Understanding the Pathogenesis of Inflammation Using Rodent Models Identification of a Transcription Factor (NF B) Necessary for Development of Inflammatory Injury

Alex B. Lentsch, Ph.D., is Assistant Professor in the Department of Surgery, University of Louisville School of Medicine, Louisville, Kentucky. Peter A. Ward, M.D., is Professor and Chairman of the Department of Pathology, The University of Michigan Medical School, Ann Arbor, Michigan. Alex B. Lentsch, Ph.D., is Assistant Professor in the Department of Surgery, University of Louisville School of Medicine, Louisville, Kentucky. Peter A. Ward, M.D., is Professor and Chairman of the Department of Pathology, The University of Michigan Medical School, Ann Arbor, Michigan. Acute inflammation is a necessary response to tissue injury, designed to maintain homeostasis by returning the tissue to its preinjury state. In general, this response can be characterized as a cascade of events that result in complex, yet coordinated, interactions between blood leukocytes, blood vessels, and cells of the tissue(s) involved. These events are directed toward removal of injurious agents and restoration of normal tissue structure and/or function. However, dysregulated inflammatory processes cause many human diseases. Thus, successful development of therapeutic strategies to suppress undesirable inflammatory responses depends on two factors: (1) knowledge of the steps leading to activation of the inflammatory response and (2) an understanding of the regulatory mechanisms that serve to control the progression and/or resolution of the inflammatory response. Much has been learned about these two factors using rodent models of inflammation. Lung inflammatory injury induced in rats by distal airway deposition of immunoglobulin G (IgG 1 ) immune complexes has been used for many years for the study of lung inflammation ( Johnson and Ward 1974 ). The inflammatory pathways in this particular model are rather similar to events related to lung injury caused by ischemia ( Caty and others 1990 ), by the presence of bacteria ( Lechner and others 1993 ), or by bacterial lipopolysaccharide ( Simons and others 1996 ) and therefore represent a model relevant to human disease. In rats and mice, hepatic ischemia and reperfusion results in both local and remote organ inflammation ( Colletti and others 1990a ; Jaeschke and others 1990 ; Lentsch and others 1998a ). These models are clinically relevant because liver resectional surgery ( Huguet and others 1994 ), liver transplantation ( Lemasters and Thurman 1997 ), and hemorrhagic shock with fluid resuscitation ( Vedder and others 1989 ) are all complicated by inflammatory organ injury stemming from ischemic insult to the liver. The inflammatory responses in lung and liver models share many similarities ( Figure 1 ). The inflammatory insult causes complement activation as well as activation of tissue macrophages ( Jaeschke and others 1993 ; Ward 1996 ). Activated macrophages generate the “early response” cytokines, tumor necrosis factor (TNF 1 )-α and interleukin (IL 1 )-1 ( Colletti and others 1990b ; Mulligan and Ward 1992 ). These cytokines stimulate vascular endothelial cells to express adhesion molecules (ICAM-1, E-selectin), which facilitate adhesion of blood neutrophils to the endothelium ( Springer 1990 ). TNFα and IL-1 also stimulate the production of neutrophil chemoattractants that are the chemokines of the IL-8 family of cytokines, from vascular endothelial cells and other tissue parenchymal cells ( Schall and Bacon 1994 ). Vascular adhesion molecules and chemokines work in concert to bring about neutrophil transmigration from blood vessels to the tissue interstitium ( Springer 1994 ). In organs such as lung and liver, the accumulation of neutrophils together with activated macrophages results in tissue injury mediated by the generation and release of oxidants and proteases ( Jaeschke and others 1996 ; Varani and others 1989 ). Common pathway of acute inflammatory tissue injury. TNFα, tumor necrosis factor-α; IL, interleukin. Recently, a considerable amount of research has been aimed at the upstream, molecular mechanisms that regulate gene expression of proinflammatory mediators. Using monocyte and macrophage cell lines in vitro, the transcription factor nuclear factor kappa B (NFκB 1 ) has been shown to be a major regulator of many functionally diverse proinflammatory mediators. NFκB is a general term used to describe a number of dimeric combinations of members of the Rel family of gene regulatory proteins that possess transcriptional activating properties ( Ghosh and others 1998 ). The most common form of NFκB consists of a heterodimer of p50 (NFκB1) and p65 (RelA) proteins. This complex has the ability to bind with promoter sequences in DNA and to inaugurate transcription (generation of mRNA) for many proinflammatory mediators. However, other combinations of Rel family members have been identified, and different configurations of Rel proteins (such as p65/p50 and p65/p52) may have preferential sensitivities to different target promoter sequences ( Perkins and others 1992 ). In unstimulated cells, NFκB is retained in the cytoplasm through interactions with inhibitory proteins of the inhibitory factor kappa B (IκB 1 ) family. To date, at least seven IκB proteins have been identified in vertebrates ( Ghosh and others 1998 ). All IκB proteins contain ankyrin repeat domains, which function to facilitate protein-protein interactions. In the case of IκB/NFκB interactions, ankyrin repeat domains of IκB proteins prevent nuclear translocation of NFκB by masking nuclear localization sequences of the NFκB (hetero)dimers. In response to a wide variety of cellular stimuli, IκB proteins are dissociated from the NFκB complex and then proteolytically degraded ( Figure 2 ). This process varies for different IκB proteins but involves phosphorylation of IκB by members of the IκB-kinase (IKK) family ( DiDonato and others 1997 ). Phosphorylation targets IκB for ubiquination and degradation by the 26S proteasome. Degradation of IκB leads to “activation” of NFκB, which is defined as translocation of the NFκB complex from the cytoplasm to the nucleus. Once in the nucleus, NFκB binds specific promoter elements of DNA and induces transcription of relevant genes. The specificity of NFκB for DNA promoter segments is dependent on nucleotide base sequences recognized by NFκB. The mediators under the control of NFκB include the proinflammatory cytokines TNFα and IL-1 ( Collart and others 1990 ; Hiscott and others 1993 ), numerous chemokines ( Widmer and others 1993 ), and many vascular endothelial cell adhesion molecules ( Collins and others 1995 ). Obviously, the next logical step pursuant to the findings described above was to determine whether NFκB was involved in inflammatory responses in vivo. Mechanism of NFκB activation. IκB, inhibitory factor kappa B; NFκB, nuclear factor kappa B. As outlined in Figure 1 , the use of rodent models has allowed identification of many of the mediators involved in the development of acute inflammatory injury. The next objective was to apply the knowledge gained from in vitro studies of NFκB to determine whether this transcription factor was involved in the inflammatory response occurring in complex organ systems. Initial studies in a rat model of systemic inflammation induced by intraperitoneal injection of bacterial lipopoly saccharide demonstrated that activation of NFκB occurred in numerous tissues ( Blackwell and others 1994 ; Essani and others 1996 ; Manning and others 1995 ). In addition, these studies showed that expression of chemokines and adhesion molecule mRNA was associated with NFκB activation. Subsequent studies of more carefully controlled models of inflammation helped characterize the precise role of NFκB during inflammation of different organs. Using a rat model of lung inflammation induced by intrapulmonary deposition of IgG immune complexes, the precise time course of NFκB activation during lung injury has been documented ( Lentsch and others 1997 , 1998b ). In this model, alveolar macrophages (obtained by bronchoalveolar lavage) are activated by the inflammatory insult (that is, IgG immune complexes) and rapidly demonstrate increased nuclear translocation of NFκB. The activation of NFκB in alveolar macrophages is associated with enhanced production of the proinflammatory cytokines TNFα and IL-1 ( Mulligan and Ward 1992 ). These cytokines are known to cause upregulation of chemokines and vascular adhesion molecules within the lung ( Mulligan and others 1993 ; Shanley and others 1997 ). When either TNFα or IL-1 was neutralized using blocking antibodies, lung NFκB activation was greatly attenuated ( Lentsch and others 1998b ). Furthermore, when alveolar macrophages were depleted using liposome-encapsulated dichloromethylene diphosphonate, lung NFκB activation was virtually abolished ( Lentsch and others 1998a ). In rats depleted of alveolar macrophages, lung instillation of TNFα caused activation of NFκB in whole lung tissues. These studies suggest that during acute inflammatory lung injury, activation of NFκB in alveolar macrophages may be responsible for proinflammatory cytokine production. These proinflammatory cytokines appear to propagate the inflammatory response in lung by activating NFκB in other lung cell types, possibly resulting in the expression of chemokines and vascular adhesion molecules in a variety of cell types. The importance of NFκB activation during lung inflammation has also been demonstrated in studies employing agents that specifically inhibit the nuclear translocation of NFκB. In vitro, antioxidants prevent the phosphorylation and degradation of IκB and limit the extent of nuclear translocation of NFκB ( Ghosh and others 1998 ). In vivo administration of the antioxidant N-acetylcysteine suppressed lung NFκB activation induced either by intraperitoneal injection of lipopolysaccharide or by intrapulmonary deposition of IgG immune complexes ( Blackwell and others 1996 ; Lentsch and others 1998b ). Interestingly, it was found that another antioxidant, catalase, was incapable of inhibiting lung NFκB ( Lentsch and others 1998b ). Because Nacetylcysteine is a very small molecule (163 d) that easily diffuses across cell membranes and catalase is very large (∼240 kDa) and probably does not gain cellular entry, it appears that only oxidants generated in the cytoplasm of lung cells are involved in the activation of NFκB. The use of antioxidants in the study of NFκB is limited due to the relatively nonspecific nature of these agents. What is unclear in these studies is how the antioxidants affect NFκB activation. Furthermore, these agents also reduce tissue injury by scavenging the oxidants released from activated phagocytic cells. Thus, information from these studies regarding the role of NFκB in inflammatory injury is somewhat speculative. In contrast, other studies employing antiinflammatory cytokines known to regulate the production of TNFα and IL-1 have suggested that NFκB may be central to the acute inflammatory response. Two of the most potent antiinflammatory cytokines, IL-10 and IL-13, greatly reduce lung inflammatory injury while almost completely suppressing activation of NFκB ( Lentsch and others 1997 ). Inhibition of NFκB activation by both IL-10 and IL-13 was accomplished by preserving the cytoplasmic expression of the NFκB-inhibiting protein IκBβ. These studies not only helped identify the in vivo antiinflammatory mechanisms of IL-10 and IL-13, but they also increased our understanding of the regulation of lung inflammatory injury. Both IL-10 and IL-13 are constitutively expressed in lung, and endogenous production of these cytokines serves as a negative feedback loop of the inflammatory response, potentially limiting the progression of inflammation by inhibiting NFκB activation ( Lentsch and others 1999a ). In other studies, a serine protease inhibitor, secretory leukocyte protease inhibitor (SLPI 1 ), was shown to suppress lung inflammatory injury as well as inhibit the activation of NFκB ( Lentsch and others 1999b ). These inhibitory effects of SLPI were associated with upregulation of the NFκB-inhibiting protein IκBβ. Blockade of endogenous SLPI with antibody augmented the lung inflammatory response and enhanced activation of NFκB ( Lentsch and others 1999b ). These studies strongly suggest that endogenous IL-10, IL-13, and SLPI regulate the inflammatory response in vivo by their effects on NFκB activation. The involvement of NFκB in acute liver inflammation induced by hepatic ischemia and reperfusion in rats and mice has also been evaluated. In these models, hepatic ischemia causes activation of liver macrophages (Kupffer cells). These cells release reactive oxygen species and proinflammatory cytokines, including TNFα, which may directly injure liver parenchymal (hepatic) cells. However, enhanced production of TNFα plays a more important role in the initiation of a cascade of events leading to the later phase of liver injury, which is mediated by neutrophils ( Figure 1 ). One of the main functions of TNFα is the hepatic upregulation of adhesion molecules and neutrophil-attracting chemokines ( Colletti and others 1995 , 1998 ; Lentsch and others 1998c ). The coordinated actions of adhesion molecules and chemokines mediate the recruitment of neutrophils into the liver. Sequestered neutrophils release proteases and reactive oxygen intermediates, which directly damage hepatocytes and endothelial cells and also contribute to capillary plugging, causing hepatic hypoperfusion ( Jaeschke and others 1996 ; Vollmar and others 1996 ). Using this model, activation of NFκB in the liver was shown to occur shortly after reperfusion ( Bradham and others 1997 ; Yoshidome and others 1999 ; Zwacka and others 1998a ). Although the details of cell-specific NFκB activation have not yet been delineated in this model of inflammation, the time course of activation is consistent with upregulation of vascular cell adhesion molecules and chemokines within the liver ( Colletti and others 1998 ; Lentsch and others 1998c ). Similar to studies of lung inflammation, treatment with antioxidants reduced liver injury in association with suppressed activation of NFκB ( Zwacka and others 1998b ). In addition, investigations of IL-10 and SLPI demonstrate that these antiinflammatory mediators also reduce hepatic ischemia/reperfusion injury through effects on NFκB ( Lentsch and others 1999c ; Yoshidome and others 1999 ). These effects are of interest because unlike the lung in which NFκB activation is associated with degradation of IκBα ( Lentsch and others 1997 , 1998b ), activation of NFκB during hepatic ischemia/reperfusion occurs without measurable degradation of either IκBα or IκBβ ( Zwacka and others 1998a ). A possible explanation is that exogenously administered IL-10 or SLPI may augment production of IκB proteins as a mechanism of their inhibitory effects on NFκB. However, whether endogenous production of these mediators regulates NFκB activation and liver inflammatory injury remains to be determined. The use of rodent models of inflammation has allowed detailed investigation into some of the earliest events in the induction of the acute inflammatory response. Because transcription factors such as NFκB control gene expression of mediators at every level of the inflammatory response (proinflammatory cytokines, chemokines, adhesion molecules), knowledge gained from work done in animal models offers valuable therapeutic potential. Furthermore, these models have provided information critical to a more complete understanding of the regulation of inflammatory processes. It is very interesting that antiinflammatory mediators as diverse in function as cytokines (IL-10 and IL-13) and a protease inhibitor (SLPI) may suppress inflammatory responses through effects on a single transcription factor (NFκB). Investigations of these mediators have also been performed in vitro, and although IL-10 and DL-13 suppress NFκB activation in monocytes and macrophages, SLPI does not ( Lentsch and others 1999b ). These types of findings illustrate the necessity for the use of rodents in inflammation research and emphasize the fact that in vitro studies are often inadequate for a reflection of the in vivo response.