Mechanisms of immune resolution

Abstract

Initially after injury, the innate/proinflammatory and some aspects of the acquired immune response are up-regulated to maintain a defense against foreign pathogens, clear tissue debris present at the wound site, and orchestrate aspects of tissue remodeling, cell proliferation and angiogenic process, associated with the wound response. However, for proper wound healing to progress, this initial inflammatory response has to be regulated or shut down so as to allow for the reestablishment of matrix, recellularization, and tissue remodeling. Inability to properly resolve the extent of innate/acquired response at a site of injury can lead to poor wound healing, immune suppression, and recurrent infectious episodes. This review attempts to summarize information on regulatory mechanisms that are thought to be involved in controlling/resolving innate or acquired immune responses so as to provide a framework for use in thinking about the impact these processes and their manipulation may have on wound healing and its potential management.

Problems associated with poor wound healing, chronic wounds, and wound care management remain a significant healthcare issue that consumes substantial financial resources on an annual basis, both in the United States and in Europe (1). Although it is clear that resolution of underlying predispositions or diseases, such as diabetes, renal disease, or malnutrition, and good clinical management of wound site problems are critical in reducing wound care problems, there remains a significant number of situations in which wound healing still proceeds poorly. This leaves the critically ill patient open to risk of recurrent infection, poor wound closure, or dehiscence/splitting. Thus, expanding our understanding of wound healing and those immune regulatory processes/cells that may effect the critically ill patient remains significant.

Historically, much effort has been directed at discerning the contributions made by components of the innate (2, 3) and, to a lesser degree, the acquired immune response (4, 5) to the wound healing process in the critically ill patient. In this respect, the most overt contributions made by cellular components of the immune system are during the initial inflammatory stage of wound healing and, to a lesser extent (although no less critical), the proliferative/reepithelialization/wound repair phase and to the subsequent tissue remodeling/granulation phase (1, 2). The initial inflammatory phase, not surprisingly, is dominated by the presence of large numbers of neutrophils at the site of injury, which serve to handle foreign materials and pathogens, utilizing their phagocytic, enzymatic, and respiratory functions (2). Inability to perform these functions can lead to opportunistic infections, whereas unrestrained activity can lead to an imbalance of proteolytic enzymes and their inhibitors associated with chronic wounds. Interestingly, although the neutrophil in the wound is thought to be important to these previously mentioned aspects, it is not itself critical to the process of wound healing because healing readily takes place in the absence of neutrophils in sterile wounds (6). Subsequently, as normal wound healing progresses into the intermediate-late inflammatory phase, neutrophil numbers begin to wane after approximately the first 48 hrs and are supplanted by increasing numbers of macrophages. These macrophages have received the greatest investigative attention, and much is now understood about their capacity to coordinate the transition from the inflammatory phase to the proliferative stage (2, 7). Macrophages serve in this capacity by not only utilizing a myriad of proinflammatory and anti-inflammatory mediators but also through their ability to communicate directly with cells of the innate system, such as the neutrophil, and the adaptive arm of the immune system, through lymphocytes. The ability of macrophages to produce a number of angiogenic, matrix, and proteolytic proteins or their inhibitors also plays an important role in the tissue remodeling phase, although these aspects are dominated primarily by fibroblasts/keratinocytes or epidermal cells. Although largely under-appreciated, T cells, as representatives of the adaptive arm of the immune system, also seem to have a regulatory impact on this wound healing process (4, 5, 8, 9). Irrespective of the immune cell considered, the role played by each cell and their coordinated activation and shut down are critical to competent wound healing in the same sense as cells and mediators interact to morphologically remodel tissue during embryogenesis or other regenerative processes. That said, most models of wound healing address this process experimentally in a setting in which few predispositional concerns are typically considered, such as the effect of ongoing infection (e.g., sepsis or previous exposure to shock hypoxic events or other insults typical of the critical ill trauma patient). Therefore, appreciation of how cells of the innate or adaptive response are controlled/resolved/suppressed are equally critical to our knowledge of the activation if we are to understand the proper vs. pathologic roles of these cells in the wound site of such patients/animals. With this in mind, it is the objective of this review to summarize salient aspects of the anti-inflammatory response or the processes by which inflammation is resolved.

DOWNREGULATORY/ANTI-INFLAMMATORY MECHANISM OF THE IMMUNE SYSTEM

Initially, we need to summarize what seems to be the primary mechanisms that are available to the host immune system and may be utilized in resolving/down-regulating an ongoing immune response. Although not an exclusive list, five general processes come to mind that directly or indirectly not only result in the arrest of activated immune cells, but frequently the death of these cells via the process of apoptosis (regulated cellular suicide as opposed to necrotic cell lysis/death).

Dissipation of Activating/Inflammatory Stimuli

First, although essentially a passive mechanism in some senses, is the loss or dissipation of activating/inflammatory stimuli that drive innate or adaptive immune cell responsiveness (Fig. 1). In the broadest sense, this can be thought of as the removal of the foreign microbes/debris in the wound or byproducts of necrotic cell death. Many of these agents, such as microbial wall components, lipopolysaccharide, muramyl dipeptide, lipoteichoic acid, activate immune cells directly via pattern recognition receptors (Toll-like receptors) (10, 11). Similarly, endogenous cellular proteins, such as heat shock proteins, various oligonucleotide forms produced by necrotic cell damage, or phospholipids/glycolipids, released in the wound by damaged tissue can interact with these Toll-like receptors (12–14). Interestingly, although not thought of as agents involved in pathogen pattern recognition, a number of the components of the complement cascade and coagulation pathway that are activated by the process of injury or infection can also serve as such immune cell activating signals (15, 16). Whether you think of these stimuli as aspects of “infectious nonself” (17) or “danger” molecules (18), their role is to activate the innate and subsequently the adaptive immune cell response. Thus, the loss of such signals should contribute to mitigation of cell activation.

Figure 1.

Postulated steps at which the process of neutrophil (A), macrophage (B) or T-cell (C) activation/differentiation might be inhibited.

Active Suppression via Anti-inflammatory Mediator Release

Second, immune cells release/express a variety of soluble mediators that in an exocrine/paracrine/autocrine fashion may actively drive the suppression of an innate or adaptive immune response. Many of these mediators are described in Table 1; however, some of the best known of these are the glucocorticoids, catecholamines, prostaglandins of the E-series, nitric oxide (NO), and the cytokines interleukin (IL)-1ra, IL-4, IL-13, IL-10, transforming growth factor (TGF)-β, and the soluble receptors for IL-1RII, TNF-RI/II, or the Duffy receptor for IL-8. Agents such as IL-1ra, soluble receptors or non-signaling receptors act primarily as competitive antagonists of proinflammatory ligand binding or signaling. Alternatively, agents like IL-4, IL-13, IL-10, and TGF-β are thought to mediate the induction of a suppressed state in specific lineages of phagocytes and lymphocytes via intracellular alterations of the activated or resting cell. The intracellular mode of action for many of these agents seems to involve the induction of signaling through janus kinases (JAK1), signal transduction and activator of transcriptions (STATs), protein kinases, phosphatases, the SMAD proteins (intracellular protein mediators of TGF-β receptor binding), certain members of the suppressor of cytokine signaling protein (SOCS) family, and alterations in mitogen activated protein kinase (MAPK) family signaling (19, 20) (Fig. 2). Interestingly, a number of these same soluble mediators have been reported to regulate apoptosis in these same cells (21–24).

Table I.

Cytokines/mediators of general interest that are thought to contribute to the resolution or the suppression of the host immune response

	Cellular (Local) Effects	Systemic Effects
IL-1ra	Competitive inhibitor of IL-1 α and β binding to IL-1 cellular receptor	Antagonizes IL-1-mediated effects
IL-4	••T-cell differentiation/stimulates Th2 (CD4 T-cell subtype)/Tc2 (CD8 T-cell subtype) phenotype commitment while suppressing Th1/Tc1 phenotype, •••cell-mediated immunity; ••B-cell activation/differentiation/proliferation; •••macrophage and natural killer cell activation, cytotoxicity; •••macrophage proinflammatory cytokine response, ••IL-1ra secretion; ••macrophage antigen presentation; •••macrophage-inducible nitric oxide production	Mediates IgE-directed immune response; some selected antitumor effects: antagonizes selected aspects of cell-mediated immune response/delay-type hypersensitivity.
IL-10	••T-cell differentiation/stimulates Th2 (CD4 T-cell subtype)/Tc2 (CD8 T-cell sub- type) phenotype commitment while suppressing Th1/Tc1 phenotype, •••cell- mediated immunity; ••B-cell differentiation; •••macrophage and natural killer cell activation, cytotoxicity; •••macrophage proinflammatory cytokine response, ••IL-1ra secretion; ••macrophage antigen presentation; •••macrophage- inducible nitric oxide production	Potentiates IL-4-mediated IgE immune response; antagonizes cell-mediated immune response/delay-type hypersensitivity.
IL-13	•••macrophage and natural killer cell activation, cytotoxicity; •••macrophage proinflammatory cytokine response, ••IL-1ra secretion; ••macrophage antigen presentation; •••macrophage inducible nitric oxide production; no effect on T-/B-cell in humans	Mediates IgE-directed immune response; antagonizes cell-mediated immune response
TGF-β	•••T-cell proliferation and ••T-cell differentiation toward Th3 (CD4 T-cell sub- type) phenotype commitment; ••B-cell differentiation, •••B-cell proliferation; •••macrophage and natural killer cell activation, cytotoxicity; •••macrophage/••monocyte proinflammatory response, ••IL-1ra macrophage secretion, ••macrophage prostanoid release; •••macrophage antigen presentation; •••macrophage-inducible nitric oxide production, chemoattractant for fibroblasts, macrophage, PMNs, T-lymphocytes; ••localized PMN activation	••Wound healing and angiogenesis; •••in vivo inflammatory response of IL-2, TNF, and IL-1; immunosuppressant; •••lymphopoiesis and hematopoiesis; ••bone and cartilage resorption; ••vasoconstriction
Soluble cytokine/decoy receptor	Bind cytokine and typically inhibit interaction with functional cell-surface receptor (e.g., TNF-RI(p55), TNFRII(p75), IL-2R, IL-1R, IL-6) or presence of nonfunctional decoy receptor (e.g., IL-8) competes with cytokine	Directly/indirectly antagonizes particular proinflammatory cytokine-mediated effects
NO	NO from iNOS •••T-cell proliferation and T-cell apoptosis; NO from iNOS ••macrophage/••monocyte proinflammatory response, ••macrophage apoptosis, •••IL-10 release; NO from iNOS ••macrophage cell activation, cytotoxicity; NO from iNOS ••local production of reactive oxygen and peroxynitrate; ••localized PMN activation, •••NO from eNOS reduces PMN interaction with vascular endothelia; NO from eNOS relaxes vascular tone	•••Wound healing (via iNOS); •••vasoconstriction (via eNOS); ••in vivo inflammatory/antimicrobial response (via iNOS); lymphoid immunosuppressant (via iNOS)

IL, interleukin; Ig, immunoglobulin; TGF, transforming growth factor; PMN, polymorphonuclear leakocytes; TNF, tumor necrosis factor; NO, nitric oxide; iNOS, inducible nitric oxide synthase; eNOS, endothelial-constitutive NOS.

Figure 2.

Diagrammatic representation of the various anti-inflammatory intracellular signaling pathways (→) and their inducers, which, for the most part, inhibit immune cell activation. Pathways that seem to be involved in transmitting signals that both stimulate and inhibit cell activation are indicated by .

Studies from our laboratory (25) and a number of others (26–30) have shown that in late sepsis or after traumatic injury or shock, there is a decrease in splenic lymphocyte IL-2 and interferon (IFN)-γ release capacity that is associated with an increased ability to produce the anti-inflammatory cytokines IL-4 and IL-10.

IL-10

Regarding IL-10’s role in the immune dysfunction seen with polymicrobial sepsis or shock, there is some controversy. Work by Hogaboam et al. (31) indicates that mice pretreated with a polyclonal antibody to IL-10 exhibit a decreased survival to subsequent cecal ligation and puncture (CLP). Along these lines, a study by Minter et al. (32) indicates that mice who have been transiently transfected in vivo with an adenovirus containing the human IL-10 gene show a marked attenuation of the proinflammatory response associated with septic challenge in the form of CLP. Alternatively, a study by Ertel et al. (33) indicates that IL-10 administered after the induction of sepsis in mice not only had the capacity to suppress the early proinflammatory cytokine response, but also suppressed their Th1 lymphokine response and decreased the animals’ overall survival. Song et al. (34) have assessed the in vitro and in vivo effects of either anti-IL-10 antibody treatment or IL-10 gene deficiency on lymphocyte Th1 cytokine release and survival after a lethal septic challenge. Their results indicate that IL-10 gene deficiency or anti-IL-10 antibody treatment prevents the sepsis-induced depression of Th1 lymphokine production. Further, they found in vivo treatment of animals with antibody to IL-10 had a salutary effect on animal survival only when administered after the proinflammatory/hyperdynamic/hypermetabolic phase, but not when given during proinflammation. A finding supported by Terrazas et al. (35) using IL-10 inhibitor ASL01.

Song et al. (36) has also shown that in vitro inclusion of antibodies to IL-10 with CLP mouse splenic lymphocytes markedly attenuated the rise in phosphorylated p38 MAPK while restoring their ability to produce IL-2 and IFN-γ (Fig. 2). The activation of p38 MAPK in both splenic T cells and splenic/peritoneal MØs from CLP mice is protracted as 4 hrs was needed to see a peak signal (37, 38). Such sustained p38 signaling has been associated with anti-inflammatory as opposed to the typical proinflammatory cytokine/mediator release (39, 40) and is evident in injury patients’ blood polymorphonuclear leukocytes (PMN) (41, 42). Conversely, we (37, 38, 43) and others (44–46) have observed that the suppression of p38 MAPK activity (by in vivo/in vitro SB203580 treatment) blocked CLP enhancement of lymphoid/MØ IL-10 production. Extracellular signal-regulated kinase inhibition with PD98059 had no such effect (43). These results suggest that IL-10 seems to mediate its anergic effects via an autocrine/paracrine loop by altering p38 MAPK signal transduction.

IL-4

The Th2 cytokine IL-4 is known to regulate Th1- and Th2-cell responsiveness primarily through the activation of the signal transducer and activation of transcription factor-6 (STAT6) pathway. Studies in our own laboratory indicate splenocytes isolated from mice after the onset of sepsis (25) or hemorrhagic shock (47) exhibited an enhanced capacity to produce IL-4. However, unlike IL-10, little evidence of this cytokine’s presence in the shocked/septic mouse’s circulation has been collected (48). This suggests that the role of IL-4 may be more localized (autocrine/paracrine) in nature (Fig. 2). In this regard, Song et al. (49) observed that when mice were posttreated with neutralizing monoclonal antibody against mouse IL-4, septic Balb/c mouse splenic lymphoid Th1 cytokine responsiveness could be restored while preventing the enhancement of Th2 cell cytokine release and associated STAT6 phosphorylation. Furthermore, neutralization of IL-4 markedly increased the survival rates in septic animals when given 12 hrs post-CLP. Taken together, these data indicate that the Th2 cytokine IL-4 can contribute to the suppression of cell-mediated immunity and death associated with polymicrobial sepsis in animals like the Balb/c mouse, which shows a predisposition for enhanced endogenous IL-4 production.

The data on other JAK/STAT family members and their contribution to immune suppression in experimental animal models of shock or sepsis has been limited. Unlike IL-4 and IL-13, which act primarily through the JAK2-STAT6 pathway, Th1 cytokines like IL-12 predominantly use JAK1/3-STAT4, and IL-2 signals through JAK1/2-STAT5 and IFN-γ via JAK1-STAT1 (19). IL-10 (a Th2 cytokine), alternatively, seems to activate JAK1-STAT3 (19). In this respect, Matsukawa et al. (50) and Godshall et al. (51), using a model of CLP that produces over 60% mortality in 48 hrs, found that deficiency in STAT6 or STAT4 gene expression improved survival. This was associated with an attenuation of acute proinflammation. Conversely, Song et al. (52), in our laboratory, using a sublethal CLP model (20% mortality at 10 days), found that although gene deficiency of STAT6 provides restoration of lymphoid Th1 responsiveness in septic mice, it does not protect these animals from septic mortality (60% mortality) (52). Deficiency of STAT4, however, not only produced a marked loss of Th1/proinflammatory immune responsiveness, but also produced an even greater reduction in survival after septic challenge than that seen with STAT6 −/− animals (95% mortality) (52). One might speculate that the difference may be due to the use of acute vs. chronic mortality models complicated by the use of the highly susceptible Balb/c mouse (53).

Activation of JAK/STAT (e.g., STAT1, STAT3, STAT5a/b, STAT6) family members also induce the synthesis of a novel group of proteins known as SOCS, which inhibit the IL-2, IL-6, and IFN signaling pathways (19, 54) (Fig. 2). Ogle et al. (55) recently showed that the SOCS3 and SOCS1 (counter-regulatory signaling) proteins seem to be markedly up-regulated in the liver and spleen after thermal injury. Although no causal relationship has yet been documented, it is tempting to speculate that induction of the SOCS proteins may be contributing to aberrant down-regulation of the innate or cell-mediated immune response. In this respect, our preliminary data examining SOCS1 and SOCS3 expression in the septic mouse indicate that change in their expression seems restricted to increases in SOCS3, but not SOCS1, in peritoneal MØs, but not in the spleen or thymus (56).

TGF-β

Studies from our laboratory (57) and others (48, 58) have also demonstrated that immunosuppression after shock or sepsis is associated with the systemic release of the anti-inflammatory cytokine TGF-β (59, 60). The role of TGF-β and IL-6 (a direct/indirect inducer of TGF-β release) as potential anti-inflammatory agents in sepsis and in trauma has been documented by the work of Miller-Graziano et al. (61) and Zhou et al. (62). In light of this, one can envision a network of systemic effects mediated by agents that induce the release of IL-6, leading to increased TGF-β levels, which in turn directly or indirectly (via agents like IL-10 (63)) induce the suppression of host responses seen during sepsis. However, little direct evidence is available concerning TGF-β–induced changes either on TGF-β target cell signaling via the SMAD family (64) or p38 MAPK (65) activation and on survival after CLP (Fig. 2).

Nitric Oxide

Another potentially important immunosuppressive agent that may contribute to the suppression/resolution of the lymphocytic-mediated immune response after sepsis is NO. Hogaboam et al. (31) have shown that NO may play a role in improving survival of female mice subjected to sepsis. However, when a nonspecific NO antagonist (L-NAME) was used, they found that such posttreatment also increased the peritoneal exudate fluid IL-10 levels but not the capacity of the peritoneal MØ to make IL-10. Chung et al. (66) studied the onset of immune suppression in intra-epithelial lymphocytes and observed that the NO release by inducible NO synthase (iNOS) does seem to play a role in the onset of immune hyporesponsiveness in the gut mucosal lymphoid system. We have also recently reported that NO derived from iNOS, by an autocrine/paracrine activation loop (Fig. 2), may also be a partial contributor to splenic lymphoid dysfunction seen in sepsis as iNOS deficiency restored IL-2, but not IFN-γ, production (67). Alternatively, Cobb et al. (68) reported that mice deficient in iNOS succumb more readily to CLP. Interestingly, like many other proinflammatory genes, iNOS expression is dependent on the activation of nuclear factor-κB (NF-κB) by upstream signals from IL-1β/TNF/Toll-like receptors–IL-1 receptor activated kinase family members (69–72) and by activation of p38 MAPK (73, 74). In this regard, much as we have found that the in vivo inhibition of NF-κB translocation by pretreatment of mice with pyrrolidine dithiocarbamate with iNOS −/− mice, it actually increased the mortality seen with CLP (75). However, because NO, produced by iNOS and NF-κB activation, contributes to aspects of innate/cellular immune functions, which are vital to fight infection, and to deleterious proinflammatory events associated with shock, this might explain why the survival of these animals declined.

In addition to the anti-inflammatory mediators discussed above, there are a number of other agents like the prostaglandins (PGE2; 318,1245,93,841,4081, 2915,3841,5038), sex hormones (66, 76–78) (prolactin, androgens, estrogens (79, 80)), glucocorticoids (81, 82), neuropeptides (i.e., substance P, adrenomedulin) (83), and catecholamines that contribute to the immune dysfunction seen in sepsis and after shock or injury. These have been reviewed extensively elsewhere, and the reader is directed to those articles.

Potential Regulatory Cell Populations Contributing to Resolution

A third mechanism, which in many ways derives from or utilizes aspects of immune suppressive mediator release, is the activation/induction of immune suppressive cell populations (Fig. 1C and 3). In this respect, the suggestion that injury could induce the activation of what was described as a “suppressor T-cell system” was proposed initially by Munster (84). Work by a number of laboratories looking at mixed T-cell responses after thermal injury reported the existence of both the CD8⁺ (Tc, cytotoxic T-lymphocyte) T-cell population, which seemed to play the role of suppressor inducer cells (85–89), and the CD4⁺ (Th, T-helper lymphocyte) T-cell population, which seems to serve as suppressor effector cells. Similar observations have also been made looking at alternative models of tissue injury (90–92) and hemorrhage (93). However, despite these observations, few investigators have been able to isolate pure populations, which were either a functionally or phenotypically unique lineage of cells and were consistently immune suppressive.

Figure 3.

Postulated interrelationship of macrophages (M), dendritic cells, neutrophils (PMN), T lymphocytes (T_h0, T_h1, T_h2, T_h3, NK-T cell or T_c1, T_c2 cells), their general activational stimuli, and the cytokines they express to the regulation/development of a competent cell-mediated immune response.

With the advances in cytokine biology and the derivation of a number of unique lymphoid cell lines in the 1980s and early 1990s, it became clear that within the CD4⁺ T-helper cell population there existed a number of unique lymphoid subpopulations (94) (Fig. 1C and 3). These subpopulations are defined primarily by their propensity to produce either cytokines like IL-2 and IFN-γ that support the development of a cell mediated immune response, or cytokines such as IL-4, IL-5, IL-6, IL-10, and IL-13 that, while driving humoral immunity, generally suppress cell-mediated immunity (94–96). Several investigators looking at divergent models of injury have since reported evidence of the development of a “shift” in the injured/infected animals’/patients’ T-helper cell response to one that seems to be dominated by the immune suppressive (Th2) lymphoid phenotype in conditions of burn (97–102), hemorrhage (47, 103), and sepsis (25, 43, 104). Interestingly, Zedler et al. (105) indicated that such “shifting” between immune enhancing and immune suppressive (Th1 to Th2) T-cell phenotypes also may be evident in the CD8⁺ T-lymphocyte population of critically injured patients. However, although the use of combined intracellular cytokine staining with various known T-cell markers, such as CD3, CD4, and CD8, has given an initial insight into the possible existence of such immune suppressive cells, it is now becoming clearer that a fair degree of plasticity exists in these phenotypes and that there may be a number of less common lymphoid subpopulations that can also be immune suppressive/regulatory.

Two additional populations that are more common to the mucosal lymphoid tissue and found less in the systemic sites (blood, spleen, peripheral lymph nodes) are the γδ-T-cell and the NK-T-cell populations (106, 107). These cells respond to endotoxin (ETX), unlike CD4/CD8-αβ-TCR⁺-T cells (40, 108). Antigenic stimulation is also poorly understood as these cells respond to nonclassic MHC-class-I-like antigens that are, interestingly, expressed during bacterial challenges and tissue injury (109). γδ-T-cells can also exhibit Th1-Th2 cell shifting in cytokine expression (110, 111). The initial studies looking at γδ-T cells’ contribution to host immunity in thermally injured (112), pulmonary pneumonia (113), and septic (114) mice suggest these cells are not involved in overt immune suppression, but seem to be critical to maintaining the innate/cellular immune response (Fig. 3).

Alternatively, NK-T cells have been found to play a role in maintaining local immune privilege in sites like the eye and, under certain circumstances, inducing systemic immune suppression via Th2-like cytokine release (115, 116). Although our understanding of the function of these cells is also incomplete, it is known that their activation is mediated via antigen (i.e., glycolipids, which may be derived both from bacteria and damaged cells) presentation by CD1d (117–120) (Fig. 3). Such NK-T-cell activation can be inhibited with a blocking antibody against CD1d (121). Recent findings by Rhee et al. (122) suggest that in some mice, their activity may reduce septic survival; however, their exact mode of contribution has yet to be determined.

The capacity of MØs and, to a lesser extent, dendritic cells to serve as immune suppressive cells has been well documented by a number of investigators (61, 104, 123–126) (Fig. 1B and 3). However, this anti-inflammatory characteristic is thought to be the product of timing or the nature of the stimulus. In this respect, it is well documented that after stimulation, MØs initially synthesize proinflammatory agents, but over time the same cells begin to produce immune suppressive agents such as IL-6 (an agent with both proinflammatory and anti-inflammatory roles), IL-1ra, IL-10, PGE2, and release various soluble cytokine receptors (44). Alternatively, the nature of the stimuli encountered during phagocytosis can affect the extent to which a MØ favors proinflammatory vs. anti-inflammatory mediator release. The phagocytosis of necrotic cell debris and apoptotic cell material seems to drive proinflammatory and anti-inflammatory cytokine production, respectively (127, 128). One could speculate that the presence of a significant component of immune cell apoptosis in the septic (129) or traumatized animal as we and others have reported on MØs from septic or traumatized subjects might serve as a stimulant for the increased release of anti-inflammatory mediators. Conversely, necrotic cell debris from damaged tissues has been shown in vitro to be a proinflammatory stimuli (13, 130, 131). However, as tissue injury of various origins induces both apoptotic and necrotic cell death, it is unlikely that in vivo effects will be so simplistic (130, 132–137).

Alterations in Cell-Associated Co-Stimulator Molecules

A fourth potential mechanism for suppressing an evolving immune response, at least in lymphocytes, is the antagonism of cell-cell mediated co-stimulatory events (Fig. 1C and 3). Because the T cell requires concomitant cell signals to become fully activated/differentiated during its response to presented foreign antigen through a variety of cell surface co-stimulatory molecules, it is possible to block such development by inhibition of co-stimulatory events (138). This can be mediated by the lack of expression of co-stimulatory signals on concomitant antigen presenting cells, such as the loss of intercellular adhesion molecule-1, B7.1, B7.2, and CD40 (139–141). Alternatively, the activation of inhibitory receptors such as cytotoxic T lymphocyte-associated antigen 4 and CD45 within the receptive T-cell can serve to block the activating signal via their capacity to dephosphorylate key components in the receptor signaling complex (142–144). Typically, such loss of co-stimulatory signaling leads to a state of anergy or tolerance to further antigenic stimulation of the T cell; however, such incomplete signaling (lack of a co-stimulant) can also directly or indirectly drive the onset of apoptosis in the tolerized cell. This is most likely due to the loss of the capacity of these cells to produce growth/differentiation factors (e.g., IL-2, IL-3, granulocyte-macrophage-colony-stimulating factor, etc.) that would normally suppress the induction of apoptosis.

With respect to the cell-associated co-stimulants, although data exist on their potential role in transplant engraftment (145–147), their contribution to the immune dysfunction associate with trauma, shock, or sepsis is essentially unknown. However, as these are effectors primarily of lymphocyte-macrophage interactions, one would anticipate that stages such as wound reepithelialization or remodeling might be most affected by co-stimulant effects/defects.

Apoptotic Processes Utilized to Resolve Immune Response in the Wound

The fifth mode by which an immune response may be suppressed or resolved is by the overt induction of the immune cells’ endogenous apoptotic (cell suicide) process. In a number of instances, apoptotic death of a given immune cell can be induced as a form of fratricide through the actions of effector cells (i.e., T-lymphocytes or macrophages) that interact with the target immune cell (148, 149). From a mechanistic sense, our understanding of the apoptotic process has grown tremendously over the last 10 yrs. Roughly speaking, apoptotic cell death now seems resultant from one of three major pathways: cell death receptor driven, the mitochondrial pathway, or endoplasmic reticular stress-induced cell death (150–154). These pathways and their regulatory mechanism or mechanisms, summarized in Figure 4, are reviewed in detail elsewhere (151, 152, 155). However, it is important to remember, as with all the previously discussed immune regulatory processes, that the apoptotic process is tightly controlled in a cell in tissue selective manner and that the mediators and regulators in one immune cell subpopulation may not only have divergent effects in another but may well be absent.

Figure 4.

Some of the key components, mediators, and pathways (mitochondrial, endoplasmic reticulum [ER], and death receptor pathways) that have been implicated in the induction and suppression of immune cell apoptosis.

Interestingly, the very process of macrophage or lymphocyte activation required to mount a competent inflammatory or adaptive immune response to a foreign pathogen sets in motion, for most cells, processes involved in mediating their own demise (156, 157). In this respect, the process by which a mature T cell is activated, via concordant T-cell receptor complex stimulation and co-stimulatory molecule engagement, simultaneously induces the up-regulation of death receptors (like Fas) that make the activated T cell more susceptible to ligands that induce apoptosis (157–159). With respect to the resolution of inflammation or local tissue injury, lymphocyte Fas/FasL–mediated fratricide is also a potentially important process (Fig. 1). Numerous examples of lymphocyte-lymphocyte/infected lymphocyte induction of cell death have been documented (160, 161). Fratricide has also been suggested as a mechanism for contributing to the maintenance of sites of immune privilege, such as the eye or the fetus during development (161, 162). However, fratricide may also have pathologic potential when directed inappropriately at cells, like epithelial or endothelial cells present at sites of injury/inflammation. Examples of this potential have been suggested for diverse tissues such as the kidney (163), pancreas, and lung (164).

With respect to traumatic shock/sepsis-induced changes in lymphoid A_o, the initial experimental studies focused on the thymus (165). This was based on this tissue’s accessibility in rodents and the knowledge that the thymus is highly susceptible to stress-induced apoptosis. Studies looking at various bacteremic/septic and shock models all consistently report that thymic apoptosis increases (165). Regarding polymicrobial sepsis induced by CLP, it was observed that increased apoptosis in the thymus of septic mice could be detected as early as 4 hrs after CLP, increasing up through 24 hrs. This increase in apoptosis in the thymi of these CLP mice seems to be primarily a response to glucocorticoids, and possibly NO, and not to endotoxin or death receptors like Fas and TNF-R (165) (Fig. 4). In addition to thymic apoptosis, studies by Hotchkiss et al. in mice (166, 167) and septic/multiple organ failure patients (168) have shown evidence of increased splenic lymphocyte apoptosis associated with increased mortality. Interestingly, in those studies, little evidence of nonlymphoid/nonimmune organ apoptosis was seen in either experimental sepsis or patients.

A number of other lymphoid tissues also seem to be actively undergoing increased apoptosis after the onset of sepsis or shock. Mixed bone marrow cells showed an increase in apoptosis at 24 hrs but not at 4 hrs after CLP (165). Although phenotypic and morphologic assessment indicated that most of the increase in apoptosis in the thymus was in the immature T-cell population (CD4⁺CD8⁺ and CD8⁻CD4⁻ cells), the increase in bone marrow cell apoptosis was associated with only the B220⁺ cells (B-lymphocytes). However, unlike thymocytes, treatment of CLP mice in vivo with antagonists of glucocorticoids or TNF failed to suppress the increased apoptosis in the bone marrow.

Gut-associated lymphoid tissues, such as the Peyer’s patches, lamina propria, and intraepithelial lymphocytes, also exhibit increased apoptosis in response to polymicrobial sepsis (169–171). Although the subpopulation of lymphocytes undergoing apoptosis varies among these gut-associated lymphoid tissue sites, many of these subsets of cells exhibited a marked increase of Fas antigen expression. Thus, these apoptotic changes would seem to be examples of activation-induced cell death (AICD) (151) in these gut-associated lymphoid tissue sites (Fig. 1C and 4). We have made similar findings after traumatic shock (165). The functional aspect of this increased in vivo apoptosis seems to be related to the endogenous stimulation of immunoglobulin-A production by B-lymphocytes and increased nuclear c-Rel expression and augment ex vivo cytokine expression by the T cells (75).

These findings correlate well with the in situ observations by Hiramatsu et al. (167) who reported evidence of increased apoptosis at 24 hrs after CLP in both mouse Peyer’s patches and in lymphoid cells lining the small and large intestine. Most recently, Hotchkiss et al. (172) documented that increased intestinal lymphoid apoptosis is a common finding in patients undergoing surgery after major trauma. Most intriguingly, it has been reported that the phenotypically distinct intestinal intraepithelial lymphocyte population also exhibits changes associated with increased apoptosis (165). This seems to be a FasL-Fas antigen–mediated process independent of ETX sensitivity and, again, may be a reflection of localized immune cell activation in response to sepsis.

To the extent that increased lymphoid apoptosis seen in these tissues might be associated with immune hyporesponsiveness (decreased proliferative response and diminished IL-2/IFN-γ release capacity), which we have eluded to earlier with respect to the hypodynamic/anti-inflammatory state of sepsis, findings indicated that mitogenic stimulation of splenocytes isolated from mice 24 hrs after CLP causes a significant increase in the rate of AICD (173). This increase in AICD also seems to be restricted to T cells of the helper (CD4⁺) lineage (173). Studies also suggest that the Fas/FasL system (173) and IL-10 (173) may be involved in this process (Fig. 4). Interestingly, Castro et al. (174), using a high-dose ETX model, showed that ETX mediates not only an increase in AICD but also a change in cytokine productive capacity. These findings are in keeping with the observations of Oka et al. (reviewed in (175)) who reported that inhibition of FasL binding decreased AICD in peripheral blood mononuclear cells obtained from surgical trauma patients, and with Papathanassoglou et al. (175) who documented an increase in peripheral blood lymphocyte Fas/FasL expression in patients with multiple organ failure. Finally, recent findings from our laboratory suggest that splenic lymphoid immune suppression (loss of IL-2/IFN-γ release) seen in CLP mice in response to in vitro stimuli seems to be mediated by an IL-10–induced increase in AICD due to increasing expression of Fas receptor, which is driven in part by Th2-cell’s action on Th1/Th0-cells (173). The pro-apoptotic effect of IL-10 on increased AICD in circulating lymphocytes of surgical trauma patients has also recently been reported (176).

With respect to the degree that the loss of lymphokines such as IL-2 or granulocyte-macrophage-colony-stimulating factor may be involved in the increase of lymphocyte apoptosis seen in polymicrobial sepsis, no clear evidence is yet available that provision of these agents to septic animals alters lymphoid apoptosis. However, such a possibility is suggested by Teodorczyk-Injeyan et al. (177), who showed that the activation-induced increases in apoptosis seen in peripheral blood lymphocytes from burn patients could be partially prevented by supplementation with exogenous IL-2 in vitro.

The strongest evidence that the apoptotic changes seen in these lymphoid tissue can contribute directly to morbidity and mortality seen in sepsis comes from survival study data. Studies by Hotchkiss et al. indicate that transgenic mice that overexpress the human Bcl-2 gene in their lymphoid compartment (166) and mice treated with pan-specific caspase inhibitor zVAD (167) showed not only in a reduction in thymic and splenic apoptosis but also had a survival advantage after CLP. We have documented that animals deficient for the ligand for the Fas receptor (FasL −/−) or competitively inhibited by administration of a Fas-receptor immunoglobulin fusion protein, FasFP, also showed a distinct improvement in overall septic survival (165, 178). Taken together, these data provide evidence that increased in vivo/ex vivo apoptosis is not only present in a variety of lymphoid tissues in the traumatized and septic mouse or patient, but also is associated with the pathobiology of organ failure and death in these individuals.

For proper wound healing to progress, the initial inflammatory response has to be regulated or shut down so as to allow for the reestablishment of matrix, recellularization, and tissue remodeling.

With respect to our knowledge of phagocyte apoptosis, the greatest wealth of data comes from studies focusing on human peripheral blood polymorphonuclear leukocytes (PMNs/neutrophils/granulocytes) (179, 180) and, to a lesser extent, macrophages (181, 182). It has been established that once PMNs are released into circulation, their apoptotic program has already been activated. Thus, the typical half-life of an unstimulated PMN, before morphologic changes and its removal from circulation, is 6–12 hrs (179). Interestingly, if PMNs from healthy volunteers are stimulated in this naive state by inflammatory agents (e.g., LPS, TNF, IL-8, IL-6, IL-1, granulocyte-macrophage-colony-stimulating factor), the onset of apoptosis can be delayed (179, 183, 184) (Fig. 1A). The hypothetical value of this delay in the apoptotic response is that it provides the PMN with a longer lifespan, allowing the PMN more time to migrate to sights of inflammation/tissue injury and microbial contamination (179). This seems to be the case in many critically ill/traumatically injured patient situations as evidenced by several investigations showing (180, 185) that peripheral blood and bronchial alveolar lavage PMNs from acute respiratory distress syndrome, trauma, and septic patients show evidence of decreasing apoptosis. Activation of these cells’ apoptotic process can be mediated directly through death receptors like FasL and TNF (186–188) and through the dissipation of growth factors/chemokine signals that lead to the initial slowing of the apoptotic process (189). Interestingly, PMNs have been shown to reactivate their own cell death program in response to the ingestion of bacteria (190). However, with respect to the reanimation of apoptosis via death receptors in the inflammatory site, this is likely not an endogenous, but an exogenous, effect mediated by macrophages and, to a lesser extent, lymphocytes present in wounds (189, 191, 192).

Regarding MØs, the majority of work has assessed the in vitro response to stimuli such as ETX (LPS), TNF, IL-1, IL-10, IFN-γ, FasL, and NO (181, 182) (Fig. 1B). As with the other immune cells mentioned above, the response to apoptotic stimuli also seems to be time dependent (181). Also, although most of the components of the Fas-FasL and TNF pathways are evident, it is less clear that a comparable series of anti-apoptotic gene products are present. The MØ response to polymicrobial sepsis seems to be tissue specific. Evidence of ex vivo apoptosis can be seen in the peritoneal (serosal) MØ isolated as early as 4 hrs after CLP and increasing over time, whereas a small but significant decrease in the basal apoptosis frequency is evident in liver macrophages. Interestingly, if subsequently challenged in vitro with an inflammatory stimulus such as LPS, both of these macrophage populations become considerably more apoptotic than comparative sham-CLP animal cells. This is associated with a functional disability in their capacity to release proinflammatory cytokines, like IL-1β and IL-6, and decreased caspase 1, but increased caspase 3, 8, and 9 activity (181). Further, these changes in peritoneal MØ apoptosis/cytokine release do not seem to be markedly affected by Fas/FasL activation, but the liver MØ changes are (178). To the extent that such changes in MØ apoptosis/function are just an aberration seen in mice, it is worth noting that Williams et al. (193) have reported similar observations in circulating monocytes derived from the blood of septic patients.

SUMMARY

Although we have discussed five independent mechanisms by which the inflammatory response may be controlled/suppressed, it has to be realized that there is significant overlap in these pathways. Furthermore, the coordinated/synergistic application by the host’s immune system is most likely required to mediate the resolution of an inflammatory response so that wound healing and tissue remodeling can take place. Thus, the host’s immune system is charged with the job of balancing the rapid expansion and maintenance of a proinflammatory/cellular immune response, which is sufficient to ward off not only foreign pathogenic challenge, but also clear the killed pathogen, cell debris, and damaged/injured tissue. Concomitantly, at the wound site, the regulatory immune cells involved in this response to infection must eventually provide an environment of minimal inflammation so that proliferation and tissue remodeling can take place. It is most likely that at least a portion of the changes seen in critically ill patients or animals, in whom shock or injury exist as predispositional components, are due to a loss of this balance required to regulate concomitant response to infection and injury (i.e., the wound response) (194–198). As our understanding of these immune regulatory processes improves, we may have a better understanding of how to manage wound healing problems that arise in the traumatized, critically ill patient.