Ischemia/reperfusion injury (IRI) is a multi-factorial antigen-independent inflammatory condition that profoundly affects both early and long-term function of the allograft as suggested by both clinical and experimental data. In recent years, the acute phase of IRI has been increasingly viewed as part of the innate immune response. Identification of novel molecular pathways and new insights into the mechanisms of known mediators of IRI have established links among innate immunity, adaptive immune responses and organ regeneration, and thus long-term graft function. This review approaches these novel aspects of IRI in the context of solid organ transplantation, presenting data on new observations with kidney, liver and heart allografts.
Ischemia/reperfusion injury (IRI) is a multi-factorial antigen-independent inflammatory condition that has both immediate and long-term effects on the allograft. According to the classic understanding, acute ischemia leads to the activation of the endothelium with an increase in permeability and expression of many adhesion molecules. These molecules are crucial for the recruitment and infiltration of effector cells into the post-ischemic tissue. Transcription factors such as nuclear factor-κB (NF-κB) are induced and activated, leading to enhanced expression of inflammatory genes. The endothelial cells lose their anti-adhesive properties and develop a thrombogenic and adhesive surface. On reperfusion, the ischemia-primed endothelial cells are prone to leukocyte and platelet adhesion, further increasing endothelial cell permeability and cell activation. The adherent leukocytes release reactive oxygen species (ROS) and a variety of cytokines, enhancing the inflammatory reaction. Subsequently, the leukocytes transmigrate and enter the subendothelial space. IRI involves loss of energy, derangement of ionic homeostasis and cell death. This initial cascade of events has a profound effect on early graft function.
Recently, the acute phase of IRI has been increasingly viewed as part of the innate immune response to the lack of vascular perfusion and oxygen. It has been long established that IRI will have lasting effects on the graft, as suggested by both clinical and experimental data. Recognition of novel molecular pathways as well as new insights in to the mechanisms of previously well-defined mediators of IRI, have established links among innate immunity, adaptive immune responses and organ regeneration, and thus long-term graft function (Figure 1). This review approaches these novel aspects of IRI in the context of solid organ transplantation, primarily concentrating on new observations with kidney, liver and heart allografts.
IRI Stimulates Innate Immunity
The immune response comprises innate and adaptive components. The innate immune system is responsible for the initial response of the organism to potentially dangerous stresses, such as pathogens, tissue injury or malignancy. The innate response is antigen independent, and uses pattern recognition receptors to recognize conserved molecular motifs. The role of innate immunity is less well understood in organ transplantation. In experimental transplant models where no microbial infection is present, the innate immune response is triggered by other non-infectious stimuli, including ischemia/reperfusion, systemic stress and surgical injury. These non-infectious stimuli contribute to the initiation of an inflammatory response in part by signaling via Toll-like receptor (TLR) mediated pathways (1).
Toll-like receptors are a family of pattern recognition receptors that are activated by specific components of microbes and certain other molecules. They comprise the first line of defense against many pathogens, and play a central role in the function of the innate immune system. They are expressed in different cell types, primarily in antigen presenting cells (APC), including macrophages, dendritic cells (DC) and B cells. In addition to hematopoietic cells, TLRs are also present on endothelial and stromal cells, including the epithelial lining of the respiratory, intestinal and urogenital tracts (2).
Signal transduction of TLRs results in activation of NF-κB and the mitogen activated protein (MAP) kinases p38 and Jun N-terminal kinase (3). Activation of these receptors results in a stimulus-specific expression of genes required to control infection and injury, including the production of inflammatory cytokines and chemokines, complement products, and the recruitment of neutrophils to the site of injury.
TLR activation has been recently demonstrated to play a role in ischemia models. TLR4 is involved in innate immunity primarily by recognizing lipopolysaccharides (LPS). Its engagement is linked to NF-κB activation in several cell types with subsequent expression of multiple pro-inflammatory cytokines (TNF-α, IL-1β) as well as inducible NO synthase and ICAM-1. Increased TLR4 expression (at both mRNA and protein levels) in Kupffer cells was demonstrated shortly after reperfusion in rats undergoing orthotopic liver transplantation with cold preserved syngeneic organs (4). TLR4 deficient mice were found to be less prone to IRI following hepatic ischemia (5). Additionally, in vivo experiments demonstrated that TLR4 but not TLR2 is required for the initiation of IRI in the liver. LPS translocated from the gut may be important in activating TLR4, but it is most likely not the initial stimulus. Liver IRI can also develop in vitro, and longer preservation results in more profound injury, despite sterile conditions. These findings suggest that endogenous TLR ligands are the primary stimuli initiating liver IRI (6,7).
A similar role for TLR4 was established in myocardial ischemia. It is constitutively expressed both in cardiomyocytes and the vasculature. Hypoxia markedly increases its expression. Several studies describe decreased infarct size in TLR4 negative mice, due to decreased lipid peroxidation and complement activation. Putative endogenous ligands for TLR activation include HSP60, which circulates due to hypoxic tissue injury. Physical alteration of the receptor-plasma membrane environment by ROS may also activate TLRs in the absence of any ligand. The significance of the activation of TLRs by endogenous ligands is being increasingly recognized. Heat shock proteins (HSPs) are intra-cellular molecular chaperones of aberrantly folded, nonnative, denatured proteins that are primarily involved in cytoprotection and adaptation for cell survival in response to stressful stimuli. Following profound stress (e.g. cell necrosis), HSPs may be released extracellularly as molecular chaperones in a cytokine-like manner. Putative endogenous ligands of TLRs were first described in studies with HSPs, and include HSP60, HSP70 and gp96. The immunostimulatory and inflammatory activity of these molecules are mediated through TLR2 and TLR4 (8,9). Additional molecules induced by injury that are able to stimulate TLR include inducible host antibiotic defensins and extracellular matrix components (Table 1).
Table 1. IRI-related endogenous ligands of TLR
Heat shock proteins: HSP60, HSP70, GSP96
DC maturation, increased cytokine production via NF-κB activation, stress responses
DC maturation, increased cytokine production via NF-κB activation, tissue repair gene induction
Inducible defensins from urogenital epithelium, skin and respiratory tract: hBD1, hBD2, hBD3
NF-κB activation, recruitment of DC and T cells
The role of the complement system in IRI is long established. Increase in the production of complement components is regulated by pro-inflammatory cytokines, primarily IL-6, TNF-α and IFN-γ. While the importance of the alternative activation pathway has been demonstrated, recent evidence from mice suggests an additional role of the lectin-binding pathway. Ischemia and subsequent reperfusion leads to renal mannose-binding lectin (MBL) depositions in the early reperfusion phase, followed by deposition of C3, C6 and C9 in the later reperfusion phase. In the kidney, the deposition of MBL-A and -C colocalized with C6, demonstrating that MBL is involved in complement activation. MBL-depositions in peritubular capillaries and tubular epithelial cells can also be observed early after transplantation of ischemically injured kidneys (10,11).
While systemically released complement from liver and vascular endothelium has been long known, the significance of locally produced complement has recently been recognized. In the kidney, endothelial cells and smooth muscle cells express C3. In addition, the complement regulatory (mainly inhibitor) proteins such as CD55 are poorly expressed on these cells. The significance of locally produced complement is clearly supported by knockout studies: C3 deficient kidney grafts survive much longer in C3 positive recipients than C3 positive grafts. When analyzing the contribution of regional complement production in the context of innate and adaptive immunity, syngeneic kidney grafts display a first phase increase in complement production/deposition that wanes by 48 h after reperfusion. Allografts also have early elevation of local complement production, with an additional second phase around day 4. Levels peak during rejection, and over 10% of total circulating complement can be released from the graft (12).
It is important to emphasize that the complement system not only is a manifestation of IRI and innate immunity, but also can regulate adaptive immunity. Binding of complement components to cell surfaces will result in a wide range of changes including the enhancement of B- and T-cell responses. In B cells, complement enhances antibody production via complement receptors (13). T cells are regulated by multiple mechanisms including modulation of cytokine release. C3a and C5a, the so-called anaphylatoxins, have opposing roles during allergen-induced T-cell polarization as C3a promotes Th2 responses, whereas C5a prevents Th2 polarization (14). Additionally, a member of the complement regulatory receptor family, CD46 or membrane cofactor protein, widely expressed on human cells, alters the activity of T cells via cytokine modulation and differentiation of regulatory T cells. Cross-linking of CD46 and CD3 on human CD4+ T cells results in induction of a regulatory T phenotype and release of IL-10 (15).
Recently, a regulatory role of decay-accelerating factor (Daf) in T-cell responses has also been established. Daf dissociates C3/C5 convertases assembled on host cell surface, and thereby prevents complement activation. The absence of Daf on APC and on T cells enhances T-cell proliferation, and augments the frequency of effector cells induced during primary T-cell activation. This effect is partly C5-dependent, indicating a role for local alternative pathway activation. Interactions between cognate T cells and APC were accompanied by rapid production of alternative pathway components and down-regulation of Daf expression, suggesting that local alternative pathway activation and surface Daf protein function as negative modulators of T-cell immunity (16).
The coagulation system coevolved with the innate immune system in eukaryotic development, to defend the host following tissue injury and/or microbial invasion. Activation of the coagulation cascade has important implications in IRI, and has been demonstrated in liver, kidney, pancreas and cardiac grafts. The instant blood-mediated inflammatory reaction (IBMIR) to human transplanted islets is considered an innate reaction, which is mediated by both clotting factors (thrombin) and components reflecting cell injury released by the islet (tissue factor). Recently, several new strategies were introduced to improve the outcome of clinical islet transplantation, including thrombin inhibitors, pre-treatment of the islets with nicotinamide to reduce tissue factor release and coculturing with endothelial cells to provide protection against IBMIR (17,18).
Changes in chemokine and chemokine-receptor expression following transplantation regulate the migration of leukocytes from the peripheral circulation into an allograft. Appropriate alloreactivity in vivo requires leukocyte interactions controlled within secondary lymphoid organs, and the migration of DC from the allograft into secondary lymphoid tissue is of paramount importance to the rejection process. In transplanted organs, ischemia/reperfusion, TLR activation, cytokine stimulation and complement activation all induce expression of multiple pro-inflammatory chemokines in the grafts. These changes escalate IRI and amplify allograft rejection. Activation of TLRs on graft stromal cells initiates an inflammatory response characterized by the recruitment of cells to the sites of injury or infection by a tightly controlled sequence of events. This includes the expression of selectin, chemokine and chemokine-receptor genes that regulate cell migration to the sites of inflammation. Key inflammatory chemokines induced include interleukin 8 (IL-8/CXCL8), growth-related oncogene (GRO/CXCL1), monocyte chemoattractant protein 1 (MCP-1/CCL2), MCP-2 (CCL8), MCP-3 (CCL7), MCP-4 (CCL13), macrophage inflammatory protein, (MIP-1/CCL3-∀), CCL4-∃ and RANTES (CCL5) (19,20).
Dendritic cells form an important connection between the innate and adaptive arms of the immune response and have been increasingly recognized to have an important role in IRI as well. The tissue damage associated with IRI leads to the release of the endogenous ligands described above. Engagement of TLR on DC induces activation and maturation. IRI-induced maturation modifies the distribution, phenotype and function of DC. Immature DC express inflammatory chemokine receptors CCR1, CCR2, CCR5 and CXCR1. During maturation, these receptors decrease in expression, while the expression of CCR4, CXCR4 and CCR7 increase, promoting migration of DC from inflamed tissue to secondary lymphoid tissue. Following ischemic stress, MHC class II+ CD45-B220+ F4/80 DC were detected in necrotic areas of the liver graft 20 h after reperfusion. DC freshly isolated from reperfused livers displayed a mature phenotype characterized by up-regulated expression of B7 costimulatory molecules, MHC-class II and CD1d. IL-10 and TGF-β mRNA expression increased, whereas IL-12p40 mRNA levels were decreased and IFN-γ mRNA levels were unchanged (21–23).
The maturational program enables DC to stimulate naive T lymphocytes. Induction of CD80 and CD86 on the DC surface is a particularly important step in the initiation of adaptive immunity. CD80 and CD86 costimulators are increased by TLR stimulation and ‘flag’ the antigenic peptide presented by MHC molecules on the same DC as being ‘microbe derived’. T cells can receive the costimulatory signal only in an antigen-specific, cognate interaction with DC. Thus, TLR-induced expression of costimulators translates the non-clonal pattern recognition signal into clonal antigen-specific immune responses (24). Ligation of TLR also contributes to T-helper cell activation by overcoming suppression mediated by regulatory CD4+ CD25+ T cells. TLR primarily control the activation of antigen-specific Th1 but not Th2 adaptive immune responses (25).
The extent of graft injury can influence the developmental stages of DC as a relative lack of endogenous TLR ligands could preclude maturation. Immature DC are incapable of initiating T-cell responses, but contribute to the expansion and differentiation of regulatory T cells. Immature DC are thus potentially tolerogenic by generating antigen-specific unresponsiveness, and may indeed prolong allograft survival (26,27). Conversely, when TLR on DC are activated, they release IL-6 that makes antigen-specific T cells refractory to suppression by regulatory T (Treg) cells. This mechanism is independent of costimulation, and it cannot be substituted by other modes of DC activation (28,29).
Innate Immunity and the Adaptive Immune Response
As discussed above, components of the innate immune response, such as complement, DC and TLR, have profound effects on the development of the adaptive immune response. Recently, evidence of reciprocal regulation of the innate responses by adaptive immunity has also been demonstrated. To elucidate these complex interactions in transplant settings, studies compared gene expression following heterotopic heart transplantation in recombinase activating gene (RAG)-deficient (alymphoid), syngeneic and allogeneic recipients. Immediately after transplantation, the host mounts a vigorous innate immune response. This innate response includes up-regulation of multiple pro-inflammatory genes, including chemokines, cytokines and cytokine receptors. No substantial differences in immediate gene expression profiles were found among the three groups, suggesting that the changes were induced by tissue damage, ischemia and surgical stress. The early phase of allograft rejection, similar to infectious models, consisted therefore predominantly of elements of the innate immune response. In extended kinetic analysis of gene expression, allogeneic recipients displayed a shift toward increased expression of defense-related genes (interferons, MHC antigens and complement) starting several days after transplantation (30). Analysis of 60 inflammatory genes including serum cytokines, intra-graft chemokines and cytokines, and their receptors confirmed a subset up-regulated only in the context of an adaptive response (lymphotoxin, RANTES and IP-10). Expression of markers induced by innate mechanisms (MIP-1, MIP-1, MIP-2 and MCP-1) was markedly amplified over time in the allogeneic, but not syngeneic or lymphocyte deficient, recipients. Thus, inflammatory mediators induced by innate mechanisms in the early phase after transplantation are further amplified by the subsequent adaptive response (31,32).
T cells—Bridging Innate and Adaptive Immune Responses in IRI?
T cells constitute one of the primary arms of the adaptive immune response, and according to classical immunological dogma, these cells were not suspected to play a role in ischemic injury. Recently, however, a T-cell contribution to IRI has been demonstrated. Knockout mice, lacking CD4/CD8 T cells, or cell adhesion receptors (ICAM-1) on T lymphocytes, are protected from renal IRI. Moreover, production of T cell associated cytokines (IL-1 and RANTES) and leukocyte adhesion molecules such as selectins, CD11/CD18 and ICAM-1, occur in IRI. Adoptive transfer of T cells from wild-type animals into the CD4/CD8 T cell knockout mice restored ischemic injury. The use of CTLA4-Ig to block B7-CD28 interaction significantly ameliorated renal dysfunction in a rat renal IRI model, inhibiting both T-cell and macrophage infiltration and activation (33). Recent observations in renal IRI suggest that IFN-γ producing Th1 cells are deleterious, while IL-4 producing Th2 are protective. The enzymes signal transducers and activators of transcription (STAT) 4 and STAT6 regulate Th1 or Th2 differentiation and cytokine production, respectively. In STAT6−/− mice IRI is worse compared with wild type, while STAT4−/− mice have mildly improved function. Post-ischemic phenotype and function in IL-4-deficient mice is similar to STAT6−/− mice. Thus, STAT6-dependent pathways have a major protective role in renal IRI, and IL-4 deficiency is the mechanism underlying the STAT6 effect. A similar role for STAT6 has been reported in hepatic ischemia.
In murine cardiac grafts, the early intra-graft inflammation is very similar between isografts and allografts during the initial 12 h following reperfusion. The main event, neutrophil infiltration, is regulated by chemokines including KC/CXCL1 and MIP-2/CXCL2. While the inflammation resolves in the isograft group, it is sustained and elevated in allografts. Differences in expression levels of chemokines and cytokines do not fully account for the kinetic changes in the extended inflammation. When recipients were depleted of certain populations of leukocytes, only depletion of CD8+ T cells attenuated the inflammatory response to allografts. A role for IFN-γ was demonstrated as anti-IFN-γ mAb treatment or use of IFN-γ−/− recipients displayed similarly attenuated inflammatory responses (34). The DC population that expresses CD8 and produces IFN-γ may participate in this early immune response in vivo, but the fact that the IFN-γ-dependent increase in neutrophil infiltration was observed in allogeneic, but not syngeneic, grafts suggests a requirement for CD8+ T cells. The same study demonstrated the ability of allogeneic endothelial cells to stimulate highly purified CD8+ T cells to induce expression of IFN-γ-dependent genes in vitro. Thus, the amplified inflammatory response observed in allografts is mediated by a population of circulating CD8+ T cells that produce IFN-γ during interaction with graft endothelial cells. This early IFN-γ-dependent CD8+ T cell-mediated immune response is detectable prior to T-cell priming (35). This suggests the presence of established memory CD8+ T cells, possibly primed to environmental antigens, that initially amplify the innate response (36).
IRI Initiates Regenerative Pathways
The ultimate fate and function of a transplanted organ depends on the balance between tissue damage inflicted by IRI and immune attacks, and subsequent restorative and regenerative processes. The cytokine-dependent pathways in IRI that damage the resident cells may also play a significant role in initiating the cell cycle, which is critical for graft recovery. Thus, while many of the responses in IRI have been viewed as harmful, some may also be beneficial to initiate repair and survival.
The liver has an enormous capacity to regenerate in response to resection or IRI. Cytokines such as TNF-α, IL-1β and IL-6 that are induced as part of the IRI related inflammatory cascade have a dual function, mediating inflammatory responses and initiating liver regeneration. Mice genetically deficient in these cytokines demonstrate a severe defect in hepatocyte replication after partial hepatectomy. The mice also lack the increase in the activation of NF-κB and STAT3, normally seen after resection. Jun B, c-fos, c-myc and c-jun are induced during liver regeneration and are associated with cell repair or initiation of DNA synthesis. Both c-fos and Jun B, which are transactivated by STAT promoter elements, are increased in livers with longer ischemic times, and also have significant regenerative activity. These interactions among cytokines, rapid activation of latent transcription factors, and expression of genes responsible for cellular repair and regeneration regulate the ability of the liver graft to repair itself following IRI.
In the kidney, regeneration and recovery follows shortly after injury as necrotic cells are replaced by replicating cells. Cell cycle surveillance mechanisms detect defects in DNA synthesis and chromosome segregation to block cycle progression at different decision points. These checkpoints insure that each phase of the cycle is irreversible, and each phase is completed before another is initiated. Failure of checkpoint surveillance results in dysregulated mitosis and often cell death. The checkpoints are controlled by cyclin-dependent kinase (cdk) inhibitors, especially p21. Ischemic injury to the kidney results in marked induction of p21 mRNA in nuclei of both distal and proximal tubule cells. Wild-type p21(+/+) mice were compared to mice homozygous for a p21 gene deletion p21(−/−). After 30 or 50 min of ischemia, p21(−/−) mice displayed more severe morphologic damage, and had higher mortality than p21(+/+) controls. In p21(−/−) mice, widespread cell death was associated with increased cell cycle activity, and increases in nuclear size (37). These data suggest that cell injury induces pathways that compete between cell death and cell cycle arrest versus coordinated cell cycle control necessary for optimum recovery.
Recent reports also suggest a role for stem cells and progenitor cells, both from local pools and from the circulation. In post-transplant regeneration there may be rapid recovery of endothelial cell loss through colonization by recipient derived endothelial cells, and this was found to be more common in patients with allograft rejection or IRI (38). Increasing evidence suggests that regeneration following ischemic injury is critically dependent on stem cell mobilization and homing (39), and several mediators of these processes have been identified. Stromal-cell-derived factor 1 (SDF-1) is a CXC chemokine required for stem cell homing to bone marrow. SDF-1 expression was found to be up-regulated immediately after both myocardial infarction and focal cerebral hypoxia. Intra-myocardial injection of SDF-1 plasmid in mice 2 weeks after surgically induced myocardial ischemia and myocardial infarction resulted in increased accumulation of labeled stem cells into the ischemic border, suggesting that gene therapy with an SDF-1 vector offers a therapeutic strategy (40). Another important factor in stem cell mobilization is endothelial nitric oxide synthase (eNOS). Mice deficient in eNOS show reduced VEGF-induced mobilization of endothelial progenitor cells. This mechanism may contribute to impaired regeneration processes in conditions associated with reduced systemic NO bioactivity (41).
Summary and Perspective
Current dogma has assumed that IRI is the non-specific result of the trauma of preservation and surgery. Most IRI related research has been concentrated on preservation techniques and elements of the inflammatory cascade and has been done primarily in syngeneic settings. Discoveries and developments during the last decade, as well as advanced surgical methods, have significantly improved early graft performance. Novel observations, however, indicate that a transplanted allogeneic organ is recognized, and under attack by the recipient immune system within minutes to hours following reperfusion, and IRI is a highly coordinated and specific process mediated by components of both innate and adaptive arms of immunity.
Further research is needed to explore the pathways and mechanisms by which the innate response is activated. The role of TLRs appears to be central, but exact interactions among different cellular and molecular components including parenchymal cells, NK cells, T cells, monocytes or complement factors are far from clear. The endogenous ligands and specific TLR involvement are not fully known. Specific knowledge regarding different organs is also lacking.
As discussed above, many lessons about the interactions between the innate and adaptive arms are not taken from transplant but rather from tumor and infectious models. These interactions should be reinvestigated in allogeneic settings, focusing on reciprocal and multi-lateral regulatory effects among DC, T cells, NK cells, NKT cells and monocytes. Future investigations must also determine the effect of IRI and innate immunity on the development of tolerance as this remains unexplained at this time. An additional promising but incompletely explored area of research is the relationship among cytoprotective genes and innate immunity, as some of the proteins (e.g. HSP70) that regulate the expression of these genes are also recognized as ligands of TLR.
While the inhibition of the production of oxygen radicals and inflammatory cytokines will remain central to strategies for treatment of IRI, certain aspects need to be reconsidered. To protect grafts from early T cell-regulated inflammatory damage, therapeutic possibilities aimed at interfering with the interaction of recipient T cells and the graft endothelium should be investigated. As the importance of organ regeneration is increasingly recognized in long-term graft performance, the possible negative consequences of immunosuppressive treatment (e.g. inhibition of stem cell proliferation and migration) to this process need to be considered.
J. S. B. was supported by grants NIH (AI41428, AI44929, AI62765) and JDRF (1-2005-16).