Downregulation of CR6-interacting factor 1 suppresses keloid fibroblast growth via the TGF-β/Smad signaling pathway

Keloids are a type of aberrant skin scarring characterized by excessive accumulation of collagen and extracellular matrix (ECM), arising from uncontrolled wound healing responses. While typically non-pathogenic, keloids are occasionally regarded as a form of benign tumor. CR6-interacting factor 1 (CRIF1) is a well-known CR6/GADD45-interacting protein, that has both nuclear and mitochondrial functions, and also exerts regulatory effects on cell growth and apoptosis. In this study, cell proliferation, cell migration, collagen production and TGF-β signaling was compared between normal fibroblasts (NFs) and keloid fibroblasts (KFs). Subsequently, the effects of CRIF1 deficiency were investigated in both NFs and KFs. Cell proliferation, cell migration, collagen production and protein expressions of TGF-β, phosphorylation of Smad2 and Smad3 were all found to be higher in KFs compared to NFs. CRIF1 deficiency in NFs and KFs inhibited cell proliferation, migration, and collagen production. In addition, phosphorylation of Smad2 and Smad3, which are transcription factors of collagen, was decreased. In contrast, mRNA expression levels of Smad7 and SMURF2, two important inhibitory proteins of Smad2/3, were increased, suggesting that CRIF1 may regulate collagen production. CRIF1 deficiency decreases the proliferation and migration of KFs, thereby inhibiting their overgrowth via the transforming growth factor-β (TGF-β)/Smad pathway. CRIF1 may therefore represent a potential therapeutic target in keloid pathogenesis.

ences in cell migration between NFs and KFs. Cells were seeded in 6-well plates and incubated for 24 h, after which a scratch was made through the center of each well. Photographs were taken at baseline and 24 h after scratching. Under basal conditions, in the absence of any exogenous modulation, the number of cells covering the wound site was significantly higher in KFs at 24 h after injury, as compared to NFs (Fig. 1E,F).
We also performed transwell assay to detect cell migration. Cells were seeded in 6-well plates and incubated for 24 h, after which they were transferred to the upper transwell chamber in serum free medium. The lower chamber contained complete medium. After 24 h of incubation, the migrated cells were stained with DAPI and counted under a microscope. To rule out cell proliferation as the cause of migration in these experiments; cells were treated with a cell proliferation inhibitor Mitomycin C. As shown in Fig. 1G,H, mitomycin C treatment slowed down fibroblast proliferation as seen by a reduction in the total cell number (data not shown), but it did not cause complete cell death to affect cell migration and therefore the graph shows that the pattern of cell migration was similar in both the conditions (with or without Mitomycin C treatment). Hence, it was confirmed that migration of KFs was higher than NFs.
The TGF-β/Smad signaling pathway is activated in keloid fibroblasts. Among the many growth factors involved in the wound healing process, TGF-β appears to play a central role. TGF-β protein expression levels were found to be significantly higher in KFs than NFs in this study ( Fig. 2A,B). The TGF-β/Smad signaling pathway plays an important role in keloid formation. Smad proteins are signaling proteins downstream of TGF-β that mediate the intracellular signaling transduction of TGF-β. TGF-β binds to high-affinity receptors that phosphorylate and activate the downstream substrates Smad2 and Smad3. Protein expression levels of phosphorylated Smad2 and Smad3, two major transcription factors of ECM, were found to be higher in KFs than NFs ( Fig. 2A,B).
Furthermore, Inhibitory Smads (Smad6 and Smad7) are antagonists for the activity of receptor-regulated Smads, leading to the termination of TGF-β signaling. Between these two proteins, Smad7 is more specific for TGF-β signaling, with lower protein expression of SMAD7 in KFs as compared to NFs ( Fig. 2A,B). Smad ubiquitination regulatory factor 2 (SMURF2), a Smad-specific ubiquitin ligase, also plays a critical role in the regulation of TGF-β/Smad signaling. Here, we found that the protein expression level of SMURF2 was higher in KFs than in NFs. At the same time, no significant difference was found in CRIF1 protein expression at the basal level between NFs and KFs ( Fig. 2A To compare the ECM synthesis of NFs and KFs in our study, mRNA expression levels of the ECM components collagen 1A1 (COL1A1) and collagen 3A1 (COL3A1) were measured by quantitative PCR (qPCR). The mRNA expression levels of both COL1A1 and COL3A1 were higher in KFs than NFs (Fig. 2C,D). At the same time, the (B) NFs and KFs were cultured in 6-well plates for 24 h. After incubation, cells were harvested and ADAM-MC kit was used for cell counting. (C) Proliferating cell nuclear antigen (PCNA) protein expression, used as a marker of cell proliferation, was measured by western blot; β-actin was used as the loading control. Full-length blots are presented in Supplementary Fig. S1. (D) PCNA density was calculated using ImageJ software. (E) NFs and KFs were cultured in 6-well plates for 24 h followed by wounding for 24 h. Images were taken using a light microscope. (F) Quantification of wound closure was performed using ImageJ software. (G) Transwell assay was conducted to determine cell migration with or without MMC treatment. Scale bar 200 μm. (H) Quantification of the number of migrated cells was performed using ImageJ software. All data are presented as the mean ± SD of three independent experiments. *P < 0.05 relative to NFs.

Figure 2.
Comparison of transforming growth factor-β (TGF-β)/Smad expression levels between normal fibroblasts (NFs) and keloid fibroblasts (KFs). (A) NFs and KFs were cultured in 6-well plates for 24 h. After 24 h incubation, cells were harvested, lysed, and analyzed by western blot to assess differences in protein levels of various proteins of the TGF-β/Smad signaling pathway and CRIF1; β-actin was used as the loading control. Fulllength blots are presented in Supplementary Fig. S2. (B) Protein density was calculated using ImageJ software. The mRNA levels of (C) collagen 1A1 (COL1A1), (D) collagen 3A1 (COL3A1) (E) Smad7 and (F) SMURF2 were determined by quantitative polymerase chain reaction (qPCR) in NFs and KFs. All data are presented as the mean ± SD of three independent experiments. *P < 0.05 relative to NFs.  www.nature.com/scientificreports/ factors that influence multiple processes including cell migration as well as cell proliferation 20 . Cell migration of NFs and KFs after CRIF1 downregulation was measured using a cell scratch assay. Cells transfected with control or CRIF1 siRNA were incubated for 24 h, after which a scratch was made through the center of each well. Photographs were taken at baseline and 24 h after scratching. NFs (Fig. 4A) and KFs ( Fig. 4B) migrated into the wound area and nearly covered the whole scratch after 24 h, while cells transfected with CRIF1 siRNA did not migrate as far suggesting that CRIF1 downregulation significantly reduced cell migration.
Cell migration was also measured by the transwell assay. Cells transfected with control or CRIF1 siRNA were incubated for 24 h, after which they were transferred to the upper transwell chamber in serum free medium. The lower chamber contained complete medium. After 24 h of incubation, the migrated cells were stained with DAPI and counted under a microscope. To rule out cell proliferation as the cause of migration in these experiments; cells were treated with a cell proliferation inhibitor Mitomycin C. As shown in Fig. 5A,B, mitomycin C treatment slowed down fibroblast proliferation as seen by a reduction in the total cell number (data not shown), but it did not cause complete cell death to affect cell migration and therefore the graph shows that the pattern of cell migration was similar in both the conditions (with or without Mitomycin C treatment). It was confirmed that migration of NFs (Fig. 5A) and KFs (Fig. 5B) was reduced by CRIF1 downregulation.
The TGF-β/Smad signaling pathway is altered by CRIF1 downregulation in normal and keloid fibroblasts. Inhibition of the TGF-β/Smad signaling pathway markedly reduces keloid formation. Protein expression levels of TGF-β, phospho-Smad2, and phospho-Smad3 were measured by Western blotting after transfection of NFs (Fig. 6A,B) and KFs (Fig. 6C,D) with control or CRIF1 siRNA, to examine the changes in the TGF-β/Smad signaling pathway. TGF-β protein expression was significantly reduced in CRIF1-downregulated cells in both NFs and KFs. Similarly, phosphorylation of Smad2 and Smad3 was also decreased in response to CRIF1 downregulation. Subsequently, qPCR was performed to measure changes in the mRNA expression levels of the ECM components COL1A1 and COL3A1; the levels of COL1A1 in NFs (  . CR6-interacting factor 1 (CRIF1) downregulation leads to decreased cell migration measured by scratch wound healing assay in normal fibroblasts (NFs) and keloid fibroblasts (KFs). (A) NFs were transfected with CRIF1 siRNA in a dose-dependent manner for 24 h, followed by wounding for 24 h. (B) KFs were transfected with CRIF1 siRNA in a dose-dependent manner for 24 h, followed by wounding for 24 h. Images were taken using a light microscope. Quantification of wound closure was performed using ImageJ software. All data are presented as means ± SD of three independent experiments. *P < 0.05 compared to siCON.

Scientific Reports
| (2021) 11:500 | https://doi.org/10.1038/s41598-020-79785-y www.nature.com/scientificreports/ Smad7 and SMURF2 are increased by CRIF1 downregulation in normal and keloid fibroblasts. Smad7 and SMURF2 are well known inhibitors of the Smad pathway. High expression of these proteins results in inhibition of the TGF-β signaling pathway, in turn causing decrease in cell proliferation and collagen deposition. As phosphorylation of Smad2 and Smad3 proteins was reduced under conditions of CRIF1 downregulation, we transfected NFs and KFs with either control or CRIF1 siRNA and examined the mRNA and protein levels of Smad7 and SMURF2. As shown in Fig. 7E,F, both Smad7 and SMURF2 mRNA expression levels were significantly increased by CRIF1 downregulation respectively in NFs as well as KFs (Fig. 7G,H). The protein expression of both Smad7 and SMURF2 was similarly found to be increased in NFs (Fig. 8A) and KFs (Fig. 8B). These results indicate that the increased expression of inhibitory proteins (Smad7 and SMURF2) may be beneficial for keloid growth.

Discussion
Keloids arise as a result of excessive wound healing after skin damage, characterized by increases in cell proliferation, migration, inflammation, and collagen production, and excessive ECM synthesis 9 . In this study, we compared NFs and KFs under specific conditions. Previous reports have suggested that the anti-apoptotic effects of KFs in combination with increased proliferation and collagen deposition were higher than seen in NFs, with ~ 15% of genes being significantly upregulated in KFs 21,22 . This suggests that persistent differences in KFs are the main cause of keloid proliferation, making KFs a distinct cell type compared to NFs 23 . The fibroblasts that were extracted from keloids and normal skin tissues in this study were confirmed to have the characteristics of each tissue (Figs. 1, 2). Mitochondria play an important role in energy metabolism via the production of ATP and other metabolites, serving as an essential source of the energy and biomolecules necessary to support tumor cell growth 24 . A large amount of reactive oxygen species (ROS) is produced as a byproduct of oxidative respiration in mitochondria, providing the energy necessary to fuel tumor cell proliferation. Vincent et al. showed that keloid metabolism is similar to that of tumor cells 15 . Keloids consumes more glucose than normal cells, and exhibits higher accumulation of lactate, although little is known regarding the effects of mitochondrial dysfunction in keloids. Previous studies have shown that deletion of CRIF1 results in poor synthesis of primary OXPHOS polypeptides and abnormal insertion into the mitochondrial inner membrane 13 . CRIF1 downregulation is a major factor underlying mitochondrial dysfunction in endothelial cells, with poor synthesis of OXPHOS subunits resulting in greater levels of mitochondrial dysfunction due to the accumulation of ROS 14 . CRIF1 has also been shown to function as an essential transcriptional coactivator, with Crif1 − / − embryos becoming inviable beyond embryonic day 6.5 due to extensive developmental arrest, accompanied by defective proliferation and extensive apoptosis 12 . In this study, The data also reveals that downregulation of CRIF1 using siRNA has similar effects in both normal and keloid fibroblasts. Since cell proliferation, migration and TGF-β signaling are higher in KFs compared to NFs at the basal level, it can be said that reducing CRIF1 in keloid cells has a higher effect on the characteristics of these cells as compared to normal cells or it makes their phenotype resemble the normal cells. TGF-β is among the most important cytokines driving the proliferation of fibroblasts and production of ECM. Smad proteins are a family of signaling proteins that participate in intracellular signal transduction of the TGF-β superfamily 25 . After TGF-β binds to its associated receptors, Smad2 and Smad3 are activated to form a trimer with Smad4. This trimer then migrates into the nucleus to regulate the transcription of target genes 26 . Smad7 is a key negative regulator of Smad, directly binding to TGF-β type I receptors resulting in the inhibition of Smad2 and Smad3 activation. Previous studies have shown that Smad7 is more specific for TGF-β signaling compared to other inhibitory Smads 27,28 . Importantly, Smad7 is downregulated in keloids, as well as other fibrotic tissues, such as the liver, lung, and kidney 29 . This low expression of Smad7 or other Smad-independent signaling pathways leads to the overproduction of collagen 30 . In our study, CRIF1 downregulation resulted in an increase in Smad7 and SMURF2 expression. Smad7 can bind tightly to the activated TGF-β-I receptor to prevent the After 48 h incubation, cells were harvested, lysed, and analyzed by western blot to assess differences in protein levels of TGF-β, P-Smad2, and P-Smad3; β-actin was used as the loading control. Full-length blots are presented in Supplementary Fig. S5a. (B) Protein density was calculated using ImageJ software. (C) KFs were transfected with CRIF1 siRNA in a dose-dependent manner for 48 h. After 48 h incubation, cells were harvested, lysed, and analyzed by western blot to assess differences in protein levels of TGF-β, P-Smad2, and P-Smad3; β-actin was used as the loading control. Full-length blots are presented in Supplementary Fig. S5b. (D) Protein density was calculated using ImageJ software. All data are presented as the mean ± SD of three independent experiments. *P < 0.05 compared to siCON.

Scientific Reports
| (2021) 11:500 | https://doi.org/10.1038/s41598-020-79785-y www.nature.com/scientificreports/ phosphorylation of Smad. Taken together, these findings show that the biological effects of TGF-β under both normal and pathological conditions are tightly regulated by the coordination of Smad proteins in the TGF-β/ Smad signaling pathway 28 . Our results suggest that ECM synthesis was affected by inhibition in phosphorylation of both Smad2 and Smad3 proteins, which are the transcription factors of COL1A1 and COL3A1 (Fig. 8C). The increase in Smad7 expression plays a key role in the decreased phosphorylation of Smad2 and Smad3, with some studies showing that COL1A1 and COL3A1 could be reduced by increasing the expression of Smad7 1,31 . Therapeutic gene delivery is one of the effective treatments for certain diseases or genetic disorders at the cellular level 32,33 . Topical application of siRNA to treat skin disorders is of particular interest. For gene suppression in the skin, topical delivery of siRNA has various benefits such as direct access to the target site, reduced adverse effects which are sometimes associated with the systemic administration and visual monitoring of the affected region for any side effects 34 . There are various approaches described in previous studies for topical delivery of siRNAs [35][36][37] . The findings of the present study can be translated as a therapy to suppress keloid formation in the form of localized patch or topical delivery system such as an ointment, which could be applied at the keloid site and would suppress keloid formation by locally reducing CRIF1 expression.
A potential limitation of this study might be that keloids are specific to humans and there are some ethical limitations to conduct the studies in human patients. Moreover, no other animal species has been found to naturally develop a scar tissue, which is similar to that of human keloids. Therefore, the development of an animal model for detailed study of keloids has been extremely difficult. Many groups have attempted to generate animal models through two main approaches: (1) by inducing a comparable degree of fibrosis in an animal that does not normally develop it, or (2) by transplanting human-derived tissue or cells into the host animal for in vivo analysis 38,39 . For practically utilizing the findings of this study in the in vivo settings, the earlier approach of inducing fibrosis (keloid-like scar) in the CRIF1 knockout mice could be followed. However, either such scars are only maintained for a limited time period or they usually develop characteristics of hypertrophic scars. Secondly, transplantation of human keloid tissue/cell into the CRIF1 knockout mice could be done. Both these approaches can give researchers a reliable way to measure and study in detail about the keloid pathogenesis and developing a therapeutic strategy. Nevertheless, as both the methods have various challenges and limitations, similar attempts in the past have not been fully successful and there is still much to be done in the future. Finally, it can be concluded that CRIF1 downregulation impairs keloid function and suppresses their overgrowth via the transforming growth factor-β (TGF-β)/ Smad pathway.

Materials and methods
Primary keloid-derived and normal skin fibroblast cultures. This study was approved by the Institutional Review Board of Chungnam National University Hospital (IRB No. CNUH 2018-07-067) and was performed in accordance with the Helsinki Declaration of 1964 and its later amendments. Written informed consent was obtained from all patients. Samples were obtained from three patients (of both sexes) aged 18-45 years, with keloids of the earlobe. Surgical procedure involved the removal of keloid tissue along with the surrounding normal skin from each patient. The keloid tissue was used for isolating keloid fibroblasts and the surrounding normal skin was used for isolating normal fibroblasts. Cultures were established from tissue specimens pro- After 48 h incubation, cells were harvested, lysed, and analyzed by western blot to assess differences in protein levels of SMAD7 and SMURF2; β-actin was used as the loading control. Protein density was calculated using ImageJ software. Full-length blots are presented in Supplementary Fig. S6a. (B) KFs were transfected with CRIF1 siRNA in a dose-dependent manner for 48 h. After 48 h incubation, cells were harvested, lysed, and analyzed by western blot to assess differences in protein levels of SMAD7 and SMURF2; β-actin was used as the loading control. Protein density was calculated using ImageJ software. Full-length blots are presented in Supplementary Fig. S6b. All data are presented as the mean ± SD of three independent experiments. *P < 0.05 compared to siCON. (C) Schematic representation of CRIF1 downregulation induced changes in the TGF-β/ Smad signaling pathway leading to the inhibition of invasive KFs. Cell scratch assay. A cell scratch assay was used to assess cell migration. Cells were transfected with negative control or CRIF1 siRNA in 6-well tissue culture plates for 24 h, after which a sterile 1 ml pipette tip was used to detach the cells from the monolayer across the center of the well. Floating cells were flushed out by gently rinsing with PBS and replaced with serum-free medium (to rule out cell proliferation as the cause of wound closure) followed by incubation for another 24 h. The total incubation time post transfection was therefore 48 h. Cell movement was monitored microscopically. Photographs were taken immediately and at 24 h after scratching. The migration capacity of cells was expressed as a percentage of wound coverage: relative migrated distance = (A 0h − A 24h )/A 0h × 100%, where A 0h and A 24h represent the wound area measured immediately and 24 h after scratching, respectively. The wound area was quantitatively evaluated using ImageJ software (NIH, Bethesda, MD, USA).
Transwell assay. Transwell assay was employed to detect cell migration. Cells were transfected with negative control or CRIF1 siRNA in 6-well tissue culture plates for 24 h, followed by transfer of 5 × 10 5 /ml cells in the upper transwell chamber (24-well plate; Corning, New York, USA) and culture with FBS free medium. Complete growth medium with 10% FBS was added to the lower chamber and incubated for another 24 h. Then, cells on the upper side (nonmigrating cells) were removed and migrated cells on the lower face were washed with PBS, fixed with 4% paraformaldehyde, stained with DAPI and counted on 5 random high-power fields (× 200 magnification) under a microscope and averaged.

Mitomycin C treatment.
To rule out cell proliferation as the cause of migration in transwell experiments; a cell proliferation inhibitor Mitomycin C (MMC) was used. 4 h after transfection of Normal or Keloid fibroblasts with CRIF1 siRNA, MMC (Sigma M4287) was treated at a dose of 0.04 mg/ml for 5 min in serum free media. After flushing out MMC, cells were replaced with fresh media.