Potential of mesenchymal stem cells as topical immunomodulatory cell therapies for ocular surface inflammatory disorders

Abstract Ocular surface inflammatory disorders (OSIDs) are a group of highly prevalent, heterogeneous diseases that display a variety of aetiologies and symptoms and are risk factors for serious complications, including ocular and cornea impairment. Corneal inflammation is a common factor of all OSIDs, regardless of their cause or symptoms. Current medications include over‐the‐counter lubricating eye drops, corticosteroids, and ciclosporin, which either do not treat the corneal inflammation or have been associated with multiple side effects leading to alternative treatments being sought. Regenerative medicine cell therapies, particularly mesenchymal stem cells (MSCs), have shown great promise for immunosuppression and disease amelioration across multiple tissues, including the cornea. However, for successful development and clinical translation of MSC therapy for OSIDs, significant problems must be addressed. This review aims to highlight considerations, including whether the source of MSC isolation impacts the efficacy and safety of the therapy, in addition to assessing the feasibility of MSC topical application to the cornea and ocular surface through analysis of potential scaffolds and cell carriers for application to the eye. The literature contains limited data assessing MSCs incorporated into scaffolds for corneal administration, thus here we highlight the necessity of further investigations to truly exploit the potential of an MSC‐based cell therapy for the treatment of OSIDs.


| INTRODUCTION
The cornea is a highly organized, transparent tissue at the ocular surface. It is comprised of three main cellular layers: the epithelium, the stroma containing the keratocytes, and the endothelium, separated by the Bowman's membrane and Descemet's membrane, respectively 1 ( Figure 1). Coating the outer mucosal surface of the cornea is the tear film, a thin, liquid layer, 2 mainly constituted of mucin and lipid. As the cornea is avascular, the tear film plays a vital role in the supplementation of nutrients and oxygen, as well as the expulsion of waste such as epithelial debris, foreign bodies, and toxins. Interactions between the ocular surface and the tear film allow for a smooth optical surface, correct functioning of limbal epithelial cells and protection from mechanical and microbial insults. 3 Additionally, healthy corneal tissue is maintained through tight immunoregulatory mechanisms at the ocular surface, modulated by both the innate and adaptive immune systems.
Ocular surface inflammatory disorders (OSIDs) occur when the tightly regulated homeostasis at the ocular surface is disturbed, and encompass a range of heterogeneous diseases with a variety of aetiologies and symptoms, where inflammation plays a critical role in pathogenesis. 4 Dry eye disease (DED), meibomian gland dysfunction (MGD), allergic eye diseases, cicatricial conjunctivitis, chemical eye burn, trauma, iatrogenic insult following corneal and/or refractive surgery, and contact lens-related complications are the common examples of OSIDs that are frequently encountered and managed in clinical practice.
OSIDs are highly prevalent in the general population. For example the global prevalence of DED has been estimated at around 5% to 50% depending on the diagnostic criteria and study population. 5 MGD, a major contributor to evaporative DED, has been shown to cause a myriad of negative impacts on the ocular surface including heightened inflammation, oxidative stress, tear hyperosmolarity, and increased corneal epitheliopathy. 6 These diseases often serve as an important risk factor for major ocular surface complications including infectious keratitis, corneal vascularization, opacity, visual impairment, corneal melt, and perforation. [7][8][9] In addition, OSIDs are regularly associated with pain and irritation, causing a considerable reduction in the patient's quality of life, activities in daily living, and work productivity. 10 Irrespective of their source, insult to the cornea ultimately results in a vicious cycle, where chronic irritation activates an immune response, augmenting the irritation. 4 Currently, treatments include over-the-counter lubricating eyedrops to alleviate disease symptoms, and corticosteroids to ameliorate the inflammation. However, these treatments require long-term topical application, multiple times a day (every hour), placing high demand on patient compliance and interfering with their day-to-day life. Furthermore, corticosteroids have been linked to severe adverse effects including increased risk of infectious keratitis, inhibition of corneal wound healing, raised intraocular pressure, and cataracts. 11,12 Ciclosporin serves as a valuable steroid-sparing immunomodulatory agent for managing a range of OSIDs, although side effects are common. 13 Lifitegrast, a recent FDA approved drug, represents another useful topical anti-inflammatory treatment for DED. However, both ciclosporin and lifitegrast are associated with a high rate, up to 70%, of side effects, including burning sensation, itching, and blurred vision, among others. 14 Because of the abundance of therapeutic factors possessed by human stem cells, regenerative medicine may hold the key to developing a superior treatment to alleviate OSIDs. This review outlines the process required for the application of stem cell therapy for OSIDs, through assessing optimum cell type and delivery method to the ocular surface. Here, we focus on the use of mesenchymal stem cells (MSCs) due to their well-accepted immunomodulatory properties and suggest that applying the cells topically, via a removable substrate or scaffold, may offer the most convenient and efficacious therapy.

| POTENTIAL SOURCES OF STEM CELLS FOR IMMUNOMODULATION OF THE INJURED OCULAR SURFACE
Inflammation is recognized as a significant feature in the etiopathophysiology of OSIDs, therefore stem cells with efficacious anti-inflammatory properties would be optimal for successful treatment. Limbal epithelial stem cell transplantation (LSCT) and cultivated corneal epithelial (CCE) sheets have shown promising therapeutic results for restoring a normal corneal epithelial phenotype in patients with severe chemical injury and dry eye. 15,16 However, the primary utilization of LSCT and CCE is to generate an entire new epithelial layer in situ or in vitro, respectively, rather than for their immunosuppressive capacity, used predominantly in cases where injury has resulted in a limbal epithelial stem cell deficiency (LSCD). Their incapacity to suppress inflammation is supported by data demonstrating contraindications of LSCT in the presence of active inflammation in bilateral diseases, including Stevens-Johnson syndrome, ocular cicatricial pemphigoid, and graft vs host disease (GVHD). In fact, the failure of LSCT is often accredited to sites of active inflammation creating a toxic microenvironment at the ocular surface. 17 Although these techniques have proven, in some cases, successful to treat injuries such as chemical burn, which are associated with high levels of inflammation, The structure of the cornea. Working from the ocular surface anterior to posterior, the cornea is made up of an epithelium; Bowman's membrane; stroma; Descemet's membrane, and endothelium

Significance statement
This is the first review focusing on the potential of engineering mesenchymal stem cell (MSC) therapies that can be applied topically to the ocular surface, in order to treat inflammatory disorders that cannot be managed through steroids or other means. This study aims to highlight different considerations, including whether the source of MSC isolation may impact the efficacy and safety of the therapy, in addition to assessing the feasibility of topical stem cell application to the ocular surface through analysis of potential scaffolds.
it is likely that some of the immunosuppression was governed and achieved by the immune-modulating, amniotic membrane (AM) scaffold the cells were applied with. 15,16 As the pros and cons of LSCT have been covered in previous reviews, 17,18 we wish in this review to highlight alternative sources of stem cells that could be considered for novel regenerative medicine therapies.
Differentiating induced pluripotent stem cells (iPSCs) into immune-mediating cells, such as regulatory T cells, 19 holds the potential to improve the inflammatory symptoms of OSIDs. However, this therapeutic strategy is limited by the high tumorigenic potential, cost, and regulation associated with the generation and application of iPSCs. 20 MSCs are best known in regenerative medicine for their ability to modulate both the innate and adaptive immune systems, 21 suggesting a potential use for the treatment of inflammation in OSIDs. Their capacity to reduce inflammation has been assessed in vitro and in vivo on multiple tissues, including the kidney, heart, cartilage, liver, brain, skin, and cornea, 22 with preclinical success demonstrated by their current use in clinical trials. 23 MSCs encompass a group of fibroblast-like, multipotent progenitor stromal cells, defined initially by their capacity to differentiate into osteoblasts, adipocytes, and chondrocytes, 24 however MSCs are now utilized primarily for the ability to elicit a therapeutic response through communication with target tissue cells.

| DIRECT COMMUNICATION OF MSCs AND TARGET CELLS
Limited evidence has demonstrated that MSCs can interact with the target tissue directly via cell-cell contacts such as gap junctions and tunneling nanotubes. 25 This has been demonstrated in cardiac tissue, where the respiratory chain in myocytes was salvaged through mitochondrial transfer. Although not investigated in the literature, hypothetically this mechanism could restore cells at the ocular surface and is therefore an area with potential for future exploration.

| PARACRINE SIGNALLING OF MSCs AND POTENTIAL EFFECT ON CORNEAL IMMUNOMODULATION
The main interest surrounding MSCs has shifted to their paracrine function, as a positive therapeutic response can be achieved irrespective of whether the cells reach the target organ. 26 There is an abundance of data demonstrating MSC secretion of anti-inflammatory factors, cell-mobilization factors, and growth factors in response to inflammatory mediators. 27 Stimulation of MSCs with interferon-y (IFN-y) has been studied abundantly in the literature, demonstrating activation of the IFN- HGF has also been implicated as a fundamental factor in immunomodulation, secreted by MSCs stimulated with IL-1ß. 38 HGF alone is powerful enough to suppress antigen presenting cell activation and to limit the generation of Th1 cells in the lymphoid tissue. Topical HGF application significantly reduced the rejection of corneal grafts in a murine model of GVHD, through suppression of immune cell infiltration, and has the potential to maintain and restore corneal transparency through the inhibition of α-SMA and its inducer TGF-ß. 31,39 Other key anti-angiogenic molecules secreted by MSCs include TNF-α stimulated gene/protein (TSG-6), demonstrated as vital in the inhibition of neovascularization, and suggested to function through the inhibition macrophage infiltration and the induction of apoptosis of vascular endothelial cells. 40 As well as macrophages, TSG-6 has been demonstrated to suppress activation and infiltration of neutrophils following chemical and mechanical corneal injuries, 41 making it a potent modulator of both angiogenesis and inflammation.
An alternative method to exploit this paracrine signaling mechanism of MSC to treat OSIDs would be through harvesting extracellular vesicles from the MSC for therapeutic application. 42 The potent therapeutic factors of MSCs packaged in small vesicles could help to overcome the safety and regulatory hurdles of cell application and have shown potential in corneal wound healing and immunomodulation in vivo. 43 Fully elucidating the pathways and interactions of different MSCs and the corneal microenvironment will help to increase the safety profile and therapeutic value of these cells for both tissue regeneration and inflammation suppression, highlighting the necessity to explore different MSC sources.

| MSC SOURCE
It is of utmost importance to consider MSC source, both tissue and donor (autologous or allogeneic). Although MSCs have previously been claimed as immune-privileged because of their lack of expression of Major Histocompatibility Class (MHC) II proteins and costimulatory molecules B7 and CD40 ligand, 44 immune rejection of MSCs derived from allogeneic sources has proven a major therapeutic challenge for application to a wide variety of conditions. 45 Similarly, the ocular microenvironment has been claimed to be immuneprivileged, with original accounts demonstrating placement of a foreign antigen in the eye did not elicit an immune response. 46 Although GVHD is a contraindication of an ocular allogeneic stem cell transplant in approximately 40% to 60% of patients, 47 the immunomodulatory properties of MSCs may give them additional protection, even if from an allogeneic source, with reports of multiple clinical trials using MSCs to both prevent and treat GVHD. 48 Although allogeneic cell therapy is beneficial for the manufacturing of the therapy, potential adverse effects of foreign cells are vital to consider.
MSCs can be isolated from most tissues in the body and cultured in vitro, however they do not all possess the same properties. For example, literature demonstrating MSC secretion of the antiinflammatory cytokine, IL-10, is highly contradictory, and could be because of the source of the cells. 49 For successful translation to clinic, it is important that multiple sources of MSCs are explored, to develop the most efficacious and cost-effective treatment for OSIDs.

| EFFECT OF CULTURE, PASSAGE, AND PRIMING OF MSCs
The effect of culture and passage must be balanced when considering MSCs as a therapeutic agent. Optimally, the maintenance of MSC phenotype and behavior is vital, however the ability to culture cells to high passage numbers allows greater opportunity for allogeneic scaleup. in vitro passage investigations have shown that ageing MSCs are subject to morphological changes and reduced immunomodulatory capacity with a significant reduction in release of trophic factors such as VEGF, 67,68 lead to the use of innovative culture techniques such as the Quantum hollow fiber bioreactor, to culture greater number of cells without adverse changes. 69 Optimization of culture medium should also be performed as different media have been shown to affect the phenotype of initially identical cell populations. 70 Priming, or "licensing" of the cells, with in vitro application of cytokines such as IFN-y has been shown to improve immunosuppressive capacity and pharmaceutical utility. 71 Although the mechanisms are not fully elucidated, suggested explanations include the upregulation of IDO, the clustering of MHC and co-inhibitory molecules, and epigenetic changes. 72,73 Additionally, priming the cells through hypoxia treatment and activation of the MSC nucleotide binding domain, as well as techniques including gene modification existing to improve therapeutic potential. 74

| APPLICATION OF MSCs TO THE OCULAR SURFACE: TOPICAL VS ALTERNATIVE METHODS
A comparison of studies demonstrating MSC efficacy in various ocular surface disease models, using different delivery mechanisms can be found in Table 1. In contrast to developing stem cell therapies for internal organs, the location of the ocular surface makes it an ideal candidate for the noninvasive topical application of stem cells. The advantages of topical application of MSCs, in a similar manner to that discussed for skin healing, 89 101 Ke et al also found that combination of a topical polysaccharide hydrogel and subconjunctival injection of BM-MSCs performed additively to enhance corneal epithelial cell recovery and corneal clarity in a rat model of alkali burn, 85 reinforcing the idea that the choice of substrate if as important as the stem cell.
Synthetic hydrogel bandage CLs are currently used to protect the corneal surface in combination with the delivery of pharmacological or biological therapeutics. 102 Most are composed of siloxane hydrogel, 103 and hold desirable qualities, while the absence of protein reduces the risk of allogeneic rejection or disease transmission, 104 and their shape allows self-maintenance on the cornea. To avoid the undesirable effects of corneal epithelial cell attachment and protein fouling when placed on the ocular surface, CL materials rarely contain cell adhesion motifs, and consequently must be functionalized to behave as a cell delivery device; these can be provided by integrin binding sites from serum, 3T3 feeder layers, and surface plasma polymerization with acid groups. [105][106][107] Three-dimensional scaffolds produced via electrospinning have a large surface area, with the nanofibers arranged to imitate extracellular matrix proteins. MSCs have been demonstrated to attach and proliferate effectively on these scaffolds, and when applied to the cornea aid healing and regeneration. 78,108 Further modification to allow for the possibility of cell detachment has also been explored with thermoresponsive, electrospun scaffolds for the culture of C-MSCs. 109,110 However, the invasive procedure of suturing the scaffold to the ocular surface seems unfavorable compared to nonsurgical alternatives.

DATA AVAILABILITY STATEMENT
Data sharing is not applicable to this article as no new data were created or analyzed in this study.