The Inflammatory Response to Ischemic Acute Renal Injury

Acute ischemia reperfusion injury (IRI) is a major cause of acute kidney injury (AKI). A major conceptual insight is that the total amount of renal injury is the result of not only the initial ischemic insult, by the ensuing responses. This simple insight has profound clinical implications: it suggests that therapy after the initial insult has already occurred may ameliorate the course of AKI. Some of these responses are maladaptive. Examples include inappropriate intrarenal hemodynamics, altered mitochondrial and other metabolic functions, endothelial dysfunction, and tubular obstruction and back-leak. Other responses are reparative. Examples include the production of growth factors such as Wnt7.

In addition to the above, IRI elicits an inflammatory response. Some components of this inflammatory response exacerbate injury, other components mediate repair. In either case, inflammation is a major determinant of the ultimate outcome of ischemic AKI. A complete understanding of the inflammatory response to IRI—the composition and functions of the leukocytes, the stimuli that elicit inflammation, and the regulation of inflammation—remain a fundamental unsolved problem in renal disease. The goal of this chapter to provide a perspective on our rapidly growing insights into the inflammatory response to IRI.

Introduction

Acute ischemia reperfusion injury (IRI) is a major cause of acute kidney injury (AKI). A major conceptual insight is that the total amount of renal injury is the result of not only the initial ischemic insult, by the ensuing responses. This simple insight has profound clinical implications: it suggests that therapy after the initial insult has already occurred may ameliorate the course of AKI (see Figure 88.1 ). Some of these responses are maladaptive. Examples include inappropriate intrarenal hemodynamics, altered mitochondrial and other metabolic functions, endothelial dysfunction, and tubular obstruction and back-leak. Other responses are reparative. Examples include the production of growth factors such as Wnt7.

In addition to the above, IRI elicits an inflammatory response. Some components of this inflammatory response exacerbate injury, other components mediate repair. In either case, inflammation is a major determinant of the ultimate outcome of ischemic AKI. A complete understanding of the inflammatory response to IRI—the composition and functions of the leukocytes, the stimuli that elicit inflammation, and the regulation of inflammation—remain a fundamental unsolved problem in renal disease. The goal of this chapter to provide a perspective on our rapidly growing insights into the inflammatory response to IRI.

In addition to its importance for injury of the native kidney, the inflammatory response to IRI is important for transplantation. Ischemic injury to the allograft inevitably accompanies organ harvesting, the cold storage, and the warm ischemia during the surgical creation of the vascular anastomoses between allograft and recipient. In addition, for brain dead donors, the donor kidneys are damaged by the hemodynamic instability resulting from the trauma that causes brain death. The leukocytes recruited to the allograft by ischemia exacerbate rejection. This is predicted by the “danger” theory of immunology where injury elicits inflammation, the leukocytes detect non-self at the site of injury (infectious pathogens or the allograft), and then develop an immune response against the non-self. The relation of this “injury response” to rejection has previously been reviewed.

An inflammatory response is also elicited by chronic injuries in common human diseases such as progressive renal failure from diabetes and hypertension. Understanding the inflammatory response to the “single hit” model of acute ischemia may aid our understanding of the more complicated chronic diseases. The goal of this chapter is to review our current understanding of this rapidly evolving field. Although complement and gene activation by hypoxia/reactive oxygen species are important, our focus will be on the nature of the inflammation and its regulation by injured and dying cells.

Most studies of the inflammatory response to renal ischemia have involved rodents. The potential difficulties in extrapolating such studies to the clinical setting have been discussed previously. Although inflammation is not a prominent feature of human ischemic acute renal failure, leukocytes are present. The susceptibility of patients to acute renal failure may correlate with polymorphisms of pro- and anti- inflammatory genes; and this further supports a role of inflammation in the pathogenesis of human disease. Biopsy studies during ischemic acute renal failure of native human kidneys are limited, but post-anastomosis biopsies of renal allografts are increasingly common. Most such biopsies indicate inflammation, particularly in deceased, compared to living, donors. Furthermore, intraoperative biopsies have also indicated expression of pro-inflammatory genes. Such inflammation may be a response to ischemic injury to the allograft due to the hypotension associated with the trauma that caused brain death, due to the cold storage, and due to the warm ischemia during creation of the vascular anastomoses. In addition, inflammation in the cadaver kidneys was also caused by neurohormonal effects of brain death. Inflammation during these intraoperative biopsies are not due to rejection because there is no time for immune recognition of the transplant. Furthermore, biopsies of kidneys between identical twins, where there is no allo-recognition, also shows inflammation that must be due to ischemic injury occurring during the transplant process.

Some argued that injury and damage to kidney is not the major contributor to most cases of human ischemic AKI. Instead the major factor is abnormal renal microvascular function and use of oxygen. Indeed, in those rare cases where the kidney is biopsied, the morphologic injury by conventional staining techniques are minimal. However, the longterm maladaptive effects of ischemic AKI suggests that injury does occur and the kidney never completely recovers from this injury. Indeed, recurrent episodes of ischemic AKI may be a major contributor to the current epidemic of CKD.

The medullary thick ascending limb and the S3 straight proximal tubule in the outer medulla are the tubules most vulnerable to ischemic injury in both rodents and humans. This is the area with the greatest inflammation. Although the outer medulla is the injured earliest and after the least ischemia, longer periods of ischemia result in the injury of the cortex also. Because different structures in the kidney may be injured depending on the intensity of ischemia, and because these structures produce different cytokines, chemokines, and other regulatory molecules in response to IRI, renal IRI may be a family of diseases rather than a single entity. Thus, the length of ischemia time, or the temporature of ischemia may modify the inflammatory response to IRI.

The inflammatory response to acute ischemia of the heart, and brain have been more intensively studied than ischemic acute renal failure because of the greater clinical incidence of coronary artery disease and stroke. Where renal studies are not available, we will review studies from these and other non-renal organs. Although the general principles may be the same in these various organs, the particular mechanisms of the inflammatory response to ischemic injury may be different in different organs. For example, blocking the pro-inflammatory cytokine interleukin 1α and β ameliorates ischemic injury of the rodent brain and heart, but has no effect on ischemic acute renal failure.

Leukocytes in Injured, Ischemic Tissues: Friend, and Foe.

Over a decade ago, anti-inflammatory agents were shown to ameliorate ischemic acute renal failure. These studies demonstrated the maladaptive effects of the inflammatory response to injury. Recent studies elucidate greater detail about which leukocytes are involved and how they regulate renal IRI.

Mononuclear phagocytes—monocytes, macrophages and dendritic cells : The relationship of the various members of the mononuclear phagocyte family—monoctye, macrophages, and dendritics, and there various subsets—is complex. Various mononuclear phagocytes participate in ischemic AKI. Some exacerbate injury. Others facilitate repair.

These leukocytes exacerbate the early phases of ischemic injury. Macrophages appear within hours after ischemic injury in both mice and rats; these macrophages are located adjacent to the vasa rectae of the outer medulla. This is the region of the rodent kidney that is most vulnerable to ischemic injury, and where there is endothelial injury and expression of both B7 and adhesion molecules.

Elimination of this early macrophage infiltrate prevented the increased interleukin 6 that occurs after renal ischemia. The former exacerbate ischemic renal injury because its elimination by transgenic knockout, or anti-interleukin-6 antibodies ameliorates renal injury. In situ hybridization shows that interleukin 6 is produced by macrophages in the ischemic kidneys; the construction of bone marrow chimeras where renal parenchymal cells or macrophages have their IL6 gene knocked out showed that the greatest injury occurred when macrophages were capable of making interleukin 6. Macrophages are also capable of producing a number of other molecules that might exacerbate ischemic acute renal failure. However, as discussed below, what cytokines are produced by which renal cells, and which cytokines have harmful versus helpful effects remain to be clearly delineated.

In these studies, there was an early infiltration of macrophages in the absence of neutrophils. This sequence of macrophages then neutrophils contradicts the classical paradigm which proposes that neutrophils infiltrate first, and produce molecules that recruit monocytes subsequently. However, recent data indicate that monocytes can infiltrate tissues early and, in some cases, in the absence of neutrophils. In the lung and liver, macrophage inflammation may precede neutrophilic inflammation. This may also occur in ischemic acute renal failure. Furthermore, the nature of renal cell death during ischemic AKI may regulate the type of inflammation. Apoptosis generally recruit macrophages, but necrosis recruits neutrophils. We discuss the various types of cell death and their effects in inflammation later in this chapter.

In addition to this early infiltrate, there is also a late infiltrate of macrophages and related dendritic cells during the first weeks after acute ischemic injury. Large numbers of these leukocytes are still present after the recovery of glomerular filtration has already occurred. The contribution of these leukocytes to renal injury and repair is not known. On the one hand, they may contribute to chronic injury. On the other hand, some macrophage subpopulations participate in tissue repair, perhaps through the secretion of growth factors such as Wnt7b or anti-inflammatory cytokines such as interleukin 10. These macrophages have many attributes of “M2” macrophages present late after infections.

Macrophage infiltration into the outer medulla is regulated by endothelia. Endothelia are the border between the vasculature and the renal interstitium. Thus, the quantity and composition of leukocyte traffic from blood into the renal interstitial spaces is regulated by proinflammatory genes expressed by endothelia. Ischemic endothelia in the outer medulla do increase their expression of pro-inflammatory ICAM-1 (CD54) and B7. 1 (CD80). In addition, endothelial expression of P-selectin (CD62P) and VCAM-1 (CD106) also contribute to the inflammatory response to renal ischemia, but the precise anatomical location is not known. Inactivation of ICAM-1 and selectins via antibodies, antisense oligonucleotides, or transgenic knockout ameliorates inflammation and injury after acute ischemia.

Furthermore, macrophage infiltration into the ischemic kidney is regulated by MCP 1, a chemokine that attracts macrophages are expressed by the ischemic kidneys. Transgenic knockout of the receptor for MCP 1 (CCR2) or administration of a truncated, inhibitory form of MCP 1 both ameliorate ischemic injury and inflammation.

In addition, macrophages may be recruited by molecules released by necrotic or apoptotic cells. This will be discussed in later sections of this chapter. Finally, blocking B7 on ischemic endothelia decreases macrophage infiltration and ischemic renal injury. CD28, the ligand for B7, is not known to be expressed by macrophages, but is expressed by T cells. This suggests a role for T cells in ischemic acute renal failure (see below).

Neutrophils : In contrast to the early infiltration of macrophages, some studies report that there is a later infiltrate of neutrophils. The role of these neutrophils is not clear. Early reports suggested that elimination of these neutrophils with antibodies ameliorated ischemic injury. But this may have reflected the use of polyclonal antibodies that actually recognized both neutrophils and macrophages. Recent data using monoclonal antibodies for neutrophils are controversial. Some, but not others, find that deleting neutrophils ameliorated ischemic injury. One difficulty in these studies is the use of the monoclonal antibody for Ly6C/G (Gr1). Although Ly6C/G is highly expressed on neutrophils, it is also expressed, albeit weakly, on some subsets of monocyte/ macrophages.

Renal parenchymal cells produce the neutrophil chemokines KC and MIP 2, the murine analogs for human interleukin 8, as well G-CSF that would stimulate neutrophil production by the bone marrow. In one study, antibody to the neutrophil chemokines KC and MIP 2 decreased neutrophilic infiltration and also ameliorated ischemic injury. However, antibody to the receptor for these chemokines CXCR2 unexpectedly exacerbated injury. These results need to be reconciled. One possibility is that these chemokines have both maladaptive and adaptive functions; in addition to regulating neutrophilic inflammation, KC and MIP 2 may also regulate the differentiation of renal tubular cells during the repair process after injury.

Those advocating a role for neutrophils point out that, in addition to releasing toxic molecules that might injure the kidney, neutrophils are now known to produce cytokines, chemokines, and other regulatory molecules. By producing these molecules, neutrophils may regulate any subsequent inflammation and repair. Those, who find no role for neutrophils in renal injury, might argue that the presence of neutrophils in tissue does not necessary indicate that these neutrophils are activated. Thus, extopic gene expression of the neutrophil chemokine KC results in a neutrophilic infiltrate but no tissue damage, presumably because the neutrophils are not activated to produce toxic molecules. This neutrophil infiltration may be regulated by NK T cells.

Lymphocytes : Small numbers of T cells are found in kidneys after renal ischemia. The role of these T cells is not understood. On the one hand, elimination of T cells via the foxn1 mutation (nude mice) ameliorated injury. FTY720, an immunosuppressive drug that traps lymphocytes in lymph nodes, inhibits ischemic acute renal injury. But, on the other hand, elimination of all classical T cells via mutation of the TCRα chain or via mutation of the rag gene (scid mice) did not inhibit. There is further controversy when elimination of specific subsets of T cells was examined. On the one hand, eliminating CD4 T cells with monoclonal antibodies did not ameliorate injury, but elimination with transgenic knockout did. Similar controversy surrounds the role of CD8 T cells. On the one hand, transgenic knockout of CD8 did not ameliorate ischemic injury; however, anti-CD8 antibodies used in combination with anti-CD4 antibodies did. B-lymphocytes may also contribute to ischemic AKI. This is discussed further in the section on complement and “natural” antibodies.

The role of lymphocytes in ischemic injury is further complicated by observations suggesting that T cells ameliorate injury in some models. These studies fall into three groups: those involving CD4 T cells, those involving interferon gamma, and those involving “unconventional” γδ T cells.

Some CD4 T cells may ameliorate injury. CD4 knockout results in decreased HGF production and increased tubular apoptosis after ureteral obstruction. Nude mice with no classical T cells have increased injury after optic nerve injury, and injection of such T cells improves repair. CD4 T cells play a dual role of exacerbating and inhibiting inflammation after ischemic hepatic injury.

Interferon gamma may play a dual function. Interferon gamma is a cytokine associated with ischemic injury, and produced in quantity by T cells. Interferon gamma exacerbates ischemic acute renal failure in some models. However, there are interferon gamma dependent pathways of tissue repair. Whether or not there are such pathways in the kidney is not known.

Recent data suggest that unconventional T cells recognize injured tissues. Thus, some T cells with γδ T cell receptors recognized stressed epithelia and release keratinocyte growth factors that facilitate repair. Whether or not such T cells are present in the ischemic kidney is not known. NK T cells also participate in ischemic AKI and some NKT cell regulate inflammation during AKI.

The Proinflammatory Effects of Injury—Damps, Sterile Inflammation, and the “Danger/ Damage” Hypothesis

Although the above shows that the inflammatory response to ischemia may an important determinant of the extent of injury and repair, how ischemic injury is translated into inflammation is a major outstanding question. Approaches to addressing this question are reviewed in the remainder of this chapter. Several concepts and terms need to be defined before beginning our discussion. A major recent concept is that injury itself is proinflammatory. The proinflammatory signals are generated in several ways. Molecules, normally residing within cells, elicit inflammation when they are released into the extracellular space or are expressed on cell surfaces. In addition, enzymes released by injured cells or leukocytes convert extracellular matrix molecules into proinflammatory signals. Finally, intracellular stress may generate proinflammatory signals. Altogether these proinflammatory molecules have been called “danger (or damage)-associated molecular pattern” molecules (DAMPS) or Alarmins. This inflammatory response to injury has been called the “danger” response or “sterile” inflammation to differentiate it from the inflammatory response to infections.

The biology of DAMPS, their receptors, and their regulation of injury, of the inflammatory response to injury, and of repair remain to be completely elucidated. The biology is complex because the receptors are promiscuous and interact with numerous DAMPS, because each DAMP interacts with multiple receptors, because DAMPS interact with each other, and because the cellular response to a DAMP is unique to the cell and its microenvironment. We review our current understanding of this rapidly evolving field with the humble realization that this understanding will certainly change in the future.

TLR4 and HMGB1

During ischemic AKI, the best characterized receptor for DAMPS is TLR4. TLR4 is best known as the receptor for endotoxin produced by gram negative bacteria, and mediates the inflammatory response against these bacteria. It is one member of a family of “pattern recognition receptors” that recognize molecules produced by pathogens. In addition to endotoxin, these molecules include viral DNA, viral RNA, and sugar molecules unique to yeast. The TLR’s are present on most leukocytes of the “innate” immune response—macrophages and neutrophils. They are critical for the immediate inflammatory response against infections.

A major discovery was the insight that TLR4 not only recognizes endotoxin, but also recognizes DAMPS. These molecules are called “endogenous” because they are produced by mammalian cells and to differentiate them from endotoxin, the “exogenous” TLR4 ligand that is produced by gram negative bacteria.

Striking confirmation for the importance of TLR4 in ischemic disease were experiments comparing inflammation and injury in wildtype mice versus TLR4 deficient mutant mice after ischemia to the heart, liver, lung, and kidney. In all of these studies, mutant mice with non-functional TLR4 are protected from ischemic injury. Furthermore, human mutations that inactivate TLR4 ameliorate post-transplant ischemic AKI.

HMGB1 is released from injured cells and is the best documented DAMP ligand for TLR4 during ischemic AKI. The HMGB1-TLR4 interaction is one of the few DAMP-TLR4 interactions that have been confirmed by biophysical studies. Furthermore, HMGB1 has a proven role in ischemic AKI. Its expression increases in both murine ischemic AKI, and biopsies of human renal transplant grafts that had ischemic AKI during the transplant process. Antibodies that inactivate HMGB1 ameliorate murine ischemic AKI. Altogether this data show HMGB1 has maladaptive functions during ischemic AKI.

Within four hours reperfusion, reactive oxygen species produced during ischemia reperfusion increase endothelial expression of TLR4. HMGB1 released from injured renal tubulesbind the endothelial TLR4, and cause increased adhesion molecule expression. These adhesion molecules facilitate the immigration of leukocytes into the injured outer medulla. HMGB1 binds to TLR4 on these leukocytes, including macrophages, and stimulates the production of maladaptive interleukin 6. In addition, the HMGB1 also stimulates TLR4 on tubules and stimulates the production of maladaptive chemokines and cytokines. See review and Figure 88.2 .