observed with a higher concentration of SIS3 (20 μM; data not shown).

Tomentosenol A inhibits the TGF-β1 stimulated decrease of SMAD3 mRNA levels in NAHDF

To determine whether tomentosenol A negatively regulates the transcription of genes encoding the canonical TGF-β1 signalling effector proteins SMAD3 or SMAD2 or positively regulates inhibitory SMAD7, the relative levels of SMAD3, SMAD2 and SMAD7 transcripts in NAHDFs and HSFs were examined. TGF-β1 decreased SMAD3 mRNA levels in NAHDFs at 48 h (20.0% of the solvent control, Fig. 2A) and HSF (50.0% of the solvent control, Fig. 2B) relative to those in the media control. This effect was partially reversed by tomentosenol A (6.25 μM; 79.7% of the solvent control), GW788388 (88.5% of the solvent control) and SIS3 (70.2% of the solvent control) in NAHDFs (Fig. 2A). However, tomentosenol A and SIS3 had no effect on the TGF-β1-mediated decrease in SMAD3 mRNA levels in HSF (45.7% and 30.0% of the solvent control, respectively), although the effect was reversed by GW788388 (86.7% of the solvent control, Fig. 2B). TGF-β1 (10 ng mL^-1), tomentosenol A (6.25 μM), GW788388 (1 μM) and SIS3 (10 μM) had no effect on SMAD2 mRNA levels in NAHDFs or HSFs (Fig. 2C and D, respectively). SMAD7 mRNA levels increased in response to TGF-β1 in NAHDFs (8.4-fold) and HSFs (6.2-fold), and this effect was reversed by coincubation of cells with tomentosenol A (6.25 μM), GW788388 and SIS3 (Fig. 2E and F).

[See PDF for image]

Fig. 2

Effects of tomentosenol A on SMAD3, SMAD2 and SMAD7 mRNA levels in NAHDFs (A, C, E, RT‒qPCR) and HSFs (B, D, F, RT‒qPCR). The cells were incubated with a solvent control or TGF-β1 (10 ng mL⁻¹) with or without tomentosenol A (6.25 μM), GW788388 (1 μM), SIS3 (10 μM) or media for 48 h. The data represent the mean ± SEM of the relative fold change in gene induction (2^−ΔΔCq) for three independent experiments (duplicated in separate wells) carried out in duplicate with the test compounds compared with the media control. The quantification cycle (Cq) was normalised to two reference genes (GAPDH and POLR2A). Data were analysed using GraphPad Prism (10.2.3), and statistical significance was determined using ANOVA with Tukey’s post hoc multiple comparisons (*p < 0.05; **p < 0.005; ***p < 0.0005; ****p < 0.0001).

Tomentosenol A inhibits the transcriptional transduction of TGF-β1 via the SMAD binding element (SBE)

To examine the effect of tomentosenol A on TGF-β1-stimulated SMAD3 transcriptional activity, we used HEK293 cells that were stably transfected with a SMAD binding element (SBE)/secreted embryonic alkaline phosphatase (SEAP) construct (Fig. 3A). In response to TGF-β1 stimulation, phosphorylated trimeric SMAD complexes translocate to the nucleus and regulate SEAP transcription by interacting with SBEs in the promoter region. TGF-β1-stimulated SEAP activity was inhibited by tomentosenol A (1.56–12.5 μM) in a concentration-dependent manner (Fig. 3B). GW788388 (1 μM) reduced the SEAP activity stimulated by TGF-β1 to the levels observed in untreated cells, whereas SIS3 (10 μM) or tomentosenol A (6.25 μM) alone only partially inhibited the SEAP activity stimulated by TGF-β1 (28.0% and 61.3% inhibition, respectively; Fig. 3C). The combination of tomentosenol A (6.25 μM) and SIS3 (10 μM) reduced the stimulatory effect of TGF-β1 on SEAP activity to levels that were observed in untreated cells (Fig. 3C). In the absence of TGF-β1, tomentosenol A (1.56–12.5 μM), GW788388 (1 μM) and SIS3 (10 μM) had no effect on SEAP activity (Supplementary Figure S2).

[See PDF for image]

Fig. 3

HEK293-blue cells stably transfected with a SMAD binding element (SBE)/Secreted embryonic alkaline phosphatase (SEAP) reporter were used to demonstrate the activation and downstream transcriptional regulation of the TGF-β/SMAD2/3 signalling pathway (A). The cells were incubated for 72 h with TGF-β1 (10 ng ml⁻¹) with or without tomentosenol A (1.6–12.5 μM) (B) or with or without tomentosenol A (6.25 μM), GW388788 (1 μM), SIS3 (10 μM) or the combination of tomentosenol A (6.25 μM) with SIS3 (10 μM) (C). The data were obtained via an EnSpire multimode plate reader at 640 nm and represent the mean ± SEM of four separate experiments carried out in duplicate. The data were analysed using GraphPad Prism (10.2.3), and statistical significance was determined by ANOVA with post hoc multiple comparisons (*p < 0.05; **p < 0.005; ***p < 0.0005; ****p < 0.0001). Panel A was created using Biorender.com.

Tomentosenol A inhibits the expression of profibrotic proteins.

Activation of TGFβRII by TGF-β1 results in the formation of a receptor heterodimer and downstream phosphorylation of receptor-activated SMADs (R-SMAD2/3). The R-SMADs form a heterotrimer with SMAD4, translocating to the nucleus and increasing the transcription of profibrotic genes (Fig. 4A). We examined whether tomentosenol A inhibited a TGF-β1-stimulated increase in FN1 mRNA levels and the expression of total cellular fibronectin protein in NAHDFs using RT‒qPCR and a human fibronectin ELISA, respectively. As expected, compared with the media control, TGF-β1 (10 ng mL⁻¹) increased FN1 mRNA levels (11.1-fold, Fig. 4B) and fibronectin protein expression (9.6-fold, Fig. 4C). This TGF-β1 response was inhibited by tomentosenol A (6.25 μM), GW788388 (1 μM) and SIS3 (10 μM). However, the inhibitory effect of SIS3 on fibronectin protein expression was significantly less than that of tomentosenol A and GW788388 (P < 0.05). The effect of tomentosenol A on the incorporation of α-SMA into cytosolic stress fibres was examined using confocal microscopy of NAHDFs. After 72 h of incubation, non-stimulated cells expressed α-SMA in the perinuclear region (Fig. 4D). TGF-β1 stimulated the incorporation of α-SMA into stress fibres (Fig. 4E), which was prevented by coincubation of cells with tomentosenol A (6.25 μM) or GW788388 (1 μM) (Figs. 4F and G, respectively).

[See PDF for image]

Fig. 4

The TGF-β/SMAD2/3 signalling pathway regulates profibrotic gene transcription and protein expression (A). NAHDFs were incubated for 48 h with solvent control or TGF-β1 (10 ng mL⁻¹) with or without tomentosenol A (6.25 μM), GW788388 (1 μM) or SIS3 (10 μM) and examined for increased FN1 mRNA levels (RTqPCR) (B) and expression of fibronectin (ELISA) (C). Data represent N = 3 experiments, reported as the mean of the relative fold gene induction (2^-ΔΔCq) for biological replicates of compounds compared with the media control ± SEM (B) or mean ± SEM (C). Cq was normalised to two reference genes (GAPDH and POLR2A). The data were analysed using GraphPad Prism (10.2.3), and statistical significance was determined via ANOVA with post hoc multiple comparisons of the fold change (B) or absorbance (450 nm, C) (*p < 0.05; **p < 0.005; ***p < 0.0005; ****p < 0.0001). Images of NAHDFs immunostained for α-SMA (green) and counterstained with DAPI nuclear stain (blue) and incubated for 72 h with solvent control (D) or TGF-β1 (10 ng/mL) (E) with or without tomentosenol A (6.25 μM) (F) or GW788388 (1 μM) (G). Images were captured using a 40 × oil immersion objective lens, using a Nikon Eclipse Ti inverted confocal microscope with a Nikon DS-Qi1Mc 1.5-MP monochrome camera attachment. Scale bar = 50 µm. Panel A was created using BioRender.com.

Tomentosenol A inhibits profibrotic mRNA transcription

RT‒qPCR was used to determine the relative mRNA levels of the profibrotic genes ACTA2 (α-SMA), COL1A1 (pro-α1 chains of type I collagen), COL3A1 (pro-α1 chains of type III collagen) and CCN2 (connective tissue growth factor, CTGF) in NAHDFs and HSFs. TGF-β1 (10 ng mL⁻¹) upregulated the mRNA levels (relative to the media control) of ACTA2 in NAHDFs (5.6-fold, Fig. 5A) and HSFs (2.8-fold, Fig. 5E), COL1A1 in NAHDFs (9.0-fold, Fig. 5B) and HSFs (16.4-fold, Fig. 5F), COL3A1 in NAHDFs (5.4-fold, Fig. 5C) and HSFs (5.5-fold, Fig. 5G) and CCN2 in NAHDFs (75.4-fold, Fig. 5D) and HSFs (37.0-fold, Fig. 5H). The increases in mRNA levels stimulated by TGF-β1 were inhibited by tomentosenol A (6.25 μM), GW788388 (1 μM) and SIS3 (10 μM) in both NAHDFs (Figs. 5A–D) and HSFs (Figs. 5E–H) (p < 0.0001).

[See PDF for image]

Fig. 5

RT‒qPCR analysis of ACTA2 in NAHDFs (A) and HSFs (E) and COL1A1 in NAHDFs (B) and HSF (F) and COL3A1 in NAHDFs (C) and HSF (G) and CCN2 in NAHDFs (D) and HSFs (H). The cells were incubated for 48 h with the solvent control or TGF-β1 (10 ng mL⁻¹) with or without tomentosenol A (6.25 μM), GW788388 (1 μM) or SIS3 (10 μM). The data represent the mean ± SEM of the relative fold change in gene expression (2^−ΔΔCq) of three independent experiments with the compounds compared with the media control. The quantification cycle (Cq) of the gene of interest (GOI) was normalised to that of two reference genes (GAPDH and POLR2A). The data were analysed via GraphPad Prism (10.2.3), and statistical significance was determined via ANOVA with Tukey’s post hoc multiple comparisons of the fold change (*p < 0.05; **p < 0.005; ***p < 0.0005; ****p < 0.0001).

Discussion

The development of hypertrophic scarring (HS) is induced by aberrant canonical TGF-β signalling associated with deep dermal wounds. We previously reported that tomentosenol A, a meroterpenoid isolated from propolis produced by Tetragonula carbonaria, has antifibrotic potential in human foetal fibroblasts¹⁶. Meroterpenoids have previously been shown to have anti-inflammatory, anti-proliferative¹⁷ and antifibrotic activities¹⁸. In this study, we characterised the in vitro response to tomentosenol A in normal adult human dermal fibroblasts (NAHDFs) and human hypertrophic scar fibroblasts (HSFs), identifying the inhibition of the canonical effector SMAD3. To refine the characterisation, two inhibitors of TGF-β signalling were included in our assays: GW788388, an inhibitor of the serine-threonine kinase TGF-βRI (ALK5)¹⁹, and SIS3, a specific inhibitor of SMAD3 phosphorylation²⁰.

In this study, 10 ng mL⁻¹ of recombinant TGF-β1 was used to stimulate the pro-scarring response in human fibroblasts. This concentration is relevant to the pathophysiology of hypertrophic scarring. Kwan et al. (2016)²¹ reported a serum TGF-β1 concentration of between 10 and 45 ng mL⁻¹ for patients who have burns covering between 0 and 55% skin surface area. The concentration of TGF-β1 used in this study was at the lower end of this range. Further to this, recombinant TGF-β1 has been used in other in vitro studies to stimulate a pro-fibrotic cellular response at a concentration similar to that used in this study (4 ng mL⁻¹ to 20 ng mL⁻¹)^{22, 23, 24–25}. Previous work by our research group²⁶ and data in the current study, demonstrates that 10 ng mL⁻¹ activates the TGF-β/SMAD3 signalling pathway in cultured human dermal fibroblasts.

Transduction of canonical TGF-β signals occurs through ubiquitously expressed receptor-activated SMAD2 and SMAD3 (rSMAD). rSMADs are regulated by SMAD7 through a negative feedback mechanism⁵. To characterise tomentosenol A, we explored whether the antifibrotic activity was associated with the inhibition of SMAD3 phosphorylation. We found that tomentosenol A inhibited the TGF-β1-stimulated phosphorylation of SMAD3 in a dose-dependent manner (Fig. 1E). SMAD3 phosphorylation at serine 423 and serine 425 initiates SMAD3 translocation to the nucleus and binding to 8-bp (5’ GTCTAGAC 3’) palindromic SMAD binding elements (SBEs)⁵. SMAD3 binds to half of the sequence (GTCT or AGAC) via the Mad Homology 1 (MH1) domain and, together with transcriptional cofactors, upregulates the expression of genes inducible by TGF-β. The interaction of SMAD3 with SMAD-binding elements (SBEs) is known to upregulate the gene expression of ACTA2²⁷, FN1²⁸, CCN2²⁹, COL1A1⁹ and COL3A1³⁰ in in vivo and in vitro models of pulmonary and cardiac fibrosis and fibroblasts, respectively. Pharmacological inhibition of SMAD3 in dermal fibroblasts is associated with downregulation of fibroblast proliferation, myofibroblast differentiation, and profibrotic protein expression^31,32. To date, the mechanism underlying the antifibrotic activity of tomentosenol A has not been investigated, nor has tomentosenol A been investigated as a potential therapy for HS. Interestingly, Ding and colleagues (2016)¹⁸ demonstrated that the meroterpenoid lingzhifuran A specifically inhibited TGF-β1-stimulated SMAD3 phosphorylation in NRK-52E (rat kidney epithelial) cells and that in vitro inhibition of SMAD3 correlated with the inhibition of renal fibrosis in vivo.

We examined mRNA levels of the TGF-β effector proteins SMAD2 and SMAD3. SMAD2 transcript levels were not influenced by TGF-β1 stimulation. In contrast, TGF-β1 reduced SMAD3 mRNA levels in normal and hypertrophic scar fibroblasts (Fig. 2A, B). This may be due to regulatory events occurring at the mRNA level to compensate for the profibrotic activity of TGF-β1. The repression of SMAD3 expression by TGF-β1 has been reported in normal and scleroderma-derived fibroblasts and is described as a feedback regulatory mechanism to reduce fibrotic activity³³. Only GW788388 reversed the downregulation in both cell types, indicating that regulatory control of TGF-β signalling may be dysfunctional at the level of SMAD3 signalling in chronic scar tissue fibroblasts. These findings also suggest that tomentosenol A regulates a target in the TGF-β/SMAD signalling pathway that is downstream of ALK5. Consistent with this hypothesis, ALK5 blockade with GW788388 successfully reversed the TGF-β1-stimulated downregulation of SMAD3 mRNA expression in HSFs, but tomentosenol A and SIS3 had no effect. Furthermore, and supporting this hypothesis, tomentosenol A had no effect on the ALK5 dependent TGF-β stimulated SMAD2 phosphorylation (Fig. 1B, D).

Increased SMAD7 mRNA levels were observed with TGF-β1 stimulation in both NAHDFs and HSFs, and this effect was inhibited by tomentosenol A, GW788388 and SIS3. SMAD7 mediates negative feedback regulation through a variety of mechanisms, including induction of TGF-βR dephosphorylation³⁴, interference in the interaction between SMAD2/SMAD3 and TGF-βR1³⁵ and suppression of TGF-β1/SMAD3-regulated microRNAs³⁶. Our data indicate that while tomentosenol A inhibits TGF-β signalling and that SMAD3 may be involved in its antifibrotic effects, the inhibition of SMAD3 activity does not occur via a reduction in SMAD3 or an increase in SMAD7 mRNA levels.

To further explore the potential antifibrotic signalling effect of tomentosenol A, we used a HEK293 TGF-β1 SEAP reporter assay. Tomentosenol A inhibited SEAP activity in a dose-dependent manner (Fig. 3B). The control samples demonstrated that SEAP was inactive when TGF-β1 was absent (Fig. 3C, Supplementary Fig. 2) confirming that SEAP activity was dependent on TGF-β1. This interpretation was supported by the abolition of SEAP activity by the TGFβRI inhibitor GW788388 (Fig. 3C). Tomentosenol A and SIS3 partially inhibited SEAP activity; however, the effect of SIS3 was less than that of tomentosenol A. When combined, these compounds abolished the effects of TGF-β1 stimulation. This observation warrants further investigation. Notably, redundancy may exist for SMAD signalling, with evidence for the formation of SMAD3-independent complexes in cells that have been exposed to TGF-β1, for example, the formation of a SMAD2/4/4 complex³⁷. Importantly, a higher concentration of SIS3 did not result in a greater response, indicating that the concentration of SIS3 used in this assay (10 µM) was maximal. SIS3 was described by Jinnen et al. (2006)²⁰ as a specific SMAD3 inhibitor. Given that the data presented in this paper show that there is no difference in the inhibitory activity of tomentosenol A and SIS3 on SMAD3 phosphorylation (Fig. 1F), the results suggest that tomentosenol A may inhibit additional targets in the TGF-β1/SMAD signalling pathway, which are yet to be characterised.

A correlation between TGF-β1-stimulated FN1 transcription and fibronectin expression was observed at 48 h in NAHDFs. Insoluble cellular fibronectin contains an alternatively spliced FN III extradomain (ED-A), which is expressed in fibroblasts during the remodelling of the extracellular matrix in response to injury³⁸. Fibronectin, particularly ED-A, has been implicated in myofibroblast differentiation, HS contracture and the progression of fibrotic disease³⁹. TGF-β1 increased the mRNA and protein levels of FN1 and fibronectin, respectively. This stimulatory effect was inhibited by tomentosenol A, GW788388 and SIS3 (Fig. 4B, C). However, the reduction in fibronectin expression after SIS3 treatment was less than that observed with tomentosenol A, while the effect on FN1 mRNA levels was not significantly different. This may indicate further posttranscriptional regulation of fibronectin expression by tomentosenol A, which has not yet been examined.

Early research into the modulation of myofibroblast activity demonstrated that the expression of fibronectin ED-A precedes and is necessary for the TGF-β1-stimulated differentiation of fibroblasts to a myofibroblast phenotype⁴⁰. This differentiated phenotype involves the incorporation of α-SMA into cytosolic stress fibres (α-SMA SF⁺), which contribute to contraction and matrix stiffness in fibrotic disorders⁴¹. Fibroblasts, as opposed to myofibroblasts, express a basal level of diffuse but not organised (α-SMA SF^-) α-SMA⁴¹, which is supported by our data. We have shown that TGF-β1 stimulates the incorporation of α-SMA into cytosolic stress fibres in NAHDFs and that tomentosenol A inhibits this incorporation (Fig. 4E, F). These findings suggest that tomentosenol A inhibits the TGF-β1-stimulated induction of myofibroblast differentiation.

We examined the transcriptional response of genes that are markers of myofibroblast differentiation (α-SMA and procollagen 1 α1)⁴² and promote fibrosis (procollagen 1 α1, CTGF, and fibronectin)^38,43. TGF-β1 induced an increased level of ACTA2, COL1A1, COL3A1, and CCN2 mRNA in normal and scar-derived dermal fibroblasts. In in vitro and ex vivo experiments, the myofibroblast phenotype is poorly maintained without continued TGF-β1 stimulation⁴⁴. Scar-derived fibroblasts exhibit increased levels of myofibroblast markers, including α-SMA and procollagen 1. Myofibroblast dedifferentiation may occur under the influence of paracrine factors secreted from adipose tissue⁴⁵. While the mechanism for this is not known, inhibition of the TGF-β/SMAD2/3 pathway has been shown to reversibly dedifferentiate myofibroblasts to a fibroblast phenotype that is characterised by loss of α-SMA expression⁴². Tomentosenol A inhibited the increase in ACTA2, COL1A1, COL3A1 and CCN2 mRNA levels induced by TGF-β1 in both NAHDFs and HSFs (Fig. 5A–H), suggesting that this compound may have the potential to inhibit myofibroblast differentiation during scar formation and promote dedifferentiation in the later stages of HS scar progression. Xu and colleagues (2023)⁴² proposed that through the dedifferentiation of myofibroblasts to a fibroblast-like phenotype, fibrotic disease can be reversed rather than only delaying progression. We demonstrated that GW788388 and SIS3 inhibited the TGF-β1-induced increases in ACTA2, COL1A1, and COL3A1 mRNA levels relative to those in the media control. Interestingly, the increase in CCN2 mRNA was abolished by SIS3, which is consistent with the findings of Purohit et al. (2017)²⁹, who demonstrated that the transcriptional regulation and expression of CCN2 is independent of SMAD2 and that SMAD3 is critical and nonredundant in transducing profibrotic TGF-β signals⁴⁶. In their seminal paper, Hamilton et al. (2022)¹⁶ reported that tomentosenol A inhibits TGF-β1-stimulated soluble collagen formation in neonatal fibroblasts. This finding is supported by our observation that the TGF-β1-stimulated transcription of COL1A1 and COL3A1 is also inhibited by tomentosenol A.

Our data demonstrate that tomentosenol A is an effective inhibitor of TGF-β1-stimulated profibrotic signalling in dermal fibroblasts. We demonstrated that suppression of TGF-β1 induced SMAD3 phosphorylation and downstream SMAD3 transcriptional activation via tomentosenol A. Importantly, tomentosenol A downregulates profibrotic TGF-β/SMAD3 target gene transcription and protein expression. Our data suggest that the antifibrotic activity of tomentosenol A occurs downstream of ALK5. Owing to the pleiotropic nature of TGF-β signalling and the critical role of growth factors in homeostasis, inflammatory responses and tissue repair, therapies targeted at profibrotic effectors downstream of receptor kinases may reduce unwanted off-target and systemic effects of pan-TGF-β inhibition⁴⁷. While our current findings are limited to investigating the effect of tomentosenol A on the inhibition of SMAD3 in vitro, they support efforts to further characterise the antifibrotic effects of the compound, both in vitro and in vivo.

Materials and methods

Materials and reagents

Human hypertrophic scar fibroblasts (HSFs, Caucasian/Eurasian/European, male, 21 years of age, right foot) were purchased from Banksia Scientific Company/Cell Research Corp. (Brisbane, Australia). Human embryonic kidney (HEK)-Blue TGF-β reporter cells and QUANTI-Blue solution were purchased from InVivoGen (California, USA). Normal adult human dermal fibroblasts (NAHDFs), FBM fibroblast growth basal medium, FGM-2 fibroblast growth medium-2 BulletKit and Dulbecco’s modified Eagle’s medium were purchased from Lonza Bioscience (Sydney, Australia). Rabbit anti-human phospho-SMAD3 (Ser423 and 425) pAb (07–1389) was purchased from Merck (Melbourne, Australia). Rabbit anti-human phospho-SMAD2 (Ser465 and 467) mAb (3108), rabbit anti-human SMAD2/3 (D7G7) XP mAb (8685), rabbit anti-human SMAD4 (D3M6U) mAb (38,454), SMAD2/3 control cell extracts (12,052) and Alexa Fluor 488 goat anti-mouse IgG Fab2 (4408) were purchased from Cell Signaling Technology (Massachusetts, USA). Rabbit anti-human/mouse glyceraldehyde 3-phosphate dehydrogenase (GAPDH) pAb (PA1-988) was purchased from InvivoGen (California, United States). IRDye 680LT goat anti-rabbit IgG secondary Ab (926–68,021) was purchased from LI-COR Biosciences (Lincoln, USA). DAPI dihydrochloride (D9542) and mouse anti-actin, α-smooth muscle mAb (A5228) were purchased from Merck (Melbourne, Australia). SIS3 was purchased from Stemcell Technologies (Vancouver, Canada). GW788388 was purchased from Adooq Bioscience (California, USA). Recombinant TGF-β1 was purchased from Sigma Aldrich (Castle Hill, Australia). AlphaLISA SureFire Ultra phosphoSMAD3 (Ser 423 and 425) was purchased from PerkinElmer (Revvity) (Macquarie Park and Mulgrave, Australia, respectively). Mini-PROTEAN TGX Stain-Free Protein Gels (10% and 12%), Precision Plus Protein Dual Colour standards, and nitrocellulose membrane (0.45 µm) were purchased from Bio-Rad (South Granville, Australia). Nunc Edge 2.0 96-well TG plates, Nunclon Delta (75 cm²) culture flasks and trypsin–EDTA (0.25%) in phenol red and human fibronectin ELISA kit (BMS2028) were purchased from Thermo Fisher Scientific (Massachusetts, USA). cOmplete mini protease inhibitor cocktail tablets were purchased from Sigma‒Aldrich (Macquarie Park, Australia). Phosphatase inhibitor cocktail II was purchased from Abcam (Melbourne, Australia). Folin-ciocalteu′s phenol reagent was purchased from Merck (Melbourne, Australia). Normocin was purchased from InVivoGen (San Diego, USA). GelRed nucleic acid gel stain was purchased from Biotium (Fremont, USA). The Isolate II RNA minikit, Sensifast cDNA synthesis kit and Sensifast SYBR No-Rox kit were purchased from Meridian Bioscience (Cincinnati, USA). Oligonucleotide primers were purchased from Bioneer Pacific (Kew, Australia).

Propolis

The propolis collected from 40 T. carbonaria hives located in southeastern Queensland, Australia, was provided by Dr. Tim Heard.

Extraction and purification of tomentosenol A

The method for preparing tomentosenol A from propolis from T. carbonaria has been reported previously¹⁶. Briefly, 100 mg of dried methanol–water extract from the propolis was reconstituted in mobile phase (MP) (1 mL of 70% MPA/30% MPB (MPA: 95% H₂O, 5% CH₃CN; MPB: 10% H₂O, 90% CH₃CN)) and fractionated using a Kinetex C₁₈ column. The biologically active fraction was dissolved in hot methanol at 60 °C and slowly cooled to room temperature to obtain pure tomentosenol A in the form of colourless needle-like crystals. The purified crystals were reconstituted in dimethyl sulfoxide (DMSO) to a stock concentration of 5.6 mg mL⁻¹ (50 mM).

Cell culture

NAHDFs and HSFs were maintained in fibroblast basal medium-2 (FBM-2) supplemented with fibroblast growth medium-2 Bullet Kit containing 0.5% recombinant human insulin, human fibroblast growth factor-B, gentamycin/ampicillin-1000 and foetal calf serum (FCS) 10 ml (total v/v 2%), (Lonza, Basel, Switzerland) and normocin (1 mg mL⁻¹). When combined, this medium is referred to as fibroblast growth medium-2 (FGM-2). The cells were incubated at 37 °C with 5% CO₂ and humidified. HEK-Blue™ TGF-β cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% FCS, penicillin (100 U ml⁻¹), streptomycin (100 mg mL⁻¹) and normocin (1 mg mL⁻¹) and incubated at 37 °C with 5% CO₂ and humidification. The cells were incubated in DMEM supplemented with 1% FCS for 18 h prior to incubation with all compounds. The stock compounds were further diluted to the required concentrations in DMEM containing 1% FCS.

The effects of tomentosenol A were compared with those of GW788388 (an ALK5 inhibitor) and SIS3 (a SMAD3 inhibitor). The TGF-β/SMAD2/3 pathway was activated with 10 ng mL⁻¹ human recombinant TGF-β1.

All assays included controls; DMEM supplemented with 1% FCS with 0.05% (v/v) DMSO and 0.4% (v/v) citrate/BSA normalised to the solvent in tomentosenol A and TGF-β1, respectively (media control); and DMSO in DMEM supplemented with 1% FCS equivalent to the highest amount of DMSO in all the compounds (solvent control), unless otherwise stated.

AlphaLISA-phosphorylated SMAD3 immunoassay

The effect of tomentosenol A on TGF-β1-stimulated SMAD3 phosphorylation was determined using an AlphaLISA pSMAD3 immunoassay (N = 3). NAHDFs were seeded at 4 × 10⁴ cells/well into 96-well plates and incubated for 48 h in FGM-2. The media was changed to DMEM supplemented with 1% FCS, and the cells were incubated for 18 h. The cells were then incubated for 1 h in triplicate with or without TGF-β1 (10 ng mL⁻¹) and tomentosenol A (0.05–6.25 μM), GW788388 (1 μM) or SIS3 (10 μM), or the combination of tomentosenol A (3.125 μM or 6.25 μM) and SIS3 (10 μM). Inhibitors were added 15 min before the addition of TGF-β1. The compounds were removed, and the cells were briefly washed with ice-cold phosphate-buffered saline (1 × PBS; 137 mM NaCl, 10 mM phosphate, 2.7 mM KCl, pH 7.4). Lysis buffer (AlphaLISA SureFire Ultra phosphoSMAD3 (Ser 423 and 425) kit, 70 μL) was added to each well, and the plate was placed on a rocker for 10 min at 20 °C. The acceptor mixture was prepared according to the manufacturer's instructions for the shallow well AlphaPlate-384 (ProxiPlate, PerkinElmer/Revvity) by combining 47% (v/v) reaction buffer 1, 47% (v/v) reaction buffer 2, 4% (v/v) activation buffer and 2% (v/v) acceptor beads. The cell lysate (10 μL) was combined with 5 μL of acceptor mixture per well. The plate was sealed, placed on a rocker for 2 min, and then incubated for 1 h at 20 °C. The donor mixture was prepared in the dark by combining 98% (v/v) dilution buffer with 2% (v/v) donor beads. The donor mixture (5 μL) was added to each well, and the plate was sealed, wrapped in foil, placed on a rocker for 2 min and incubated overnight at 20 °C. The alpha signal was measured via an EnSpire Multimode Plate Reader (PerkinElmer), with an excitation/emission wavelength of 680/615 nm (well diameter of 3.3 mm, well depth of 5.3 mm and volume of 28 μL).

Protein electrophoresis and Western blot

NAHDFs (N = 3) were seeded into 6-well plates at 2 × 10⁵ cells/well and incubated to ~ 80% confluence in FGM-2. The media was replaced with DMEM supplemented with 1% FCS, and the cells were incubated for 18 h. The media were removed, and the cells (total surface area 28.8 mm²) were incubated with the solvent in media (DMSO v/v 0.05%) or with TGF-β1 (10 ng mL⁻¹) with or without tomentosenol A (6.25 μM) for 1 h. Tomentosenol A was added 15 min prior to the addition of TGF-β1 where applicable. The compounds were then removed, and the cells were washed with ice-cold phosphate-buffered saline (1 × PBS, 500 μL volume; 137 mM NaCl, 10 mM phosphate, 2.7 mM KCl, pH 7.4). Radioimmunoprecipitation assay buffer (RIPA, 100 μL volume; NaCl 50 mM, SDS 0.1% (w/v), Tris–HCl (pH 7.4) 50 mM, Na deoxycholate 24 mM, and NP40 1.0% (v/v)) supplemented with cOmplete protease inhibitor (1 tablet per 7 mL) and phosphatase inhibitor (20 μL.mL^-1) was used to obtain cell lysates. The plates were rocked on ice for 10 min, scraped with a cell scraper and collected in 2 mL Eppendorf tubes. The lysate was then sonicated twice at 4 °C for 10 s, with a 1 min pause in between, and then centrifuged at 4 °C and 10,000 × g for 15 min. The supernatant was collected and stored at −80 °C in 20 μL aliquots. Equal amounts (9.5 μg) of protein were separated on precast 10% polyacrylamide gels under reducing conditions. Chameleon Precision Plus Protein Dual Colour standards (2.5 μL) and SMAD2/3 positive control cell extract (Cell Signaling Technology, 4 μL) were added to separate lanes. Proteins were then transferred to a nitrocellulose membrane and incubated overnight at 10 V and 4 °C. The membranes were washed for 1 min in Tris-buffered saline (TBS; Tris 20 mM, NaCl 150 mM, pH 7.6) and allowed to dry for 10 min at 37 °C before being blocked with 5% (w/v) low-fat skim milk powder diluted in TBS for 1 h on a rocker at 20 °C. The membranes were washed in TBS with Tween 0.1% (v/v) (TBS-T) for 3 × 5 min before hybridisation with primary antibodies (rabbit anti-human phospho-SMAD3 (Ser423 and 425) pAb (1:1000, MW 52 kDa), rabbit anti-human phospho-SMAD2 (Ser465 and 467) mAb (1:1000, MW 60 kDa), rabbit anti-human SMAD2/3 mAb (1:1000, MW 58/50 kDa), rabbit anti-human SMAD4 mAb (1:1000, MW 60 kDa) or rabbit anti-human glyceraldehyde 3-phosphate dehydrogenase (GAPDH, 1:1000, MW 36 kDa)) diluted in TBS-T with 5% (w/v) BSA and incubated overnight at 4 °C. The membranes were washed in TBS-T for 3 × 5 min and then incubated in the dark for 1 h at 20 °C in IRDye 680LT goat anti-rabbit IgG (LICOR, 1:10,000) diluted in TBS-T with low-fat skim milk powder 5% (w/v). The membranes were washed with TBS-T for 3 × 5 min and TBS for 1 × 5 min. The blots were imaged via an Odyssey CLx scanner at 700 nm, and the fluorescence was analysed via Image Studio Software version 6.0 (LICORbio™, Lincoln, USA).

HEK-Blue TGF-β1 SMAD3/SEAP reporter assay

HEK-Blue TGF-β cells were seeded in 24-well plates in DMEM supplemented with 10% FCS at 2.5 × 10⁵ cells per well (500 μL volume) and incubated for 24 h with 5% CO₂ at 37 °C (N = 4). The cells were then further incubated for 18 h in DMEM supplemented with 1% FCS (500 μL volume). The media was then replaced with TGF-β1 (10 ng mL⁻¹) with or without tomentosenol A (1.6–12.5 μM), GW788388 (1 μM) or SIS3 (10 μM), and the mixture was incubated for 72 h with 5% CO₂ at 37 °C. Inhibitors were added 15 min before the addition of TGF-β1. The conditioned media was centrifuged at 10,000 × g at 4 °C for 10 min, and the supernatant was collected. The Quanti-Blue reagent was combined with Quanti-Blue buffer and Milli-Q H₂O at a 1:1:98 ratio, vortexed to mix and incubated at 20 °C for 10 min. The Quanti-Blue solution (180 μL) was combined with the supernatant (20 μL) in a 96-well plate (in duplicate) and incubated at 37 °C for 30 min. The optical density was measured at 640 nm.

Reverse transcription, real-time quantitative polymerase chain reaction (RTqPCR)

NAHDFs or HSFs (N = 3) were seeded into 6-well plates at 2 × 10⁵ cells/well and incubated to ~ 80% confluence in FGM-2. The medium was replaced with DMEM supplemented with 1% FCS, and the cells were incubated for 18 h. The reduced-serum medium was removed, and the cells were incubated for 48 h with TGF-β1 (10 ng/mL) with or without tomentosenol A (6.25 μM), GW788388 (1 μM), SIS3 (10 μM), or media and solvent controls in duplicate. Inhibitors were added 15 min before the addition of TGF-β1. Total RNA was extracted via an ISOLATE II Meridian Bioscience RNA Mini Kit (Cincinnati, USA) according to the manufacturer's instructions. Total RNA (150 ng) was reverse transcribed for the synthesis of first-strand cDNA. with a SensiFAST cDNA synthesis kit. The primer sequences (Table 1) were analysed for GC content, melting temperature, product length and specificity to the target via Primer-BLAST software (National Centre for Biotechnology Information)⁴⁸. The primer efficiency and sensitivity were determined by preparing a standard curve, analysis with a melt curve and agarose gel electrophoresis of standard dilution amplicons. The quantification cycles (Cq) of the target genes were detected on a Bio-Rad CFX96 thermocycler using SensiFAST SYBR No-ROX. The relative gene expression of the target genes was normalised to that of two concurrently measured reference genes, GAPDH (glyceraldehyde-3-phosphate dehydrogenase) and POLR2A (RNA polymerase II), and the data are presented as the fold change, 2^ΔΔCq, of the treated sample compared with the untreated control (media control)⁴⁹. Statistical analysis was performed on the 2^ΔΔCq data⁴⁹.

Table 1. Oligonucleotide primers used for RT‒qPCR.

Gene and NCBI(‡) reference	Sequence(†) and size (nt)	Product length (bp)	GC content (%)	Melting temperature (Tm) °C	Self 3’ complementary	Exon‒exon spanning	Efficiency	R2	Refs.
Genes of interest (GOI)
H. sapiens ACTA2 NM_001406462.1	F. CTATGCCTCTGGACGCACAACT—22 R. CAGATCCAGACGCATGATGGCA—22	115	54.55 54.55	62.62 63.01	1.00 3.00	Yes	1.03	0.98	51
H. sapiens COL1A1 XM_054315083.1	F. TCTGCGACAACGGCAAGGTG—20 R. GACGCCGGTGGTTTCTTGGT—20	146	60.00 60.00	63.59 60.60	1.00 0.00	Yes	0.98	0.99	52
H. sapiens COL3A1 NM_000090.4	F. TTGAAGGAGGATGTTCCCATCT—22 R. ACAGACACATATTTGGCATGGTT—23	83	45.45 42.13	58.81 58.91	2.00 2.00	No	0.97	0.99	53
H. sapiens SMAD7 XM_047437509.1	F. TTCCTCCGCTGAAACAGGG—19 R. CCTCCCAGTATGCCACCAC—19	116	57.89 63.16	59.63 59.78	3.00 1.00	No	1.07	0.99	54
H. sapiens SMAD3 NM_005902.4	F. GCCTGTGCTGGAACATCATC - R. TTGCCCTCATGTGTGCTCTT -	127	55.00 50.00	59.26 59.89	2.00 0.00	No	0.98	0.99	55
H. sapiens SMAD2 NM_001135937.3	F. ACCGAAATGCCACGGTAGAA—20 R. TGGGGCTCTGCACAAAGAT—19	123	50.00 52.63	59.68 59.23	3.00 2.00	Yes	1.01	0.99	55
H. sapiens CCN2 NM_001901.4	F. CCAATGACAACGCCTCCTG—19 R. TGGTGCAGCCAGAAAGCTC—19	159	57.89 57.89	58.83 60.75	1.00 2.00	Yes	1.02	0.99	56
H. sapiens FN1 NM_001365522.2	F. ACTGCGAGAGTAAACCTGAAGC—22 R. GCGGTTTGCGATGGTACAGCT- 21	164	50.00 57.14	60.61 63.86	2.00 4.00	Yes	0.95	0.99	57
Reference genes
H. sapiens GAPDH NM_001357943.2	F. GGGGGAGCCAAAAGGGTCATCATCT—25 R. GAGGGGCCATCCACAGTCTTCT—22	236	56.00 59.09	66.43 63.75	1.00 0.00	Yes	1.01	0.98	58
H. sapiens POLR2A NM_000937.5	F. ACCTGCGGTCCACGTTGTGT—20 R. CCACCATTTCCCCGGGATGCG—21	132	60.00 66.67	64.72 66.01	1.00 3.00	Yes	0.98	0.99	59

^‡National Centre for Biotechnology Information, ^†F- forward (5’ to 3’), R- reverse (5’ to 3’), nt- nucleotides, bp- base pairs.

Fibronectin enzyme-linked immunosorbent assay (ELISA)

NAHDFs (N = 3) were seeded at 1 × 10⁴ cells/well into a 96-well plate and incubated for 48 h in FGM-2. The media was changed to DMEM supplemented with 1% FCS, and the cells were incubated for 18 h. The reduced-serum media was removed, and the cells were incubated for 48 h with TGF-β1 (10 ng/mL) with or without tomentosenol A (6.25 μM), GW788388 (1 μM) or SIS3 (10 μM), or media, solvent or no-cell controls, in duplicate. The conditioned media were collected in 1.5 mL Eppendorf tubes and centrifuged at 805 × g for 10 min at 4 °C. The supernatants were stored at −20 °C. The wash and assay buffers, human fibronectin standard (0.31–20 ng mL⁻¹), biotin conjugate, streptavidin-HRP and stop solution were prepared and used according to the manufacturer’s protocol (Human Fibronectin ELISA Kit, Cat# BMS2028; Thermo Fisher Scientific). The supernatants (samples) were diluted 1:100 with assay buffer. Standards (100 μL), the assay buffer control (100 μL) or equal volumes of assay buffer and diluted samples (50 μL) were added to sample wells with biotin conjugates (50 μL), which were covered with a plate seal and incubated at 20 °C on an orbital shaker (IKA®, KS 130 basic) at 540 rpm for 2 h. The sample wells were subsequently washed six times for 15 s each. Streptavidin-HRP (100 μL) was added to the sample wells, which were covered with a plate seal and incubated at 20 °C on a plate shaker at 540 rpm for 1 h. The sample wells were washed six times for 15 s each. Substrate solution (100 μL) was added to the wells for 9 min prior to the addition of stop solution (100 μL). Optical density was measured using an EnSpire Multimode Plate Reader at 450 nm.

Confocal microscopy

NAHDFs were seeded into 8-well chamber slides at 8.4 × 10³ cells per well¹⁶ and incubated in a 5% CO₂ humidified incubator at 37 °C for 48 h (to ~ 80% confluence) in FGM-2 (200 μL). The media was replaced with DMEM supplemented with 1% FCS, and the cells were incubated for 18 h. The media were removed, and the cells were incubated for 72 h with TGF-β1 (10 ng mL⁻¹) with or without tomentosenol A (6.25 μM), GW788388 (1 μM) or controls, with the exchange of fresh solutions at 24 h intervals to minimise the concentration of the compounds associated with evaporation (400 µL/well). With 24 hourly solution exchange, evaporation accounted for a negligible (< 2.5%) reduction in total volume within the wells.

The cells were fixed with 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS) and incubated for 17 min at 20 °C⁵⁰. The cells were rinsed with PBS three times, permeabilized with 0.1% Triton X-100 in PBS for 5 min and washed with PBS three more times. The cells were then blocked with 1% bovine serum albumin (BSA) in PBS, incubated at 20 °C for 1 h, washed with 0.1% BSA in PBS three times and hybridised with a monoclonal mouse anti-human α-SMA clone 1A4 antibody (1:200 in 0.1% BSA in PBS) overnight at 4 °C. The cells were washed for 3 × 1 min and 3 × 5 min in 0.1% BSA in PBS at 20 °C and then incubated with Alexa Fluor 488 (H + L) secondary antibody, goat anti-mouse IgG (1:1,000 in 0.1% BSA in PBS) and 4',6-diamidino-2-phenylindole (DAPI, 2 μg mL⁻¹) for 1 h in the dark. The cells were washed for 3 × 1 min and 3 × 5 min in 0.1% BSA in PBS. The slides were coverslipped with glycerol, sealed with nail varnish and then viewed with a Nikon Eclipse Ti inverted confocal microscope. Images were captured via a Nikon DS-Qi1Mc 1.5-MP monochrome camera attachment.

Data analysis

The data were analysed via Excel software (Microsoft 365, Albuquerque, USA), GraphPad Prism version 10.2.3 (GraphPad Software, San Diego, USA) and SPSS version 27 (IBM, Armonk, USA). The normality of the distribution was determined via the Shapiro‒Wilk test. Parametric data were analysed via one-way ANOVA with post hoc multiple comparisons, and the means of each data set (Tukey’s correction) were compared with an alpha threshold and confidence level of 0.05. Nonparametric data were analysed via the Kruskal‒Wallis test to compare the mean rank with Dunn’s test for statistical hypotheses. The data are displayed as *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. The error bars represent the standard error of the mean (SEM).

Acknowledgements

This research is supported by an Australian Government Research Training Program (RTP) Scholarship. The Authors thank Dr Tim Heard for providing the raw propolis material. The authors would like to acknowledge Lucas' Papaw Foundation for its financial support.

Author contributions

LJR, FDR; concept design, article drafting, data curation and formal analysis, methodology, validation, visualization. FDR, project administration. SB, TDT, RJH, FDR; supervision, article review and editing. TDT; isolation and preparation of tomentosenol A. All the authors revised the manuscript and approved the final version.

Data availability

Raw data is available on Figshare. https://doi.org/10.6084/m9.figshare.27850359.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval

This article does not contain any studies with human or animal subjects. Ethical approval or informed consent is not applicable for this article.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1038/s41598-025-07918-2.

Publisher's note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1. Čoma, M et al. Molecular changes underlying hypertrophic scarring following burns involve specific deregulations at all wound healing stages (inflammation, proliferation and maturation). Int J Mol Sci.; 2021; 22, pp. 897-917.1:CAS:528:DC%2BB3MXltlGjsb8%3D [DOI: https://dx.doi.org/10.3390/ijms22020897]

2. Mony, MP; Harmon, KA; Hess, R; Dorafshar, AH; Shafikhani, SH. An updated review of hypertrophic scarring. Cells; 2023; 12, pp. 678-700.1:CAS:528:DC%2BB3sXls1altL0%3D

3. Nabai, L; Pourghadiri, A; Ghahary, A. Hypertrophic scarring: Current knowledge of predisposing factors, cellular and molecular mechanisms. J Burn Care Res.; 2020; 41, pp. 48-56. [DOI: https://dx.doi.org/10.1093/jbcr/irz158]

4. Shirakami, E; Yamakawa, S; Hayashida, K. Strategies to prevent hypertrophic scar formation: A review of therapeutic interventions based on molecular evidence. Burns & Trauma; 2020; 8, 3. [DOI: https://dx.doi.org/10.1093/burnst/tkz003]

5. Deng, Z et al. TGF-β signaling in health, disease, and therapeutics. Signal Transduct Target Ther.; 2024; 9, pp. 61-101. [DOI: https://dx.doi.org/10.1038/s41392-024-01764-w]

6. Li, B; Wang, JH. Fibroblasts and myofibroblasts in wound healing: Force generation and measurement. J Tissue Viabil.; 2011; 20, pp. 108-120. [DOI: https://dx.doi.org/10.1016/j.jtv.2009.11.004]

7. Prunotto, M et al. Stable incorporation of α-smooth muscle actin into stress fibers is dependent on specific tropomyosin isoforms. Cytoskeleton (Hoboken); 2015; 72, pp. 257-267.1:CAS:528:DC%2BC2MXht1yqtr3I [DOI: https://dx.doi.org/10.1002/cm.21230]

8. Massagué, J; Sheppard, D. TGF-β signaling in health and disease. Cell; 2023; 186, pp. 4007-4037.1:CAS:528:DC%2BB3sXhvFegsb3N [DOI: https://dx.doi.org/10.1016/j.cell.2023.07.036]

9. Fan, C; El Andaloussi, S; Lehto, T; Kong, KW; Seow, Y. Smad-binding decoy reduces extracellular matrix expression in human hypertrophic scar fibroblasts. Mol Med Rep.; 2020; 22, pp. 4589-4600.1:CAS:528:DC%2BB3cXisVyru7fJ [DOI: https://dx.doi.org/10.3892/mmr.2020.11549]

10. Chen, SJ et al. Stimulation of type I collagen transcription in human skin fibroblasts by TGF-β: Involvement of Smad 3. J Invest Dermatol; 1999; 112, pp. 49-57.1:CAS:528:DyaK1MXjvFWitA%3D%3D [DOI: https://dx.doi.org/10.1046/j.1523-1747.1999.00477.x]

11. Yan, X et al. Smad7 protein interacts with receptor-regulated smads (R-Smads) to inhibit transforming growth factor-β (TGF-β)/smad signaling. J Biol Chem.; 2016; 291, pp. 382-392.1:CAS:528:DC%2BC28XhsFSlug%3D%3D [DOI: https://dx.doi.org/10.1074/jbc.M115.694281]

12. Pasupuleti, VR; Sammugam, L; Ramesh, N; Gan, SH. Honey, Propolis, and royal jelly: A comprehensive review of their biological actions and health benefits. Oxid Med Cell Longev.; 2017; 1259510–1259532, 2017.1:CAS:528:DC%2BC1cXitlKrsbfK [DOI: https://dx.doi.org/10.1155/2017/1259510]

13. Massaro, FC; Brooks, PR; Wallace, HM; Russell, FD. Cerumen of Australian stingless bees (Tetragonula carbonaria): gas chromatography-mass spectrometry fingerprints and potential anti-inflammatory properties. Naturwissenschaften; 2011; 98, pp. 329-337.2011NW...98.329M1:CAS:528:DC%2BC3MXktlOqtLg%3D [DOI: https://dx.doi.org/10.1007/s00114-011-0770-7]

14. Hamilton, KD; Brooks, PR; Ogbourne, SM; Russell, FD. Natural products isolated from Tetragonula carbonaria cerumen modulate free radical-scavenging and 5-lipoxygenase activities in vitro. BMC Complem Altern Med.; 2017; 17, 232.1:CAS:528:DC%2BC1cXhvFOitL7O [DOI: https://dx.doi.org/10.1186/s12906-017-1748-6]

15. Massaro, FC et al. Effect of Australian propolis from stingless bees (Tetragonula carbonaria) on pre-contracted human and porcine isolated arteries. PLoS ONE; 2013; 8, e81297.2013PLoSO..881297M1:CAS:528:DC%2BC2MXlsVGitb8%3D [DOI: https://dx.doi.org/10.1371/journal.pone.0081297]

16. Hamilton, KD et al. Anti-fibrotic potential of tomentosenol A, a constituent of cerumen from the Australian native stingless bee. Tetragonula Carbonaria Antioxid.; 2022; 11, pp. 1604-1628.1:CAS:528:DC%2BB38Xit1KnsbvM

17. Li, CS et al. Meroterpenoids with antiproliferative activity from a Hawaiian-plant associated fungus Peyronellaea coffeae-arabicae FT238. Org Lett.; 2016; 18, pp. 2335-2338.1:CAS:528:DC%2BC28XmvFGjsbk%3D [DOI: https://dx.doi.org/10.1021/acs.orglett.6b00685]

18. Ding, W-Y et al. Isolation of lingzhifuran A and lingzhilactones D-F from Ganoderma lucidum as specific Smad3 phosphorylation inhibitors and total synthesis of lingzhifuran A. RSC Adv.; 2016; 6, pp. 77887-77897.2016RSCAd..677887D1:CAS:528:DC%2BC28Xhtlagtb%2FE

19. Petersen, M et al. Oral administration of GW788388, an inhibitor of TGF-β type I and II receptor kinases, decreases renal fibrosis. Kidney Int.; 2008; 73, pp. 705-715.1:CAS:528:DC%2BD1cXis1WmtLc%3D [DOI: https://dx.doi.org/10.1038/sj.ki.5002717]

20. Jinnin, M; Ihn, H; Tamaki, K. Characterization of SIS3, a novel specific inhibitor of Smad3, and its effect on transforming growth factor-β1-induced extracellular matrix expression. Mol Pharmacol.; 2006; 69, pp. 597-607.1:CAS:528:DC%2BD28XhtVCgtbs%3D [DOI: https://dx.doi.org/10.1124/mol.105.017483]

21. Kwan, PO; Ding, J; Tredget, EE. Serum decorin, interleukin-1β, and transforming growth factor-β predict hypertrophic scarring postburn. J Burn Care Res.; 2016; 37, pp. 356-366.

22. Ding, H; Chen, J; Qin, J; Chen, R; Yi, Z. TGF-β-induced α-SMA expression is mediated by C/EBPβ acetylation in human alveolar epithelial cells. Mol Med.; 2021; 27, 22.1:CAS:528:DC%2BB3MXlvVajt7Y%3D [DOI: https://dx.doi.org/10.1186/s10020-021-00283-6]

23. Midgley, AC et al. Transforming growth factor-β1 (TGF-β1)-stimulated fibroblast to myofibroblast differentiation is mediated by hyaluronan (HA)-facilitated epidermal growth factor receptor (EGFR) and CD44 co-localization in lipid rafts. J Biol Chem.; 2013; 288, pp. 14824-14838.1:CAS:528:DC%2BC3sXot1Wiur0%3D [DOI: https://dx.doi.org/10.1074/jbc.M113.451336]

24. Zhang, G-Y et al. Activation of peroxisome proliferator-activated receptor-γ inhibits transforming growth factor-β1 induction of connective tissue growth factor and extracellular matrix in hypertrophic scar fibroblasts in vitro. Arch Dermatol Res.; 2009; 301, pp. 515-522.2009geki.book...Z1:CAS:528:DC%2BD1MXptFGlsbY%3D [DOI: https://dx.doi.org/10.1007/s00403-009-0959-1]

25. Vivar, R et al. Role of FoxO3a as a negative regulator of the cardiac myofibroblast conversion induced by TGF-β1. Biochim Biophys Acta Mol Cell Res.; 2020; 1867, 118695.1:CAS:528:DC%2BB3cXlslGqtLk%3D [DOI: https://dx.doi.org/10.1016/j.bbamcr.2020.118695]

26. Hamilton, KD et al. Anti-fibrotic potential of tomentosenol A, a constituent of cerumen from the Australian native stingless bee. Tetragonula Carbonaria. Antioxidants (Basel); 2022; 11, pp. 1604-1619.1:CAS:528:DC%2BB38Xit1KnsbvM [DOI: https://dx.doi.org/10.3390/antiox11081604]

27. Hu, B; Wu, Z; Phan, SH. Smad3 mediates transforming growth factor-beta-induced alpha-smooth muscle actin expression. Am J Respir Cell Mol Biol; 2003; 29, pp. 397-404.1:CAS:528:DC%2BD3sXntFynu74%3D [DOI: https://dx.doi.org/10.1165/rcmb.2003-0063OC]

28. Zhao, J et al. Smad3 deficiency attenuates bleomycin-induced pulmonary fibrosis in mice. Am J Physiol Lung Cell Mol Physiol.; 2002; 282, pp. L585-593.1:CAS:528:DC%2BD38Xit1GhtLo%3D [DOI: https://dx.doi.org/10.1152/ajplung.00151.2001]

29. Purohit, T; Qin, Z; Quan, C; Lin, Z; Quan, T. Smad3-dependent CCN2 mediates fibronectin expression in human skin dermal fibroblasts. PLoS ONE; 2017; 12, pp. e0173191-0173208.1:CAS:528:DC%2BC2sXhtVGjsbnE [DOI: https://dx.doi.org/10.1371/journal.pone.0173191]

30. Khalil, H et al. Fibroblast-specific TGF-β-Smad2/3 signaling underlies cardiac fibrosis. J Clin Invest; 2017; 127, pp. 3770-3783. [DOI: https://dx.doi.org/10.1172/jci94753]

31. Peterson, JM et al. Galunisertib exerts antifibrotic effects on TGF-β-induced fibroproliferative dermal fibroblasts. Int J Mol Sci; 2022; 23, pp. 6689-6701.1:CAS:528:DC%2BB38Xhs1OjsLzL [DOI: https://dx.doi.org/10.3390/ijms23126689]

32. Zhang, YF et al. Baicalein attenuates hypertrophic scar formation via inhibition of the transforming growth factor-β/Smad2/3 signalling pathway. Br J Dermatol.; 2016; 174, pp. 120-130.1:CAS:528:DC%2BC28Xhtlyltrk%3D [DOI: https://dx.doi.org/10.1111/bjd.14108]

33. Mori, Y; Chen, SJ; Varga, J. Expression and regulation of intracellular SMAD signaling in scleroderma skin fibroblasts. Arthritis Rheum.; 2003; 48, pp. 1964-1978.1:CAS:528:DC%2BD3sXmtFartb4%3D [DOI: https://dx.doi.org/10.1002/art.11157]

34. Shi, W et al. GADD34-PP1c recruited by Smad7 dephosphorylates TGFbeta type I receptor. J Cell Biol.; 2004; 164, pp. 291-300.1:CAS:528:DC%2BD2cXmsV2isQ%3D%3D [DOI: https://dx.doi.org/10.1083/jcb.200307151]

35. Hayashi, H et al. The MAD-related protein Smad7 associates with the TGF-β receptor and functions as an antagonist of TGF-β signaling. Cell; 1997; 89, pp. 1165-1173.1:CAS:528:DyaK2sXkt1Kks70%3D [DOI: https://dx.doi.org/10.1016/s0092-8674(00)80303-7]

36. Chung, AC et al. Smad7 suppresses renal fibrosis via altering expression of TGF-β/Smad3-regulated microRNAs. Mol Ther.; 2013; 21, pp. 388-398.1:CAS:528:DC%2BC38XhslKqu7%2FO [DOI: https://dx.doi.org/10.1038/mt.2012.251]

37. Lucarelli, P et al. Resolving the combinatorial complexity of Smad protein complex formation and its link to gene expression. Cell Syst.; 2018; 6, pp. 75-89.e11.1:CAS:528:DC%2BC1cXisVCgtbo%3D [DOI: https://dx.doi.org/10.1016/j.cels.2017.11.010]

38. Klingberg, F et al. The fibronectin ED-A domain enhances recruitment of latent TGF-β-binding protein-1 to the fibroblast matrix. J Cell Sci.; 2018; 131, pp. 201293-201305.1:CAS:528:DC%2BC1cXhslartrvF [DOI: https://dx.doi.org/10.1242/jcs.201293]

39. Ishise, H et al. Hypertrophic scar contracture is mediated by the TRPC3 mechanical force transducer via NFkB activation. Sci Rep.; 2015; 5, pp. 11620-11635.2015NatSR..511620I1:CAS:528:DC%2BC2MXhtFKjur7P [DOI: https://dx.doi.org/10.1038/srep11620]

40. Serini, G et al. The fibronectin domain ED-A is crucial for myofibroblastic phenotype induction by transforming growth factor-β1. J Cell Biol.; 1998; 142, pp. 873-881.1:CAS:528:DyaK1cXlsVWltLk%3D [DOI: https://dx.doi.org/10.1083/jcb.142.3.873]

41. Hillsley, A; Santos, JE; Rosales, AM. A deep learning approach to identify and segment alpha-smooth muscle actin stress fiber positive cells. Sci Rep.; 2021; 11, 21855.2021NatSR.1121855H1:CAS:528:DC%2BB3MXisVegt7zO [DOI: https://dx.doi.org/10.1038/s41598-021-01304-4]

42. Xu, R et al. Mesenchymal stem cells reversibly de-differentiate myofibroblasts to fibroblast-like cells by inhibiting the TGF-β-SMAD2/3 pathway. Mol Med.; 2023; 29, pp. 59-74.1:CAS:528:DC%2BB3sXos1Crsbc%3D [DOI: https://dx.doi.org/10.1186/s10020-023-00630-9]

43. Abreu, JG; Ketpura, NI; Reversade, B; De Robertis, EM. Connective-tissue growth factor (CTGF) modulates cell signalling by BMP and TGF-β. Nat Cell Biol.; 2002; 4, pp. 599-604.1:CAS:528:DC%2BD38Xls1OrtLo%3D [DOI: https://dx.doi.org/10.1038/ncb826]

44. Hecker, L; Jagirdar, R; Jin, T; Thannickal, VJ. Reversible differentiation of myofibroblasts by MyoD. Exp Cell Res.; 2011; 317, pp. 1914-1921.1:CAS:528:DC%2BC3MXotVGgtbw%3D [DOI: https://dx.doi.org/10.1016/j.yexcr.2011.03.016]

45. Higginbotham, S et al. Inhibition and reversal of a TGF-β1 induced myofibroblast phenotype by adipose tissue-derived paracrine factors. Stem Cell Res Ther.; 2024; 15, 166.1:CAS:528:DC%2BB2cXhtlyhtrfO [DOI: https://dx.doi.org/10.1186/s13287-024-03776-3]

46. Flanders, KC. Smad3 as a mediator of the fibrotic response. Int J Exp Pathol.; 2004; 85, pp. 47-64.1:CAS:528:DC%2BD2cXlt1Wqsrg%3D [DOI: https://dx.doi.org/10.1111/j.0959-9673.2004.00377.x]

47. Shi, N et al. Research progress on drugs targeting the TGF-β signaling pathway in fibrotic diseases. Immunol Res.; 2022; 70, pp. 276-288.1:CAS:528:DC%2BB38XptFaqtr0%3D [DOI: https://dx.doi.org/10.1007/s12026-022-09267-y]

48. Ye, J et al. Primer-BLAST: a tool to design target-specific primers for polymerase chain reaction. BMC Bioinform.; 2012; 13, pp. 134-145.1:CAS:528:DC%2BC38Xht1eqsr3I [DOI: https://dx.doi.org/10.1186/1471-2105-13-134]

49. Schmittgen, TD; Livak, KJ. Analyzing real-time PCR data by the comparative C(T) method. Nat Protoc.; 2008; 3, pp. 1101-1108.1:CAS:528:DC%2BD1cXmvVemt7c%3D [DOI: https://dx.doi.org/10.1038/nprot.2008.73]

50. Cheng, R et al. Influence of fixation and permeabilization on the mass density of single cells: A surface plasmon resonance imaging study. Front Chem.; 2019; 7, 588.2019FrCh..7.588W1:CAS:528:DC%2BB3cXmvVarsLc%3D [DOI: https://dx.doi.org/10.3389/fchem.2019.00588]

51. Monsuur, HN; van den Broek, LJ; Koolwijk, P; Niessen, FB; Gibbs, S. Endothelial cells enhance adipose mesenchymal stromal cell-mediated matrix contraction via ALK receptors and reduced follistatin: Potential role of endothelial cells in skin fibrosis. J Cell Physiol.; 2018; 233, pp. 6714-6722.1:CAS:528:DC%2BC1cXpt1aksLk%3D [DOI: https://dx.doi.org/10.1002/jcp.26494]

52. Hayashi, M et al. Identification of the collagen type 1 α1 gene (COL1A1) as a candidate survival-related factor associated with hepatocellular carcinoma. BMC Cancer; 2014; 14, pp. 108-118.1:CAS:528:DC%2BC2cXhvVGmt7vI [DOI: https://dx.doi.org/10.1186/1471-2407-14-108]

53. Elkouris, M et al. SET9-mediated regulation of TGF-β signaling links protein methylation to pulmonary fibrosis. Cell Rep.; 2016; 15, pp. 2733-2744.1:CAS:528:DC%2BC28Xps1Sktbo%3D [DOI: https://dx.doi.org/10.1016/j.celrep.2016.05.051]

54. Li, S et al. RS4651 suppresses lung fibroblast activation via the TGF-β1/SMAD signalling pathway. Eur J Pharmacol.; 2021; 903, pp. 174135-174145.1:CAS:528:DC%2BB3MXhtVCrurbK [DOI: https://dx.doi.org/10.1016/j.ejphar.2021.174135]

55. Kuo, FT et al. Relative expression of genes encoding SMAD signal transduction factors in human granulosa cells is correlated with oocyte quality. J Assist Reprod Genet; 2011; 28, pp. 931-938. [DOI: https://dx.doi.org/10.1007/s10815-011-9609-6]

56. Kong, HJ et al. Crosstalk between YAP and TGFβ regulates SERPINE1 expression in mesenchymal lung cancer cells. Int J Oncol.; 2021; 58, pp. 111-121.1:CAS:528:DC%2BB3cXis1eis7%2FO [DOI: https://dx.doi.org/10.3892/ijo.2020.5153]

57. Tunali, G et al. A positive feedback loop driven by fibronectin and IL-1β sustains the inflammatory microenvironment in breast cancer. Breast Cancer Res.; 2023; 25, pp. 27-40.1:CAS:528:DC%2BB3sXlsFWms7o%3D

58. Meital, LT et al. Omega-3 fatty acids decrease oxidative stress and inflammation in macrophages from patients with small abdominal aortic aneurysm. Sci Rep.; 2019; 9, 12978.2019NatSR..912978M1:CAS:528:DC%2BC1MXhslOhtLzN [DOI: https://dx.doi.org/10.1038/s41598-019-49362-z]

59. Giebe, S et al. Cigarette smoke extract counteracts atheroprotective effects of high laminar flow on endothelial function. Redox Biol.; 2017; 12, pp. 776-786.1:CAS:528:DC%2BC2sXmtlahtb0%3D [DOI: https://dx.doi.org/10.1016/j.redox.2017.04.008]