Human dermal fibroblasts (hDFs; C0135C, Thermo Fisher Scientific, Waltham, MA, United States) were cultured in Eagle’s minimum essential medium alpha modification (α-MEM; Sigma-Aldrich, St. Louis, MO, United States) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Euroclone, Milan, Italy), 2 mM L-glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin (Sigma-Aldrich) up to the third passage. hDFs cultured upon passage 3 (P3; early culture passage) and passage 12 (P12; late culture passage) were then seeded at a cell density of 3,000 cells/cm². When reaching 60% confluence, hDFs, at both culture passages P3 and P12, were stimulated with the addition of human recombinant TGF-β1 (10 ng/mL; Abbkine, Wuhan, China) for 3 and 7 days, to induce the fibroblast-to-myofibroblast conversion.

To evaluate the effects of BMP2 on TGF-β1-pre-stimulated hDFs (P3), 10 ng/mL rhTGF-β1 and 50 ng/mL rhBMP2 (OriGene, Rockville, MD, United States) were added, according to previous findings (Kheirollahi et al., 2019) for further 4 days.

The study was performed in accordance with the recommendations of Comitato Etico Provinciale—Azienda Ospedaliero-Universitaria di Modena (Modena, Italy), which provided the approval of the protocol (ref. number 3299/CE, September 2017).

The dental pulp was harvested from third molars of adult subjects (n = 3; 18–35 years), after obtaining their written informed consent, in compliance with the Declaration of Helsinki. The dental pulp was digested in α-MEM containing 3 mg/mL type I collagenase plus 4 mg/mL dispase (all from Sigma-Aldrich) and then was filtered onto 100-μm Falcon Cell Strainers to obtain a cell suspension. Cell suspension was then plated in 25-cm² culture flasks and expanded in a culture medium at 37°C and 5% CO₂. Following cell expansion, human dental pulp cells underwent magnetic cell sorting, using the MACS^® separation kit, against the stemness surface markers STRO-1 and c-Kit, which allow to obtain a purer stem cell population within human dental pulp. hDPSCs were sequentially immune-selected as previously described (Di Tinco et al., 2021). Mouse IgM anti-STRO-1 and rabbit IgG anti-c-Kit primary antibodies (Santa Cruz Biotechnology, Dallas, TX, United States) were used and subsequently revealed by magnetically labeled anti-mouse IgM and anti-rabbit IgG secondary antibodies (Miltenyi Biotec, Bergisch Gladbach, Germany). The obtained STRO-1⁺/c-Kit⁺ hDPSCs were expanded and used for any experimental evaluation at passage 3.

The transwell co-culture system between hDFs (P3 and P12), previously stimulated with TGF-β1, and hDPSCs (seeded in 1:1 ratio) was set up in six-well insert plates. Briefly, hDFs were first seeded on the bottom of the culture plate at 3,000 cells/cm²; then, 24 h later, cells were stimulated with 10 ng/mL rhTGF-β1 (Abbkine) and cultured under these conditions for 3 and 7 days. The TGF-β1 stimulus was administered twice a week. After 3 and 7 days of culturing hDFs with TGF-β1 stimulation, hDPSCs were seeded on transwell inserts, i.e., 0.4 µm polycarbonate membranes (Corning, New York, NY, United States) and maintained in co-culture for 4 days with TGF-β1-treated hDFs (3 and 7 days), still under TGF-β1 exposure. hDPSCs and hDFs cultured alone and without exposure to rhTGF-β1 were used as controls.

Real-time PCR analyses were carried out to evaluate the mRNA levels of ACTA2, FN1, COL1A1, HDAC4, TNC, SPARC, BMP2, BMP4, and BMP6. hDFs and hDPSCs were homogenized, and total RNA was extracted and purified using the PureLink RNA columns (Thermo Fisher Scientific). cDNA synthesis was performed using Maxima First Strand cDNA Synthesis Kit with DNase I treatment (Thermo Fisher Scientific). Quantitative real-time PCRs were performed using SYBR Green Master Mix (Bio-Rad) on the CFX Connect Real-time PCR instrument (Bio-Rad, Hercules, CA, United States) with the oligonucleotides. The oligonucleotides used in this study are listed in Table 1 (all obtained from Sigma-Aldrich).

Relative quantification was calculated from the ratio of the cycle number (Ct) at which the signal crossed a threshold set within the logarithmic phase of the given gene to that of the reference hRPLP0. Mean values of the duplicate results of three independent experiments for each sample were used as individual data for 2^−ΔΔCt statistical analysis.

Immunofluorescence analyses were performed on hDPSCs and fibroblasts as previously described (Pisciotta et al., 2022). Cells were fixed in 4% paraformaldehyde in pH 7.4 phosphate-buffered saline (PBS) for 20 min and washed in PBS. Then, where needed, cell permeabilization was performed by using 0.1% Triton X-100 (Sigma-Aldrich) in PBS for 5 min. Samples were washed thrice with PBS and blocked with 3% BSA in PBS for 30 min at room temperature and then were incubated with the primary antibodies [mouse anti-α-SMA (Invitrogen, Waltham, MA, United States), mouse anti-Coll-I (Abcam, Cambridge, UK), mouse anti-fibronectin (Invitrogen), rabbit anti-PPARγ (Cell Signaling Technology, Danvers, MA, United States), rabbit anti-GLI1 (Invitrogen), and rabbit anti-PDGFRβ (Cell Signaling Technology)] diluted 1:50 in PBS containing 3% BSA for 1 h at room temperature. After rinsing with PBS containing 3% BSA, cells were incubated for 1 h at room temperature with secondary antibodies diluted 1:200 in PBS containing 3% BSA (goat anti-mouse Alexa 488, goat anti-rabbit Alexa 488, and donkey anti-mouse Alexa 546; Invitrogen). Finally, after rinsing with PBS, cell nuclei were stained with 1 μg/mL DAPI in PBS for 7 min, and then samples were mounted with Fluoromount anti-fading medium (Invitrogen). Samples were observed using a Nikon A1 confocal laser scanning microscope (Nikon, Minato, Tokyo, Japan). The confocal serial sections were processed using ImageJ software to obtain three-dimensional projections, and image rendering was performed using Adobe Photoshop software (Pisciotta et al., 2022).

Whole-cell lysates were obtained as previously described (Di Tinco et al., 2021). A measure of 30 µg of protein extract for each sample was quantified by Bradford protein assay (Bio-Rad), and separation was performed by SDS-polyacrylamide gel electrophoresis on Mini-PROTEAN^® TGX™ Stain-Free Precast gels. Gels were then UV-activated using a ChemiDoc MP Imaging System (Bio-Rad), and proteins were subsequently transferred to 0.2-μm nitrocellulose membranes (Bio-Rad). The membranes were then imaged using the ChemiDoc Imaging System (Bio-Rad) for total protein normalization. Membranes were incubated overnight with the following primary antibodies: mouse anti-α-SMA (Invitrogen), mouse anti-fibronectin (Invitrogen), rabbit anti-PPARγ (Cell Signaling Technology), rabbit anti-PDGFRβ (Cell Signaling Technology), and rabbit anti-GLI1 (Invitrogen), all diluted 1:1,000 in 0.1% TBS-Tween 20 (Sigma-Aldrich) and then were incubated with HRP-conjugated anti-rabbit and anti-mouse secondary antibodies (1:3,000; Thermo Fisher Scientific) for 30 min at room temperature. The membranes were visualized using a ChemiDoc Imaging System (Bio-Rad). Finally, the relative expression levels of each evaluated marker were then obtained by normalizing the density of the protein bands to the corresponding stain-free blot image using Image Lab software (Version 6.1, Bio-Rad).

All the experiments were performed in triplicate. Data were expressed as mean ± standard deviation (SD). One-way ANOVA followed by the Newman–Keuls post hoc test was performed to analyze differences among three or more experimental groups. Differences between co-culture and control groups were analyzed by one-way ANOVA followed by Dunnett’s post hoc test (GraphPad Prism software version 8 Inc., San Diego, CA, United States). In any case, statistical significance was set at p-values < 0.05.

hDFs P3 were cultured under TGF-β1 stimulation for 3 and 7 days to mimic different times of exposure toward a pro-fibrotic microenvironment, and then cells were evaluated for the expression of pro-fibrotic markers, before and after culturing with hDPSCs for 4 days through a transwell co-culture system (Figure 1). Data from real-time PCR analyses showed that, after 3 days of stimulation, TGF-β1 induced an upregulation of the myofibroblast-associated genes, as revealed by a statistically significant increase in mRNA levels of ACTA2, HDAC4, and SPARC (**p < 0.01, ***p < 0.001 vs ctrl hDFs, Figure 1A). After a prolonged exposure to the pro-fibrotic stimulus (7 days), hDFs revealed a statistically significant increase in mRNA levels of ACTA2, COL1A1, FN1, HDAC4, SPARC, and TNC, when compared to the control group (*p < 0.05, **p < 0.01, ***p < 0.001 vs ctrl hDFs, Figure 1A) with COL1A1, FN1, and TNC being further increased in a statistically significant manner when compared to the 3-day stimulation group (°p < 0.05, °°p < 0.01, °°°p < 0.001 vs hDFs + TGF-β 3 days, Figure 1A), which demonstrated the ability of TGF-β1 to induce a fibroblast-to-myofibroblast conversion.

After transwell co-culture with hDPSCs, hDFs pre-stimulated with TGF-β1 for 3 and 7 days showed a downregulation of all the pro-fibrotic genes, except for COL1A1 and TNC, which showed statistically significant decreased mRNA levels only after 7 days of pre-stimulation with TGF-β1 (^§ p < 0.05, ^§§ p < 0.01, ^§§§ p < 0.001 vs hDFs + TGF-β 3 days and hDFs + TGFβ 7 days, Figure 1A).

Western blot analyses further confirmed the ability of TGF-β1 to induce statistically significant increased expression levels of α-SMA and fibronectin after either 3 or 7 days of stimulation (**p < 0.01, ***p < 0.001 vs ctrl hDFs, Figure 1B), with a further statistically significant increase in expression of both markers in the 7-day group, when compared to the 3-day group (°°°p < 0.001 vs hDFs + TGFβ 3 days), showing that a longer stimulation induced a stronger commitment toward the myofibroblast phenotype (Figure 1B). After co-culturing with hDPSCs through the transwell system, hDFs pre-stimulated with TGF-β1 for 3 days showed, at the protein level, a statistically significant increased expression of α-SMA with respect to the experimental counterpart hDFs cultured alone (^§§ p < 0.01 vs hDFs + TGFβ 3 days), whereas the opposite trend was revealed in 7-day TGF-β1^stim hDFs after co-culture with hDPSCs (^§ p < 0.05 vs hDFs + TGFβ 7 days). Similar data were obtained by WB analysis of fibronectin in the same experimental groups (^§§§ p < 0.001 vs hDFs + TGFβ 3 days and vs hDFs + TGFβ 7 days). Moreover, after co-culture with hDPSCs, the hDFs pre-stimulated with TGF-β1 for 7 days showed statistically significant lower levels of fibronectin, when compared to the counterpart hDFs pre-stimulated with TGF-β1 for 3 days (^### p < 0.001 vs 3-day TGFβ^stim hDFs after hDPSC co-culture).

Confocal immunofluorescence analyses further proved that the fibroblast-to-myofibroblast conversion was successfully triggered by TGF-β1 and that co-culture with hDPSCs was able to induce a reduction in the myofibroblast-associated markers α-SMA, Coll-I, and fibronectin (Figure 2). In particular, a consistent remodeling of the ECM-associated component fibronectin is shown in Figure 2, besides a morphological shift/cytoskeletal rearrangement linking to a decreased expression of α-SMA. These data suggest that hDPSCs are able to modulate the phenotype of early culture passage fibroblasts, after short and prolonged times of pro-fibrotic stimulations.

Peroxisome proliferator-activated receptor γ (PPARγ) has been demonstrated to play important roles in regulating processes related to fibrogenesis—including cellular differentiation, inflammation, and wound healing—by counteracting TGF-β1 pro-fibrotic effects (Uto et al., 2005; Lakatos et al., 2007). Notably, as shown in Figure 3, Western blot analysis highlighted that TGF-β1 induced a decreased expression of PPARγ in hDFs, and that it was instead increased in a statistically significant manner, in both experimental groups of TGF-β1-pre-stimulated hDFs after co-culture with hDPSCs (^§§§ p < 0.001 vs hDFs + TGFβ 3 days and vs hDFs + TGFβ 7 days; Figure 3A).

In particular, a strong nuclear immunolabeling against PPARγ was observed in TGF-β-pre-stimulated hDFs after co-culturing with hDPSCs, hinting the activation of anti-fibrotic mechanisms contrasting the fibrogenic outcome of TGF-β1 stimulation (Figure 3B).

hDPSCs were co-cultured through a transwell system with hDFs pre-stimulated by TGF-β1 for 3 and 7 days, still in the presence of TGF-β1, to mimic a fibro-inflammatory microenvironment in vitro. As shown in Figure 4, Western blot analyses revealed that a statistically significant increased expression of α-SMA was induced in hDPSCs under these co-culture conditions in both the experimental groups, when compared to control hDPSCs cultured alone (*p < 0.05, **p < 0.01 vs. hDPSCs, Figure 4A), whereas no statistically significant difference was detected among the two co-culture groups.

In parallel, PDGFRβ expression was statistically significantly higher in hDPSCs after co-culture with hDFs pre-stimulated with TGF-β1 for 7 days (*p < 0.05, vs hDPSCs), whereas no statistically significant difference was detected in GLI1 expression among the three experimental groups (Figure 4A). Confocal immunofluorescence analyses further confirmed these findings, besides revealing no immunolabeling against GLI1 and Coll-I (Figure 4B). These data highlight that after culturing under pro-fibrotic conditions, hDPSCs enhanced their pericyte-like features, without being induced toward a pro-fibrotic/myofibroblast phenotype.

The modulation of BMP2, BMP4, and BMP6 in hDPSCs was evaluated by real-time PCR analyses (Figure 5A). Histograms in Figure 5A show a statistically significant fold increase in BMP2 mRNA levels after co-culture with 3 and 7 days TGF-β1-pre-stimulated hDFs, when compared to control hDPSCs cultured alone (*p < 0.05 vs hDPSCs). At the same time, mRNA levels of BMP4 were decreased in hDPSCs after co-culture with hDFs under both pre-stimulation conditions (**p < 0.01 vs hDPSCs) with respect to the control group, whereas mRNA levels of BMP6 revealed a statistically significant fold decrease only in hDPSCs after co-culture with 7-day TGFβ^stim hDFs (*p < 0.05 vs hDPSCs; Figure 5A).

Based on these data, hDFs (P3) pre-stimulated with TGF-β1 for 3 and 7 days were cultured in the presence of rhBMP2 for 4 days; then, real-time PCR analyses were carried out on hDFs from each experimental group to evaluate the effects of rhBMP2 on the expression of myofibroblast-associated genes ACTA2, COL1A1, FN1, HDAC4, SPARC, and TNC. Figure 5B shows that a statistically significant reduction in mRNA levels of COL1A1, FN1, and SPARC was observed in 3-day TGF-β1-pre-stimulated hDFs + BMP2 (^§§ p < 0.01, ^§§§ p < 0.001 vs hDFs + TGFβ 3 days), whereas only ACTA2, FN1, and SPARC showed a statistically significant decrease in mRNA levels in 7-day TGF-β1-pre-stimulated hDFs + BMP2 (^§§ p < 0.01, ^§§§ p < 0.001 vs hDFs + TGFβ 7 days).

Moreover, confocal immunofluorescence analyses revealed that immunolabeling against PPARγ was stronger and specifically localized at the nuclear level in TGF-β1-pre-stimulated hDFs exposed to rhBMP2, suggesting that this stimulus induced a myofibroblast-to-lipofibroblast conversion (Figure 5C).

These data may support the hypothesis that BMP2 pathway activation is one of the mechanisms leveraged by hDPSCs to exert an anti-fibrotic action in modulating the myofibroblast phenotype induced by TGF-β1 stimulation.

hDFs at the late culture passage (P12) were cultured under the TGF-β1 stimulation for 3 and 7 days, to mimic pro-fibrotic stimulation on fibroblasts that are more committed toward a myofibroblast phenotype, and then transwell co-culture systems were set up with hDPSCs, still in the presence of TGF-β1 for further 4 days. Data from real-time PCR analyses (Figure 6A) showed that after co-culture with hDPSCs, only 3-day TGF-β1-pre-stimulated fibroblasts reduced their mRNA levels of ACTA2, COL1A1, FN1, and SPARC in a statistically significant manner (^§§ p < 0.01, ^§§§ p < 0.001 vs hDFs + TGFβ 3 days). Conversely, in 7-day TGF-β1-pre-stimulated fibroblasts, the mRNA levels of expression of all these pro-fibrotic markers were statistically significantly increased after co-culture with hDPSCs, when compared to the counterpart cultured alone (^§ p < 0.05, ^§§ p < 0.01, ^§§§ p < 0.001 vs hDFs + TGFβ 7 days).

Confocal immunofluorescence analyses highlighted that hDFs (P12) already expressed α-SMA and the pro-fibrotic ECM markers, i.e., Coll-I and FN, whose expression was maintained after TGF-β1 stimulation. Furthermore, immunofluorescence analyses also clearly showed that hDPSCs co-culture was not able to induce a decrease in expression of such markers in late passage TGF-β1-pre-stimulated fibroblasts (Figure 6B). The effects of such a pro-fibrotic co-culture system were investigated in hDPSCs as well. Figure 7 shows that real-time PCR analyses revealed that the mRNA levels of fibrotic ECM-associated markers was statistically significantly higher in hDPSCs after co-culturing with TGF-β1-pre-stimulated fibroblasts P12, in the presence of TGF-β1 stimulation (**p < 0.01, hDPSCs after 3-day^stim TGFβ vs hDPSC ctrl; *p < 0.05, ***p < 0.001 hDPSCs after 7-day^stim TGFβ vs hDPSC ctrl).