Abstract

Introduction

Physiologic wound healing is a multi-focal, highly orchestrated process that involves four distinct but overlapping phases; coagulation, inflammation, tissue replacement, and resolution. Each one of these phases is dependent on the active involvement and contributions of various cells. In the coagulation phase, clotting of serum and degranulation of platelets release matrix components and a host of growth factors and chemokines. These growth factors and chemokine promotes the immigration and activation of hematopoietically-derived inflammatory cells followed by fibroblasts, endothelial cells, and keratinocytes. During the tissue replacement phase, these signals promote the dedifferentiation of the formed cells to enable rapid motility and production of the granulation tissue. The neo-vasculature forms at this stage via angiogenesis to supply oxygen and nutrients to the highly metabolic environment. During the resolution phase, the fibroblasts and myofibroblast contract the matrix and promote collagen deposition and collagen fibril organization. Throughout the first three phases of wound healing; platelets, inflammatory cells and the provisional matrix foster angiogenesis through the secretion of growth factors, cytokines and extracellular matrix (ECM) components. However, during the last phase of wound resolution, up to 90% of the neo-vasculature involutes, in response to not only clearance of trophic factors but active signaling from late-stage chemokines via the CXCR3 G protein coupled receptor (1–3). It is during this last phase that the matrix matures and the skin reverts to a pauci-cellular state.

The vascular system plays a critical role in promoting the progression of the wound healing process. However, its dysfunction, either too much or too little early angiogenesis or conversely late regression, leads to a host of problems including dehiscence, ecchymoses and ulcer formation. As the vasculature is regulated by pericytes, attention has recently turned to these ill-defined but ubiquitous cells in controlling wound repair. These cells have been found to contribute to more than regulation of the neo-vasculature during repair, and are the focus of this perspective.

During wound repair, pericytes interact with platelets, inflammatory cells and connective tissue cells in addition to endothelial cells. Although no direct interaction between platelets and pericytes has been identified, pericytes have been implicated in modulating the extrinsic clotting cascade (4). Pericytes have been observed to possess phagocytic properties and may function to assist macrophages during the inflammatory phase of wound healing (5). Under inflammatory condition, pericytes have been observed to act as antigen presenting cells and promote the activation of T-cells (6). Pericytes secrete several cytokines including stromal-cell derived factor (SDF-1), CXCL9 and CXCL12, which was found to enhance lymphocyte infiltration (7, 8). There is emerging evidence-indicating pericytes as a source of myofibroblast and transition could occur through a PDGF-dependent mechanism (9). And lastly, the role for which pericytes are most known, communication with endothelial cells controls neo-vascularization during wound healing by regulating vessel formation and stabilization (1, 10, 11). These finding highlight the diverse role of pericytes in mediating cellular response during wound healing. Herein we will discuss the new evidence demonstrating that pericytes are dynamic regulators of the wound healing cascade.

Pericytes

Pericytes are considered to act as the vascular smooth muscle cells of the microvasculature providing tone and integrity to the capillary bed. This is in part based on their location. They are found surrounding the arterioles, venules and capillaries. Pericytes may not be restricted to the microvasculature as pericyte-like cells having characteristics of both pericytes and smooth muscle cells have been found in small blood vessels (12). Their coverage, consisting of a single layer, of the microvasculature is variable with density being dependent on the organ. The ratios of pericytes to endothelial can be as high as 1:1 in the retina and central nervous system (CNS) cells to as low as 1:100 in capillary bed of skeletal muscle of the upper extremities (10, 13, 14). Pericyte density appears to correlate with blood pressure. Vessels with higher blood pressure gradients such as the retina and lower extremities have higher pericytes coverage (15), supporting this view of pericytes as microvascular support structure.

Morphologically, pericytes tend to have a large cell body with an elongated membrane having numerous secondary branching processes, which allows a single pericyte to interact with and regulate many endothelial cells. Pericytes are typically embedded in the basal membrane of the microvessel allowing for direct interaction with endothelial cells. There are three principal connections made between endothelial cells and pericytes; gap and adherence junctions, and adhesion plaques, which contribute to transmission of biological and mechanical signals between the cells (10, 16, 17). Gap junctions allow for the exchange of ions and small molecules allowing direct pericyte-endothelial cell communication. The adherence junctions and adhesion plaques support the transmission of contractile forces allowing for the regulation of vessel diameter and blood flow. Because of the heterogeneity of pericytes (15, 18), their identity is typically determined through morphological features along with distinct molecular markers.

There are a number of cell surface markers that define pericytes but these are not unique to pericytes and can be dynamically expressed on pericytes and other cells (10, 19)(Table 1). Smooth muscle cells, endothelial cells and fibroblasts also express a number of these markers complicating identification of these cells. The most common markers used to identify pericytes include PDGFRβ, chondroitin sulfate proteoglycan 4 (NG2), alpha smooth muscle actin (αSMA), desmin, CD146/MUC18, and desmin regulator of GTPase signaling 5 (RGS5). Although these markers are expressed on subsets of pericytes, expression is not ubiquitous across all pericytes. Expression of these markers on pericytes can vary depending on the type of vessel they surround and pericyte phenotype. Studies have shown pericytes surrounding arterioles are generally NG2⁺/αSMA⁺, venules NG2⁻/αSMA⁺, and capillaries NG2⁺/αSMA⁻ (20). These markers may provide a method to identify pericyte location within the vasculature. Levels of αSMA expression have been found to correlate with increased and decreased vasoconstriction of vessels and could be used to determine an activation state (21). Desmin expression may be a marker for maturity as immature pericytes express a high level (22). Typically CD146 and PDGFRβ are ubiquitously expressed. However, it must be appreciated that these molecules are present in numerous non-pericyte cells. In addition, pericytes also express a number of mesenchymal stem cell (MSC) markers, which include CD105, CD73 and CD90 (20). The use of pericyte morphology combined with cellular marker expression can be used to identify both vessel and pericyte phenotype. Still, the lack of even relatively specific or selective markers limits the ability to determine the source of pericytes prior to the interaction with the microvasculature. This hampers investigations into pericyte function in the earlier stages of wound healing when the microvasculature is poorly organized.

Pericyte Origin

A further question relates to the origin of pericytes. The embryonic linage, for the most part, is thought to be of a mesodermal origin. These may be part and parcel due to pericytes having a mesenchymal stem cells/multipotent stromal cells (MSC) phenotype (23). Pericytes and MSC reside in the same perivascular niche (24–26). Both MSC and pericytes share common markers (CD44, CD73, CD90, CD105) and exhibit multi-linage potential leading to pericytes having been viewed as a subset of the MSC population (25). Pericytes as a MSC phenotype is under much debate due to insufficient understanding of pericyte origin. It has been observed that a subpopulation of pericytes lack expression of αSMA, a protein necessary for cell contraction, suggesting that two distinct populations of pericytes exist in the vasculature (27) a smooth muscle (αSMA⁺) and MSC (αSMA⁻) phenotype. In support of pericyte variability, two distinct pericyte populations have been identified as nestin⁻/NG2⁺ (type-1) and nestin⁺/NG2⁺ (type-2) (28, 29). The type-1 pericyte are αSMA⁺ pericytes and may play a role in regulating vessel tone, and thus are mainly found around arterioles and venules. The type-2 pericytes participate in angiogenesis and have a MSC-like behavior in being able to differentiate into a neuronal linage (30, 31). These findings posit that pericyte phenotype and function may be regulated by its association with the vasculature; pericytes function as smooth muscle cells when associated with the vasculature and dissociation promotes dedifferentiation to a MSC phenotype. Thus, pericytes isolated and grown in culture revert to an MSC phenotype, having a regenerative or reparative capacity. Additionally, their location within all tissues make them ideal candidates for local recruitment for tissue repair. Additionally, there is some evidence that pericytes may come from bone marrow-derived cells in the circulation (32), though this may relate more to vasculogenesis with angiogenic support originating in neighboring tissue.

Pericyte Function

The well-described function of pericytes is to regulate vessel tone and permeability of mature vessels as well as modulating angiogenesis. By this alone, pericytes would impact wound healing. But even more so, in these functions we see hints of how pericytes may alter repair beyond the vasculature. Pericytes promote endothelial migration and cell survival (33, 34) in part by responding to hypoxia with upregulation of matrix metalloproteinases (MMP) 2, 3 and 9 (35). These proteinases help clear the tissue debris during early stages of wound healing but prolonged expression leads to non-healing ulcers.

In addition to promoting vessel maturation and stabilization, pericytes may also regulate vessel guidance and promote regression of non-essential nascent vessels. Upon initiation of angiogenesis pericytes detach from the angiogenic vessel and migrate into the parenchyma forming a guidance tunnel for the angiogenic sprout (36, 37). During development individual pericytes have been found associated with tip cells of the angiogenic sprout (38). These cells were observed to guide the vascular sprout and thus may play a role in determining the formation and location of a new vessel (39). We have shown that pericytes have the ability to negatively regulate angiogenesis through activation of CXCR3 on endothelial cells (1). Activated pericytes secrete the CXCR3 ligands IP-9 (CXCL11) and IP-10 (CXCL10). This signaling limits endothelial cell migration in response to growth factors by inhibiting m-calpain-mediated rear release (40) and triggers anoikis of endothelial cells via calpain-mediated cleavage of the β3 integrin (41) (Figure 1). The involution of such vessels is in large part responsible for the loss of the majority of microvasculature during the resolution phase of wound healing (42).

Open in a separate window

Figure 1

CXCR3 signaling leading to anoikis in endothelial cells. The ligands for CXCR3 (CXCL4/PF4, CXCL9/Mig, CXCL10/IP-10, and CXCL11/IP-9/I-TAC), derived from pericytes and other cells, lead to calcium influx. In the calcium sparks at the membrane, the calcium concentration can be sufficient to activate μ-calpain which then cleaves the β3 integrin. As β3 integrin is the main isoform in endothelial cells, this leads to anoikis and endothelial cell death.

It needs to be noted that along with CXCR3 activation on endothelial cells, vessel regression can also be mediated by additional factors, which include angiopoietins and withdrawal of trophic factors. The decrease in VEGF and other trophic factors later in wound healing would act as if ‘taking the foot off the gas’ and limiting the driving force for angiogenesis. Angiopoietins are complex having both vessel stabilizing properties as well as anti-angiogenic signaling (43). This is further complicated by angiopoietins impacting pericytes in addition to acting on endothelial cells directly (44, 45).

How the pericytes can play a dual role of both stabilizing vessels while pruning excess microvasculature may be explained by the maturation-dependent nature of CXCR3 expression. CXCR3 expression is nearly absent on mature vessels but expression is regained during wound-induced angiogenesis (41, 46). During wound healing, initial angiogenesis in the tissue replacement phase occurs in the absence of pericytes (1). The endothelial cells that venture into the provisional matrix without pericytes express CXCR3. As the vessel begins to mature, the endothelial cells adhere to each other and quiesce; concomitantly CXCR3 expression on endothelial cells diminishes. This would allow for the association of pericytes with the microvasculature without involution. Consistent with this, pericyte-stabilized vessels in mature wounds failed to express CXCR3 (1, 41, 47).

Pericytes in Wound Healing

From the foregoing, pericytes could play numerous roles in wound healing – angiogenic modulation, paracrine effectors (factors and enzymes), and stem cell contribution to tissue. The main function by which pericytes were recognized, vascular development and stability, is central to wound healing and pathologies thereof (Figure 2). The process of wound healing is a complex and dynamic process involving various players for secretion of soluble mediators and deposition of extracellular matrix along with migration of various cell types, including fibroblasts, keratinocytes, macrophages, leukocytes, endothelial cells and pericytes.

Open in a separate window

Figure 2

Pericytes regulate the angiogenic response during wound healing. As the tissue replacement phase initiates, the wound bed is largely anoxic, with the VEGF response driving angiogenic sprouting; the pericytes are driven to express TGFβ1 being both angiogenic and pro-fibrotic. These immature endothelial cells express high levels of CXCR3. Upon establishing functional vascular conduits, the endothelial cells mature, concomitantly downregulating CXCR3 expression but producing CXCR3 ligands that attract pericytes. The pericytes also express CXCR3 ligands, with these ligands driving anoikis of nearby immature endothelial cells resulting in vascular involution. Late in wound healing the pericytes coat and stabilize the remaining, CXCR3-negative endothelial cell tubes.

Pericytes in tissue regeneration

Pericytes have been used in a various animal models for wound repair, muscle, central nervous system (CNS), cartilage, cardiac and bone therapy (24, 61–65). The use of pericytes for skeletal muscle repair has been well demonstrated in vivo. Pericytes have been shown to colonize host muscle upon transplantation and can spontaneously fuse into myotubes promoting muscle regeneration (62). Injection of muscle-derived pericytes into injured skeletal muscle improved function better than injected myoblasts (19). When used in a mouse model of muscular dystrophy, pericyte transplantation shows an enhancement in skeletal muscle function, with the use of genetically-corrected pericytes ameliorating the disease phenotype (62, 74). Given the myotube-intrinsic nature of the dystrophy phenotype, it is assumed that the pericytes directly contributed to the myotube as a stem cell. However, secreted factors, soluble and matrix, may also contribute to the survival of the existing myotubes and limiting the inflammatory damage.

Another current area investigating the use of pericytes is myocardium repair. Initial cell therapy used MSC that were directly injected into the heart (75–77). These studies provided modest benefit but significant improvement in cardiac function was still lacking. A comparative study between bone marrow MSCs and saphenous vein pericytes for cardiac repair after myocardial infarct MI, showed that the pericytes were superior in efficacy of cardiac function with no myocardial calcification in MI (78). This was attributed to the paracrine signaling from the pericytes improving revascularization of the tissue. Pericytes also were found to home back to perivascular sites and this homing was suggested to support the long-term survival of the transplanted pericytes (61). These studies suggest that pericyte treatment attributes multiple cardioprotective mechanisms for cardiac repair rather than acting as stem cells directly contributing the repair (79). While not directly applicable to skin wound healing, these findings suggest similarities in supporting the neo-angiogenic functioning.

Role of Pericytes in Skin Fibrosis

The excessive accumulation of extracellular matrix (ECM) components, including collagen I, fibronectin, and hyaluronan, in injured tissue results in fibrosis. Unchecked fibrosis can cause permanent damage to the tissues and may result in morbidity. Despite their heterogeneous origins, different fibrotic disorders, including those of the liver, heart, lungs, skin and kidney all involve the activation of fibroblasts and myofibroblasts (80–84). Multiple origins have been hypothesized for myofibroblasts (81); the principal sources postulated being pre-existing fibroblasts, bone marrow-derived circulating mesenchymal progenitors (bone marrow stromal cells and fibrocytes) and pericytes (32, 85). It has been shown that myofibroblasts are capable of synthesizing essential components of the ECM such as collagen types I, III, IV and V (81). It has been described in kidney fibrosis that pericytes, detach from the vessel walls, migrate, acquire a fibroblast-like phenotype where they differentiate into myofibroblasts (activated fibroblasts) and act as collagen-producing cells (86).

Functional changes in endothelial cells and pericytes have been associated with excessive skin fibrosis resulting in hypertrophic scars and keloids. Kischer et al. (87) noted that microvessels in hypertrophic scar and keloid are occluded or partially occluded, apparently owing to an excess of endothelial cells (87, 88). Microvessels of hypertrophic scars and keloids appear to be surrounded by multiple satellite cells (pericytes), and the innermost satellite cells closely resemble myofibroblasts. These pericapillary myofibroblasts can migrate from their perivascular position to the interstitial area. The authors also stated that the observed extent of microvascular occlusion supports a previously published theory that hypoxia is involved in hypertrophic scarring and keloids (88). Hypertension has also been cited as a risk factor for keloids and hypertrophic scars (Reviewed extensively by (89)). The response to hypertension by pericytes/myofibroblasts in the fibrotic state is mediated by hypoxia. Pericytes seem to be sensitive to hypoxia, and their contraction during hypoxia may promote blood flow impairment (90). The pericyte contraction as well as pericapillary myofibroblast contractions can aggravate the occlusion of the blood vessel lumen, leading to a vicious cycle resulting in ever-increasing hypoxia and results in fibroproliferation. As the pericytes produce numerous MMPs that would lead to matrix degradation products that are chemotactic for cells of the innate immune system that in turn drives more myofibroblast-mediated fibrosis. A recent study by Dulauroy et al. postulated the role for pericytes as myofibroblast precursors of dermal fibrosis (91). These studies showed evidence that the majority of collagen-producing, α-SMA+ myofibroblasts formed following acute dermal or muscle injury are generated from tissue-resident ADAM12⁺ (ADAM12⁺ is a matrix metalloprotease-12 and is expression is restricted to embryogenesis and fibrosis) cells and furthermore from a PDGFRβ⁺ NG2⁺ perivascular subpopulation (92). Pericytes isolated from skeletal muscle that were nestin⁻/NG2⁺ (type-1) were found to express αSMA; when these were exposed to transforming growth factor-β (TGF-β) they became fibrogenic (93). In addition, these type-1 pericytes were also observed to express collagen I (93). Taken together expression of both nestin and ADAM12 may be key regulators that determine whether pericytes can be transformed to a myofibroblast phenotype.

Systemic sclerosis (SSc), a complex autoimmune disease is characterized by fibrosis of the skin and visceral organs. SSc is a vascular disease that is marked by chronic and prolonged tissue injury and inflammation (reviewed by (94)). Endothelial and pericyte activation has been noted as one of the contributing factors for initiation of vasculopathy and pathological accumulation of excessive ECM proteins in SSc (95, 96). Microvascular endothelium activation and damage is shown to be always present followed by dysregulation in vascular remodeling in samples from patients affected with SSc ((97, 98)). Among the many factors that result in SSc fibrotic lesions, abnormal characteristics of microvascular pericytes are cited as a major source as pericytes have important functional roles in the regulation of vascular development, maturation and remodeling. Studies suggest that endothelial cells and pericytes both undergo changes in the early stages of SSc (99). In SSc fibrotic lesions, pericytes showed markers of activation such as platelet-derived growth factor receptor beta and high-molecular-weight melanoma associated antigen (100). Strong evidence shows convergence of microvascular pericytes and resident fibroblasts to a myofibroblast lineage and thereby contributing to SSc by synthesizing excessive ECM components (101).

Interestingly, it has been reported that there is a difference in the properties of bone marrow mesenchymal stem cells (BM-MSCs) derived from healthy controls and SSc patients. SSc-derived MSCs behave as pericytes and display a more mature and myofibroblast-like phenotype probably due to changing microenvironmental cues during disease progression (102). Considering the impactful role played by pericytes in SSc pathology, molecular targets against PDGF signaling using pharmacological agents such as Imatinib mesylate, sunitinib, and sorafenib are currently under investigation.

Pericytes in Non-Healing (Chronic) Skin Wounds

In chronic metabolic diseases, diabetes in particular, the healing process is defective in not going to completion, leading to what is termed a “chronic wound” (103). Diabetic complications impair both macro-and micro-vascular structures causing stroke, atherosclerosis, impaired wound healing, retinopathy, neuropathy and nephropathy (104). It has been estimated that between 3–6 million people in the United States suffer from one of the three main types of chronic wound -- diabetic, venous or pressure ulcers (103). Microvascular pericytes act as a crucial interface between blood-borne and connective tissue signals. In normal vascular physiology, pericytes are closely associated with a stable endothelium and upon injury, are encountered by other cell types and environmental signals.

In chronic wounds, the microenvironment is hostile with an imbalance in protease activity and chemokine release. We have seen that pericytes produce many of the enzymes whose persistent presence not only mark chronic wounds but also contribute to the stalled healing process. Thus, in this situation the feed-forward cycle of matrix remodeling leading to a chronic inflammation is coupled with paracrine signals such as TGFβ and CXCR3 ligands that impair neovascularization. This provides for a model in which pericyte presence would prolong the non-healing state.

An early investigation by Laaff et al. (105) on patients with chronic venous insufficiency (CVI) that lead to leg ulcers demonstrated that most of the patients (31 out of 42) with CVI showed pericyte changes. The investigators attributed these changes to the destruction of the pericyte envelope, which might lead to microcirculatory dysfunction. However, how the pericyte dysfunction relates to failure to heal will require new explorations.

Pericytes as Therapeutic Agents

Patients suffering from diabetes, vasculitis and ischemic diseases are prone to develop chronic ulcers that are difficult to treat due to many of the patients having co-morbidities, which further hinders the healing process. Chronic wounds fail to heal at a rate up to 33% with implementation of current wound management strategies (106). Current strategies have focused on the use of cell therapy to enhance the healing of chronic wounds. The focus has been on tissue engineering due to the potential of returning tissue to it original state. Classical cell based therapies use fibroblasts or keratinocytes sheets from autologous and allogenic tissue. The use of cell sheets has improved wound healing but fail to fully incorporate the tissue graft into the new tissue. Furthermore, the tissue grafts are difficult to maintain and are expensive.

Experimental therapies have used mesenchymal stem cells (MSCs) to enhance wound healing in preclinical models (107–110). MSC use has been found to have limitations for their ability to enhance wound healing (110). They require expansion prior to use risking mutational events, and exhibit low levels of long-term survival. Thus, the noted therapeutic benefit mainly from the release of tropic factors rather than a direct contribution (110, 111). These limitations make MSC less than ideal for therapeutic use.

Recent studies have identified pericytes as a cell with mesenchymal differentiation potential (25). These cells are currently being tested in experimental models of tissue regeneration because of their plasticity, regenerative capacity and availability (24, 61–65). The data from these studies indicate that pericytes may be a more viable option then MSC for tissue regeneration.

Pericytes are easier to obtain, since they can be harvested from more accessible tissue including, subcutaneous fat, umbilical cord and placenta. Sufficiently large numbers of pericytes can be isolated from subcutaneous fat enabling transplantation without expansion. The use of pericytes for tissue regeneration has gained substantial momentum in the last 10 years. Pericytes isolated from human umbilical cord were found to improve re-epithelialization and reconstitute the dermis (24). In this study, pericytes were administered in a fibrin gel on full thickness wounds. The pericyte-treated full thickness wounds had a higher tensile strength, denser and organized collagen fibers, enhanced angiogenic response and a more organized dermis (24). In an excisional wound model, pericytes isolated from adipose tissue were found to increase vascularization, enhanced remodeling and promote earlier collagen deposition (112). Using an organotypic skin culture, pericytes were found to enhance epidermal regeneration through secretion of laminin α5 (59). These studies show that pericytes display mesenchymal stem cell properties, making them a potential cell for therapy in treating non-healing wounds.

Although these initial studies are promising, new data suggest the possibility of two subtypes of pericytes. Type 1 (Nestin⁻/NG2⁺) and type 2 (Nestin⁺/NG2⁺) pericytes were found to play diverse roles in tissue repair (28, 29, 93). The type 2 pericytes were observed to promote angiogenesis, found to be associated with vessels and possessed myogenic potential (28, 113). On the other hand, type 1 pericytes were found to promote fibrosis in injured skeletal muscle through the deposition of collagen (93). Since pericytes have been implicated in contributing to fibrosis or even chronic wounds, these finding may help in understanding the possible role pericytes may play (114). Understanding to what extent type 1 pericytes may contribute to fibrosis and to what extent type 2 pericytes may contribute to tissue regeneration will be necessary to design effective therapeutic approaches to limit undesirable side effects. Taken together, pericytes have the potential to be used as a cell therapy for non-healing wound but further understanding the function of these cells is necessary to reduce unwanted outcomes.

Conclusions

Pericytes perform multiple functions that are critical during wound healing. Their regulation of the angiogenic response, including both growing and pruning nascent vessels, and then stabilizing the surviving ones, alone would constitute a major role in the repair process. However, beyond this well-documented function, their ability to act akin to MSC as both a source of paracrine/matricrine signals and potential stem cell opens up new avenues that require exploration. The reports to-date offer tantalizing glimpses of a world of various pericyte functions that can be exploited to control healing, to promote closure or reducing scarring. However, our ability to redirect these processes are limited by the diversity of pericytes that posses these different functions and thus lack the ability to exquisitely target the interventions.

Acknowledgments

Footnotes

References