Mesenchymal stromal cells in wound healing applications: role of the secretome, targeted delivery and impact on recessive dystrophic epidermolysis bullosa treatment

Abstract

Mesenchymal stromal cells (MSCs) are multi-potent stromal-derived cells capable of self-renewal that possess several advantageous properties for wound healing, making them of interest to the field of dermatology. Research has focused on characterizing the unique properties of MSCs, which broadly revolve around their regenerative and more recently discovered immunomodulatory capacities. Because of ease of harvesting and expansion, differentiation potential and low immunogenicity, MSCs have been leading candidates for tissue engineering and regenerative medicine applications for wound healing, yet results from clinical studies have been variable, and promising pre-clinical work has been difficult to reproduce. Therefore, the specific mechanisms of how MSCs influence the local microenvironment in distinct wound etiologies warrant further research. Of specific interest in MSC-mediated healing is harnessing the secretome, which is composed of components known to positively influence wound healing. Molecules released by the MSC secretome can promote re-epithelialization and angiogenesis while inhibiting fibrosis and microbial invasion. This review focuses on the therapeutic interest in MSCs with regard to wound healing applications, including burns and diabetic ulcers, with specific attention to the genetic skin disease recessive dystrophic epidermolysis bullosa. This review also compares various delivery methods to support skin regeneration in the hopes of combating the poor engraftment of MSCs after delivery, which is one of the major pitfalls in clinical studies utilizing MSCs.

Introduction

Mesenchymal stromal cells (MSCs) are of therapeutic interest for chronic wound healing because of their extensive regenerative and immunomodulatory capacities. They demonstrate unique features, including facile culture expansion, multi-potency, low immunogenicity and positive impact on the wound healing processes of re-epithelialization and angiogenesis. Efforts to fully elucidate characteristics of MSC subtypes and capitalize on the promise of MSCs in clinical applications for chronic wound healing are ongoing.

Chronic or recurrent cutaneous wounds are the phenotypic hallmark of the heterogeneous inherited skin disease epidermolysis bullosa (EB). In a severe generalized form of genodermatosis, recessive dystrophic epidermolysis bullosa (RDEB), biallelic mutations in the COL7A1 gene result in absent or dysfunctional type VII collagen (C7). In healthy individuals, secreted C7 homotrimerizes to form anchoring fibrils in the basement membrane zone (BMZ) connecting the epidermal and dermal layers of the skin. Without these protein anchors to hold the two layers together, the skin easily separates into either blisters or wounds with mild mechanical trauma. Therapeutic options for wound management in RDEB are limited to palliative care with analgesia and protective bandaging. Based on their wound-healing properties, MSC-based therapies may represent a promising therapy in the treatment of the skin manifestations of RDEB.

This review aims to characterize the therapeutic advantages of MSCs in the context of wound healing with special attention to RDEB. RDEB is unique, in that it has characteristics of a burn yet is non-transient in nature. Treatment of burns and diabetic ulcers with MSCs will also be discussed, as this has been thoroughly described in the literature and the findings may be relevant to RDEB. The interplay of stromal and epithelial, mesenchymal and ectodermal elements in RDEB wounds makes this an opportunity to gain new knowledge about the regenerative response in genodermatoses, thermal/chemical burns and diabetic ulcers alike.

In this review, the authors first present a brief summary of MSC history and then outline the role of the MSC secretome in wound healing; provide background on recent MSC therapies for burns, diabetic wounds and EB; analyze the modes of cellular delivery; and discuss safety and optimization of MSCs for clinical use.

History and Definition of MSCs

In the 1960s and 1970s, Friedenstein et al. [1,2] discovered MSCs after finding that fibroblastic cells derived from mouse and guinea pig bone marrow were capable of producing clonal colonies of bone and reticular tissue following heterotopic transplantation. It was later discovered that these cells could also generate cartilage and adipose tissue [3]. In 1999, a subpopulation of stromal cells identified in human bone marrow was concluded to represent multi-potent stem cells with trilineage mesenchymal potential [4,5]. MSCs are generally characterized by their spindle shape and ability to adhere to tissue culture surfaces, but a specific MSC marker has yet to be defined, resulting in conflicting definitions of what constitutes an MSC. Consequently, the International Society for Cell & Gene Therapy MSC committee designated minimum criteria to define human MSCs as multi-potent mesenchymal stromal cells [6]. Importantly, the term “mesenchymal stem cells” is retained for the subpopulation of bone marrow cells that displays the two key properties of stem cells: self-renewal and multi-potency [6]. In 2019, the International Society for Cell & Gene Therapy MSC committee further clarified the nomenclature of MSCs and recommends that the acronym be supplemented with the tissue source origin of the cells to bring attention to tissue-specific properties [7].

Key properties of MSCs, such as their low levels of class 1 and class 2 HLA antigen expression, have made them an attractive clinical tool, allowing their use in transplantation without HLA matching. To date, the immunomodulatory properties of MSCs have been most prominently demonstrated as a therapy for graft-versus-host disease (GVHD). GVHD is a common complication after hematopoietic stem cell transplant in which the newly engrafted immune system (mainly CD4 and CD8 cells) attacks the host. A comprehensive review by Zhao et al. [8] outlines the clinical trials that have used MSCs for treatment of or prophylaxis against GVHD. In the context of GVHD, MSCs increase the prevalence of regulatory T cells that control GVHD by decreasing the number and pathological effects of CD4 and CD8 T cells, although there may be additional local immunomodulatory mechanisms at play [9].

MSCs have been successfully isolated from a wide range of human tissues, including bone marrow, adipose tissue and skin [4,10,11]. MSC isolation involves collection and digestion of tissue, expansion in culture and removal of non-adherent cells [12]. MSCs isolated from different tissue sources display heterogeneity in their ability to differentiate and proliferate, suggesting variable therapeutic relevance [13]. Such source-based variations in differentiation potential may be explained by different transcriptome, proteome, immunophenotype and immunomodulatory activities [14]. For example, proteomic analysis has shown that the MSC secretome is influenced by donor age and cellular niche, suggesting the therapeutic potential of MSCs could depend on tissue of origin [15]. Further research is needed to thoroughly assess the impact of cellular source and niche on MSC functions for clinical applications.

MSC Sources for Wound Healing

Although MSCs have been isolated from a variety of tissues, the most advantageous MSC source for wound healing has yet to be determined. The vast majority of evidence indicating that MSCs promote wound healing has been obtained using bone marrow-derived MSCs (BMSCs), mainly from data obtained in pre-clinical studies, with limited validation in clinical studies [16-19]. However, the use of BMSCs in a clinical capacity is not without its limitations. BMSCs must be obtained through aspiration, a generally safe but invasive procedure [20]. Furthermore, bone marrow cellularity declines with age, raising concern for the long-term differentiation potential of in vitro BMSCs [21,22]. Continued research into additional MSC sources that are more accessible than BMSCs and equally as efficacious is warranted.

Other examples of MSC sources that have been studied in the context of wound healing include adipose tissue, skin, umbilical cord and placenta [23-27]. Of these, adipose-derived MSCs (ADSCs) and dermal MSCs are both very accessible and do not have ethical controversies concerning their use. ADSCs have been shown to promote wound healing both in vitro and in vivo in the context of clinical trials for treatment of burns [28-30]. In terms of biological properties, BMSCs, ADSCs and dermal MSCs have similar immunogenicity and differentiation potentials [11,31-33]. However, these cells display differing paracrine expression profiles that can affect wound healing capacities [34]. The therapeutic potential of human BMSCs, ADSCs and amnion-derived MSCs in a murine model of cutaneous wounds was directly compared, and it was found that ADSCs had the most pronounced effect on wound closure [35]. Dermal MSCs have also demonstrated promising effects on wound healing in pre-clinical in vitro and animal studies and in several clinical trials [33,36,37].

In conclusion, the source of MSCs most likely has an impact on the ability of the cells to home to wounded tissue, engraft and exert anti-inflammatory wound healing effects [38-41]. These distinct differences, however, have not been well studied, and further pre-clinical and clinical studies are needed to determine the optimal MSC tissue source for wound healing.

Potential Role of MSCs in Cutaneous Wound Healing

MSCs have the capacity for self-renewal; can migrate to sites of injury; and provide key factors, such as cytokines and growth factors, via paracrine secretion to hasten healing [42]. Several studies have explored the immunomodulatory properties and proposed mechanisms of MSCs [43-47]. These attributes, including immunomodulation of the local microenvironment and ability to attract reparative cells and home to wounded tissue, have generated interest in the use of MSCs in cutaneous wound healing in the past decade because of lack of curative treatments for chronic wounds [42]. MSC therapies for skin wounds are based on repairing and replacing cellular substrates, attenuating inflammation, increasing angiogenesis and enhancing migration of reparative cells, although the mechanisms of distinct differentiation, mobilization and homing of MSCs are highly complex and require further study [42,48,49]. The following sections will focus on the roles of MSCs and their paracrine secretion, known as the secretome, in each stage of cutaneous wound healing. Although it has been shown that MSCs can home to wounded skin, the homing mechanism and induction of wound healing by MSC-secreted factors are complex and not fully understood. Therefore, the authors will focus on the known mechanisms of MSCs, with the goal of supplementing these cells exogenously to hasten and improve wound healing.

The skin is the body’s first line of defense against pathogens and the external environment. Epidermal keratinocytes maintain skin homeostasis by sensing tissue damage and foreign pathogens via innate signaling receptors, which enable the interface of a dynamic mechanical environment with an expansive skin-resident immunological network. Upon injury, such as in wounds, burns and dermatological diseases like EB, keratinocytes, dermal fibroblasts and immune cells coordinate wound healing in three stages: (i) inflammation, (ii) proliferation and (iii) remodeling (Figure 1).

Fig. 1.

MSC recruitment to wounded skin and the inflammatory phase, and known and potential roles of MSCs in each phase of wound healing. (A) Skin injury and hemostasis. (B) Inflammation. (C) Proliferation. (D) Remodeling. ADM, adrenomedullin; KGF, keratinocyte growth factor; MMPs, matrix metalloproteinases; TIMPs, tissue inhibitors of metalloproteinases.

When skin is injured, MSCs can infiltrate the wound in response to local inflammatory mediators such as TNF-α and IL-1β, secreted by neutrophils, which are the first to infiltrate the skin wound [48,50]. It may be advantageous to exogenously supplement these cells when they are dysfunctional or missing. When endogenous or exogenous MSCs reach the wound, they can secrete pro-inflammatory cytokines such as CSF2, IL-6, IL-8, CCL2 and CCL3, recruiting neutrophils and pro-inflammatory macrophages (Figure 1A,B) [51]. The neutrophils and macrophages then activate the adaptive immune system, distinguish commensal skin microbiota from foreign pathogens and begin to degrade damaged tissue. MSCs can also recruit plasmacytoid dendritic cells and limit the number of activated T cells and other innate immune cells to mediate the transition from the inflammatory phase to the proliferative phase [48].

MSCs in the proliferative phase of wound healing

In the proliferative phase of wound healing, MSCs secrete proteins to promote re-epithelialization (epidermal growth factor [EGF], keratinocyte growth factor and hepatocyte growth factor [HGF]) and angiogenesis (vascular endothelial growth factor [VEGF] A, ANGPT1 and platelet-derived growth factor B [PDGF-B]) (Figure 1C) [52]. They also prime fibroblasts to produce matrix proteins via fibroblast growth factor (FGF) [53].

MSCs can skew tissue-resident macrophages from an inflammatory phenotype (M1) toward a wound healing (M2) phenotype [54]. M2 macrophages are characterized by expression of IL-1RA, IL-4, IL-13, IL-10 and TGF-β. These macrophages promote a pro-reparative environment and help clear cellular debris. Local injection of MSCs has been shown to hasten wound healing and decrease infiltration of inflammatory M1 macrophages in pre-clinical models [54].

MSCs in the remodeling phase

In the final stage of wound healing, MSCs contribute to matrix remodeling through secretion of matrix metalloproteinases to induce matrix deposition and tissue inhibitors of metalloproteinases to abrogate deposition of extracellular matrix (ECM) proteins, which is vital for wounds to heal without scarring (Figure 1D) [45]. MSCs have also been thought to prevent hypertrophic scarring through secretion of HGF, FGF, adrenomedullin and TGF-β3 [55].

MSC secretome in wound healing

The secretome is defined as the set of molecules a cell secretes into the extracellular space. The MSC secretome has been shown to be involved in many different processes, such as paracrine signaling, stem cell differentiation, chemoattraction, inhibition of fibrosis and apoptosis, secretion of pro-angiogenic factors and immunomodulation [15,56-58]. This is an area of active research aimed at improving clinical outcomes in patients with chronic wounds and dermatological conditions that result in delayed or aberrant wound healing.

As mentioned previously, the MSC secretome most likely varies based on the tissue of origin but contains a core group of cytokines, including CCL2, CCL5, EGF, FGF, IL-6, IL-8, IL-10, TGF-β and VEGF-A [56]. The MSC secretome in human skin during wound healing lacks characterization, but recent work has provided some important insights. Components of the MSC secretome that have been shown to play a role in wound healing are described in Table 1.

Table 1.

Known components of the MSC secretome and their roles in wound healing.

	Wound healing phase affected
Component of MSC secretome	Inflammatory	Proliferation	Remodeling	Chronic wounds	Known or potential role in wound healing
CCL2, also known as MCP-1 and CCL3	x				Chemokines, monocyte chemoattractants. Recruits and promotes essential macrophage functions in initial stages of skin wounding [59].
CCL5	x				Released by neutrophils and macrophages to recruit inflammatory cells [60]. Recruits bone marrow-derived epithelial progenitor cells.
GM-CSF, also known as CSF2	x			x	Increases inflammatory cytokine IL-1 and MCP-1 to improve wound healing in diabetic ulcers via increased neovascularization and macrophage/neutrophil infiltration [61].
EGF		x			Stimulates epidermal and dermal regeneration [48].
FGF-1, FGF-2, FGF-5 and FGF-7		x			Stimulate proliferation of various cells of mesodermal, ectodermal and endodermal origin [62]. Can stimulate angiogenesis. FGF-7, also known as KGF, is specific for epithelial cells.
Inflammatory cytokines IL-6 and IL-1	x			x	Mobilize immune cells to wound. Keratinocyte, fibroblast proliferation and migration [58]. Can cause damage at high concentrations.
IL-8	x			x	Cell migration and chemotaxis of neutrophils to wound. Potential role in keratinocyte migration in skin wounds [63].
TGF-β1 and TGF-β3			x		Promote ECM synthesis and wound contraction; anti-inflammatory in some contexts [58].
VEGF		x	x		Collagen deposition, angiogenesis and epithelialization. Potential mitogenic, chemotactic and permeability benefits in non-healing wounds [64].
ANGPT1	x	x		x	Enhanced angiogenesis [65], lymphogenesis and blood flow in treated diabetic mice [66].
PDGF-BB	x		x		Stimulates chemotaxis of fibroblasts and proliferation and influences gene expression in macrophages and fibroblasts for tissue repair [67, 68].
HGF			x		Secreted by mesenchymal cells and acts as a multi-functional cytokine on cells of mainly epithelial origin; plays a role in angiogenesis, tumorigenesis and tissue regeneration [69]. Potentially prevents fibrosis [70].
MMPs and TIMPs			x		Promote ECM degradation (MMPs) and deposition (TIMPs) during wound healing, present in acute and chronic wounds, correct regulation necessary for wound resolution [71].
ADM			x		Accelerates wound healing by promoting angiogenesis, collagen deposition and remodeling. Treatment shortened wound closure in rat model [72].

ADM, adrenomedullin; GM-CSF, granulocyte-macrophage colony-stimulating factor; KGF, keratinocyte growth factor; MMPs, matrix metalloproteinases; TIMPs, tissue inhibitors of metalloproteinases.

Another avenue of exploration in MSC-based therapies for wound healing is the use of MSC-derived exosomes. Exosomes are small extracellular vesicles (30–100 nm in size) derived from endosomal compartment cells, which contain biomolecules such as proteins, messenger RNA and microRNA. Exosomes are of specific interest in transplantation and wound healing treatments because they contain bioactive molecules similar to those found in MSCs but are “cell-free” and therefore not limited by the potential host rejection seen with “whole-cell” treatments [73-77]. Exosomes can be harvested from conditioned MSC medium under standard cell culture conditions. Hu et al. [78] recently published a comprehensive review of MSC exosomes for wound healing applications.

To summarize, MSC exosomes contain many components found in the MSC secretome and are modulated by the specific microenvironment. A recent study found that MSC exosomes derived from bone marrow, adipose tissue and umbilical cord contained wound healing-mediated growth factors (VEGF-A, FGF-2, HGF and PDGF-B), whereas TGF-β was found only in exosomes from umbilical cord MSCs [62]. This result stresses the importance of characterizing exosomes by MSC source before clinical application to ensure the presence and quality of exosomes can be replicated with the end application in mind (burns, diabetic ulcers or genetic diseases) [79].

In conclusion, exosomes provide a promising way to harness some of the potential of MSCs without the need for whole-cell treatments. The drawbacks to exosomes are their heterogeneity based on culture conditions and lack of complete mechanism of action of whole MSCs, including paracrine secretion of growth factors and the ability to persist and differentiate at the site of transplantation.

Wound Etiopathogenesis and Clinical Applications of MSCs

MSCs have gained traction in clinical applications because of their role in tissue regeneration and inflammation modulation. In the dermatological field specifically, MSCs have been used to improve resolution of diabetic wounds [80]; burns [81]; GVHD [82-86]; and inflammatory skin diseases [87-89] such as atopic dermatitis [90], psoriasis [91], lupus erythematosus [92] and genodermatosis, including EB [93,94]. Importantly, these cutaneous afflictions showcase a broad variety of wound etiologies, which vary in the depth of the wound and the host’s ability to heal. They can be broadly broken up into two categories, extrinsic (burns, pressure, injury) and intrinsic (nutrition status, diabetes, genetic disease), which will be discussed in this section [95]. When designing MSC therapies, the wound etiopathology should be carefully considered to dictate the best strategy for repair.

A concise review by Golchin et al. [89] outlines many MSC clinical trials for skin diseases. Despite a plethora of studies currently underway, many have not shown tangible improvements for patients. This furthers the need for pre-clinical studies to focus on individual mechanisms of immunomodulation and interaction with other cell types in the skin. A similar MSC review by Lee et al. reported that although multiple clinical studies have shown MSCs to enhance wound healing by decreasing time to wound closure, the mechanism of this benefit in vivo is unclear, with potentially multiple factors at play [97]. It is hypothesized that MSCs travel to wounded sites, differentiate into epithelial progenitors and secrete paracrine factors depending on their microenvironment [96]. To date, in vitro studies have found that MSCs can enhance migration of fibroblasts and keratinocytes either with direct co-culture or with the use of MSC-conditioned media [97-100]. The authors will now discuss previous clinical and pre-clinical studies of MSC treatments on specific wound etiologies with specific attention to RDEB and how targeted MSC therapy could best be used or adjusted to fit the needs of each type of wound.

Genetic Skin Wounds: EB

Genetic skin diseases such as the heterogenous EB group are characterized by skin fragility and caused by mutations in one of 19 different skin proteins [101,102]. These patients have missing or dysfunctional skin structural proteins in the epidermis (EB simplex) or BMZ between the epidermis and dermis, as seen in junctional EB and RDEB. Wounds caused by these genetic diseases have difficulty healing correctly because of the disrupted architecture and loss of skin structural proteins necessary for skin repair. RDEB, which is caused by mutations in the ECM protein C7, is the most severe of this group, with even slight touch causing blistering of the skin. Many patients have sites of recurrent skin separation that turn into chronic wounds.

Alterations in wound repair in RDEB include increased and persistent inflammation; reduced proliferation and migration of dermal fibroblasts and epidermal keratinocytes; and, potentially, pathological remodeling, resulting in wounds healing by scarring [102]. Specifically, chronic wounds in RDEB allow for bacterial colonization and infection, in turn aggravating the immune response and further delaying wound healing. IL-4R; IL-13RA1; CXCL13; MCP-1; regulated on activation, normal T-cell expressed and secreted; and IFN-α are all overexpressed in RDEB skin, revealing a dysfunctional immune response [102]. Additionally, a pre-clinical study in Col7a1-deficient mice found that loss of C7 alters keratinocyte migration by disrupting laminin-332 deposition and impairs fibroblast migration and granulation tissue formation after cutaneous insult [103]. These studies exemplify a profound perturbation of normal wound healing in RDEB and the need for targeted therapies to address the specific needs of this patient population. As MSCs have the capability of producing growth factors needed for proper wound healing, improving angiogenesis and modulating inflammation, they could be a viable treatment option for genodermatological diseases such as RDEB, especially if non-autologous MSCs are used, and could provide the missing or dysfunctional protein to the skin.

Pre-Clinical and Clinical Use of MSCs in RDEB

RDEB is a systemic disorder. In addition to the readily observable skin wounds in RDEB patients, there are other sites of basement membrane disruption. Blistering in the mouth or fusion of the tongue to the floor of the mouth can make eating almost impossible. Esophageal erosions can lead to webs and strictures that can cause severe dysphagia. Consequently, malnutrition and vitamin and mineral deficiencies may lead to growth restriction in young RDEB patients. Finally, it is thought that repeated cycles of inflammation and fibrosis lead to a lifetime risk of aggressive squamous cell carcinoma that is higher than 90%, with a 5-year survival rate near zero [104,105].

There is currently no cure for RDEB, but strides to improve treatment have been made with bone marrow transplant and use of MSCs to dampen aberrant inflammatory responses and improve wound healing [106-111]. A drawback to these therapies is that the beneficial effects have shown to be only transient. Repeated injections throughout a patient’s life would likely be necessary for MSC-based treatments. Additionally, the mechanism of improved patient phenotype and decreased blistering is unclear. It is hypothesized that MSCs migrate to wounded tissue and then exert anti-inflammatory effects, providing benefit via the MSC secretome. Therefore, scientific focus should be on and research efforts should be geared toward pre-clinical studies to dissect out the concrete mechanisms that are at play when MSCs are delivered systemically or locally. Pre-clinical studies are limited and have focused on in vitro MSC interactions with skin cells (keratinocytes and fibroblasts) and using MSCs in wound healing models of humanized mice or mice without competent immune systems.

An in vitro study using BMSC exosomes on RDEB fibroblasts showed that exosomes can influence RDEB fibroblasts by transporting C7 to the extracellular space directly and providing fibroblasts with messenger RNA coding for C7, allowing for the fibroblasts to translate and synthesize C7 themselves [112]. Another group generated MSC-like cells from induced pluripotent stem cells derived from keratinocytes [113]. The researchers detected C7 at the dermal–epidermal junction after subcutaneous and intravenous injection of these cells into mice.

An in vivo murine study demonstrated that MSCs can be engineered to overproduce C7 [114]. These MSCs were incorporated into human bioengineered skin (using RDEB keratinocytes and fibroblasts) and grafted onto a NOD-scid IL2Rgamma^nu11 mouse. Engineered MSCs were locally retained, deposited C7 at the dermal–epidermal junction and restored anchoring fibril density. An additional study reported similar results in a tamoxifen-inducible Col7a1 knock-out mouse, showing that local injection of congenic MSCs restored the anchoring fibrils at the dermal–epidermal junction and partially reversed the RDEB phenotype [115]. Pre-conditioning of mice with cytokines TGF-β or TNF-α for 24–72 h before harvesting BMSCs showed increased Col7a1 expression, providing additional support for MSC pre-conditioning as a mechanism to increase C7 production in MSCs before clinical application [116]. Another in vivo study found bone marrow transplant to restore C7 in the cutaneous BMZ, specifically with PDGFRa+ bone marrow-derived cells migrating to skin grafts from Col7a1^null mice that had been grafted on to green fluorescent protein (GFP) bone marrow-transplanted C57BL6/N wild-type mice [117].

A small percentage (1–2%) of dermal cells in the skin express the marker ABCB5 and have been shown to display protein expression and anti-inflammatory benefits similar to those of BMSCs [54,118]. The use of dermal-derived ABCB5+ MSCs provided systemically to neonatal Col7a1^−/−RDEB mice significantly prolonged survival compared with phosphate-buffered saline-injected control animals, with ABCB5+ dermal MSC-treated mice showing decreased M1 macrophage skin infiltration [119]. Additionally, recent work in a wounded mouse model has shown that ABCB5+ MSCs exhibit superior homing to skin wounds compared with BMSCs after peripheral injection [5]. RNA sequencing of ABCB5+ MSCs compared with BMSCs showed expression of major histocompatibility complex class II and homeobox (Hox) genes, specifically HOXA3, in ABCB5+ MSCs. Critical to inducing migration of endothelial and epithelial cells for wound repair, HOXA3 is a promising gene candidate for investigating mechanisms of wound homing and local wound healing [5]. These cells are being evaluated in early-phase clinical trials, including local therapy for chronic venous ulcers and systemic administration for treatment of RDEB (NCT03529877, EudraCT 2018-001009-98). Results from the first human trial in which 12 wounds from nine patients with chronic venous ulcers were treated with local administration of ABCB5+ dermal MSCs showed safety, tolerability, median wound size reduction of 63% and perceived decrease in pain [120].

MSCs have shown promise in clinical trials for improving wound healing and quality-of-life outcomes in dermatological diseases such as RDEB, although the mechanism of action of MSCs in the disease process remains unclear. It is possible that BMSCs are recruited to wounded skin by the release of HMGB1 by apoptotic keratinocytes. This has been shown in a pre-clinical murine model of skin grafting, and HMGB1 is the only known biomarker of RDEB, with serum levels correlating with severity of disease [67,121]. A comprehensive list of MSC-based therapies injected either locally or systemically and their outcomes in RDEB is shown in Table 2, excluding case reports.

Table 2.

Clinical RDEB studies and their outcomes using MSCs as therapy.

Authors	Study type and ID number	RDEB patients, n	Intervention	Findings
El-Darouti et al. [122]	Double-blind	7 BMSCs + cyclosporine, 7 BMSCs no cyclosporine	One single injection of BMSCs ± cyclosporine	• Number of new blisters decreased significantly in both groups (P = 0.003 and 0.004). • Rate of healing of new blisters significantly faster in both groups (P < 0.001).
Petrof et al. [123]	Unblinded, prospective phase 1/2 study; EudraCT 2012-001394-87, ISRCTN46615946	10	Three intravenous infusions of non-HLA-matched BMSCs (day 0, day 7 and day 28)	• No increase in C7 and no new anchoring fibrils at day 60. • Patients reported decrease in disease severity. • Improved wound healing (increased suction blister time and decreased blister counts) for up to 6 months.
Rashidghamat et al. [121]	Phase 1/2 clinical trial; EudraCT 2014-004500-30	10	Two intravenous infusions of BMSCs (day 0 and day 14)	• Reduction in disease activity scores (eight of 10 subjects). • Significant reduction in pruritus. • Decrease in biomarker HMGB1 at day 28 and day 60.
Conget et al. [124]	Prospective study of two patients, unblinded	2	Intradermal allogeneic BMSC injection	• C7 protein detected along BMZ day 7 after injection. C7 present for 4 months after injection of BMSCs.

ID, identifier. BMZ, basement membrane zone

In conclusion, although in vivo studies have shown promise for restoring C7 in the BMZ, clinical studies have shown variable and transient improvement in patients with RDEB. Additional investigation of the specific immunomodulatory properties of MSCs and their impact on the skin microenvironment demands further study in vitro.

Burns

Burns are the fourth leading cause of injury in the world and can leave permanent functional damage [125]. Treatment of deep (second-degree) or full-thickness (third-degree) burns relies on excision of necrotic tissue followed by replenishment of healthy tissue, often with the use of autologous skin grafts. Burns differ from other types of skin wounds, in that they can progress and deepen with time and, similar to RDEB, can heal with scarring. A recent review by Rangatchew et al. [126] highlights the use of MSCs for burn therapy, summarizing three human and a multitude of murine and porcine studies. The three human studies included two single case reports and a larger study with 60 patients randomized to one of three groups (n = 20 per group): (i) excision of burn with autologous skin graft, (ii) excision with BMSC injections 2 days and 10 days post-burn and (iii) excision and injection or topical spraying of umbilical cord MSCs 2 days and 10 days post-burn [127-129]. Results from the human studies found increased vascularization of granulation tissue, as indicated by visual assessment and decreased percentage of burn area after MSC treatment. Importantly, aside from infection at the burn site in 25% of patients (not likely attributed to MSC application), no additional adverse events were reported [126]. Although these studies are initially promising, further studies with more rigorous, objective measures of wound healing, such as total body surface area of wound, skin biopsies to show granulation tissue formation, analysis of inflammatory cytokines and infections in wound beds or time to complete wound closure, are needed.

Animal studies using MSCs to treat burns are more numerous, showing that local treatment improves body weight and activity level of mice while significantly reducing local and systemic inflammation, apoptosis and burn wound progression. The limitations of these studies include great heterogeneity in animals and burn models as well as the type of MSCs used. Because of the progressive nature of burns, treatment with MSCs should focus on prompt delivery to combat progression of the wound and induce the formation of vascularized granulation tissue to hasten wound healing.

Diabetic Wounds

In diabetes, skin ulcerations are common on the lower extremities, affecting 15% of patients, 14–24% of whom require amputation as a result of inadequate wound healing and infection [130]. Diabetic skin ulcers result from poor vascular perfusion and are characterized by disintegration of the dermis and epidermis. Diabetic wounds are similar to RDEB, in that they are remarkably resistant to healing and frequently become colonized with pathogens. Additionally, treatment for diabetic wounds focuses on non-adherent dressings, wound gel and prevention of additional trauma, with minimal targeted therapies to address the disrupted microenvironment [95]. The specific impact of MSCs on the improvement of angiogenesis could be valuable in the treatment of diabetic wounds, as outlined in the review by Maranda et al. [81]. In chronic diabetic wounds, angiogenic factors such as VEGF-A, IL-6, FGF and TGF-β1 are decreased, impeding wound healing [131]. Additionally, decreased vascular flow can cause a hypoxic wound bed. Interestingly, MSCs cultured in hypoxic conditions demonstrate increased production of anti-inflammatory cytokines and growth factors such as TGF-β and VEGF-A and expression of ECM molecules (type III collagen, fibronectin and elastin) and improve the viability and migration of dermal fibroblasts, potentially increasing their therapeutic efficacy in this type of wound [132-135].

Pre-clinical studies in rats have found that BMSCs promote the production of MMP-2, EGF and IGF-1 by human keratinocytes and enhance the migration and proliferation of keratinocytes by downregulating TIMP1 and TIMP2 and initiating Erk signaling [136]. A review of the trials using MSCs as therapy in diabetic wounds is outlined by Cao et al. [137]. These studies ranged from 15 to 96 patients, and the outcomes showed positive results, including increased pain-free walking distance, improvement in leg perfusion and ankle-brachial index and increase in the vascularity of the dermis. Interestingly, a study of 96 patients with diabetic foot ulcers in whom autologous BMSCs were injected into the tibial artery showed 79% limb salvage [138]. Overall, these results are promising and allow for objective, quantitative measurements of efficacy; however, the mechanism of MSC migration and local contribution to wound healing is still uncertain.

Chronic Wounds

If resolution of the inflammatory phase of wound healing does not occur because of infection, disrupted skin architecture or poor circulation, a chronic wound can ensue. This is common in both RDEB and diabetic ulcers [102]. This state of chronic inflammation can cause increased levels of metalloproteinases that disrupt normal remodeling of the skin after injury by destroying components of the ECM and abrogating the secretion of growth factors needed for resolution of wound healing [48]. The traditional treatment for chronic wounds relies on debridement (removing foreign debris and contaminated/necrotic tissues), compression, pressure relief, hyperbaric oxygen, antimicrobials and moist dressings to hasten wound healing. None of these treatments target the aberrant inflammatory environment of a chronic wound [139]. MSCs have been found to not only support wound healing, as described earlier, but also combat harmful infections. Direct antimicrobial activity was demonstrated by BMSCs against Escherichia coliand Staphylococcus aureus, both common wound pathogens, in in vitro and in vivomodels [140]. A more recent study found that ADSCs effectively decreased the growth of E coli in vitro, demonstrating significant antimicrobial activity [141]. Additionally, ADSCs along with umbilical cord-derived MSCs have been shown to inhibit the growth of Pseudomonas aeruginosa, another type of bacteria common to skin wounds [142,143]. The antimicrobial effects observed by MSCs further justify their therapeutic use in wound healing.

In conclusion, wounds vary widely depending on the origin of insult. This should be noted when developing therapeutic strategies to target the disruption by focusing on attenuating inflammation, increasing angiogenesis, genetically modifying cells to increase expression of missing/dysfunctional proteins (genetic diseases), pre-conditioning them to increase expression of immunomodulatory factors or using specific culture conditions to achieve desired MSC efficacy. The downfalls of using MSCs in clinical trials may result from poor engraftment of MSCs after delivery. Therefore, the mode of delivery must be carefully considered and will be discussed in the following section.

Mechanisms of MSC Delivery to Skin

Topical delivery

Topical delivery can include both intradermal injection and cutaneous application. Local injection of MSCs to the wound site allows for a greater number of cells to be introduced to the area, but results in past studies have been transient, lasting up to 4 months [124]. BMSCs have also been delivered to human diabetic wounds in a fibrin spray [17]. This study found the application of the cells was safe, but a large number of cells were needed (>1 × 10⁶ per cm² of wound area) to decrease wound size (P = 0.0058). The same delivery method using GFP+ BMSCs was tested in a diabetic mouse model, and GFP+ cells were found near blood vessels, suggesting that these cells may persist, although the time point of analysis was 5 days after treatment.

Systemic delivery

Systemic injection of MSCs has the advantage that the cells can travel to multiple sites of disruption, although the mechanism involved and the ability of MSCs to home to wounded skin are still being studied. MSCs express several chemokine receptors, principally CCR7, which is a receptor of CCL21. Previous work has shown that this plays a role in MSC migration, and CCL21 expression is increased in irritated skin [144]. MSC migration to the skin was increased after intradermal injection of SLC/CCL21 and also resulted in accelerated wound repair [59]. Drawbacks to systemic injection of MSCs include infusion reactions such as fever and death of cells in the small vasculature, specifically of the lungs, without ever reaching a cutaneous destination. This is termed the “pulmonary first-pass effect,” which is a major barrier in the ability of MSCs to home to cutaneous wounded tissue. Animal studies in rats have estimated that approximately 4% of systemically injected MSCs are likely to survive passage through the pulmonary vasculature, although studies in humans are lacking [145,146].

Scaffolds and Three-Dimensional Bioprinting

Incorporating biomaterials into MSC delivery has been shown to be advantageous in wound healing. Seeding of MSCs on a collagen sponge prior to application to human skin wounds demonstrated therapeutic efficacy, as determined by healing of the wounds and increased epithelialization of surrounding tissue [18]. When applied to wounds of nude mice, the collagen sponges containing MSCs facilitated marked tissue regeneration after 2 weeks, whereas this was not detected when MSCs were injected without the collagen scaffold. This use of biomaterials to aid in cell delivery is an example of the “biological wound dressing concept” [147]. Though only a limited number of biological wound dressings have demonstrated significantly superior healing outcomes, the delivery of MSCs within a scaffold composed of biomaterials is a promising avenue for MSC therapies.

MSCs can be seeded within a scaffold of biomaterials, achieving both physical and biomechanical integrity. Yannas et al. [148] determined specific structural features, including highly cross-linked collagen glycosaminoglycan matrices, to be the foundation of a “template” for skin regeneration following injury. This acellular scaffold was shown to delay the onset of wound contraction by about 10 days in an in vivo experiment on adult guinea pigs [149]. Inclusion of cells such as MSCs in a bioengineered scaffold is hypothesized to be superior to the use of either component independently. In an in vivo rat wound model, human MSCs were seeded on a medical-grade polycaprolactone dressing. Compared with control groups, this combined MSC and medical-grade polycaprolactone dressing demonstrated significantly improved skin regeneration and decreased wound contraction [150]. Theoretically, bioactive materials secreted from MSCs could be delivered to the wound site through the porous structure of the scaffold to promote wound healing. Additionally, MSC scaffolds may contain multiple cell types and ECM components that produce skin-like structures and promote skin regeneration. MSCs have been shown to differentiate into an epidermal lineage when co-culturing MSCs and fibroblasts on a collagen gel and adding EGF and vitamin D3 [151].

Although scaffolds allow for the delivery of MSCs as well as possibly advantageous biomaterials and cell types to accelerate wound healing, they are laborious to produce and lack precise cell placement. Three-dimensional (3D) bioprinting, or printing of biomaterials into a specific shape, can be theoretically fully standardized and automated, resulting in control over cell and biomaterial organization. The route of 3D delivery of MSCs is thought to more closely replicate the natural in vivo environment of MSCs and thus foster a natural cellular response. Culturing MSCs within a pelleted cell spheroid, for instance, has been shown to improve MSC differentiation into both the typical lineages and hepatocytes [152]. MSCs have been bioprinted in structures replicating tissues such as cartilage and tumors [153,154]. Furthermore, with their ability to differentiate into endothelial-like cells, MSCs are being investigated for their use in vasculature bioprinting [155]. Replicating complex structures such as blood vessels is important when considering bioprinting for skin regeneration. Although 3D bioprinting has potential as a cell and tissue delivery system, this technology is still very much in its infancy and requires incredible optimization. At the current stage, functional skin cannot be bioprinted; rather, cells can be bioprinted to mimic skin structure. Importantly, the biological process that follows the actual bioprinting is crucial for the development of skin-like structures and the overall success of this method.

A comparison of the advantages and disadvantages of each aforementioned MSC delivery method is shown in Table 3. The favored mechanism of MSC delivery to skin wounds requires additional investigation, with the safety and efficacy of systemic versus topical administration of MSCs as critical variables in MSC therapy for chronic wounds in patients with RDEB-associated or other chronic wounds.

Table 3.

A comparison of various methods of MSC delivery to skin wounds.

Delivery method	Advantage	Disadvantage
Topical spray/local injection	Delivery directly to affected areas; avoids injection reactions.	Transient improvement in phenotype.
Systemic injection	Delivery to all affected areas, potential for long-term persistence.	Potential risk of infusion reactions, cell death in small vasculature; efficacy is dependent on homing ability of cells.
Scaffold	Addition of biomaterials and other cell types to promote healing, template for MSCs to expand and secrete wound healing biomaterials.	Laborious to produce and customize; lacks precision of cell placement.
3D bioprinting	Standardized and automated process; allows for precise cell and biomaterial placement	Technology is in very early phases; bioprinted skin-like structures lack vasculature and complex multi-layered skin structure.

Genetic Modifications of MSCs

Genetic modification of MSCs has been used to promote skin regeneration and increase MSC-mediated immunomodulatory effects [156]. Developing MSC-based therapies for genodermatological skin diseases like RDEB is a continuing challenge. Although MSCs demonstrate innate wound healing properties, genetic modification of these cells is an exciting technique for enhancing their therapeutic potential. MSCs can be genetically modified to overexpress or inhibit gene expression or modify certain pathways [157,158]. Additionally, modification of MSCs with CRISPR/Cas-based non-viral gene editing has been shown to increase the viability of transplanted MSCs by increasing HIF1a gene levels [159]. In RDEB, genetically modified MSCs have the potential to not only promote wound healing but also act as a vehicle for gene, protein or drug delivery. Over a decade ago, gene therapy was one of the most obvious avenues to investigate in the treatment of RDEB and other forms of EB. In a recent study, human umbilical cord-derived MSCs were transduced with a self-inactivating lentiviral vector encoding COL7A1 transgene to express C7 [114]. The engineered MSCs were delivered intradermally into human RDEB grafts pre-established on immunodeficient mice. These MSCs persisted, deposited C7 and restored anchoring fibril density. Alternative approaches using retroviral gene-corrected keratinocytes have been successful in individual cases of EB [160-162]. However, all gene therapy approaches for EB remain experimental [163]. The use of genetic modification to improve the natural properties of MSCs or target diseases like RDEB specifically greatly increases the possible utility of MSCs.

Approaches to Improving the Potential of MSCs for Wound Treatment

Because of their broad potential and impact on immune modulation, MSCs are one of the most studied cellular applications for the treatment of human disease. Clinical studies have fallen short of promising pre-clinical results, and several studies have examined this history [164,165]. In 2019, biomedical research as a whole was criticized for having a “replication crisis” in pre-clinical studies [166]. MSC research similarly demonstrates inconsistent outcomes. Poor reproducibility in the field may result from differences in MSC tissue source, culture conditions or MSC handling before their use in clinical applications.

Typically, human subjects enrolled in clinical trials receive one or multiple doses of allogeneic MSCs that have been previously cryobanked under Good Manufacturing Practice (GMP) conditions or following cell culture with or without pre-conditioning the cells with cytokine stimulation, serum deprivation, 3D culture conditions, hypoxic conditions or another form of cellular manipulation. For previously cryobanked MSCs that are thawed directly prior to clinical use, there have been observations of molecular signatures of cell injury, such as a decrease in suppressive functions and shortened persistence in vivo, which could be related to freezing-induced senescence and could account for the varied efficacy of MSCs used in clinical trials. This is outlined further in the review by Moll et al. [167].

There have been multiple attempts to improve the ability of MSCs to exert clinical benefit in patients. These have included augmenting MSC survival with ECM proteins or hypoxic conditioning [132]. For example, ECM proteins of interest include tenascin C and decorin, which have been shown to decrease inflammation, improve tissue repair during wound healing and abrogate skin scarring [132].

One of the most studied mechanisms by which inflammation triggers MSC activity involves IFN-γ. This cytokine is typically produced during Th1 immune responses, which are associated with autoimmunity mediated by cellular means, such as CD8 T cells and natural killer cells. Exposure of MSCs to INF-γ has been demonstrated by numerous groups to increase immune suppressive activity via stimulation of the enzyme indoleamine-2,3-dioxygenase [168]. MSC exposure to indoleamine-2,3-dioxygenase induces the production of other inhibitors of inflammation by MSCs, including the complement inhibitor factor H and the immunomodulatory molecules TGF-β and HGF. TNF-α pre-treatment of MSCs increases angiogenic activity in vitro, as assessed by expression of VEGF, and serum deprivation in MSC cultures enhances expression of VEGF, angiopoietins, IGF-1 and HGF, which are markers of angiogenesis [169].

Finally, another approach could be culturing MSCs in 3D scaffolds, which upregulates the production of anti-inflammatory molecules compared with monolayer culture [168]. Although there are many possible ways to enhance the in vitro suppressive and favorable wound healing activity of MSCs, a systematic study of these approaches needs to be tested in GMP-cultured MSCs with clinical applications in mind. Most importantly, criteria for clinical MSC preparations should be matched to the hypothesized mechanism of action and tested pre-clinically, and clinical handling of the MSCs prior to use should depend on the tissue source, type of disease and desired recipients [165].

Safety and Manufacturing of MSCs for Wound Healing Applications

The use of MSC-related therapies is not without risk, and the safety of these cells must be considered fully when developing clinical applications. Primary safety concerns regarding MSC administration include tumorigenicity, fibrosis and pro-inflammation. MSCs have the capacity to develop into tumors and have been shown to actually promote tumor development by releasing certain chemokines and growth factors [14]. Additionally, it has been reported that the immunosuppressive qualities of MSCs support tumorigenicity and metastasis of tumor cells [170,171]. In certain animal models, MSCs have been reported to home to tumors and promote tumor development [172,173]. To safely administer MSCs, the potential for malignant transformation must be ruled out with strict cell quality control measures. Furthermore, it has been found in ex vivo expansion that MSCs develop genomic instability during later passages [174,175]. This finding stresses the importance of determining appropriate cell passage numbers when generating clinical therapies. Although these findings are very concerning, at the time of this review, current in vivo clinical trials involving MSC therapies have not reported adverse cases of tumorigenicity [176].

MSC-based therapies also pose a risk of fibrosis [177]. Though MSCs do have immunosuppressive effects, this has been shown only following exposure to very high levels of pro-inflammatory cytokines. In the presence of low levels of TNF-α and IFN-γ, MSCs have actually been shown to have an inflammatory response [178]. The environment of MSCs is crucial to their effect on the immune system, and this must be considered when developing MSC therapies. Importantly, thrombogenic events have been reported in several clinical trials following MSC infusion [179-182]. In addition to mechanical vessel obstruction by MSCs (reported to be as large as 30 μm in diameter), in vitro and in vivo studies have shown that MSCs display pro-coagulant activity [183-185]. The use of anti-coagulant drugs like heparin or bivalirudin may counter the hypercoagulable milieu induced by MSC infusion [186]. Continued research into assessing and limiting coagulation following MSC therapies is vital.

Although MSCs are thought to be generally safe, caution and protective measures are vital and also warrant further investigation into the use of MSC-conditioned medium therapeutically as opposed to administering MSCs as a cellular product. As explained previously, MSCs promote wound healing mainly through paracrine action of the secretome. MSC-conditioned medium contains the secreted factors that are thought to cause the advantageous effects, such as re-epithelialization, angiogenesis and anti-fibrosis. Furthermore, Yagi et al. [141] demonstrated that MSC-conditioned medium alone was effective at significantly suppressing S aureus bacterial growth. Clinical therapies utilizing conditioned medium rather than MSCs may be an advantageous option and may possibly reduce the risk of tumorigenicity. Animal studies implementing this strategy to promote wound healing have verified it as a promising technique [64,70,187,188].

Finally, it is vital that MSCs for clinical application in wound healing be cultured and expanded in reproducible conditions for consistent characterization–such as retention of MSC markers or functional analyses–prior to administration. Therefore, much effort has focused on creating GMP standards to allow for clinical use of MSCs. Factors that need to be considered are standardization of culture conditions (incubator temperature, oxygen and carbon dioxide levels); cell seeding density; length of culture; and culture medium, specifically the use of xenogeneic additives such as fetal bovine serum (FBS) [189,190]. FBS use in the clinical context brings a variety of risks, such as prion and viral transmission or adverse immunological reactions to xenogeneic components as well as extreme variability of FBS batches [191]. A potential alternative is the use of human serum or human platelet lysate. Comparison of human serum with FBS at 10% v/v has shown higher proliferation capacity of MSCs in human serum. Mushahary et al. [192] discuss the impact of using human components on MSC differentiation and expansion in further detail in their review.

In conclusion, comprehensive studies directly comparing human components with FBS in MSC expansion and differentiation from various tissue sources are lacking. This leaves many questions unanswered regarding the best way to prepare MSCs for optimal clinical results. Therefore, further studies and research in this area are urgently needed.

Discussion

MSC-based therapies have shown promise in aiding wound healing. The advantageous effects of the MSC secretome make them a strong candidate for a novel wound therapy, although to create a treatment targeted to the specific wound microenvironment, the wound etiology must be taken into account. Furthermore, although MSCs have been shown to play a beneficial role in the inflammation, proliferation and remodeling phases of wound healing, the logistics of catering to each of the different wound healing phases must be contemplated. The immunomodulatory and antimicrobial effects of MSCs make them an especially promising therapy for genetic skin wounds, such as those seen in RDEB, burns and diabetic ulcers. Several critical challenges need to be addressed when developing MSC-based therapies, including choice of cell source; delivery method; in vitro pre-conditioning; and, depending on the application, overall safety profile. Continued research–specifically pre-clinical studies focusing on the mechanisms of immunomodulation of the surrounding microenvironment and ability of MSCs to engraft and persist in wounds–is recommended to further elucidate the current limitations and unknown mechanisms involved in MSC orchestration of wound healing, especially their differing effects depending on specific niche and environmental stimuli. Current data, including pre-clinical and clinical trials, however, do support the future realities of MSC-based therapies for wound healing, although the mode of therapy and pre-conditioning regimen will need to be standardized for each application.