Growth factor mimetics for skin regeneration: In vitro profiling of primary human fibroblasts and keratinocytes

INTRODUCTION

Skin wound healing represents a significant clinical and socioeconomic challenge globally. In Europe, chronic wounds affect approximately 1.5–2 million people, with the costs associated with wound management often reaching up to 3% of total healthcare expenditure. Despite recent advancements, including the development of several bioengineered skin substitutes and the availability of medical devices in clinics, none of these approaches result in complete regeneration. Even with ongoing efforts to develop innovative regenerative approaches, nonhealing wounds remain costly and burdensome. Patients with complex wounds, such as burns, experience healing processes that lead to repair rather than true regeneration. Key limitations include poor integration, insufficient vascularization, scar formation, wound contraction, and immune rejection.

Inclusion of signaling molecules like growth factors has emerged as a promising strategy to enhance skin graft performance. Clinical trials suggest that growth factors can positively influence hard-to-heal wounds. However, the risk-benefit ratio of using these biological agents remains a topic of debate. Like other biologicals, growth factors have intrinsic pharmaceutical and pharmacokinetic shortcomings, including low stability, short half-lives, high production costs, and batch-to-batch variations in quality. Small molecular modalities have gained attention as a more stable and less complex alternative in the field of regenerative medicine. However, the utility of small molecules in skin regeneration has not been extensively explored. In this study, we evaluated several compounds for their potential to directly or indirectly mimic three distinct signaling pathways known to play critical roles in skin regeneration (Figure 1).

Sonic hedgehog (SHH) signaling is a key regulator of skin tissue homeostasis, driving skin cell proliferation and promoting angiogenesis. Following injury, it stimulates hair follicle neogenesis. Purmorphamine, a purine derivative, activates SHH signaling by binding and activating Smoothened, an essential component of the hedgehog pathway. Fibroblast growth factor-2 (FGF-2) is also recognized for its significant ability to improve wound healing and enhance scar quality in various types of wounds. SUN11602 is a small molecule that mimics FGF-2 functions through activation of fibroblast growth factor receptor 1 (FGFR1), followed by downstream activation of the MEK/ERK cascade and expression of signaling-related target genes. Although these FGF-2-mimetic features have been explored primarily in the context of neuroprotection, their potential in skin regeneration remains largely untested. Prostacyclin (PGI2), a prostanoid with well-established roles in angiogenesis and regenerative responses in tissues such as skeletal muscle, bone, and nerves, exhibits angiogenic and antifibrotic activities in preclinical studies. ONO-1301, a synthetic PGI2 mimetic, acts as a PGI2 receptor agonist, activating the cAMP/PKA signaling cascade and triggering the release of proangiogenic factors like vascular endothelial growth factor (VEGFA).

Given their distinct growth factor-mimetic features, the three small molecules were applied to primary human skin cells to study their effects on proliferation, differentiation, and migration, with the aim of assessing their potential as putative regenerative modalities in engineered skin substitutes.

RESULTS AND DISCUSSION

Fibroblast proliferation is a critical process during the proliferative phase of wound healing. After determining safe dose ranges for all compounds on commonly used fibroblast and keratinocyte cell lines (Figure S1), we evaluated three selected small molecules for their effects on the proliferation of primary normal human fibroblasts (NHF) from three different donors (Figure 2a). Time-dependent proliferation profiles were generated for each compound (Figure S2). After 5 days of exposure, ONO-1301 (2 µM) and purmorphamine (0.5 µM) significantly increased the number of viable NHF cells from all donors by twofold, whereas SUN11602 had no effect.

Next, we performed gene expression analyses using reverse transcription-quantitative polymerase chain reaction (RT-qPCR) on compound-treated primary NHFs (Figure 2b) to assess markers of proliferation (CCND1, MKI67), differentiation (ACTA2), overall fibroblast activation (COL1A1, ELN), and relevant healing response markers (VEGFA, MMP1). Data in Figure 2b show that, after 48 hours of compound exposure, only ONO-1301 (2 µM) stimulated the expression of VEGFA and MMP1; SUN11602 (100 µM) did not significantly alter any of the monitored target genes, and purmorphamine (0.5 µM) reduced the expression of VEGFA, COL1A1, and ELN. Full dose-dependent expression profiles after 48 hours are shown in Figure S3.

Additionally, we evaluated the effect of all compounds on the migratory activity of primary NHF cells using a wound healing assay (Figure 2c). High serum (10% fetal bovine serum [FBS]) served as a positive control, increasing migration by approximately twofold compared to the low serum (0.1% FBS) basal medium control. None of the tested compounds dramatically affected scratch wound closure by cell migration. Only ONO-1301 exhibited a mild but significant effect, reducing the migratory capacity of these cells. These data suggest that ONO-1301 efficiently stimulates fibroblast proliferation and reduces cell migration, consistent with a reported PKA-dependent inhibition of fibroblast migration by prostacyclin analogs. This activity may contribute to ONO-1301′s general antifibrotic mode of action. Moreover, ONO-1301 was the only compound that triggered the expression of MMP1 and VEGFA, indirectly acting as a small molecule “VEGF mimetic” on primary human fibroblasts. High VEGF levels typically accelerate and improve skin healing by promoting angiogenesis and restoring microcirculation. Increased MMP1 expression further supports the antifibrotic and wound remodeling potential of ONO-1301.

We then focused on the functional profiling of the three small molecule growth factor mimetics on skin keratinocytes. Differentiation and cellular activity states of primary normal human keratinocytes (NHKs) were captured by RT-qPCR-based gene expression analysis (Figure 3a), while full dose-dependent expression profiles after 48 hours are shown in Figure S4. In analogy to our studies in fibroblasts, VEGFA and MMP1 expression levels were monitored as markers for wound healing-relevant regenerative cell responses. Keratins-10 and -14 (KRT10, KRT14), involucrin (IVL), loricrin (LORICRIN), and transglutaminase 1 (TGM1) were analyzed as differentiation markers for keratinocytes.

Interestingly, only the putative FGF-2 mimetic SUN11602 (100 µM) increased keratinocyte differentiation, as nearly all expression markers (KRT10, IVL, LORICRIN, TGM1) were significantly altered. Both VEGFA and MMP1 expression were enhanced, suggesting that SUN11602 mimics typical FGF-2 activity on skin cells. The threefold increase in VEGFA levels was statistically significant, and the extent of MMP1 expression stimulated by SUN11602 (approximately 16-fold) was unparalleled among the compounds tested, potentially beneficial for wound remodeling. In contrast, ONO-1301 (2 µM) and purmorphamine (0.5 µM) only weakly enhanced MMP1 expression. Consistent with these results, both ONO-1301 and purmorphamine were inactive regarding the migratory capacity of NHKs, as verified by our wound healing assay (Figure 3b). However, SUN11602 inhibited keratinocyte migration by approximately twofold, aligning with its potent stimulation of cell differentiation. During later stages of wound healing, keratinocytes must cease migration to initiate differentiation. Taken together, SUN11602 appears to uniquely affect keratinocyte phenotypes that may be beneficial for skin regeneration.

Stimulation of keratinocyte differentiation is widely recognized as a critical component of late-stage wound healing, particularly through the restoration of the skin barrier and the formation of impermeable external epidermal layers. These features, coupled with VEGF-triggered angiogenesis and MMP-1-mediated promotion of the remodeling phase of wound healing via the degradation of excessive extracellular matrix, position SUN11602 as a promising tool for advancing the later stages of wound healing.

Large, difficult-to-heal, and chronic wounds, regardless of their underlying causes, remain a significant therapeutic challenge. While state-of-the-art wound care and split-thickness autografts represent the current standards for managing complex wounds, novel approaches from regenerative medicine, such as the development of bioengineered skin grafts, are gaining traction. However, the success and effectiveness of these skin substitutes depend on the inclusion of molecules that support tissue regeneration across multiple levels. Replacing growth factors and cytokines with highly potent small molecule mimetics or activators of the underlying signaling events could offer favorable pharmacokinetic and pharmacodynamic profiles for next-generation, chemically defined engineered skin grafts. Even when genuine, fully functional growth factor mimicry is not feasible, a small molecule-mediated enhancement or sustained downstream signaling activity could still hold therapeutic value.

In this study, we evaluated three small molecules for their potential to mimic or mediate the biological activities of SHH, FGF-2, or VEGF signaling, which are key pathways involved in skin regeneration. Phenotypic and functional analysis of cultured primary skin fibroblasts and keratinocytes revealed unique, cell-specific, and complementary activity profiles for the FGF-2 mimetic SUN11602 and the putative VEGF mimetic ONO-1301, while purmorphamine showed limited activity. The prostacyclin receptor agonist ONO-1301 primarily affected skin fibroblasts, promoting their proliferation, inhibiting migration, and stimulating the expression of MMP1 and VEGFA. In contrast, SUN11602 selectively stimulated skin keratinocyte differentiation and their expression of MMP1 and VEGFA. These activities align with processes essential for efficient skin regeneration following injury. A combined small molecule FGF-2/VEGF mimicry approach could enhance angiogenesis-related microcirculation, reduce early excessive fibrosis (via ONO-1301), and facilitate wound remodeling at later stages (via SUN11602), offering a promising strategy for improving outcomes in wound healing.

CONCLUSION

Concluding from our findings, SUN11602 and ONO‐1301 together may act as a valuable tool for improving the quality of wound healing. Building on these results, additional studies will have to further delineate the best application methods for these molecules, as well as their stability and compatibility with commonly applied biomaterials in tissue engineering.

EXPERIMENTAL

Small molecules

SUN11602 (Catalog No. 4826) and Purmorphamine (Catalog No. 4551) were obtained from Tocris Bioscience. ONO-1301 (Catalog No. O2264) was purchased as a mixture of E/Z-isomers from Sigma-Aldrich. Stock solutions were prepared by dissolving the compounds in dimethyl sulfoxide (DMSO) (Boom B.V. Meppel), as recommended by the manufacturers. All cell groups were treated with the same concentration of the vehicle (0.45% DMSO).

General cell culture methods

Primary normal human keratinocytes (NHKs) were isolated from skin explants obtained after abdominoplasty, as previously described. These cells were cultured using keratinocyte growth medium (KGM; Lonza) and differentiated by depleting growth factors, as outlined in earlier studies. Primary normal human dermal fibroblasts (NHF) from healthy donors were isolated according to previously established protocols and maintained in Dulbecco’s Modified Eagle’s Medium (DMEM; Gibco, Life Technologies, Inc.) supplemented with 10% fetal bovine serum (FBS; Gold, Australian origin; PAA Laboratories GmbH) and 1% penicillin/streptomycin (10,000 U/100 mg/mL; Gibco, Life Technologies, Inc.). Cells were cultured at 37°C in a humidified atmosphere containing 5% CO2.

Cell proliferation assay

NHF cells were seeded at a density of 2000 cells/well in 96‐well plates and allowed to adhere for 24 h at 37°C. Standard medium was replaced by low‐serum medium containing 0.1% FBS and the plate was incubated for 16 h, after which different concentrations of small molecules were added to the cells: SUN11602 (25–100 μM), ONO‐1301 (0.5–2 μM), and purmorphamine (0.1–0.5 μM). The plates were incubated for a maximum period of 120 h and cell viability was assessed every 24 h using Alamar Blue® Cell Viability Assay (Thermo Fisher Scientific, Inc.), following the manufacturer’s instructions. Briefly, culture medium was replaced by Alamar Blue solution and plates were incubated for 4 h. After incubation, fluorescence was assessed at an excitation wavelength of 540 nm and emission of 620 nm in a microplate reader (Synergy 2; BioTek).

Quantification of messenger RNA expression levels

For gene expression analysis, NHF cells were seeded in 24-well plates at a density of 3 × 10⁴ cells per well in DMEM with 10% FBS and incubated at 37°C. After 24 hours, the standard medium was replaced with DMEM containing 0.5% FBS and supplemented with SUN11602 (25–100 µM), ONO-1301 (0.5–2 µM), or purmorphamine (0.1–0.5 µM). NHK cells were seeded in 24-well plates at a density of 2 × 10⁴ cells per well in keratinocyte growth medium (KGM; Lonza) and cultured until they reached confluence. The medium was then replaced with keratinocyte basal medium containing SUN11602 (25–100 µM), ONO-1301 (0.5–2 µM), or purmorphamine (0.1–0.5 µM). Control cells were exposed to medium containing 0.45% DMSO (vehicle). Cells were harvested for RNA isolation after 48 hours of treatment. For keratinocytes, the expression levels of the following genes were assessed: VEGFA, MMP1, KRT14, KRT10, IVL, LORICRIN, and TGM1. For fibroblasts, the genes analyzed included VEGFA, MMP1, CCDN1, MKI67, ACTA2, COL1A1, and ELN (Table S1).

RNA isolation and quantitative polymerase chain reactions

To isolate RNA, the cells were washed with phosphate-buffered saline (PBS) and then 250 µL of TRIzol reagent (Invitrogen) was added. After vigorous pipetting to lyse the cells, the lysate was transferred to 1.5 mL Eppendorf tubes. Next, 1/5 of the lysate volume (chloroform, Boom B.V. Meppel) was added, followed by vortexing for 30 seconds. The samples were centrifuged at 17,000g for 20 minutes at 4°C. From the aqueous layer, 150 µL was carefully transferred to a new 1.5 mL Eppendorf tube. An equal volume of 70% ethanol (Boom B.V. Meppel) was added to the aqueous layer and mixed briefly by inversion. The samples were then purified using RNeasy mini kit columns (Qiagen GmbH) according to the manufacturer’s protocol.

Following purification, an on-column DNase I (RNase-free DNase I; Qiagen) treatment was performed to eliminate any contaminating genomic DNA. The RNA yield was quantified spectrophotometrically at 260 nm using the NanoDrop (Thermo Fisher Scientific). Gene expression levels were measured using quantitative polymerase chain reactions (qPCR). For complementary DNA (cDNA) synthesis, a 20-µL reaction volume was prepared using 50 ng of RNA (in 15 µL RNase-free water) and the iScript cDNA synthesis kit (Bio-Rad), as per the manufacturer’s instructions. As a no-reverse transcription control, 15 µL of water was used. The synthesized cDNA was diluted 20-fold in RNase-free water and used as a template for qPCR amplification (CFX real-time detection system; Bio-Rad) of the genes listed in Table S1.

Each qPCR reaction was carried out in a total volume of 12.5 µL, containing 2.5 µL of cDNA (1.6 ng), 0.6 µM primers (obtained from Biolegio, Nijmegen), and 6.25 µL of iQ SYBR Green SuperMix (Bio-Rad). The reaction was initiated by incubation at 95°C for 3 minutes, followed by 40 cycles of denaturation at 95°C for 15 seconds and annealing/extension at 60°C for 30 seconds. To ensure the purity of the qPCR product, the melting temperature was determined by heating the samples from 65°C to 95°C with 0.5°C increments every 10 seconds. The quantification cycles (Cq) for the genes of interest were normalized to the Cq value of the reference gene ribosomal protein large P0 (RPLP0). The relative quantity was calculated using the 2^(-ΔΔCt) method, where cells treated with only the vehicle DMSO (0.45%) served as the control. No-template controls (2.5 µL of water instead of cDNA) and no-reverse transcription controls were included on each plate for each gene. Cq values higher than 38 were excluded from analysis. All primers were validated to ensure PCR amplification efficiency (E) fell within the acceptable range of 100% ± 10%, as recommended in the MIQE guidelines. Primers meeting this criterion were used for subsequent experiments.

Statistical analysis

Independent triplicates (\( N = 3 \) were performed for all experiments, except for gene expression analysis, which utilized Student’s t-test for statistical evaluation. A \( p < 0.05 \) was considered significant, and all data were analyzed using GraphPad Prism version 6.00.

The results were expressed as mean ± SD, and all values were obtained from the no-template controls (2.5 µL of water instead of cDNA) and no-reverse transcription controls were used. Quantification cycles (Cq) higher than 38 were not used for analysis. Primers meeting the required efficiency criteria were used for the subsequent experiments.

To calculate the relative quantity, the 2^(-ΔΔCt) method was applied, following the manufacturer’s protocol. The samples were purified using RNeasy mini kit columns (Qiagen GmbH). The RNA yield was quantified spectrophotometrically at 260 nm using the NanoDrop (Thermo Fisher Scientific). Gene expression levels were measured using quantitative polymerase chain reactions (qPCR). For complementary DNA (cDNA) synthesis, a 20-µL reaction volume was prepared using 50 ng of RNA (in 15 µL RNase-free water) and the iScript cDNA synthesis kit (Bio-Rad) as per the manufacturer’s instructions. The samples were purified using RNeasy mini kit columns (Qiagen).

The melt temperature was measured by increasing the temperature from 65°C to 95°C with 0.5°C increments every 10 seconds. Quantification cycles (Cq) higher than 38 were not used for analysis. Primers with an acceptable efficiency of 100% ± 10% were used for the subsequent qPCR experiments. The relative quantity was calculated using the 2^(-ΔΔCt) method, where cells treated with only the vehicle DMSO (0.45%) served as the control. The samples were purified using RNeasy mini kit columns (Qiagen GmbH) according to the manufacturer’s protocol. The \( p < 0.05 \) was considered significant, and all data were analyzed using the \( N = 3 \) independent triplicates, as recommended in the MIQE guidelines.