The Role of Chemokines in Mesenchymal Stem Cell Homing to Wounds

Bone marrow-derived MSCs improve wound healing

MSCs are multipotent adult stem cells that were first isolated from bone marrow but have since been detected in all adult tissues.^1,2,4,5 Clinical interest in MSC is based on their ability to enhance repair and regeneration of injured tissue and to modulate the immune system in inflammatory diseases. For therapeutic use, MSCs are being harvested from a variety of tissues including adipose tissue, umbilical cord, and bone marrow. This review will primarily focus on bone marrow-derived MSCs given both endogenous and systemically administered bone marrow-derived MSC home to sites of tissue injury.²

Bone marrow-derived MSCs are a heterogeneous population of plastic adherent cells that represent 0.001–0.01% of total marrow.^1,2,4 In 2006, the International Society for Cellular Therapy outlined three minimal criteria for defining a bone marrow-derived MSC⁵ (Fig. 1). Bone marrow-derived MSC must be (i) plastic adherent; (ii) positive for the stromal cell surface markers CD73, CD90, and CD105 while negative for the hematopoietic lineage markers CD45, CD34, CD14, CD11b, CD79a, and HLA-DR; and (iii) able to differentiate into bone, fat, and cartilage in vitro. Although these criteria have helped to standardize the definition of MSC, it is important to note that they are based on in vitro properties and may not necessarily reflect the in vivo phenotype. Resolving this issue is an important goal for the field. Also important is characterization of the subpopulations of MSC. These studies are needed to clarify whether MSC populations from different tissues are the same and whether there are discrete subpopulations of true stem cells and/or specific subpopulations with enhanced therapeutic effects. It is clear that the definition of MSC phenotype will change as these studies progress.

Preclinical human studies have investigated the effect of MSC treatment on chronic wounds,^6,7 burns,⁸ radiation wounds,⁹ and blistering skin disorders.¹⁰ The results from these studies and animal models have been promising with positive outcomes on wound repair. In general, MSC treatment results in accelerated wound closure, increased angiogenesis, and increased granulation tissue formation.¹ There is also evidence of increased wound tensile strength, reduced scarring, and regeneration of epidermal appendages such as sweat glands, sebaceous glands, and hair follicles.^1,11,12 In addition, MSCs also have the potential to restore the structural integrity of the skin by producing extracellular matrix proteins. For example, exogenous MSCs have been shown to provide collagen VII to restore the damaged epidermal–dermal junction in the blistering disorder dystrophic epidermolysis bullosa.^10,13

MSC mediate their beneficial effects on wound repair through paracrine signaling, immune modulation, and differentiation.¹ MSC paracrine signaling regulates local cellular responses to injury. The MSC-derived signals include growth factors, cytokines, chemokines, and prostaglandins.¹⁴ Interestingly, exosomes (microvesicles that transfer proteins and RNA between cells) have recently been implicated in MSC paracrine signaling: in a rat model of burn injury, MSC-derived exosomes increased cell proliferation and accelerated reepithelialization in wounds.¹⁵ MSC have also been reported to modulate the inflammatory response to injury. MSCs regulate macrophage activation¹⁶ and promote the conversion of M1 proinflammatory macrophages to M2 anti-inflammatory macrophages in wounds.¹⁷ Additionally, MSC differentiation contributes to wound repair by replacing damaged tissue. In wounds, MSCs have been observed to differentiate into keratinocytes, endothelial cells, and pericytes.¹ The impact of MSC paracrine signaling, immune modulation, and differentiation on wound repair is currently limited by the small number of MSCs that actually home and engraft to wound.^1,2 Strategies to optimize stem cell recruitment to sites of injury are clearly needed to maximize the therapeutic benefits of MSC therapy.

Chemokines, their receptors and chemotaxis

Chemokines are positively charged short peptides ranging in size from 7 to 13 kDa.¹⁸ The three-dimensional shape of the peptide is due in large part to the disulfide bonding between four amino terminal cysteines (Fig. 2). The arrangement of these four cysteines has been used to group chemokines into structurally related families. To date, four families have been identified: the CCL family with adjacent cysteine residues; the CXCL family with the cysteines separated by a single amino acid; the CL family with two instead of four cysteines; and the CX3CL family with the cysteines separated by three amino acids. The CCL and CXCL families are the largest with 26 and 17 members respectively.

Chemokines bind G protein-coupled receptors, which are seven-transmembrane receptors signaling via heterotrimeric G proteins (Fig. 2).¹⁸ The chemokine receptors are classified according to their interaction with specific chemokine families so CXCR receptors bind CXC chemokines, whereas CCR receptors bind CC chemokines. It should be noted that there are fewer chemokine receptors than chemokines. As a result a single chemokine receptor may bind several different chemokines and a single chemokine may bind more than one receptor.

Chemokines and their receptors were first defined by their role in mediating chemotaxis, which is directed migration in response to a chemical stimulus. To date, most of our knowledge about the mechanism of mammalian chemotaxis has been from the study of neutrophil and lymphocyte trafficking.¹⁹ These circulating leukocytes sense chemokine gradients immobilized via glycosaminoglycans on the endothelium, and polarize to form leading and trailing ends. This gradient sensing and cytoskeletal reorganization is mediated by chemokines interacting with specific chemokine receptors on the leukocyte cell surface.

Bone marrow-derived MSC home to cutaneous wounds

Bone marrow-derived MSCs have an inherent capacity to home to sites of injury and inflammation.^1,2,20 Endogenous and systemically administered MSC have been detected in injured skin, eye, vasculature, and heart. Systemically administered MSC also home to injured brain, lung, kidney, and liver. This homing is a multi-step process involving directed migration to the site of injury where MSCs adhere to the vasculature and then transmigrate from the vasculature into the injured tissue (Fig. 3). In the case of endogenous MSCs, homing also involves mobilization from the bone marrow niche into the peripheral circulation. A growing list of factors have been identified that participate in MSC homing including cytokines, chemokines, growth factors, bioactive lipids, and adhesion molecules.²⁰ Despite this general understanding of MSC homing, much work remains to be done to elucidate the molecular mechanisms responsible for each step.

Chemokines regulate MSC homing to cutaneous wounds

Endogenous bone marrow-derived MSCs home to cutaneous wounds. We and other investigators using bone marrow chimeras have observed MSCs localized to the dermis and blood vessels in wounds.^21–24 Additional evidence that MSCs are recruited to wounds is provided by studies injecting MSCs intravenously.^25–27 These systemically administered MSCs home to wounds and enhance wound healing. For bone marrow-derived MSCs to home to wounds, MSCs must be expressing a chemokine receptor that recognizes chemokines produced at the site of injury.

Gene expression analyses show that murine bone marrow-derived MSCs express the following chemokine receptors: CCR1, CCR2, CCR3, CCR4, CCR5, CCR6, CCR7, CCR9, CCR10, CXCR3, CXCR4, and CXCR7,²⁸ whereas human bone marrow-derived MSCs express all the members of the CCR and CXCR families plus CXC3R and XCR.^29,30 Interestingly, the MSC chemokine receptor profile has recently been reported to be sensitive to time in culture. For example, the expression levels for CCR1, CCR2, and CXCR4 in murine MSCs in long-term culture (>10 passages) were at least fivefold lower than MSC in short-term culture (<4 passages).²⁸ It is possible that time in culture selects growth of a discrete MSC subpopulation with a unique chemokine receptor repertoire. Further investigation is warranted to determine whether this sensitivity to time in culture explains the divergent receptor profiles that have been previously reported for human MSCs.²⁹ These results underscore the need to assess the effects of MSC isolation and expansion on chemokine receptor profile. Also necessary are studies to determine whether the chemokine receptor profile is different between MSCs in vivo and in vitro. Again, having specific lineage markers for MSCs would greatly facilitate characterization of the MSC chemokine receptor profile. Understanding the regulation of the chemokine receptor profile is clearly important for optimizing both endogenous and exogenous MSC homing to wounds.

Cutaneous wounds produce a plethora of chemokines that may serve as chemoattractants for MSCs. The chemokine expression profile of a wound is regulated temporally and spatially during the repair process.¹⁸ Each chemokine has a specific expression pattern and unique role in inducing the directed migration of cells into and within the wound. The inflammatory response to injury provides an example of the sequential expression of chemokines to attract specific immune cell types.¹⁸ CXCL8 (also known as interleukin-8) produced soon after injury attracts neutrophils to the wound. CCL2 is then released by wound neutrophils to recruit monocytes, which are macrophage precursors. Lastly, expression of CCL3, CCL4, and CCL5 in the granulation tissue mediates B and T lymphocyte migration into the wound. Also pertinent to our understanding of chemokine function during wound repair is that a single chemokine may regulate the migration of multiple different cell types. CXCL8 illustrates this point as it serves as chemoattractant for neutrophils, endothelial cells, and dermal fibroblasts.¹⁸ A recent review by Martins-Green et al. summarizes the chemokines participating in each of the key cellular responses to injury.¹⁸ There have been a number of chemokine–chemokine receptor axes implicated in homing of bone marrow-derived MSCs to cutaneous wounds. These axes include CXCL12-CXCR4, CCL27-CCR10, and CCL21-CCR7 (Fig. 4).

Designing a chemokine-based therapeutic to enhance MSC homing to wounds

To date, strategies to enhance chemokine-mediated MSC homing to wounds are primarily targeting the CXCL12 (SDF-1)–CXCR4 axis. A variety of approaches are being used to amplify the SDF-1 signal at the wound site and/or overexpress the CXCR4 on the surface of the MSC. Increased SDF-1 levels in wounds has been achieved by localized intradermal injection of either SDF-1 recombinant protein or viral vectors expressing SDF-1 mRNA.³¹ Furthermore, both a hydrogel and an alginate scaffold have also been successfully used to deliver SDF-1 to wounds.^37,46 In all cases, administration of exogenous SDF-1 resulted in increased MSC recruitment to the wound.³¹ However, there are likely additional benefits given that diabetes and age significantly reduce SDF-1 levels in wounds.^35,47 An alternative approach to SDF-1 administration is the use of the CXCR4 antagonist, AMD3100 (Plerixafor). This antagonist is already being used in the clinic to stimulate mobilization of hematopoietic stem cells from the bone marrow into the blood in cancer patients.⁴⁸ A recent study demonstrates that local injection of AMD3100 at the site of injury increases recruitment of endothelial progenitor cells to wounds in rodent models.⁴⁸

An important advantage of administering either SDF-1 or AMD3100 to the wound is that it directly targets endogenous stem cells and circumvents the challenges with isolating, expanding, and then delivering stem cells to the wound. However, if the goal is to also recruit exogenous MSC then recognized challenges associated with MSC delivery to the wound need to be overcome. These challenges include entrapment of systemically administered MSCs in the lung⁴ and sensitivity of the MSC chemokine receptor profile to in vitro culture conditions.²⁸ There are also safety concerns as elevated levels of SDF-1 have been correlated with cancer and inflammatory diseases.⁴⁹

Direct targeting of CXCR4 levels on MSC is also being used to maximize homing to sites of injury. MSC are being genetically engineered to overexpress CXCR4 via viral vectors or cationic liposomes²⁰ and these cells are then topically delivered to the wound. An alternative to genetic engineering is to induce the MSCs to upregulate CXCR4 expression prior to delivery to the wound. This approach is referred to as preconditioning or priming. A number of preconditioning factors have been identified including hypoxia, insulin-like growth factor 1, inhibitors of glycogen synthase kinase 3-β, and valproic acid²⁰. All of these factors have been reported to increase CXCR4 levels by MSCs. Interestingly, preconditioning with SDF-1 may also induce MSCs to upregulate CXCR4 expression given it has been reported to elicit this effect on endothelial progenitor cells.⁴⁹ There are added benefits to SDF-1 preconditioning as a number of studies demonstrate that SDF-1 increases MSC engraftment and survival at the site of injury.⁵⁰ Collectively, these studies modulating the CXCL12 (SDF-1)–CXCR4 axis demonstrate feasibility and the promise of chemokine-based therapeutics for enhancing stem cell recruitment to wounds.