LIMK2 inactivation suppresses mechanical stimulation-induced dermal fibroblast differentiation and resistance to apoptosis

Abstract

Abnormal scar formation is a clinical challenge with limited therapeutic options. Mechanical stimulation is implicated in abnormal scarring. Accordingly, the present study investigated the role of LIMK2, a component of the Rho/ROCK/LIMK/cofilin pathway, in cell differentiation and apoptosis in response to mechanical stimulation and proliferation in human dermal fibroblasts (HDFs). In normal HDFs, expression of α-smooth muscle actin (α-SMA), a marker of differentiation into myofibroblasts, significantly increased with mechanical stimulation; however, this change was not observed when LIMK2 was inactivated. Mechanical stimulation increased expression of the anti-apoptotic protein Bcl-2 and decreased that of the pro-apoptotic protein BAX in controls, but these effects were not observed with LIMK2 inactivation. Moreover, fibroblast proliferation was inhibited with LIMK2 inactivation. These findings suggest that LIMK2 inactivation suppresses mechanical stimulation-induced myofibroblast differentiation and resistance to apoptosis, and also inhibits HDF proliferation, highlighting LIMK2 as a potential therapeutic target for the prevention and treatment of abnormal scars.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-37610-y.

Introduction

The wound healing process is activated upon skin injury, initiating a cascade of cellular events involving proliferation, migration, and interaction with the extracellular matrix. This intricate and regulated sequence of events ultimately results in scar formation¹. The scar develops from granulation tissue, with myofibroblasts contributing to scar contraction². Disruptions in this process can lead to fibroproliferative abnormal scar formation such as hypertrophic scars and keloids, characterized by enhanced expression and deposition of the extracellular matrix¹. Although the precise mechanisms underlying the formation of abnormal scars are not fully elucidated, increased differentiation of fibroblasts into myofibroblasts³, dysregulated apoptosis⁴, excessive fibroblast proliferation¹, and abnormalities in signaling pathways involved in wound healing⁵ are observed in their formation.

In wound healing, fibroblasts differentiate into myofibroblasts. These cells produce extracellular matrix and play a pivotal role in abnormal scar formation¹. As wound epithelialization advances and scar formation occurs, there is a subsequent depletion of cells, including myofibroblasts. Apoptosis plays a key role in this process, avoiding hypertrophic scar and keloid formation^6,7. In dermal fibroblasts, mechanical stimulation positively regulates α-smooth muscle actin (α-SMA), a marker of fibroblast differentiation into myofibroblasts⁸, by promoting actin assembly through the phosphorylation of cofilin^9,10, and also reduces the expression of apoptosis-promoting genes while increasing the expression of anti-apoptotic genes in fibroblasts¹¹. This suggests that mechanical stimulation may inhibit apoptosis, potentially causing abnormal scar formation. In fact, dermal fibroblasts derived from keloids show stronger resistance to apoptosis than those from normal scars¹².

The Rho/Rho kinase (ROCK)/LIMK/cofilin pathway is an important signaling pathway in fibrosis¹³, as it regulates the actin cytoskeleton and influences various cellular functions, including cell morphology, motility, and proliferation^14,15. Some studies have reported that inhibition of the Rho pathway inhibits cell growth, reduces proliferative capacity, and induces apoptosis in prostatic smooth muscle cells, bladder cancer cells, urothelial cancer cells, and small cell lung cancer cells^16–18. LIMK2, a protein kinase downstream of the Rho pathway, is expressed in a wide range of tissues and phosphorylates the cofilin family of actin depolymerization factors to regulate the actin cytoskeleton¹⁹. A recent study indicated that LIMK2 contributes to fibrosis in wound healing and its deletion inhibits scar expansion and detrimental remodeling in myocardium²⁰. These findings suggest that LIMK2 may be a potential therapeutic target in fibrotic disorders including abnormal scar formation in the skin.

Based on findings from these studies, the present study aimed to examine the proliferative potential (with a focus on LIMK2) and apoptosis of dermal fibroblasts, as well as their differentiation into myofibroblasts in response to mechanical stimulation.

Results

Inactivation of LIMK2 inhibits mechanical stimulation-induced differentiation of fibroblasts into myofibroblasts

The Rho/ROCK/LIMK/cofilin pathway is a major mechanical signaling pathway that plays a pivotal role in α-SMA gene expression. Rho/ROCK activation causes phosphorylation of LIMK and cofilin, which in turn increases actin assembly and positively controls α-SMA expression^9,13. Therefore, we examined whether inactivation of LIMK2 affects mechanical stimulation-induced α-SMA expression in human dermal fibroblasts (HDFs). Given that a LIMK2 mutant in which threonine 505 is replaced with valine abolishes LIMK2 activity¹⁹, we created HDFs expressing inactive LIMK2 by infecting HDFs with an AAV2 vector expressing LIMK2-inactive mutant fused to GFP, in which threonine 505 was mutated to valine (LIMK2-IA-mutant-GFP). We confirmed that the infected cells express exogenous LIMK2 by Western blotting (Fig. 1a), and that LIMK2-IA-mutant-GFP expression reduced the baseline level of cofilin phosphorylation (Fig. 1b). Similarly, HDFs expressing active LIMK2 were created with a vector expressing LIMK2-active mutant fused to GFP, in which threonine 505 was mutated to glutamate (LIMK2-A-mutant-GFP), as described previously¹⁰.

Fig. 1.

(a) AAV2 expressing LIMK2-inactive (IA)-mutant-GFP and AAV2 expressing GFP (control) were used to infect normal HDFs. Expression levels of LIMK2 and GFP were analyzed by Western blotting. The original blots are presented in Supplementary Fig. 7. (b) Decreased baseline cofilin phosphorylation 48 h after LIMK2-IA-mutant-GFP infection. In HDFs and GFP-HDFs (both controls), mechanical stimulation for 30 min increased p-cofilin levels, whereas in LIMK2-IA-mutant-GFP-HDFs, p-cofilin levels remained unchanged. (c) In LIMK2-IA-mutant-GFP-HDFs, α-SMA expression was not changed at 24 h after mechanical stimulation. In contrast, in HDFs and GFP-HDFs, α-SMA expression was significantly increased by mechanical stimulation. (d) LIMK2 activation increased, whereas LIMK2 inactivation decreased, cofilin phosphorylation in HDFs and KDFs. (e) LIMK2 activation increased α-SMA expression in both HDFs and KDFs, whereas LIMK2 inactivation decreased α-SMA expression only in KDFs. (f) LIMK2 activation increased, whereas LIMK2 inactivation decreased, collagen type I expression in HDFs and KDFs. **p < 0.01. *p < 0.05. ^##p < 0.01 vs HDFs and GFP-HDFs. The original blots are presented in Supplementary Figs. 8, 9, 10, 11 and 12.

HDFs infected with AAV expressing LIMK2-IA-mutant-GFP, HDFs infected with AAV expressing GFP only (control), and non-infected HDFs (another control) were mechanically stretched, and phosphorylated cofilin (p-cofilin) and α-SMA levels were detected by Western blotting. Zhao et al. reported that treatment of cells with 10–30 min of mechanical force increased cofilin phosphorylation⁹. Accordingly, we assessed p-cofilin levels after applying 30 min of mechanical stimulation. Since α-SMA expression was upregulated 24 h after mechanical stimulation in previous reports^10,21, cells were subjected to mechanical stimulation for 24 h to assess α-SMA expression. Both p-cofilin levels and α-SMA expression were significantly increased by mechanical stimulation in HDFs and HDFs expressing only GFP (Fig. 1b and c). In contrast, HDFs expressing LIMK2-IA-mutant-GFP showed no alteration in p-cofilin levels (Fig. 1b) and α-SMA expression (Fig. 1c) after mechanical stimulation. These results suggest that mechanical stimulation promotes the differentiation of fibroblasts into myofibroblasts as reported previously¹⁰, and that inactivation of LIMK2 inhibits differentiation of fibroblasts by inhibiting cofilin phosphorylation following mechanical stimulation.

We also examined the influence of activation and inactivation of LIMK2 on cofilin phosphorylation and α-SMA baseline expression in untreated HDFs and HDFs derived from keloid tissues (KDFs). In both HDFs and KDFs, LIMK2 inactivation reduced baseline cofilin phosphorylation (Fig. 1d) and α-SMA expression (Fig. 1e), and, in contrast, LIMK2 activation increased cofilin phosphorylation (Fig. 1d) and α-SMA expression (Fig. 1e). Since excessive collagen type I synthesis is known to occur in keloids²², we evaluated collagen type I expression by Western blot analysis. Collagen type I expression decreased following LIMK2 inactivation and increased following LIMK2 activation in both HDFs and KDFs (Fig. 1f). Since the Rho/ROCK/LIMK/cofilin pathway is associated with cell contractility and motility²³, we investigated those functions in HDFs expressing inactive LIMK2. We performed a collagen gel contraction assay, an in vitro method that models myofibroblast-promoted wound contraction in a collagen gel matrix²⁴. Although TGF-β1 treatment (positive control) significantly reduced the gel size compared to untreated control HDFs (Fig. 2a) and KDFs (Fig. 2b), inactivation of LIMK2 inhibited gel contraction, suggesting that LIMK2 inactivation inhibits cell contractility. We next performed a wound healing assay to examine the influence of LIMK2 inactivation on cell motility. A significant decrease in cell motility was observed in both HDFs and KDFs with LIMK2 inactivation compared to untreated HDFs and KDFs (Fig. 2c and d), demonstrating that LIMK2 inactivation decreases cell motility. Notably, however, LIMK2 inactivation inhibits cell proliferation (as described below) and may thus influence the outcome of the wound healing assay.

Fig. 2.

(a) Images showing results of gel contraction experiment in HDFs. Bar = 1 mm. Gel size in each condition was measured at 72 h. Size of gel containing untreated HDFs was used as the control. Each bar shows the ratio to the control. TGF-β1 treatment significantly reduced the gel size. LIMK2 inactivation resulted in less contraction compared to the control. (b) Images showing results of gel contraction experiment in KDFs. Size of gel containing untreated KDFs was used as the control. Each bar shows the ratio to the control. Similar trends to those observed in HDFs were also observed in KDFs. (c) HDF cell monolayers at 0 and 12 h after wounding. The lines indicate the wound edge at the start of the experiment (0 h). Bar = 500 µm. The migration rate was expressed as migration distance (µm/h). LIMK2 inactivation reduced the migration rate compared to the control. (d) KDF cell monolayers at 0 and 12 h after wounding. Similar to HDFs, LIMK2 inactivation reduced the migration rate. **p < 0.01. *p < 0.05.

To further elucidate the involvement of the pathway linking LIMK2/cofilin to actin dynamics, we used latrunculin A, a pharmacological inhibitor that binds to G-actin and inhibits actin polymerization, to mimic the condition resulting from LIMK2 inactivation²⁵. Latrunculin A treatment reduced baseline α-SMA (Supplementary Fig. 1) and collagen type I (Supplementary Fig. 2) expression both in HDFs and KDFs, similar to the results obtained with LIMK2 inactivation, although inactivation of LIMK2 did not decrease baseline α-SMA expression in untreated HDFs (Fig. 1e). These results suggest that decreased α-SMA and collagen type I expression due to LIMK2 inactivation may be implicated in the pathway linking LIMK2/cofilin to actin dynamics.

LIMK2 inactivation influences apoptosis of HDFs under mechanical stimulation

When culturing the cells, we noticed that HDFs infected with LIMK2-IA-mutant-GFP showed a minimal increase in cell number. Therefore, we speculated that LIMK2 inactivation could have an effect on apoptosis and/or proliferation. Indeed, an important factor associated with keloid formation is dysregulation of apoptosis, and it has been demonstrated that the percentage of apoptotic cells is higher in dermal fibroblasts derived from normal skin than those derived from keloids¹². Previous studies also have reported that mechanical stimulation of fibroblasts prevents apoptosis²⁶, upregulates Bcl-2 (an anti-apoptotic gene), and downregulates BAX (a proapoptotic gene)¹¹.

The Rho/ROCK/LIMK/cofilin pathway is also involved in apoptosis signaling^17,18. Specifically, Rho inhibition reduces the expression of Bcl-2 in endothelial cells²⁷, and increases the expression of BAX in vascular smooth muscle cells and myocytes^28,29. Therefore, we further examined the effects of LIMK2 inactivation on the regulation of apoptosis following the mechanical stimulation of HDFs. Referring to a protocol for mechanical stimulation used for atrial fibroblasts in a previous study³⁰, mechanical stimulation was applied for 6 h to assess Bcl-2 expression. With respect to BAX, a DNA microarray analysis previously showed a change in BAX expression in HDFs after 24 h of mechanical stimulation¹¹. Therefore, 24 h of mechanical stimulation was applied to assess BAX expression. Mechanical stimulation significantly increased Bcl-2 expression (Fig. 3a), while BAX expression was significantly reduced following mechanical stimulation (Fig. 3b) in control HDFs and control GFP-HDFs. However, mechanical stimulation of LIMK2-IA-mutant-GFP-HDFs did not result in altered expression of Bcl-2 or BAX (Fig. 3a and b). Moreover, Bcl-2 expression was significantly decreased in LIMK2-IA-mutant-GFP-HDFs before mechanical stimulation (Fig. 3a). These results suggest that mechanical stimulation induces anti-apoptotic signaling in fibroblasts, but LIMK2 inactivation suppresses baseline anti-apoptotic protein expression and may prevent the activation of anti-apoptotic signaling by mechanical stimulation. Regarding LIMK2 activation, a baseline increase in Bcl-2 expression was observed in HDFs and KDFs infected with AAV expressing the LIMK2-A-mutant (Fig. 3c), while baseline BAX expression was not altered (Fig. 3d). Furthermore, to investigate the involvement of actin dynamics in the apoptosis of fibroblasts, we examined the response of fibroblasts to the actin polymerization inhibitor latrunculin A. Latrunculin A treatment decreased Bcl-2 expression (Supplementary Fig. 3) and increased BAX expression in both HDFs and KDFs (Supplementary Fig. 4), suggesting that the pathway linking LIMK2/cofilin to actin dynamics is implicated in the apoptosis of HDFs.

Fig. 3.

(a) Mechanical stimulation for 6 h increased Bcl-2 expression in the two control groups. In LIMK2-IA-mutant-GFP-HDFs, Bcl-2 expression was decreased before mechanical stimulation, and mechanical stimulation did not alter Bcl-2 expression. (b) Mechanical stimulation for 24 h decreased BAX expression in the two control groups. However, BAX expression did not change in LIMK2-IA-mutant-GFP-HDFs. (c) LIMK2 activation increased Bcl-2 expression both in HDFs and KDFs. (d) LIMK2 activation did not alter BAX expression. **p < 0.01. ^#p < 0.05 vs HDFs and GFP-HDFs. The original blots are presented in Supplementary Figs. 13, 14, 15 and 16.

We also investigated whether LIMK2 inactivation promotes apoptosis by fluorescence staining of annexin V and propidium iodide (PI). Early apoptotic cells are annexin V-positive and PI-negative, whereas late (end-stage) apoptotic cells are annexin V/PI-double-positive³¹. Fluorescence staining images after 24 h of mechanical stimulation are shown in Fig. 4a. Mean percentages of apoptotic cells (annexin V-positive, PI-negative cells, and annexin V/PI-double-positive cells) for unstretched and stretched HDFs were 4.59 ± 0.53% and 3.40 ± 0.23%, respectively; percentages for unstretched and stretched GFP-HDFs were 5.01 ± 0.55% and 3.65 ± 0.81%, respectively; and percentages for unstretched and stretched LIMK2-IA-mutant-GFP-HDFs were 6.88 ± 0.90% and 7.82 ± 0.42%, respectively (Fig. 4b). Mechanical stimulation significantly reduced apoptosis in both control HDFs. In contrast, baseline apoptosis was significantly higher in LIMK2-IA-mutant-GFP-HDFs, and mechanical stimulation did not reduce apoptosis, unlike that observed in the control groups (Fig. 4b). These results suggest that mechanical stimulation has an anti-apoptotic effect in fibroblasts, and that inactivation of LIMK2 induces apoptosis of fibroblasts and abolishes the anti-apoptotic effect of mechanical stimulation.

Fig. 4.

(a) Fluorescence staining with annexin V (yellow), propidium iodide (PI) (red), and Hoechst 33,342 nuclear staining dye (blue). (b) The percentage of apoptotic cells (annexin V-positive, PI-negative cells, and annexin V/ PI-double-positive cells) was calculated by observing under a microscope after fluorescent staining. Mechanical stimulation reduced apoptosis in the two control groups. Baseline apoptosis was high in LIMK2-IA-mutant-GFP-HDFs, and mechanical stimulation did not reduce apoptosis. *p < 0.05. ^##p < 0.01 vs control HDFs and GFP-HDFs. (c, d) Quadrants showing annexin V/PI flow cytometry plots: viable cells (annexin V − /PI −), early apoptotic cells (annexin V + /PI −), late apoptotic cells (annexin V + /PI +), and necrotic cells (annexin V − /PI +). The percentage of apoptotic cells was quantified using flow cytometry: (c) HDFs; (d) KDFs. LIMK2 inactivation increased apoptosis both in HDFs and KDFs. **p < 0.01.

To support the data obtained from fluorescence staining of apoptotic cells, we further assessed apoptosis by flow cytometry and observed the baseline apoptotic ratio in HDFs and KDFs by transducing either active or inactive LIMK2. LIMK2 inactivation increased apoptosis in both HDFs and KDFs (Fig. 4c and d), while LIMK2 activation resulted in no significant difference compared to untreated HDFs or KDFs (Fig. 4c and d). Regarding the influence of actin polymerization inhibition, latrunculin A treatment also increased apoptosis, similarly to the findings obtained with LIMK2 inactivation (Supplementary Fig. 5a and b).

LIMK2 inactivation suppresses HDF proliferation

LIMK2 regulates actin filament dynamics, which in turn impacts cell proliferation³². LIMK inhibitors have been shown to inhibit cell proliferation in several proliferative diseases, including breast cancer, prostate cancer, and brain tumors^33–35. On the other hand, histological analyses of keloids have shown increased fibroblasts density and proliferation^1,22. Therefore, we investigated whether LIMK2 inactivation can suppress fibroblast proliferation.

First, we counted the number of untreated control HDFs, GFP-HDFs, LIMK2-A-mutant-GFP-HDFs, and LIMK2-IA-mutant-GFP-HDFs for up to 96 h after seeding the cells. Fold changes in cell numbers over time are shown in Fig. 5a. Significant differences were observed in fold changes in cell numbers between the two control groups and LIMK2-IA-mutant-GFP-HDFs at 24 h after seeding the cells and thereafter. There were no significant differences between LIMK2-A-mutant-GFP-HDFs and the two control groups.

Fig. 5.

(a) Fold change in cell number. The cell count ratio was significantly lower in LIMK2-IA-mutant-GFP-HDFs compared to the two control groups at 24 h post-culture and thereafter. LIMK2 activation did not affect cell proliferation. ^##p < 0.01. ^#p < 0.05 vs LIMK2-IA-mutant-GFP-HDFs. (b) Fluorescence staining images with BrdU incorporation (red) and DAPI nuclear staining (blue) 24 h after BrdU incorporation. (c) The percentages of BrdU-positive cells were calculated. Significant differences were observed between the two control groups and LIMK2-IA-mutant-GFP-HDFs. LIMK2-A-mutant-GFP-HDFs did not show significant differences compared to the two control groups. **p < 0.01. *p < 0.05.

Next, we assessed cell proliferation with a BrdU incorporation assay (Fig. 5b), which detects DNA-synthesizing cells³⁶. The BrdU-positive cell rate 24 h after BrdU incorporation was 12.15 ± 3.52% in untreated control HDFs, 13.38 ± 3.11% in GFP-HDFs, 11.01 ± 2.15% in LIMK2-A-mutant-GFP-HDFs, and 5.16 ± 1.06% in LIMK2-IA-mutant-GFP-HDFs (Fig. 5c). Significant decrease in % BrdU positive cells was observed in LIMK2-IA-mutant-GFP-HDFs compared to untreated HDFs, GFP-HDFs, and LIMK2-A-mutant-GFP-HDFs. A similar trend was observed in KDFs infected with AAV expressing the active or inactive mutant (Supplementary Fig. 6a). These results indicate that LIMK2 activation does not affect proliferation, whereas LIMK2 inactivation inhibits cell proliferation. Latrunculin A treatment reduced cell proliferation, similarly to the results obtained with LIMK2 inactivation (Supplementary Fig. 6b and c).

Discussion

Abnormal scars, such as hypertrophic scars and keloids, present significant clinical challenges due to their complex pathology and limited treatment options. The mechanism underlying the formation of abnormal scars is not yet fully understood, but these scars are often found in areas which are frequently mobile or subjected to high stretching tension such as the anterior chest, scapular, and suprapubic regions, suggesting the involvement of mechanical stimulation^11,37. Little is known about how dermal fibroblasts respond to mechanical stimulation, although the Rho pathway is considered to be implicated in the mechanical stimulation-induced promotion of α-SMA expression¹³ and thus myofibroblast differentiation. Factors such as apoptosis^12,26 and cell proliferation¹ are also involved in abnormal scar formation. In the present study, we investigated the differentiation of dermal fibroblasts into myofibroblasts and apoptosis in response to mechanical stimulation, with a focus on LIMK2. We previously reported that the differentiation of fibroblasts into myofibroblasts was unresponsive to mechanical stimulation in fibroblasts derived from neurofibromatosis type 1 (NF1) patients¹⁰, who are known to rarely form abnormal scars³⁸. In NF1-derived fibroblasts, the unresponsiveness of myofibroblasts to differentiate upon mechanical stimulation was considered to result from dysfunction of actin polymerization, which is downstream of the Rho/ROCK/LIMK2/cofilin pathway. Therefore, in the present study, we investigated how LIMK2 inactivation might affect myofibroblast differentiation in response to mechanical stimulation, as well as apoptosis and the proliferation of fibroblasts, which is crucial for abnormal scar formation.

Mechanical stimulation has been reported to increase α-SMA expression and promote myofibroblast differentiation³. However, LIMK2 inactivation did not increase α-SMA expression following mechanical stimulation in the present study (Fig. 1c), suggesting the possibility of inhibiting the differentiation of dermal fibroblasts into myofibroblasts, a cause of abnormal scar formation^1,3. On the contrary, LIMK2 activation increased baseline α-SMA in both HDFs and KDFs (Fig. 1e). Intracellular collagen type I expression was increased by LIMK2 activation, but was decreased by LIMK2 inactivation (Fig. 1f). These results suggest that LIMK2 may influence collagen synthesis. Although further analysis is necessary, given that fibroblasts are the primary source of extracellular matrix proteins³⁹, the excessive matrix deposition seen in abnormal scars may be suppressed by targeting LIMK2, as indicated by the reduction in intracellular collagen associated with LIMK2 inactivation. In addition, LIMK2 inactivation reduced collagen gel contraction and motility of HDFs and KDFs (Fig. 2). Since keloid fibroblasts are known to have enhanced contractility and motility⁴⁰, LIMK2 inactivation may represent a potential therapeutic approach to suppress their pathological behavior.

We found that LIMK2 inactivation enhanced apoptosis. In endothelial cells, Bcl-2 is downregulated by inhibition of the Rho pathway²⁷. We similarly demonstrated that LIMK2 inactivation suppresses Bcl-2 expression in HDFs. There were no significant differences in baseline expression of BAX between the three assessed groups, although previous studies reported that inhibition of ROCK increases BAX expression in vascular smooth muscle cells and myocytes^28,29. However, while BAX (a pro-apoptotic protein) decreased and Bcl-2 (an anti-apoptotic protein) increased in response to mechanical stimulation in the control groups, which is consistent with previous reports¹¹, these changes were not observed when LIMK2 was inactivated. In addition to alterations in apoptosis-related protein expression, we compared changes in apoptosis induced by mechanical stimulation and LIMK2 inactivation. In the control groups, mechanical stimulation decreased apoptosis, consistent with previous reports^11,26. However, in cells expressing inactive LIMK2, apoptosis was significantly elevated compared to the control groups before mechanical stimulation, and mechanical stimulation did not decrease apoptosis. These findings suggest that LIMK2 inactivation increases baseline apoptosis and counters the anti-apoptotic effects of mechanical stimulation. LIMK2 activation increased baseline Bcl-2 expression, with no change in BAX expression (Fig. 3c and d). Flow cytometry analysis revealed that, while LIMK2 inactivation enhanced apoptosis, LIMK2 activation resulted in no significant change both in HDFs and KDFs (Fig. 4c and d), despite the increase in Bcl-2 expression. This may be explained by the fact that the baseline apoptotic rate in control fibroblasts is inherently low (2.11 ± 0.54%), leaving little room for further reduction even when Bcl-2 is elevated by LIMK2 activation.

We also found that LIMK2 inactivation did not increase cell number. Since cell number is dictated by a delicate balance between cell proliferation and survival⁴¹, we investigated these processes further. LIMK2 inactivation significantly reduced cell proliferative capacity, as evidenced by a reduction in BrdU-positive cells (Fig. 5). Since fibroblasts are a key source of extracellular matrix proteins that drive keloid formation¹, the inhibition of fibroblast proliferation through LIMK2 inactivation may help limit the formation of these abnormal scars.

Collectively, these results suggest that LIMK2 inactivation decreases cell proliferation and increases apoptosis, leading to a reduction in cell number.These effects were consistent across both HDFs and KDFs, highlighting LIMK2 as a potential therapeutic target for abnormal scar formation.

This study extended analyses using the actin polymerization inhibitor latrunculin A, which acts downstream of the Rho/ROCK/LIMK/cofilin pathway⁴². Latrunculin A treatment reproduced the changes observed with LIMK2 inactivation: differentiation and proliferation were suppressed, and apoptosis was increased (Supplemental Figures 1, 5 and 6). These findings suggest that fibroblast differentiation, proliferation, and apoptosis are regulated by actin dynamics downstream of LIMK2.

A limitation of this study is that our analyses were performed only in cultured fibroblasts and not in clinical scar tissue. Future studies examining LIMK2 activation directly in patient-derived scar tissue will be essential to confirm and extend our findings.

In conclusion, we demonstrated that LIMK2 inactivation reduces differentiation of dermal fibroblasts into myofibroblasts, inhibits fibroblast proliferation, promotes apoptosis, and ameliorates changes induced by mechanical stimulation that possibly lead to keloid formation. Similar results were also observed upon inhibition of actin polymerization by latrunculin A, suggesting that LIMK2 functions are the key regulator within the cytoskeletal pathway. In addition, LIMK2 inactivation reduced fibroblast contractility and motility, as observed in keloid fibroblasts. Therefore, our findings highlight LIMK2 as a promising therapeutic target for the prevention and treatment of hypertrophic scars and keloids, offering potential advancements in the management of these challenging conditions.

Materials and methods

Cell culture

Normal human dermal fibroblasts (HDFs) were purchased from Zen-Bio Inc. (Research Triangle Park, NC, USA) and cultured in DMEM (Gibco, Life Technologies, CA, USA) containing 10% FBS (Gibco, Life Technologies) and 1% antibiotic–antimycotic solution (Gibco, Life Technologies). Cells were seeded in an incubator at 37 °C with 5% CO₂. Cells were passaged when they reached 90% confluence. Passages 3–8 were used for the experiments.

Primary keloid dermal fibroblasts (KDFs) were obtained from a keloid patient who underwent keloid resection. Informed consent was obtained from the patient. All procedures were approved by the Institutional Review Board of The University of Osaka Hospital (Approval No. 13442) and were performed in accordance with the approved guidelines. KDFs were isolated from the keloid sample using the explant method⁴³. Briefly, excised tissues were washed with PBS containing antibiotics several times, minced into pieces, and placed onto culture dishes. KDFs were cultured in DMEM (Gibco, Life Technologies) containing 10% FBS (Gibco, Life Technologies) and 1% antibiotic–antimycotic solution (Gibco, Life Technologies). Cells were seeded in an incubator at 37 °C with 5% CO₂. Cells were passaged when they reached 90% confluence. Passages 3–5 were used for the experiments.

Infection of HDFs with AAV2 vectors expressing GFP, LIMK2-inactive-mutant-GFP and LIMK2-active-mutant-GFP

The pAAV-CMV-LIMK2-GFP vector and AAV2-LIMK2-inactive (IA)-mutant-GFP vector expressing an inactive form of LIMK2 (505th base threonine (ACG) was mutated to valine (GTG)) was purchased from Takara Bio (Shiga, Japan). The pAAV-CMV-GFP vector was created by cloning GFP amplified from pAAV-CMV-LIMK2-GFP into the linearized pAAV-CMV vector. The 505th base (threonine (ACG)) of LIMK2 in the pAAV-CMV-LIMK2-GFP vector was mutated to glutamate (GAG) using KOD-Plus-Mutagenesis kit (Toyobo) to construct a pAAV-CMV-LIMK2-active (A)-mutant-GFP vector expressing an active form of LIMK2. The purified pAAV-CMV-GFP vector and pAAV-CMV-LIMK2-mutant-GFP vectors were each transformed into E. coli (DH5α high Champion™; Cosmo Bio, Tokyo, Japan), and the plasmid was purified according to the protocol of Midiprep (NucleoBond® Xtra Midi; MACHEREY–NAGEL, Düren, Germany).

HEK293 cells were transfected with pRC2-mi342 vector and pHelper vector (Takara Bio) in conjunction with either the pAAV-CMV-LIMK2-IA-mutant-GFP vector, the pAAV-CMV-LIMK2-A-mutant-GFP vector, or the pAAV-CMV-GFP vector. After 72 h, each AAV2 vector was purified using the AAVpro® Purification Kit (AAV2) (Takara Bio).

AAV2-LIMK2-IA-mutant-GFP, AAV2-LIMK2-A-mutant-GFP, or AAV2-GFP (control) was added to HDFs, which were incubated in DMEM containing 10% FBS at 37 °C for 48 h for infection.

Mechanical stimulation

As described previously¹⁰, 0.05% type1 collagen (Nitta Gelatin, Osaka, Japan) was applied to PDMS elastic cell-culture chambers (STB-CH-10, 32 × 32 × 10 mm, Strex, Osaka, Japan). Cells were seeded in the chamber at an initial density of 5 × 10⁵ cells/ chamber and incubated in 2 mL of DMEM containing 10% FBS and 1% antibiotic–antimycotic solution. Before applying mechanical stimulation, incubations were performed in DMEM containing 1% FBS for serum starvation for a 24-h period. Chambers were then set on the stretching device (STB-140, Strex) and uniaxial stretching (10 cycles/min, 15% of length) was performed in an incubator (37 °C, 5% CO₂).

The experimental group was divided into six groups: stretched and unstretched LIMK2-IA-mutant-GFP-HDFs, stretched and unstretched GFP-HDFs, and stretched and unstretched HDFs. Unstretched HDFs were cultured and underwent serum starvation in the same manner in the chamber, but were not stretched. Unstretched cells were collected at the same time as the stretched HDFs.

Western blot analysis

Following two PBS washes, collected cells were lysed in RIPA buffer (Fujifilm Wako Pure Chemical Co., Osaka, Japan) and protease inhibitor cocktail (Roche Diagnostics, Basel, Switzerland). Proteins were subjected to 5–20% gradient SDS-PAGE (e-PAGEL, ATTO Co., Tokyo, Japan) and electro- transferred onto PVDF membranes (Thermo Fisher Scientific Inc, MA, USA).

After blocking, membranes were incubated with the following primary antibodies: mouse anti-α-smooth muscle actin (α-SMA) antibody (Sigma-Aldrich Japan, Tokyo, Japan), mouse anti-β-actin antibody (Cell Signaling Technology, MA, USA), rabbit anti-LIMK2 antibody (Cell Signaling Technology), rabbit anti-phospho-cofilin (Ser3) antibody (Cell Signaling Technology), rabbit anti-cofilin antibody (Cell Signaling Technology), rabbit anti-collagen I antibody (abcam, Cambridge, UK), rabbit anti-BAX antibody (Cell Signaling Technology), mouse anti-Bcl-2 antibody (Cell Signaling Technology), and/or rabbit anti-GFP antibody (Santa Cruz Biotechnology, TX, USA). The membranes were then incubated with HRP-conjugated anti-mouse (Cell Signaling Technology) or anti-rabbit IgG antibody (Roche Diagnostics).

Amersham ECL Western blotting detection reagents (GE Healthcare, IL, USA) were used to visualize bands. The total protein content of each sample was measured using the DC™ protein assay kit (Bio-Rad, CA, USA) and sample volumes were adjusted accordingly. Quantification of relative band densities was performed by scanning densitometry and ImageJ software (version 1.54f., National Institutes of Health, MD, USA, https://imagej.net/ij/index.html).

Contraction assay

The stress-relaxed collagen gel contraction assay, a widely used method that reflects three-dimensional cell contractility⁴⁴, was performed. Approximately 3 × 10⁵ cells were embedded in a collagen matrix using a Collagen Gel Culturing kit (Nitta Gelatin), as follows. Cells were mixed with reconstituted collagen solution, which consists of eight volumes of type I collagen solution and one volume of reconstituted buffer (50 mmol/L NaOH, 260 mmol/L NaHCO3, 200 mmol/L HEPES), on ice. Next, 0.5 mL aliquots of this reconstituted collagen solution were placed on the bottom of 24-well culture plates and immediately warmed at 37 °C to allow gel formation. After gel formation, the rims of the gels were detached, and gels were overlaid with DMEM containing 10% FBS with or without 10 ng/mL TGF-β1 (Fujifilm Wako Pure Chemical Co.) for 72 h. The gel size was measured with ImageJ software (version 1.54f., National Institutes of Health).

Wound healing assay

The wound healing assay is known as a suitable method for evaluating migration in fibroblasts⁴⁵. Confluent cells were cultured in DMEM containing 10% FBS and then wounded with a linear scratch (500 µm) by a sterile 1000 µL pipette tip. Images of the wounded cell were taken using a microscope (BZ-X800, Keyence) at 0 and 12 h after wounding. The distance of the wound edges was measured by ImageJ software (version 1.54f., National Institutes of Health).

Measurement of apoptosis

After 24 h of mechanical stimulation for stretched HDFs, cells in all groups were collected. Collected cells were washed twice with PBS and stained with Apoptosis detection reagent (GFP-Certified Apoptosis/Necrosis Detection Kit, Enzo Life Sciences, NY, USA) and Hoechst 33,342 solution (Nacalai Tesque, Kyoto, Japan). Apoptotic cells were fluorescently labeled with annexin V-EnzoGold and propidium iodide (PI) (Nacalai Tesque), and nuclei were stained with Hoechst. Cells were observed under a fluorescence microscope (BZ-X800, Keyence, Osaka, Japan) and the apoptosis rate was measured. Collected cells were also analyzed by annexin V/PI flow cytometry using an SH800 cell sorter (Sony, Kanagawa, Japan).

Assessment of cell proliferation

Cells were seeded in 24-well plates and incubated in DMEM with 10% FBS at 37 °C. Cells were then cultured for an additional 0, 24, 48, 72, and 96 h. Cell nuclei were then fluorescently stained with Hoechst 33,342 solution (Nacalai Tesque) and observed with a fluorescence microscope (BZ-X800, Keyence) to compare cell proliferation rates.

The ratio of proliferating cells to total cells was also evaluated. 5-Bromo-2′-deoxyuridine (BrdU, Selleck, TX, USA) is a thymidine analog that is incorporated into DNA during replication and was used to label proliferating cells. Cells were incubated for 48 h, followed by 24-h BrdU incorporation. BrdU-positive cells were detected with anti-BrdU antibody (GeneTex, CA, USA). Nuclear staining with DAPI (Nacalai Tesque) was also performed, and the percentage of proliferating cells was determined using a fluorescence microscope (BZ-X800, Keyence).

Inhibition of actin polymerization

To assess the role of actin dynamics independently of LIMK2, cells were treated with 75 nM latrunculin A (Fujifilm Wako Pure Chemical Co.) for 24 h. Cells treated with vehicle (DMSO) were used as controls.

Statistical analysis

Experimental data are presented as mean ± standard deviation. Results were analyzed by Student’s t-test for comparisons between two groups. For comparisons involving three or more groups, one-way analysis of variance (ANOVA) followed by Tukey’s test was performed. JMP software (version 18.2.1, SAS Institute, NC, USA, https://www.jmp.com/en/software/data-analysis-software) was used for these analyses. A p-value < 0.05 was considered statistically significant.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

The first and second authors equally contributed to this work. M.I., K.Ku., S.M. and T.K. designed, performed and interpreted all the experiments. K.Ku., K.Ka. and S.M. developed AAV vectors and performed experiments with the vectors. M.I., N.O., R.N. and T.O. performed mechanical stimulation experiments. K.Ka. and T.K. supervised and interpreted the data. M.I., K.Ku. and T.K. wrote the manuscript. All authors reviewed the manuscript.

Data availability

The datasets generated and/or analyzed during the present study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.