The inflammatory response following tissue injury has important roles in both normal and pathological healing. Immediately after injury, the innate immune system is activated, setting in motion a local inflammatory response that includes the recruitment of inflammatory cells from the circulation. This rapid response begins with the degranulation of platelets that arrive at the site and the injury-induced degranulation of resident mast cells. Local immune cells, including resident macrophages, are activated by proinflammatory mediators released in response to injury, as well as damage-associated molecular pattern molecules (DAMPs) (Ref. Reference Zhang and Mosser1). The hypoxic environment of the wound also promotes inflammation, because hypoxia stimulates numerous cell types, including macrophages, to produce mediators that are important for inflammation (Ref. Reference Sen2). In response to these many signals, the levels of leukocyte chemoattractants increase substantially, further enhancing leukocyte recruitment.
As the recruitment of leukocytes from the circulation begins in earnest, a pattern of leucocytic infiltration into the wound develops that is similar to other acute inflammatory conditions (Fig. 1). Neutrophils, the most abundant white cell in the circulation, infiltrate the wound quickly and are the dominant leukocyte in the earliest stages (Ref. Reference Ross and Odland3). Concomitantly with the influx of neutrophils, circulating monocytes enter the wound and differentiate into mature tissue macrophages (Ref. Reference Ross and Odland3). Mast cell numbers in the wound also increase, with most of the infiltrating mast cells originating in the adjacent tissue (Ref. Reference Artuc4). In the late inflammatory phase of wound repair, T cells appear in the wound bed, and may influence the resolution and remodelling of the wound (Refs Reference Barbul5, Reference Barbul6). As inflammation resolves and the number of leukocytes diminishes, the wound undergoes a lengthy period of remodelling and resolution. Although inflammation is not prominent during this resolution phase, many studies suggest that the events of the inflammatory phase have profound effects on the final wound outcome (Refs Reference Cowin7, Reference Martin and Leibovich8). Studies in many different anatomical systems suggest that scar formation and fibrosis may derive from inflammatory cell activity (Ref. Reference Martin and Leibovich8).
Pattern of leukocyte infiltration into wounds. Inflammatory cells are present during each of the phases of wound repair, represented here as haemostasis (yellow panel), early inflammation (light pink panel), late inflammation (dark pink panel) and resolution/remodelling (blue panel). The relative density of the four most prominent types of leukocytes in wounds (mast cells, neutrophils, macrophages and T cells) is depicted. Whereas neutrophils and lymphocytes disappear, low numbers of resident mast cells and macrophages are present during the lengthy remodelling phase.
Among immune cells in the wound, the role of the macrophage has been the subject of intensive investigation, yielding more than 600 published articles on the topic within the past 5 years. The emerging picture demonstrates that wound macrophages are multifunctional and able to influence nearly all phases of repair. Modulation of macrophage function, then, is a logical and rapidly emerging target for wound therapeutics.
Role of macrophages in wound healing
Landmark studies in the early 1970s and 1980s demonstrated that macrophages are critical to wound healing, and the ability of macrophages to produce factors that stimulate angiogenesis and fibroplasia has been firmly established (Refs Reference Leibovich and Ross9, Reference Polverini10, Reference Hunt11, Reference Kovacs and DiPietro12). Early studies used guinea pigs depleted of macrophages by treatment with both antimacrophage antiserum and glucocorticoids to study the role of this cell in the healing wound (Ref. Reference Leibovich and Ross9). Because glucocorticoids have several additional effects that might influence repair, these early observations were limited in interpretation. Recent advances in the use of genetically modified mice have overcome this limitation. These techniques allow a highly selective and specific depletion of macrophages in wounds and have confirmed a crucial role for macrophages in wound healing. Two separate groups have used murine strains bearing macrophage-restricted expression of the human receptor for diphtheria toxin to effect a toxin-mediated selective depletion of macrophages before the placement of wounds (Refs Reference Goren13, Reference Mirza, DiPietro and Koh14). The wounds of mice depleted of macrophages in this manner exhibited delayed wound closure, decreased granulation tissue formation and angiogenesis, decreased collagen synthesis, and decreased levels of growth factors, including vascular endothelial growth factor (VEGF) and transforming growth factor-β (TGF-β). In addition, depletion of macrophages resulted in reduced levels of myofibroblasts, which are contractile cell types that are important for wound closure.
More recently, another group, again using a diphtheria toxin system, undertook a temporal selective depletion of macrophages at sequential times during the healing process (Ref. Reference Lucas15). The depletion of macrophages during the early inflammatory phase resulted in impaired wound closure and granulation tissue formation. Depletion during the early proliferative stage caused severe haemorrhage and curbed later wound closure and tissue maturation. Macrophage depletion that was limited to the tissue maturation phase had no significant effect on healing. These studies are some of the first to delineate the different roles of macrophages during the phases of the healing process and support the concept that this cell exhibits different functional phenotypes as repair progresses.
Recruitment of monocytes and macrophages at sites of injury
Similarly to leukocyte migration at almost any site of inflammation, monocyte recruitment into wounds involves the sequential steps of endothelial cell activation, cell-to-cell interaction, and transmigration through the endothelium into the extravascular space. Because monocytes represent only about 3% of circulating leukocytes, the rate of monocyte influx into wounds is far from stoichiometric. Initially, monocytes may be recruited by factors produced quickly after injury, such as split products from the coagulation cascade, factors released from platelet degranulation and activated complement components. But most monocyte infiltration occurs later, and the preferential infiltration of monocytes represents the response to a locally produced chemotactic gradient that specifically favours these cells. One important group of chemoattractants produced within the wound are the chemokines, which are a group of related small proteins that display highly conserved cysteine amino acid residues. Chemokines can be produced by many cell types, and individual chemokines may preferentially recruit particular populations. Several studies have examined the expression and function of chemokines in healing wounds, and the general patterns of chemokine expression correlate with the movement of leukocytes, including monocytes, into wounds (Refs Reference Fahey16, Reference DiPietro17, Reference DiPietro18, Reference Wetzler19, Reference Engelhardt20). The role of chemokines in the recruitment of leukocytes is complicated, because more than 40 chemokines have been identified. The complexity of this process has been well described by others (Ref. Reference Bonecchi21).
Specific macrophage functions in wounds
Promotion of inflammation
Resting macrophages produce only low levels of proinflammatory mediators. On exposure to proinflammatory cytokines, interferons, lipopolysaccharides or other microbial products, or DAMPs (such as heat-shock proteins, high-mobility group box proteins and molecular fragments of the extracellular matrix), macrophages acquire a proinflammatory or ‘classically activated’ phenotype (Ref. Reference Mosser22). Following activation, proinflammatory macrophages themselves produce a large number of mediators and cytokines, including interleukin (IL)-1, IL-6, IL-12, tumour necrosis factor-α (TNF-α) and inducible nitric oxide synthase (Refs Reference Barrientos23, Reference Diegelmann and Evans24). Because many of these mediators have been shown to be present in the early wound environment, macrophages seem to be a probable source (Ref. Reference Barrientos23). Macrophages also produce chemoattractants, including chemokines, that recruit additional leukocytes (Ref. Reference DiPietro25).
Reparative and anti-inflammatory function of wound macrophages
In vitro studies suggest that macrophages are capable of transitioning from a proinflammatory to an ‘alternatively activated’, or reparative, phenotype (Refs Reference Brancato and Albina26, Reference Stein27). The alternative phenotype is characterised in part by expression of anti-inflammatory mediators such as IL-1 receptor antagonist, decoy IL-1 receptor type II and IL-10, and by the production of growth factors such as TGF-β, VEGF and insulin-like growth factor-1. The transition of macrophages to an alternative phenotype has been assumed to be requisite for the switch from inflammation to proliferation in the healing wound (Ref. Reference Mosser and Edwards28).
The canonical factors for inducing the alternative phenotype are IL-4 and IL-13. Curiously, however, recent studies suggest that IL-4 and IL-13 are not requisite for the modulation of macrophage phenotypes in wounds in vivo (Ref. Reference Daley29) Anti-inflammatory cytokines, glucocorticoids and modulators of glucose and lipid metabolism induce a broad spectrum of ‘alternative’ macrophage phenotypes, including those that exhibit non- or anti-inflammatory and pro-tissue repair functions. Recent studies suggest another pathway in which M1 macrophages are induced to develop into a novel M2-like phenotype in a manner that is independent of IL-4 and IL-13 (Ref. Reference Pinhal-Enfield30). In these in vitro studies, initial activation occurred by the engagement of toll-like receptors: an action that induces expression of the adenosine 2A receptor (A2AR). Subsequent interaction of A2AR with adenosine completes the activation, leading to an M2-like cell. This M2-like cell, dubbed the M2d macrophage, has been suggested to be important to wound-healing outcomes. As mentioned above, other factors, including IL-10, glucocorticoids, prostaglandins, metabolites and the process of efferocytosis (discussed in detail below), might also induce M2-like phenotypes. In the context of the healing wound, the role of each potential stimulus to the phenotypic switch and resolution of inflammation is not completely understood.
Until recently, the existence of discrete macrophage phenotypes within wounds was largely assumed rather than proven. However, recent studies of macrophages derived from skin wounds, as well as sponges implanted subcutaneously in mice, demonstrate that macrophages do exhibit multiple phenotypes that change over time (Ref. Reference Daley29). These studies suggest that M1-like cells, which are characterized by the production of TNF-α, IL-1 and IL-6, are common in the early phases of repair, whereas M2-like cells, with less proinflammatory cytokines and elevated markers of alternative activation, including CD206 and arginase 1, are common in the later stages of repair. However, these in vivo phenotypes appear far from simple, and do not completely mimic the previously described in vitro macrophage classifications. For example, populations of macrophages were found to exhibit both markers of alternatively activated macrophages, such as mannose receptors and cytokines (TNF-α, IL-6) associated with a classically activated phenotype. At any single time point, then, the wound bed may contain several discrete macrophage phenotypes or hybrid macrophage phenotypes.
Several studies now suggest that alterations in macrophage phenotypes have a critical role in the pathogenesis of chronic wounds. For example, in a murine wound model, iron overloading of macrophages has been shown to create dominant proinflammatory M1-like macrophage populations, with resultant impaired wound healing (Ref. Reference Sindrilaru31). This study has significant clinical implications because iron overload has been previously associated with human chronic venous ulcers and iron-overloaded macrophages have been identified in chronic venous ulcers (Refs Reference Sindrilaru31, Reference Zamboni32). Thus, at least for venous ulcers, iron might create a persistant proinflammatory macrophage phenotype that is critical to the failure to heal. In addition, our preliminary studies indicate that the transition from a proinflammatory to a prohealing phenotype is impaired in excisional wound macrophages from diabetic db/db mice. However, the factors involved in maintaining the persistent proinflammatory wound macrophage phenotype in diabetic mice remain to be elucidated.
Removal of neutrophils and reduction of apoptotic load in the wound
One important function of wound macrophages is the capacity to facilitate the nonphlogistic removal of neutrophils. Neutrophils are abundant in early wounds and are essential for effective decontamination. Yet a large body of evidence suggests that neutrophils negatively influence repair, probably because this cell type is capable of destroying normal tissue (Ref. Reference Dovi, He and DiPietro33). Neutrophil proteases, such as elastase and cathepsin G, can degrade most components of the extracellular matrix as well as proteins as diverse as clotting factors, complement, immunoglobulins and cytokines (Refs Reference Briggaman34, Reference Dovi, Szpaderska and DiPietro35, Reference Li36). Because the extracellular matrix serves as a supporting scaffold for infiltrating cells, modification of the extracellular matrix by neutrophils could have important consequences for repair. Neutrophils also produce a large number of free oxygen radicals, and thus are capable of inducing considerable oxidative stress on the wound. Mediators such as superoxide and hydrogen peroxide can cause additional tissue damage, delaying the repair process and modifying healing outcomes (Ref. Reference Wilgus37). The large load of neutrophils is removed primarily by apoptosis. The removal of apoptotic cells by phagocytosis, a process that is termed efferocytosis, prevents secondary necrosis of these cells and is thought to be essential for complete repair (Ref. Reference Khanna38).
Macrophages are unique in wounds, because they represent the single most effective means of neutrophil removal. Macrophages assist in the removal of neutrophils from sites of injuries in several ways. They respond to neutrophils and their products, and can induce apoptosis in neutrophils (Ref. Reference Meszaros, Reichner and Albina39). Perhaps more importantly, macrophages recognise and actively ingest apoptotic neutrophils, thus helping to resolve wound inflammation (Refs Reference Khanna38, Reference Meszaros, Reichner and Albina40, Reference Daley41, Reference Savill42). Several studies suggest that the phagocytosis of neutrophils influences macrophage phenotype, causing a switch from proinflammatory to a growth-promoting, reparative phenotype (Ref. Reference Fadok43).
Recent studies suggest that failure to remove inflammatory cells, such as neutrophils, has a role in the pathogenesis of nonhealing wounds (Refs Reference Khanna38, Reference Acosta44). A deficit in the capability of macrophages to effectively remove neutrophils has recently been reported to be a critical component of the impaired healing seen in diabetes (Ref. Reference Khanna38). Macrophages derived from sponges implanted in diabetic mice showed a significant impairment in the phagocytosis of apoptotic cells. This deficit was associated with higher levels of apoptotic cells and proinflammatory mediators in wounds, a feature that was further validated in wound tissues of diabetic patients. A deficit in macrophage phagocytic capability has also been associated with the delayed repair that occurs with ageing (Ref. Reference Swift45). Successful efferocytosis by macrophages therefore may be requisite for appropriate wound healing, both for removal of apoptotic neutrophils and for the generation of a macrophage phenotype that supports the proliferative aspects of repair.
Promotion of angiogenesis, fibroblast proliferation and ECM synthesis
Macrophages influence wound healing through the generation of growth factors that promote cell proliferation and protein synthesis (Ref. Reference Rappolee46), as well as by the production of proteases and their inhibitors that influence ECM content and remodelling. Several factors that are known to be present in healing wounds have been shown to be produced by macrophages (Refs Reference Barrientos23, Reference Diegelmann and Evans24, Reference Burke and Lewis47). In general, this information has been considered presumptive evidence for a macrophage-based influence on the healing process. However, direct evidence for the role of the macrophage as the critical source of these factors is difficult to obtain. Within the healing wound, macrophages are rarely the sole source of any of these described factors, and many other cell types within the wound, including other immune cells, keratinocytes, fibroblasts, endothelial cells and adipocytes, also produce the same factors.
One excellent example of this conundrum is the capability of wound macrophages to influence angiogenesis by the production of VEGF. VEGF is a potent proangiogenic factor that has been shown to contribute 50% or more of the proangiogenic activity in wounds (Ref. Reference Nissen48). Macrophages certainly have proangiogenic capabilities, and are well documented to produce abundant amounts of VEGF (Refs Reference Hunt11, Reference Nissen48, Reference Knighton49). However, in epithelial wounds, keratinocytes also produce plentiful amounts of VEGF, making it difficult to determine the relative contribution of macrophages versus keratinocytes in dictating the angiogenic phenotype (Ref. Reference Brown50). A definitive answer to the question of the importance of macrophage VEGF during wound healing would require the selective temporal depletion of VEGF from the wound macrophage. Such experiments are increasing in feasibility because of the development of genetic mutants with selective deficiencies. A recent study using mice with a deletion of VEGF solely from cells of myeloid origin demonstrated that this deficiency yields delayed excisional wound healing, with little impact on incisional healing (Ref. Reference Stockmann51).
Another caveat to the interpretation of the role of macrophage-derived factors is that many are known to have both direct and indirect effects on repair outcomes. For example, PDGF from wound macrophages might assist in the recruitment of progenitor cells and additional inflammatory cells (Ref. Reference Reuterdahl52). More recently, PDGF has also been shown to cause fibroblasts to produce osteopontin, a factor that critically influences wound healing through an autocrine effect that promotes scar formation (Ref. Reference Mori, Shaw and Martin53).
A full understanding of the complex role of macrophage-derived factors is likely to benefit from emerging technologies. Future investigations to untangle the web of macrophage-derived factors might include global descriptions of macrophage mediator production patterns in healing wounds, along with analysis of direct and indirect effects of macrophage products in vivo and in vitro. Such large data sets, once generated, might benefit from advanced biostatistical analysis in order to develop models to explain these complex interactions.
One approach to the study of wound repair that continues to provide relevant information is the use of genetically tractable organisms, such as zebrafish and Drosophila. Zebrafish, owing to their near transparence, provide the additional advantage of allowing real-time live imaging of leukocyte infiltration into sites of inflammation and injury. Previous studies have examined macrophage infiltration into zebrafish wounds and have documented some of the cytoskeletal requirements for this migration (Ref. Reference Redd54). A more recent study has demonstrated the simultaneous tracking of neutrophils and macrophages labelled with differential fluorescent labels into a tail fin injury site (Ref. Reference Gray55). In this study, the two cell types showed marked differences in migration speed and kinetics of recruitment to the injury site. Notably, the study revealed a preferred pathway for macrophages along the abluminal surface of endothelial cells. The powerful approaches available in the zebrafish, including tracking of several inflammatory cell types, specific ablation of cell types, and the ability to perform mutagenesis and transgenesis, suggest that many difficult questions about macrophage function in wounds can be approached in this model system.
Clinical implications
Improving macrophage function to improve healing outcomes
An estimated 6 million people in the US have problems related to inadequate wound healing, and nonhealing ulcers remain a serious problem that greatly affects human health (Ref. Reference Sen56). The improvement of wound healing therefore continues to be the target of many therapeutic strategies. Many attempts to augment the healing process have used single growth factors or cytokines, mostly with limited success. When using single factors, difficulties with optimum delivery systems, timing and concentration are daunting tasks. In addition, the chronic wound environment can be highly proteolytic, limiting the half-life of topically applied molecular factors.
One alternative to the use of molecular therapeutics for wounds is in situ activation, recruitment or addition of exogenous macrophages. Because macrophages are a source of growth factors, augmented macrophage activity may stimulate cellular proliferation and angiogenesis in nonhealing wounds. Increasing the number of macrophages in the wound might also influence the protease imbalance that occurs in some nonhealing wounds, because macrophages can produce protease inhibitors. Finally, the addition of more macrophages might provide an increased efferocytosis capacity.
In support of the therapeutic potential for increased macrophage activity in the healing wound, several early studies document that the topical treatment of wounds with the macrophage-activating agent glucan (Refs Reference Browder57, Reference Leibovich and Danon58) improves healing outcomes. Glucans are polymers of β-1,3-linked glucose that interact with polysaccharide receptors on the macrophage, causing cell activation. Glucan treatment of wounds has been shown to increase the number of macrophages, and promote fibroplasia, re-epithelialisation and wound strength. Similarly, the application of chemoattractants such as MCP-1 that recruit monocytes to wounds has been shown to promote healing (Ref. Reference DiPietro59).
Conceptually, the supplementation of wounds with exogenous macrophages could promote repair, particularly when macrophage function is deficient. Compelling evidence in favour of such a strategy comes from Danon and colleagues (Ref. Reference Danon, Kowatch and Roth60), who demonstrated that the injection of macrophages into the wounds of aged mice could correct the age-related deficit in wound healing. Studies in human subjects have shown that monocyte-derived macrophages, obtained from peripheral blood, can improve the healing of pressure ulcers, as well as sternal chest wounds, in patients undergoing cardiac surgery (Refs Reference Danon61, Reference Orenstein62).
Taken together, these findings suggest that therapeutic strategies that increase macrophage accumulation could accelerate the wound-healing process, and might be particularly helpful in situations of impaired healing, such as ageing and diabetes. In particular, agents that could recruit and drive macrophages towards a reparative phenotype would promote tissue regeneration in the absence of destructive inflammation. The clinical implications of adding such a tool to the therapeutic arsenal could be great. As mentioned above, nonhealing wounds are an immense problem, and current treatments are frequently ineffective. Moreover, evidence indicates that macrophage dysfunction has a role in the impaired healing seen in diabetic patients and in the elderly, who are at great risk for the development of nonhealing ulcers (Refs Reference Khanna38, Reference Swift45).
Research in progress and outstanding research questions
In the absence of other inflammatory cells, are macrophages essential for healing?
Although the studies described above suggest that macrophages are generally beneficial to repair, some controversy remains over the role of macrophages in the healing wound. Studies of healing in the early- to mid-gestation fetus show that early fetal wounds exhibit very little, if any, inflammatory response while healing in a scarless fashion (Refs Reference Cowin7, Reference Hopkinson-Woolley63). Intrinsic differences in growth factor production, levels of stem cells and cellular proliferation capacity probably support fetal wound repair. However, the observed scarless repair in the absence of inflammation suggests that macrophages are not an essential feature of adequate repair, particularly if other aspects of the inflammatory response are suppressed. Several recent studies on mice genetically deficient in specific immune cells and molecules also support the concept that inflammatory cells are not needed for efficient tissue repair, as long as microbial contamination is controlled. In neonatal Sfp1 –/– mice (previously known as PU.1 –/–), which lack both macrophages and functioning neutrophils, little inflammation occurs at the wound site, and repair appears to be scar free, similarly to that in the fetus (Ref. Reference Martin64). In addition, mice with a deletion of the gene encoding Smad3, a molecule critical to the intracellular signalling of TGF-β1, exhibited accelerated cutaneous wound healing and significantly reduced influx of inflammatory cells (Ref. Reference Ashcroft65). Likewise, studies in a number of other systems have shown that soluble factors that reduce inflammation in wounds, such as the cytokine IP-10 and heparin-binding epidermal growth factor, are often beneficial to wound-healing outcomes in the adult organism (Refs Reference Luster66, Reference Xia67). Together, these findings suggest that in the face of normal inflammation, which includes oedema, mast cell degranulation and neutrophil ingress, macrophages have an important balancing role. By contrast, when the complete inflammatory response is severely suppressed and bacterial contamination is controlled, wounds appear to heal well. Thus, the role of macrophages must always be considered in the context of the specific wound environment in question. Reductionist approaches have provided a wealth of information about the function of wound macrophages; however, studies that focus on the integration of macrophage function with those of the many other cell types in the wound are required.
Macrophage phenotypes in the healing wound
As mentioned above, abundant in vitro studies suggest that macrophages can adopt discrete phenotypes in response to environmental signals (Ref. Reference Stout68). By contrast, there are few studies of the actual macrophage phenotypes that exist in the in vivo wound and the phenotypes described so far appear highly complex (Ref. Reference Daley29). Moreover, several markers of macrophage phenotypes have been proposed, and there is a lack of agreement over which are the most critical or informative. Finally, investigations of wound macrophage phenotypes are limited by technical difficulties in obtaining truly representative sample populations of cells from specific locations in wounds.
Relevant to the question of wound macrophage phenotypes is the question of the origin of wound macrophages. Most wound macrophages are thought to be derived from circulating monocytes. These cells in mice exist in two main subsets: an ‘inflammatory’ subset expresses Ly6C at high levels and a ‘noninflammatory’ subset expresses Ly6C at low levels (Refs Reference Geissmann, Jung and Littman69, Reference Sunderkotter70). Recently, similar populations of proinflammatory CD14lo and noninflammatory CD14hi monocytes have been described in humans (Ref. Reference Cros71). Our preliminary research on macrophages isolated from excisional wounds in normally healing mice shows that wound macrophages exhibit high-level Ly6C expression early following wounding, which is associated with a proinflammatory phenotype. As healing progresses, wound macrophages exhibit a transition to low-level Ly6C expression, which is associated with a prohealing phenotype. Whether the observed phenotype transition is derived from an in situ response to differential environmental cues or from selective recruitment of monocyte populations that are already predisposed to proinflammatory or prohealing phenotypes remains to be determined.
Another question is the role of macrophages in mediating the resolution phase of healing. During this final phase, capillary regression and collagen remodelling are dominant features. Macrophages can produce factors that are antiangiogenic (such as thrombospondin-1 and IP-10) (Refs Reference DiPietro72, Reference Bodnar73) and others, including CXCR3 ligands, that direct the termination of the repair response in several ways (Refs Reference Engelhardt20, Reference Yates74). Other cells within the wound may also produce these concluding signals, but because macrophages can theoretically be a major source of these types of factors, these cells might have an active role in the termination of the wound-healing process. Little is known about how macrophages might shut down the healing response, and much remains to be learned about this phase.
A composite of macrophage functions throughout the time course of the healing wound is suggested from the current literature (Fig. 2). Whether wound macrophages fall into specific discrete phenotypes remains to be determined, yet the significance of gaining a true understanding of macrophage phenotypes within wounds could be huge. Macrophages are probably important contributors to pathophysiology in nonhealing wounds, delayed healing, and fibrosis and scar formation, thus the macrophage remains an attractive target for therapeutic strategies.
Likely pattern of macrophage function during the course of wound healing. In the early wound, monocytes and resident macrophages become activated, undertake phagocytosis of microbes and perhaps early neutrophils, and produce proinflammatory mediators and chemoattractants. Macrophages also assist in the induction of apoptosis in neutrophils, thus steering the wound towards a noninflammatory, reparative state. In the later phases of wound repair, macrophages ingest apoptotic neutrophils, producing growth factors to support tissue restoration. In the very late stages, as the wound resolves, macrophages may guide tissue remodelling by producing factors to promote capillary regression and collagen remodelling.
Acknowledgements and funding
Wound-healing research in L.A.D.'s laboratory is supported by National Institute of Health Grants RO1-GM50875 and P20-GM078426. Tissue repair research in T.J.K.'s laboratory is supported by the United States Army Medical Research and Materiel Command #W81XWH-05–1-0159. The authors thank the reviewers for their critical comments and suggestions.
