Introduction

Wound healing is a complex cellular process that leads to tissue repair following injury. There are four overlap** phases of wound healing, starting with hemostasis which recruits platelet cells to the injured site for vasoconstriction, coagulation, and blood clot formation. This is followed by the inflammatory phase, which activates neutrophils and macrophages to clear microbes to prevent infection. The third phase is proliferation, which involves proliferation and migration of epithelial keratinocyte cells and fibroblasts to the injured site for re-epithelialization as well as formation of granulation tissue. The fourth phase involves tissue remodeling, which activates synthesis and deposition of the new extracellular matrix (ECM) by fibroblasts for wound contraction and scar formation [1]. This process requires tight regulation as dysregulation leads to the onset of chronic wounds. Chronic wounds do not progress through the healing process in a timely manner of 4–6 weeks, but prolong healing for up to 12 months and longer [2]. It has been reported that chronic wounds are a burden to the healthcare system as they are estimated to affect 10.5 million individuals in the United States of America [3]. Among the different types of chronic wounds, diabetic foot ulcers (DFUs) are estimated to affect 15% of the population in Africa and South America [4]. Studies have shown that chronic wounds fail to complete the wound healing process due to a prolonged inflammatory phase as a result of the increased recruitment of pro-inflammatory macrophages, and increased secretion of pro-inflammatory cytokines such as interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α) at the wound site [5, 6]. Furthermore, current strategies such as wound dressing and wound debridement are reported to be less effective for treating chronic wounds [3, 7], which indicates the need for new advanced treatment modalities.

Regulation of the wound healing process is mediated by several signaling pathways, which include the transforming growth factor-beta (TGF-β), Notch, and Wnt/β-catenin signaling pathway [8]. These pathways are involved in activating the expression of target genes as well as the synthesis and secretion of soluble proteins that mediate cell activation and transition through the healing phases [8,9,10]. Moreover, these signaling pathways interact with one another, promoting the advancement of the wound healing process [9, 11]. Dysregulation of these signaling pathways during wound healing delays tissue repair, leading to the onset of chronic wounds. This review will discuss the role of the TGF-β, Notch, and Wnt/β-catenin signaling pathways in wound healing, with a special focus on the Wnt/β-catenin signaling pathway in the different phases of wound healing. We will then define the target genes regulated by the Wnt/β-catenin signaling pathway in the cell types involved in wound healing, namely macrophages, epithelial cells (keratinocytes), and fibroblasts. We will further discuss the crosstalk between Wnt/β-catenin signaling with the Notch and TGF-β signaling pathways during wound healing, and its modulation in chronic wounds, with the primary focus on diabetic wounds/ulcers. Lastly, we will also discuss prospective therapies for the treatment of chronic wounds which target the activation of Wnt/β-catenin signaling, with emphasis on DFUs.

Cellular signaling in wound healing

Following injury, the hemostasis phase is activated by tissue factor (TF), a membrane glycoprotein that forms part of the clotting cascade that activates platelet cells, and monocytes upon exposure to blood [12]. TF together with damage-associated molecular pathogens (DAMPs) such as cell debris, RNA, and pathogen-associated molecular patterns (PAMPs) (e.g. bacterial lipopolysaccharides) activates the inflammatory phase which overlaps with hemostasis and initiates clot formation [13]. The blood clot fills the wound bed and forms a provisional wound matrix for the migration of leukocytes and platelet cells [6]. Platelet cells further secrete platelet-derived growth factor (PDGF) and the TGF-β1 cytokine which activate the TGF-β signaling pathway during inflammation [14].

The proliferative phase, which focuses on re-epithelialization of keratinocytes, angiogenesis, and formation of granulation tissue, is initiated by the release of cytokines (e.g. IL-4 and IL10) and growth factors such as basic fibroblast growth factor (bFGF) released by the reparative anti-inflammatory (M2) macrophages [6, 15]. Macrophages further release nitric oxide (NO) and TGF-β cytokines, which activate the proliferation and migration of fibroblast cells [16]. NO released by macrophages also activates existing endothelial cells to proliferate and secrete vascular endothelial growth factor (VEGF) for angiogenesis [17]. Cells at the edge of the wound are also activated and release EGF, keratinocyte growth factor (KGF) and insulin growth factor-1 (IGF-1), which induce the proliferation and migration of keratinocytes, endothelial cells, and fibroblasts. Mast cells, which are found in connective tissue of the skin and mucosa also secrete IgE antibodies, histamine, and cytokines such as IL-6 and IL-8 during the overlap between the inflammatory and proliferative phases [18]. They also secrete proteases such as chymase and tryptase, which breakdown the basement membrane and old ECM for the formation of granulation tissue [18]. Mast cells are further suggested to activate the proliferation of fibroblasts and endothelial cells by secreting IL-4 and VEGF during the proliferative phase [18, 19]. Activated fibroblasts begin to express alpha smooth muscle (α-SMA) and transdifferentiate into myofibroblasts for migration and deposition of ECM proteins at the wound site [20]. The provisional wound matrix is replaced by granulation tissue, which is largely composed of fibroblasts and myofibroblasts, M2 macrophages and new blood vessels to provide a scaffold for cell adhesion, migration, and cell differentiation during wound repair [6]. Keratinocytes and fibroblasts secrete matrix metalloproteinases (MMPs) such as MMP-2 and MMP-9, which are known to degrade the provisional matrix for deposition of new ECM rich in fibronectin, type I and type III collagen required for cell migration and the formation of granulation tissue [21, 22]. TGF-β signaling is active at the remodeling phase, which involves the maturation of granulation tissue where there is an increased number of myofibroblasts for ECM deposition and wound contraction [23]. At this phase, myofibroblasts and macrophages release MMPs and tissue inhibitor metalloproteinases (TIMPs) to resolve the immature ECM found in granulation tissue, and deposit increased levels of type I collagen, which has a high tensile strength [22, 24].

The role of TGF-β and Notch signaling pathways in wound healing

The activation and role of TGF-β signaling is well-characterized in wound healing. Briefly, the binding of TGF-β ligands (TGF-β1, TGF-β2, and TGF-β3) to the TGF-β receptor I/II (TGFβRI/ TGFβRII) heterodimeric complex activates the signaling pathway. This leads to the phosphorylation of the TGF-β receptor complex, which subsequently phosphorylates receptor SMADs (R-SMADs) (SMAD2/3) proteins that bind to SMAD4 for nuclear translocation and transcriptional activation of target genes [25, 26]. Secretion of TGF-β1 also activates polarization of macrophages to the M2 phenotype, which mediates progression from the inflammatory phase to the proliferative and remodeling phases [27]. TGF-β1 also leads to epithelial-mesenchymal transition EMT in epithelial cells for re-epithelialization, as well as the transdifferentiation of fibroblasts into myofibroblasts [9]. The TGF-β/SMAD2/3 signaling pathway activates the expression of genes that encode collagens I, III and IV, as well as α-SMA, fibronectin, MMPs and tissue-inhibitors of metalloproteinases (TIMPs) in fibroblasts (Table 1) [43]. In macrophages, TGF-β/SMAD3 signaling also targets the expression of IL-10 and mediates progression from the inflammatory phase to the latter phases of wound healing [44].

Table 1 List of Notch and TGF-β target genes expressed during wound healing

Like the TGF-β pathway, the Notch signaling pathway, is an evolutionarily conserved signaling pathway, which plays a role in embryonic development, tissue homeostasis, and tissue repair [45,46,47]. Notch signaling controls cell fate, proliferation, differentiation, and cell survival [48]. Notch signaling is activated by the binding of its ligands such as Jagged (JAG) 1 and − 2, and delta-like (DLL)-1, − 3, and − 4 to the Notch receptors (Notch 1–4). Upon ligand binding to the Notch receptor between adjacent cells, the Notch receptor is cleaved by γ-secretase (e.g. presenilin), leading to the translocation of the notch intracellular domain (NICD) to the nucleus where it functions as a transcription factor [49]. Transcriptional activation is mediated by binding of the NICD to recombinant binding protein-J (RBPJ) and mastermind-like (MAML) transcription co-activators, which activate the expression of genes such as the Hairy/Enhancer of Split 1 (HES1) and Hairy/E(spl)-related with YRPW (HEY), which are transcriptional repressors of the basic helix-loop-helix (bHLH) family, which regulate proliferation and differentiation of epidermal stem cells. Notch ligands are known to be highly expressed in epidermal cells, endothelial cells, keratinocytes, fibroblasts, and macrophages to activate angiogenesis and keratinocyte differentiation, and regulate inflammation [48, 133, 135]. Studies have also reported that DM causes reduced production of Wnt3A and Wnt4 ligands, which impair the function of pancreatic beta cells. Other ligands such as Wnt5A are reported to be low at the onset of type 2 diabetes mellitus (T2DM), but increase overtime and contribute to chronic low-grade inflammation [136]. Furthermore, the secretion of Wnt5A by macrophages also causes vascular endothelial dysfunction, which impairs angiogenesis [137]. DM further contributes to the downregulation of Wnt/β-catenin signaling by preventing the stabilization and nuclear translocation of β-catenin for the expression of genes (e.g. MYC, CCND1, and MMPs) that are required for the proliferation and remodeling stages of healing [1, 138]. High ROS levels in DM are suggested to also induce competitive binding of the limited β-catenin to other transcription factors, such as FOXO, instead of binding to TCF/LEF transcription factors, which alters the proliferation of pancreatic beta cells and insulin synthesis [139]. A study in a diabetic wounded mouse model also indicated that CXXC-type zinc finger protein 5 (CXXC5), a Wnt/β-catenin suppressor, is overexpressed in diabetic wounds, thereby downregulating Wnt/β-catenin activation and preventing angiogenesis during wound repair. Treatment with the KY19334 small molecule inhibited CXXC5 binding to Dvl, leading to Wnt/β-catenin activation and improved wound healing [140]. There is also increased levels of GSK-3β in DFUs, however, current molecules such as Thiazolidinediones that have inhibitory effects on GSK-3β are associated with a high risk of heart failure [141]. These studies indicate the need to identify improved therapies that will modulate chronic inflammation and activate Wnt/β-catenin signaling to improve the healing of diabetic wounds.

Current treatments and future strategies targeting signaling pathways for healing of chronic wounds

Standard and emerging wound care therapies

Conventional wound care strategies involve debridement, wound dressing, infection control, and pain management [142]. It was also recommended by the Wound Healing Foundation that wound care must be simplified for patients to do it themselves or easily assisted by a family member [142]. The importance of wound debridement is to remove non-viable and dead tissue [142, 143]. Infection control is also critical to prevent occurrence of drug-resistant microbial biofilm by treatment with topical antibiotics [2, 142]. Varying wound dressings which aim to manage wound moisture and pain have also been discussed. These include dressings that can deliver antimicrobial agents and debridement [142, 143]. The challenge of wound dressing is that it requires repeated application [142]. Other treatment options such as skin grafts and flaps are used for wound cover and blood supply. Negative pressure wound therapy and hyperbaric oxygen therapy, which are used to remove wound exudate and improve the formation of granulation tissue, wound perfusion, and contraction have been applied for the treatment of chronic wounds, however, the treatment cost, particularly for hyperbaric oxygen therapy is high [142, 143]. While standard wound care has shown to improve healing of chronic wounds, their effectiveness is moderate as they do not prevent the reoccurrence of chronic wounds. Other strategies such as treatment with growth factors and the use of ECM scaffolds have also been developed [1], but are moderately effective.

Emerging therapies such as stem cell therapy have entered early clinical trial stages. These include a phase I/II clinical trial investigating the safety and efficacy of allogeneic mesenchymal stem cells (MSCs) for the treatment of chronic venous ulcers [144]. In this clinical trial, dermal mesenchymal cells that express the ATP binding cassette subfamily B member 5 (ABCB5) were administered to patients with venous ulcers. There was a decrease in IL-1β-mediated inflammation, as well as a shift from M1 to M2 macrophages, and a reduction in wound size in the treatment group [144, 145]. MSCs can differentiate into other cell types such as skeletal muscle, bone, and adipose tissue, but their benefit in cell therapy for wound healing is suggested to be attributed to their ability to produce biomolecules such as KGF, VEGF, and IGF that are involved in re-epithelialization and neovascularization [146]. The current limitation of MSCs for cell therapy is overcoming the microenvironment in chronic wounds, which may require repeated cell therapy to overcome the hypoxic, high ROS and high inflammatory microenvironment that may affect their survival and proliferation upon treatment. Another advancing therapeutic strategy involves the use smart bandages. A preclinical study in a mouse model showed that using a wireless, closed-loop smart bandage with multimodal sensors stimulates the proliferation of monocyte/macrophage cell populations and improves healing of cutaneous wounds [147]. The main limitation of smart bandages is the high cost for large-scale production.

Targeting signaling pathways for treatment of chronic wounds

Current diabetic treatments include insulin injection and exercise for managing T1DM and T2DM respectively, however, the complication of non-healing wounds is still a matter to be addressed. Natural compounds such as the Chinese traditional herb Centella asiatica (C. asiatica) has been shown to promote fibroblast proliferation and ECM synthesis in wound healing. This extract of C. asiatica include triterpenoids, asiaticoside (AC) and madecassoside, which have been reported to promote collagen synthesis in human fibroblasts [148]. A study by Nie et al. [148] prepared a gel compound using C. asiatica and NO for application on diabetic cutaneous ulcers in a mouse model, and showed improved wound healing by activating the Wnt/β-catenin signaling pathway, which increased the expression of Wnt1 and β-catenin. A phase 3, randomized clinical study showed that asiaticoside extract (ON1O1) improved healing of DFUs by activating the switch from M1 to M2 macrophage phenotype [149].

Photobiomodulation therapy activates signaling pathways for wound healing

Photobiomodulation (PBM), previously known as low-level light therapy (LLLT), which utilizes light devices such as lasers and light emitting diodes (LEDs), has been identified as a potential therapeutic modality for treating cutaneous wounds, alopecia, atopic dermatitis, and other inflammatory conditions [150,151,152]. This discovery was made by Mester [150], who showed that laser treatment stimulates cellular proliferation as well as hair regeneration in a wound healing mouse model. The light from PBM devices interacts with photosensitive receptors and chromophores in the mitochondria and human skin, thus inducing a photochemical action and activating cellular signals that lead to the transcription of target genes associated with wound healing [153]. Excitation of cytochrome C oxidase in the mitochondria modulates the electron transport chain, which increases the production of adenosine triphosphate (ATP) and ROS, leading to downstream activation of signaling pathways [154]. Wavelengths ranging from 420 nm to 830 nm have been shown to modulate oxidative stress and accelerate wound healing [155]. Furthermore, PBM has been shown to improve” wound” closure in an in vitro diabetic wounded model [156]. PBM has also been shown to mediate macrophage polarization from M1 to the reparative M2 macrophage at the red and near infrared spectrum (660-1000 nm), and modulate the production of cytokines such as IL-6 and TNF-α [157, 158]. Another study showed, in injured skeletal muscle of Wistar rats, that PBM decreased the number of M1 macrophages (CD68+) 2 days post-PBM at the wavelength of 660 nm, and increased M2 macrophages (CD163+ and CD206+) 7 days post-PBM at the wavelength of 780 nm [159]. PBM has also been shown to induce proliferation and migration of keratinocytes and fibroblasts in normal and diabetic cellular models [160, 161]. Few clinical studies have shown great promise in the effect of PBM in DFUs. For instance, a study by Mathur et al. [162] showed that the application of PBM in combination with standard DFU treatment reduced wound size after 2 weeks of treatment. Another study showed that PBM accelerated wound healing in DM patients with grade 3 burn ulcers 8 weeks after treatment [163]. Preclinical studies have shown that PBM therapy in combination with mesenchymal stem cell engraftment can accelerate wound healing in a diabetic murine model [164]. The mechanisms of action in PBM-induced wound healing include activation of signaling pathways associated with wound healing such as the TGF-β, PI3K/AKT, MAPK, and the Wnt/β-catenin pathways, to mention a few [156, 160, 165, 166].

Some studies have shown that PBM activates the Wnt/β-catenin signaling pathway in outer root sheath cells and in hair follicle stem cells. For instance, a study by Kim et al. [167] showed that PBM of human outer root sheath cells at the wavelengths of 660 nm and 830 nm increased their cell proliferation and migration. Furthermore, they showed that PBM activated both the Wnt/β-catenin and ERK/MAPK signaling pathways for proliferation and migration, which suggests that PBM can activate multiple pathways at a single wavelength and dose. Another study by ** et al. [165] showed that PBM at the wavelength of 635 nm activates a new hair cycle in hair follicle stem cells by upregulating β-catenin gene expression in β-catenin transgenic mice. Interestingly, another study showed in a mouse model that PBM at the wavelength of 535 nm and power density ranging from 0.1 W/cm2 to 0.5 W/cm2 induced transcriptional activation of genes associated with Wnt/β-catenin, Notch, TGF-β, and the JAK/STAT signaling pathways [168]. These studies thus indicate that PBM can induce activation of multiple signaling pathways in cutaneous tissue. It is unclear however, if PBM activates multiple signaling pathways simultaneously, or whether there is co-activation of these signaling pathways.

While this therapeutic approach has demonstrated positive preclinical findings, some variations in the experimental and clinical parameters have also been reported. These include variations in the wavelength, radiation exposure (fluence), and irradiance. Some studies have reported variations in the effect of PBM in cell proliferation and wound healing when using the same parameters as previous studies [152, 169]. Also, studies have indicated that tissues with high mitochondrial content (e.g. muscle, brain, and heart) require low light dosage compared to tissues with low mitochondria (e.g. skin, tendon, and cartilage), which require a higher light dosage [152, 153]. Furthermore, it is suggested that different fibroblast subtypes in the skin respond differently to PBM due to the heterogeneity of these cellular subtypes [169].

Conclusion

The role of Wnt/β-catenin signaling in embryonic development is well known. But its effect in disease progression is still under investigation. We have shown in this review the importance of this pathway in wound healing. Furthermore, we have highlighted the effects of its dysregulation in chronic wounds, including diabetic ulcers. Moreover, we also discussed its crosstalk with the TGF-β and Notch signaling pathways, which is critical for wound healing. We thus recommend that future therapies investigate strategies that induce the (re)activation of the Wnt/β-catenin signaling pathway, especially for treatment of chronic ulcers that remain persistently in the inflammatory phase of healing. PBM remains one of the promising non-invasive therapeutic strategies that has the potential to improve the healing of chronic wounds via activation of the Wnt/β-catenin signaling pathway, as well as other signaling pathways critical for wound healing. Moreover, PBM can activate skin stem cells, as well as epithelial cells to augment the healing of chronic wounds. Future studies will need to further investigate optimal parameters for the clinical application of PBM therapy in different chronic wounds, and to determine if PBM is most effective alone or in combination with standard or other emerging therapies.