Trauma equals danger—damage control by the immune system

Review on how traumatic injuries influence immune system phenotypes and functions.

Abstract

Traumatic injuries induce a complex host response that disrupts immune system homeostasis and predisposes patients to opportunistic infections and inflammatory complications. The response to injuries varies considerably by type and severity, as well as by individual variables, such as age, sex, and genetics. These variables make studying the impact of trauma on the immune system challenging. Nevertheless, advances have been made in understanding how injuries influence immune system function as well as the immune cells and pathways involved in regulating the response to injuries. This review provides an overview of current knowledge about how traumatic injuries affect immune system phenotype and function. We discuss the current ideas that traumatic injuries induce a unique type of a response that may be triggered by a combination of endogenous danger signals, including alarmins, DAMPs, self-antigens, and cytokines. Additionally, we review and propose strategies for redirecting injury responses to help restore immune system homeostasis.

OVERVIEW OF TRAUMA EFFECTS ON THE IMMUNE SYSTEM

The immune system responds rapidly to traumatic injuries by reacting to tissue damage. Following this initial response to injury, cells and mediators of the innate and adaptive immune systems undergo temporal change that have been categorized into proinflammatory and counterinflammatory immune responses and are commonly referred to as SIRS, CARS, or MARS [1, 2]. Although these descriptive terminologies for this complex host response were first suggested over 20 years ago, they persist and remain useful as terms for delineating the clinical manifestations of trauma [3].

Research efforts have addressed fundamental reasons for the development of SIRS, CARS, and MARS following trauma. The aim of these efforts was to better understand the cellular and molecular mechanisms responsible for the development of these postinjury syndromes. Studies examining the effects of injury on innate and adaptive immune system phenotypes have shown that traumatic injuries induce inflammatory- and counterinflammatory-type immune responses. We believe that these changes in innate and adaptive immune system phenotype and function ultimately disrupt immune system homeostasis, and it is the loss of homeostasis that predisposes trauma patients to opportunistic infections and complications. We also suggest that injuries induce a specific type of an immune response that may have evolved to protect the injured host from opportunistic infections and excessive reactivity to damaged tissues and cells. Similar to other diseases, there are genetic and individual variables that contribute to the severity of trauma effects on the morbidity and mortality associated with trauma and trauma-induced complications. However, these variables are not well-defined. Future efforts to identify some of these variables will be a formidable task but will surely help advance treatment strategies for trauma.

One serious complication of trauma is the development of the two-hit response phenotype, which was first defined by Moore et al. [4] as an amplified response to a secondary insult in patients who survive traumatic injuries. In trauma patients who develop infections, this “second hit”, which is usually an opportunistic infection, can cause severe septic shock and MOF. Thus, the combination of increased risk of developing infections and the development of the two-hit response is a principal complication of traumatic injuries. Research findings suggest that the two-hit response occurs, as cells of the innate immune system, primarily macrophages and neutrophils, become hyper-reactive to bacteria and bacterial toxins [5–8]. Once primed, macrophages and neutrophils produce heightened levels of inflammatory cytokines and ROS, causing systemic inflammation and detrimental cell-mediated tissue destruction.

This priming effect of injuries on the immune system may be detrimental or beneficial. The excessive reactivity could trigger excessive inflammatory cascades in injured patients who develop opportunistic infections. However, the primed innate immune system may also be better prepared for rapid and strong antimicrobial immune defense. A report by Maung et al. [9] supports this latter hypothesis. In that study, burn-injured mice were found to resist i.p. Escherichia coli infections significantly better than sham-injured control mice. In marked contrast, the detrimental impact of heightened innate immune system reactivity was shown by the high mortality in injured mice given a polymicrobial challenge by the CLP technique [10, 11]. An explanation for these contrasting findings may be that the CLP model more closely mimics clinical sepsis after traumatic injuries with a mixed microbial infection, bacteremia, excessive systemic inflammation, and the development of septic shock. Moreover, the difference between trauma effects on antimicrobial immunity and the two-hit sepsis response phenotype highlights the idea that injury disrupts immune system homeostasis. For example, a compartmentalized or low-level infection can be controlled more effectively by the innate immune system following injury, but if the infection is not controlled as a result of other influences, the heightened innate immune response to infection sets in motion the two-hit response, putting the injured patient at risk of “secondary” SIRS, septic shock, and MOF.

Another observed complication of traumatic injuries is the development of a counterinflammatory-type immune response, which is believed to be a compensatory response to injury-induced inflammation. The counterinflammatory response to injury has historically been referred to as trauma-induced immune suppression and is defined clinically as the CARS or MARS phase of the trauma response. It was labeled as an immune-suppressive response, as T cell-mediated responses show anergic-like properties including: low proliferative responses to mitogens and specific antigens; suppressed skin DTH responses; suppressed Th1-type immune reactivity with increased Th2-type cytokine production; and increased Treg activity [12–24]. In addition, there is a specific increase in a subpopulation of macrophages referred to as MSCs that occurs coincident with these counterinflammatory-type T cell responses to trauma [25, 26]. These innate cells also arise in clinical conditions, which involve chronic or unresolved immune responses, such as cancer and parasitic infections [27–30]. Interestingly, the counterinflammatory or CARS phase of the injury response becomes pronounced at later time-points after injury, which suggests that it is a developed response to the cellular and molecular triggers that initiates the injury response. We suggest that a logical reason for the counterinflammatory response to traumatic injuries is to control immune reactivity to tissue damage, to suppress trauma-induced inflammation, and to promote a natural healing response to help restore immune system homeostasis. On the other hand, a primary complication of the counterinflammatory response to injuries is that it can actively suppress antimicrobial immunity and may be responsible for the increased susceptibility of trauma patients to opportunistic infections that often occur during the counterinflammatory phase of the injury response. Thus, there are likely beneficial and detrimental influences of counterinflammatory responses to traumatic injuries.

A schematic diagram summarizing these general effects of injuries on the immune system responses to trauma and their relationship to SIRS and CARS is shown in Fig. 1. This diagram illustrates that the inflammatory response to injury is driven by the innate immune system and that the counterinflammatory response is mediated by the adaptive immune system. These opposing responses appear to occur in concert, as evidenced by the kinetics of injury-induced changes in innate and adaptive immune system phenotypes. The end result is that injury disrupts immune system homeostasis represented by SIRS and CARS, shown at the peaks or the “football-shaped” curve. If opportunistic infections arise at time-points when there is a wide imbalance in immune response homeostasis, the injured host will be at high risk of developing trauma-associated complications, such as sepsis, septic shock, or MOF. We propose that these opposing immune responses to injuries represent a natural healing-type host response aimed at controlling trauma-induced tissue damage and at restoring immune system homeostasis following a traumatic event.

Figure 1. Traumatic injuries induce an imbalance in immune system homeostasis.

This diagram summarizes generalized effects of injury on the immune system and the relationship to the development of SIRS and CARS phenotypes described in trauma patients. As illustrated, injury induces progressive enhancement of inflammatory and anti-inflammatory immune responses. The inflammatory response is driven by cells and mediators of the innate immune system, and the counterinflammatory response is regulated by the adaptive immune system. These opposing injury responses disrupt immune system homeostasis and lead to the development of SIRS and CARS in trauma patients. If opportunistic infections arise at time-points when there is a wide imbalance in immune homeostasis, the injured host will be at high risk of developing trauma-associated complications such as sepsis, septic shock, or MOF. Resolution of the injury response equals restoration of immune system homeostasis in patients who survive.

INITIATORS OF THE INJURY RESPONSE

According to the CDC, accidental injuries tops the list as the leading cause of death in the United States for people under 44 years of age (http://www.cdc.gov/nchs/data/databriefs/db64.pdf). This statistic does not include deaths from complications of injuries, such as sepsis, so this statistic may be higher than calculated by the CDC. Accidental injuries can be organized into general groups to include hemorrhage, crush injuries, blunt-force trauma, burn injury, bone fractures, chemical injury, radiation injury, and combination injuries, such as trauma with hemorrhage or radiation with combined trauma. It is not challenging to pair the above-mentioned types of injury with traumatic scenarios, such as motor vehicle accidents, earthquakes, fires, explosions, shootings, stabbings, major surgeries, and radionuclear accidents or events. War and terrorism contribute significantly to some of these types of injuries, and this has alerted nations to develop emergency preparedness strategies and medical countermeasures to help protect civilians and war fighters in harms way [31].

Although injury can occur in many ways, there are some common features to trauma that allow for a reductionist, scientific approach to study how injury influences the immune system. One common component of trauma is that it causes tissue damage. Although the level and type of tissues that are damaged by injury vary, it is generally agreed that injured tissue initiates the early host response to trauma. Another feature of traumatic injuries is that they occur by accident or without prior warning. Therefore, research into how to pretreat the physiological and immunological consequences of traumatic injuries will not provide clinically useful findings. We also posit that injury is a natural type of an immune response that has undergone evolutionary changes in a manner similar to antimicrobial immune responses or adaptive immune regulation in mammals. Thus, the initiating factors and mediators of injury responses may use the same immune cells and signaling pathways for normal immune responses against pathogens but will also stimulate unique types of reactions. In addition to specific effects of trauma on immune cell responses, there are physiological events associated with the host response to trauma, such as stress, shock, and neurological responses, which add to the complexity of the trauma response. A future challenge will be to work at the interface between the physiology and immunology of trauma to provide insights about how to restore immune homeostasis following traumatic injuries.

TRAUMA EQUALS DANGER

The substantial tissue destruction and necrotic cell death that is induced by traumatic injuries, lead to the release of previously sequestered antigens and factors. These antigens alert the immune system that tissue damage is present and are appropriately called alarmins [32, 33]. Alarmins are detected by immune cells and initiate an immune response by interacting with PRRs, such as TLRs, which are present on innate or adaptive immune cells [34]. Once detected, alarmins specifically alert immune cells to initiate inflammatory responses, chemotaxis, antimicrobial defenses, and adaptive immune cell responses. The detection of alarmins by cells of the immune system forms the basis for the “danger” theory, as proposed by Matzinger in 1994 [35].

Danger is clearly a central component of the immune response to trauma. Following trauma, the immune system is exposed to a large amount of tissue damage and endogenous antigens, including alarmins, and cell debris are likely released from necrotic cells and tissue. There may also be significant exposure to microbial molecules that are part of the microbiome [36]. Immune cells would then respond to injury antigens by PRRs that recognize DAMPs of microbial or exogenous origin (PAMPs) or endogenous origin (alarmins) [37, 38]. Whereas PAMPs alert the immune system to activate antimicrobial defenses, alarmins alert the immune system to tissue damage and danger. The shared recognition of danger by PAMPs and alarmins suggests that a primary function for PRRs might be to signal antimicrobial immunity against PAMPs to trigger an antipathogen immune response or against future opportunistic pathogens that could infect damaged tissues. The ability of PRRs to recognize PAMPs and alarmins suggests an evolutionarily conserved function for these receptors as innate immune recognition systems [38].

Among the expanding list of PRRs, TLRs, especially TLR4, have emerged as promiscuous sensors for alarmins following trauma. However, other TLRs and innate immune recognition molecules have also been shown react to alarmins and signal danger responses [39, 40]. We provide an updated list of some of the known trauma-induced alarmins, as well as some basic information about these molecules in Table 1. A more in-depth overview of other potential trauma-associated alarmins was published recently by Manson et al. [56]. It is clear from this list of alarmins that TLR4 has emerged as a primary receptor for danger. However, it is important to note that PAMPs, like bacterial LPS, can also bind to TLR4, which has generated controversy over whether DAMPs may carry microbial molecules to PRRs[57]. Examples of this scenario include reports that HMGB1 binds LPS, bacterial DNA, as well as IL-1 [58, 59]. Moreover, some authors have recently called into question whether DAMPs, such as HMGB1 and HSPs, actually have cytokine-like properties [60, 61]. They report that highly purified HMGB1 and HSPs have low biological activity. Nevertheless, whether by indirect or direct mechanisms, HMGB1 and HSPs appear to act as alarmins, as they are released following cell injury and tissue trauma [62, 63]. For example, Cohen et al. [42] detected HMGB1 in the plasma of trauma patients early after the injury and reported that higher plasma levels of HMGB1 were associated with the severity of injury, organ dysfunction, post-trauma coagulopathy, and adverse outcome. Moreover, HMGB1 has been implicated in the pathogenesis of a variety of human inflammatory diseases that also cause substantial tissue injury, such as sepsis, arthritis, cancer, and type I diabetes [41, 44–46]. As HMGB1 signals through TLR4, it is believed that it causes complications in trauma by initiating or amplifying inflammatory cytokine production by macrophages [64]. Other receptors activated by HMGB1 include TLR2 and RAGE [65]. Members of the HSP family also act as alarmins, as they are released extracellularly by damaged or stressed cells or tissues following trauma [49]. Similar to HMGB1, HSPs have been shown to trigger TLR responses, in particular, TLR2 and TLR4 pathways, and HSPs have been shown to mediate detrimental effects in trauma, e.g., in the cardiovascular system [48]. Moreover, like HMGB1, HSPs have also been shown to act as “carrier” molecules for PAMPs [57, 66–68]. HMGN1 was characterized recently as an alarmin for antigen-specific immune responses, as well as LPS-induced innate immune responses. Like other alarmins, HMGN1 was found to signal through the TLR4/MyD88 pathway [47]. The role of HMGN1 in trauma is yet unknown. Finally, Zhang et al. [50] reported that systemic inflammatory responses to trauma can be initiated by mitochondrial danger molecules—formyl peptides and mtDNA—which are released into the bloodstream following trauma. The recognition of mitochondria debris as alarmins is thought to be a result of molecular similarities between bacteria and mitochondria, which are obligate bacterial endosymbionts.

Table 1. List of Injury-Induced Alarmins.

Alarmin	Receptor	Pathways activated	Pathophysiological importance	References
HMGB1	TLR2, TLR4, RAGE	MyD88, TRIF	Trauma, sepsis, arthritis, cancer, type I diabetes	[41–46]
HMGN1	TLR4	MyD88	Unknown	[47]
HSPs	TLR2, TLR4	MyD88, TRIF	Trauma, cardiovascular diseases	[48, 49]
Mitochondria/mtDNA	TLR9 and FPR1	MyD88, TRIF, G-coupled receptor signaling pathways	Trauma	[50]
Uric acid	NLRP3, IL-1R	Caspase-1, MyD88	Lung fibrosis, gout	[51–53]
ATP	P2X7	Caspase-1	Lung fibrosis	[54]
S100	TLR4, RAGE	MyD88	Autoimmune diseases, cancer	[55]

NLRP3, Nucleotide-binding oligomerization-like receptor family, pryin domain-containing 3.

INFLUENCES OF TRAUMA ON INNATE IMMUNE SYSTEM RESPONSES AND FUNCTIONS

Our group has shown that TLR4 reactivity is enhanced following trauma and that this increase in TLR4 responsiveness contributes to the development of the two-hit response phenotype in a mouse injury model [8, 69]. The primary complication of the two-hit response is the development of a systemic inflammatory response against sepsis-causing opportunistic pathogens, such as Pseudomonas aeruginosa, E. coli, and Staphylococcal infections. A clinically relevant example of the two-hit response to trauma is the high mortality that occurs in injured mice, given a sepsis-causing infection by the CLP approach at 1–2 weeks after the injury [11, 70, 71]. Importantly, the kinetics of the development of the two-hit response phenotype in mice overlaps with clinical observations in trauma patients [72, 73]. Another feature of the two-hit response to trauma is that macrophages appear to be the primary innate immune cell type showing enhanced innate reactivity [8, 74, 75]. Our group [8] showed that macrophages from burn-injured mice produce and express high levels of IL-1β, IL-6, and TNF-α following LPS challenge. Schwacha et al. [76] found similar results in a mouse burn injury model and also demonstrated that unresponsiveness to IL-10 by macrophages may be one mechanism responsible for increased inflammatory cytokine production by macrophages following trauma. Moreover, it has been shown recently that burn injury or trauma causes a progressive increase in splenic macrophages with a phenotype similar to MSCs that were first identified in tumor-bearing mice. Myeloid suppressor-type macrophages have also been identified in mice challenged with parasite antigens or cannabinoid receptor activation or in mice subjected to CLP-induced sepsis [26, 29, 30, 77–79]. The trauma-induced macrophages express cell-surface F4/80, CD11b, and GR-1 markers and appear to be the primary source of TLR4-induced TNF-α production [25, 80]. Macrophages display heighted inflammatory reactivity in other injury models, including in mouse models for femur fracture, trauma-hemorrhage, and spinal cord injury [81–83]. This proinflammatory behavior of GR-1⁺CD11b⁺ macrophages may be unique to injury responses. Thus, it appears that trauma-induced changes in macrophage phenotype may be a common response to traumatic injuries.

It is important to note that several studies have shown a diminished production of proinflammatory cytokines when whole blood or mononuclear blood cells from trauma patients were stimulated with LPS ex vivo [84, 85]. The findings in these patient studies contradict the priming effects of injury on the innate immune system, which we discussed above. One potential explanation for this controversy could be a result of the cellular compartments that were tested in patients versus in animal studies. Patient studies were performed using circulating immune cells, whereas animal studies used cells prepared from peripheral immune organs. A recent publication from the Inflammation and Host Response to Injury Large-Scale Collaborative Research Program reported on gene-expression profiles in blood leukocytes prepared from trauma patients. In this comprehensive study, it showed up-regulation of many proinflammatory innate immune cell response genes, as well as counterinflammatory genes [86]. Thus, there remains controversy over whether trauma induces suppression of innate immune cell responses.

Future work should focus on studying specific molecular triggers responsible for the development of the two-hit response phenotype following trauma and should focus primarily on macrophage biology. Alarmins are strong candidates for these molecular triggers. Furthermore, as alarmins, such as HMGB1, signal inflammasome activation, this pathway might also be a part of the signaling network for cellular responses to injury. In agreement with this, a recent report by Osuka et al. [87] showed that burn injury triggered inflammasome activation by 1 day after injury in innate and adaptive immune cell subsets. Interestingly, blocking caspase-1 activation in injured mice caused heightened inflammation and mortality, which suggests a protective role for inflammasome activation as a protective response to injury. A similar protective role for caspase-1 and inflammasome activation in the liver was shown to occur in a mouse model for hemorrhagic shock [88].

Although we do not yet know the molecular signals for increased TLR signaling by macrophages following trauma, the pathways that signal TLR responses are well known and provide some clues about how trauma augments innate immune responses and inflammation. Thirteen different TLRs have been identified in mice and 11 in humans [89]. These receptors share similarity to Drosophila Toll protein containing leucine-rich repeats and TIR domains for recognition of alarmins and intracellular signaling. Two adaptor proteins, also with TIR domains—MyD88 and TRIF—control the initiation of TLR responses and are adaptors for the MyD88-dependent and -independent (TRIF-dependent) signaling pathways [90]. Activation of the MyD88 pathway leads to immune cell activation and cytokine production through the NF-κB pathway, and the TRIF-dependent pathway induces type I IFN production through the transcription factor IFN regulatory factor 3 [90]. We identified that the p38 MAPK signaling pathway was amplified significantly in LPS-stimulated macrophages from injured mice [69]. Other reports confirmed this finding and also suggest a role for the p38 MAPK signaling pathway as mediating the amplified TLR reactivity that has been described to occur after injuries [91–94]. It is not yet known whether the LPS-enhanced p38 MAPK signaling depends on the MyD88 or TRIF pathway, and as suggested by Triantafilou et al. [95], receptor crosstalk among CD14, CXCR4, and TLRs could be involved in amplifying TLR signaling. Nevertheless, therapies that specifically target the p38 MAPK signaling pathway in trauma patients showing clinical signs of septic shock or SIRS, might help mollify the two-hit response and its associated complications.

DAMAGE CONTROL: THE ADAPTIVE IMMUNE RESPONSE TO TRAUMA

Cells and mediators of the innate and adaptive immune system do not function in isolation. Their interaction is needed to initiate specific adaptive immune responses that then feed back on innate immune cells to augment antimicrobial immunity and resolve inflammation. APCs provide one direct link between innate and adaptive immune cell activation during infections and vaccinations. For example, infections caused by bacterial pathogens or viruses evoke a strong antimicrobial-type T cell response, as APCs are activated in a manner to skew CD4 T cell differentiation toward Th1 or Th17 CD4 T cell subsets [96–99]. These CD4 T cell subsets produce cytokines, which then provide activating signals to innate cells, such as neutrophils, macrophages, and NK cells, to boost the antimicrobial effector activity. On the other hand, certain types of immune responses lead to the development of Th2-type or counterinflammatory-type immune responses. Th2 responses occur with infections caused by some parasites or allergic responses and tend to suppress or antagonize Th1- or Th17-type immune responses [96]. The balance between Th1- and Th2-type immune response phenotypes can be an important determinant of infection, disease, and clinical outcomes.

It is generally agreed that trauma suppresses CD4 T cell responses [19, 100–102]. For example, Th1-type immune responses have been shown to be reduced markedly following trauma in patients and in mouse injury models [13, 103, 104]. Trauma has also been described to promote Th2-type immune system responses and T cell anergy in trauma patients and in mice [105, 106]. Interestingly, exercise can also skew T cell-mediated immune responses, suggesting that injury-induced changes in the immune system can be induced by activities that have health benefits [107, 108]. We [105] and other research groups [109, 110] have focused significant efforts on identifying how trauma might suppress CD4 T cell activation and push the adaptive immune response toward a counterinflammatory phenotype. Early studies suggested that the counterinflammatory behavior of that adaptive immune system may be a result of a shift toward high Th2 and low Th1 responses in mice and in patients. Other investigators [13] argued that as Th1- and Th2-type immune responses were found in trauma patients, T cell anergy may be another mechanism contributing to suppressed CD4 T cell responses following trauma. Our research group [103] performed immunization studies in mice as an attempt to determine whether injury drives higher Th2-type reactivity or suppresses Th1-type responses. This was accomplished by measuring antigen-specific antibody isotype formation in mice immunized with a T cell-dependent antigen at the time of injury and measuring CD4 T cell help for antibody production. We observed that Th1 responses were suppressed dramatically in immunized burn mice. The suppressive effect on injury on Th1 responses occurred even though the immunogen was given in a strong adjuvant, CFA. This adjuvant works by providing TLR-activating signals to APCs, yet Th1 responses remained suppressed for at least 10 days after injury and immunization. This dominant and strong Th1-suppressive activity caused by injury was confirmed in subsequent immunization and adoptive transfer studies using CD4 T cells from DO-11 TCR transgenic mice [23]. DO-11 CD4 T cells express TCRs that are specific for OVA peptide (aa sequence 323–339) and can be detected in OVA peptide-immunized mice to track expansion and phenotypic influences of injury on adaptive immune responses. In these studies, burn injury suppressed antigen-driven CD4 T cell expansion and OVA peptide-specific Th1 cytokine production. Taken collectively, the results of these immunization studies suggest that trauma can actively suppress CD4 T cell activation, Th1 differentiation, and proliferation without an increase in Th2-type responses. The biological reasons for trauma effects on adaptive immune responses were not well understood at the time that these studies were published. However, we now propose that trauma-induced changes in adaptive immune responses might occur to protect the injured host from alarmin-induced innate responses and to prevent possible self-antigen reactivity to injury antigens released by tissue damage.

Tregs CONTROL INJURY RESPONSES

Tregs are a subset of thymic-derived CD4 T cells, which have been shown to be central to the maintenance of immunological tolerance and to controlling inflammation in autoimmune and inflammatory disease mouse models [111, 112]. This unique CD4 T cell subset, which makes up ∼10% of the peripheral CD4 T cell population, expresses TCR repertoire diversity similar to conventional CD4 T cells [113, 114]. However, in contrast to conventional CD4 T cells, some Tregs are self-reactive, and this self-reactivity plays a role in their immune tolerance function [115–117]. Their development is dependent on the FoxP3 transcription factor, as mice lacking the FoxP3 gene do not have detectable Tregs [118]. FoxP3^−/− mice develop fatal autoimmune disease, providing more support for the concept that Tregs are a distinct CD4 T cell lineage and that they are needed to maintain immune tolerance [119–121]. Additional subsets of Tregs, called T regulatory 1, Th3, or induced Tregs, arise in the periphery in response to IL-10 or TGF-β signaling and have been characterized as being important in controlling inflammation, especially in a mouse colitis experimental model and EAE [122–125]. Tregs are often distinguished from conventional CD4 T cells by the expression of cell-surface CD25 molecules. Although not as definitive a marker of Tregs as FoxP3, CD25 expression on CD4 T cells is widely used as a tool to purify mouse Tregs and to target the depletion of Tregs in mice [112]. Purified Tregs markedly suppress the proliferation of TCR-activated CD4 T cells by cell contact [126]. Tregs also have the capacity to produce TGF-β1, IL-9, and IL-10 and to up-regulate the cell-surface expression of several negative costimulatory molecules, such as CTLA-4, ICOS, and PD-1, as well as TNFR2 [24, 127–130]. As all of these factors have been described as potent suppressors of inflammation or T cell reactivity, it is possible that these factors contribute to Treg function during trauma or other host responses, which involve substantial tissue damage, such as severe infections, transplantation, cancer, and autoimmune reactions.

Several biological features of Tregs suggest that they may control the adaptive immune response to injury. First, their capacity to suppress CD4 T cell activation and proliferation suggests that they could drive trauma-induced suppression of Th1 responses and T cell anergy. Second, the tissue damage associated with injuries could activate self-reactive Tregs. The first evidence supporting this idea came from a mouse injury study, showing that by 12 h after burn injury, there was a significant increase in CD4 T cells coexpressing CD25 and CTLA-4 in the LNs draining the site of injury [131]. Recently, our group confirmed that burn injury triggers TCR signaling in CD4⁺FoxP3⁺ T cells but not CD4⁺FoxP3⁻ or CD8⁺ T cells in the LN draining the injury site [132]. Finally, as trauma can induce systemic immune suppression in patients, it is possible that overly active Tregs could inadvertently promote or amplify immune suppression by blocking antimicrobial immunity, while attempting to control the inflammatory response to injury.

To determine whether injury affects Tregs, we compared the regulatory activities of CD25⁺CD4⁺ T cells purified from the LNs or spleens of sham- and burn-injured mice at 7 days after injury. We discovered that Tregs from injured mice were significantly more potent at blocking CD4 T cell proliferation than sham Tregs [24]. Moreover, we found that LN but not splenic Tregs expressed higher Treg activity, which provided new evidence to suggest that Treg activation by injury may be compartmentalized to injury-site LNs. Closer investigation into injury effects on Tregs showed that it caused increased expression of costimulatory molecules, indicative of increased Treg function or survival, CTLA-4, PD-1, ICOS, and CD28 [24]. The enhanced expression of these costimulatory molecules could function to amplify Treg activity following traumatic injuries. For instance, the increase in CTLA-4 expression by Tregs at 7 days after injury could help increase Treg potency by suppressing DC APC function through an IDO mechanism or by competitively blocking CD80 and CD86 access to CD28 to suppress T cell activation [133]. ICOS and PD-1 have been shown to play a role in mediating Treg responses and CD28 signaling and also appear to be required to maintain Treg proliferation in peripheral immune tissues [134–137]. Finally and most important, conventional CD4 T cells showed no injury-induced changes in costimulatory molecule expression, suggesting that only Tregs are activated in response to injury. Similar type changes in Tregs have been reported to occur in trauma and sepsis patients, which strengthens the clinical importance of these findings reported in mouse injury studies [138, 139]. Taken together, these results provide additional evidence supporting the idea that tissue injury from trauma delivers some type of an activation signal to Tregs and that Treg activation may be a natural element of host response to traumatic injuries.

The discovery that injury augments Treg activity prompted experiments to test whether Tregs might regulate trauma-induced changes in innate and adaptive immune functions. Based on what is known about how Tregs modulate immune responses, our research group examined whether Tregs might contribute to the suppression of Th1-type immune reactivity and the control of the two-hit response phenotype in trauma. To address these questions in vivo, we optimized a Treg-depletion protocol using a moderate dose of anti-CD25 antibody, which removed CD25⁺ T cells from the peripheral immune tissues of mice but did not prevent T cell activation. T cell responses were measured in vivo by immunizing Treg-depleted mice with the T cell-dependent antigen, TNP-haptenated OVA, and measuring antibody isotype formation, as well as OVA-induced IFN-γ production. Our results indicated that sham or burn Treg-depleted mice produced higher titers of the Th1-type antibodies and higher levels of IFN-γ, which indicated that Tregs were actively suppressing Th1-type reactivity in injured mice [24]. A subsequent report using the DO-11 TCR transgenic adoptive-transfer experimental system showed that injury in Treg-depleted mice greatly enhanced antigen-specific CD4 T cell expansion as well as Th1 reactivity [140]. Importantly, the results of this study showed for the first time that Tregs may be activated by trauma and tissue damage in an antigen-specific manner. The reason for antigen-specific control of antigen-specific T cell responses may be to protect the host from developing potentially harmful Th1-type immune responses to self-antigens released by the tissue damage associated with the injury. Interestingly, findings reported by Zang et al. [141] showed that high-level T cell activation at early time-points after burn injury, induced by bacterial superantigen, led to high mortality, whereas there was no mortality when mice were challenged with superantigen at 1 week after injury—a time-point when Treg numbers are increased, and Tregs can potently suppress CD4 T cell activation [142]. Similar immune regulatory functions for Tregs have been suggested by other studies, showing that Tregs protect the host from excessive self-T cell reactivity in inflammatory disease models, such as T cell-mediated colitis and EAE [123, 143–146]. Thus, Tregs appear to control CD4 T cell responses in trauma or related inflammatory-type diseases by reduction Th1-type inflammation and self-antigen reactivity (Fig. 2).

Figure 2. Immune cell subsets that react to injury and control the injury response.

Tissue damage caused by trauma leads to the release of alarmins and self-antigens into extracellular compartments, which could activate macrophages (Macϕ) and Tregs. In macrophages, injury triggers a number of changes in phenotype and function, including enhanced TLR4 reactivity and antimicrobial responses, as well as increased appearance of GR-1⁺/CD11b⁺ macrophages. Tregs are then activated and have been shown to act as “master regulators” of the injury response by suppressing innate and adaptive cellular responses to trauma.

The potential that Tregs might also control the inflammatory response to injury was first suggested by the results of experiments examining how Rag1^−/− mice, which lack T and B cells, respond to injury. When injured, spleen cells from Rag1^−/− mice showed heightened reactivity to TLR2 and TLR4 stimulation, as compared with spleen cells from WT control mice [11]. This observation suggested that the adaptive immune system must actively suppress or control the inflammatory behavior of the innate immune response to injury. To address which adaptive immune cell type might be needed to suppress injury-induced inflammation, CD4 and CD8 T cell^−/− mice were compared directly with Rag1^−/− and WT mice for changes in TLR4 responsiveness after burn injury. The results of these experiments demonstrated that CD4 T cell^−/− mice resembled Rag1^−/− mice, showing higher TLR4 responses, whereas CD8 T cell^−/− mice resembled WT mice [147]. When Rag1^−/− mice were given WT CD4 T cells by adoptive transfer, the injury-induced increase in TLR4 responsiveness was suppressed. Further separation of CD4 T cells into CD25⁻ and CD25⁺ subsets prior to adoptive transfer into CD4^−/− mice showed that CD25⁺CD4⁺ T cells suppressed this injury-induced increase in TLR4 responsiveness [147]. These findings provided strong evidence to indicate that Tregs play an active role in controlling the inflammatory response to injury.

As it is know that injured mice and humans can develop heightened systemic inflammatory reactivity after severe injury, which can promote the development of MOF, we tested recently whether Treg deficient mice might be more susceptible to this two-hit response. Using a burn injury, followed by a LPS challenge two-hit mortality model, we directly compared the two-hit response in Treg deficient mice with control non-Treg-depleted mice [69]. We found that Treg deficient burn-injured mice were significantly more susceptible to LPS challenge, showing 100% mortality versus 50% mortality for burn WT mice (unpublished results). Importantly, Treg deficient sham mice given LPS did not die, suggesting that it is the combination of Treg deficiency and injury that enhanced the two-hit response. Moreover, LPS-challenged Treg deficient burn-injured mice showed higher levels of TNF-α and IL-6 in lungs, liver, and kidney than control mice, further suggesting that Tregs have the capacity to control inflammatory responses to traumatic injuries (Fig. 2). These findings showed that Tregs played a protective role in controlling the magnitude of the injury response to trauma. The clinical significance of this observation is that individuals with lower levels of Tregs or dysfunctional Tregs could be at higher risk for developing inflammatory complications to trauma or other inflammatory-type diseases. For example, Treg numbers are decreased in patients with systemic lupus erythematosus and rheumatoid arthritis [148, 149].

Although a clinical relationship between the development of postinjury sepsis and altered Treg numbers or function is unknown, there is evidence that injury and sepsis can influence Tregs in patients. We reported recently [20] that circulating Tregs from trauma patients demonstrate enhanced Treg activity by 5–7 days after injury as compared with Tregs prepared from patients at 1 day after injury or from normal individuals. This finding validates the results of experimental studies in our mouse burn injury model. Because of this, we have subsequently examined changes in mouse blood Treg function and found that they also displayed enhanced Treg potency by 7 days after injury (unpublished observation). Another report showed that septic patients develop higher numbers of circulating Tregs, suggesting that enhanced Treg presence or responsiveness may also be part of the host response to sepsis and strong inflammatory responses [150]. Experiments performed in a mouse sepsis model, CLP, support this hypothesis. Two independent reports indicate that CLP induces an increase in Treg numbers and function [151, 152]. Whether Tregs are beneficial or harmful during a sepsis response remains controversial. Two studies indicate that the CLP-induced survival of mice, made Treg deficient by anti-CD25 antibody treatment, was not different than control mice [151, 152]. However, one study showed that transferring CD25⁺CD4⁺ T cells into CLP-challenged mice was protective [153]. One potential reason for these opposing findings is that mice that lack Tregs may develop enhanced inflammatory reactivity following sepsis and could concomitantly develop a stronger, protective Th1 response to boost T cell-mediated immunity. These offsetting responses might result in no apparent survival differences in CLP-induced survival between mice, with or without functional Tregs. The differences in CLP survival responses between these reports could also be a result of the severity of the CLP challenge.

POTENTIAL TREATMENT STRATEGIES TO RESTORE IMMUNE SYSTEM HOMEOSTASIS FOLLOWING TRAUMA

The results of basic and clinical studies support the overall hypothesis that traumatic injuries disturb normal immune system homeostasis and that interventional approaches to reduce this imbalance could reduce the morbidity and mortality rates associated with traumatic injuries. One of the roadblocks in developing trauma-directed therapies is the heterogeneity in the injury response among trauma patients. Thus, efforts to classify patients into trauma response phenotypes by genetic or biological markers will greatly augment the potential for developing more effective treatment approaches for this complex disease process. Furthermore, the limited understanding of trauma and sepsis in the past had a large negative impact on advancing drug development and treatments for this complex disease as a result of failed clinical trials. Nevertheless, some studies in preclinical animal models have shown that it is possible to change the clinical trajectory of traumatic injuries using cytokines, sex hormones, and immunonutrition, among others, to restore immune system homeostasis following trauma. The available information gained from these studies suggests that treatments, which reduce the extremes of innate and adaptive immune system responses to injuries, help protect trauma patients from developing infectious and inflammatory complications from trauma. These encouraging preclinical findings, along with the better present day understanding of how trauma influences the immune system, suggest that clinical trials to evaluate new and safer approaches should be explored.

Cytokines are made by immune and nonimmune cell types and can transmit local and systemic signals to initiate and regulate immune responses. Early clinical studies tested the potential of blocking inflammatory cytokines, such as TNF and IL-1, as they were shown to be increased in patients who showed worse outcomes in trauma and sepsis. Unfortunately, results of these clinical trials showed that anti-inflammatory approaches were not beneficial [154]. Another cytokine therapy that was tested clinically was GM-CSF treatment for lung injury and sepsis. GM-CSF is produced by activated macrophages, T cells, endothelial cells, and fibroblasts and acts on hematopoietic stem cells to promote development of neutrophils, monocytes, and macrophages. When given to patients with acute lung injury, the treatment showed no significant benefit, but meta-analysis of its effects in treating septic shock showed it reduced infection [155, 156]. This suggests that there may be a use for GM-CSF as an immune-enhancing therapy for traumatic injuries to prevent opportunistic infections.

Advances for treating traumatic injuries have been made in preclinical animal studies. One successful approach was to use IL-12 to promote Th1-type immune responses in mice to heighten antimicrobial immunity. In one study, IL-12 treatment, given at early time-points after injury, restored resistance to CLP sepsis challenge in burn-injured mice [157]. Another report indicated that IL-12-treated burn mice were protected from herpes simplex infection when treated early after injury [158]. The clinical relevance of these studies was shown by the observation that PBMCs from trauma patients were found to produce significantly less IL-12 in response to LPS stimulation as compared with healthy controls [159]. As IL-12 is known to potently induce IFN-γ production and Th1-type immune responses, these findings provide evidence to suggest that redirecting the counterinflammatory T cell response to injuries could be a useful approach to restore immune homeostasis. However, clinical trials for IL-12 treatment in trauma patients have not yet been pursued as a result of concerns over the toxicity observed in human cancer trials in patients given high doses of IL-12 [160]. Regardless, there is ongoing interest in IL-12 in cancer research, and it may be important to revisit IL-12 as a therapy for traumatic injuries [161]. Another cytokine, which has been shown to restore immune function in injured mice, is IL-18, which induces IFN-γ production by NK cells; this could promote Th1-type responses after traumatic injuries. Treating mice with IL-18 after burn injury protected mice from CLP sepsis and also increased antimicrobial immunity in injured mice [162]. Another report showed that blocking IL-10 activity in vivo by anti-IL-10 antibody treatment could enhance Th1 immunity in mice and also protect burn-injured mice from CLP sepsis [163, 164]. A commonality among all of these cytokine-based treatments, which restore immune system function in injured mice, is that they all enhance early inflammatory-type immune reactions. Thus, it seems possible that developing treatments, which boost early inflammatory-type reactions rather than suppressing inflammation after trauma, may help redirect the immune system responses to trauma and restore immune system homeostasis. In support of this hypothesis, we found recently that giving CpG oligodeoxynucleotide at 1 day after injury could restore Th1-type responses in mice and protect them from succumbing to two-hit sepsis at 1 week after burn injury and treatment (unpublished results).

Other cytokines, which have been used to treat the immunological complications of trauma, are IL-15 and Flt3L. IL-15 is produced by activated macrophages and stimulates NK cell and CD8 T cell activation. Flt3L is a hematopoietic growth factor, which stimulates DC growth and differentiation [165]. Treating mice with IL-15 was found to restore depressed, splenic DC function in mice following trauma-hemorrhage [166]. Similarly, Flt3L ameliorated burn injury-induced immune defects and increased resistance to infection in mice [167–169]. These findings provide evidence supporting a role for cytokines as effective treatments to redirect or restore immune system function in trauma by stimulating hematopoiesis or to reprogram immune cell functions.

Gender is a central determinant of clinical outcomes for trauma and sepsis, with male gender having a negative influence on trauma complications [170–173]. Therefore, a different approach to treating the detrimental effects of traumatic injury on the immune system is modulating sex hormones. Many studies from the Chaudry group [174, 175] showed that 17β-estradiol is the key gender factor responsible for the finding that proestrus female mice have enhanced immune function following trauma as compared with male mice. Accordingly, treatment with 17β-estradiol has been shown to be beneficial in trauma-hemorrhage and burn injury in mice [176]. In a more recent study, 17β-estradiol effects on restoring immune system function in a trauma-hemorrhage mouse model were shown to be dependent on peroxisome proliferator-activated receptor γ [177]. Taken together, the results of these and other experimental studies suggest great potential for modulating sex hormones to transiently treat the immune consequences of trauma [178]. Various other approaches, which have been tested for improving immune system function after trauma, include immunonutrition. Examples of immunonutrition as treatment for critically ill patients and trauma include glutamine, arginine, and omega-3 fatty acid feeding [179]. For example, one clinical study supports using glutamine feeding to reduce clinical complications in trauma patients [180]. In summary, the results of experimental studies testing cytokines, sex hormones, and immune-enhancing nutrition as treatments for the clinical complications of traumatic injuries are encouraging. Nevertheless, translating these results from the bench to the bedside remains a challenge for all investigators in the field.

CONCLUDING REMARKS

Substantial progress has been made in understanding how traumatic injuries influence the immune system, and the information gained supports the hypothesis that trauma shifts normal immune system balance toward proinflammatory and counterinflammatory phenotypes. Data support the idea that these opposing influences of injuries on the immune system can be protective, can occur in concert, and can also contribute to the pathophysiological complications of traumatic injuries. For example, the development of the two-hit response phenotype represents a complication of trauma, whereas this same priming effect of injuries on the innate immune system can boost protective antimicrobial immune function. Moreover, research findings have uncovered the principal immune cell subsets that mediate the inflammatory and counterinflammatory responses to traumatic injuries. We propose that tissue macrophages mediate the innate inflammatory response to injuries, whereas CD4⁺ Tregs control the adaptive immune response. Furthermore, there is significant crosstalk between these immune cells, as injury in the absence of Tregs skews the injury response to a more proinflammatory-type response. The versatile regulatory activities of Tregs make them well-suited to act as master regulators of the injury response. We believe that alarmins, including DAMPs and PAMPs, will likely emerge as molecules that trigger immune responses to injuries. As such, alarmins could trigger a priming effect on the immune system to promote inflammation, antimicrobial immunity, and Treg activation. Available data support the ideas that Treg activation by alarmins could occur by direct signaling, by indirect activation via alarmin-activated APCs, or by self-antigen recognition. Unraveling these at-present, unknown mechanisms will be a significant advancement for injury research and the development of treatments to restore immune system homeostasis in trauma patients.

AUTHORSHIP

This review was prepared in collaboration among all of the listed authors (V.M.S., A.O., and J.A.L.).