News and Publications

Molecular Mechanisms of Host-Pathogen Interactions

R.J. Ulevitch, P.S. Tobias, J.D. Lee, V.V. Kravchenko, J.C. Mathison, J. Bohuslav, C. Fearns, W. Guo, Y. Kato, J. Lee, L. Mira-Arbibe, S. Orr, G. Sanna, D. Schiff, J. Schimke, R. Tapping , C. Werts, O. Ducrey,* T. Kirkland,** T.R. Martin***

* University of Fribourg, Fribourg, Switzerland
** University of California, San Diego, CA
*** University of Washington, Seattle, WA

Attacks by pathogens often set in motion chains of events that cause severe injury to the host, and nowhere is this phenomenon illustrated more dramatically than in the response by humans to infection by gram-negative bacteria. In his book Lives of a Cell, Lewis Thomas characterizes the host response to the endotoxin, or lipopolysaccharide (LPS), of gram-negative bacteria as being "read by our tissues as the very worst of bad news. . . . There is nothing intrinsically poisonous about endotoxin, but it must look awful, or feel awful, when sensed by cells. Cells believe that it signifies the presence of gram-negative bacteria, and they will stop at nothing to avoid this threat."

Clearly, much human suffering could be eased if such overzealous host responses could be tempered. However, such responses, when not overzealous, are a normal part of the host's homeostatic mechanisms, designed to respond to the threat of gram-negative bacterial infection. Accordingly, we are attempting to provide a molecular description of the events that allow cells to (1) recognize LPS and (2) learn how to control these responses without compromising host defenses against pathogens. Because LPS is a prototypic activator of cells of the immune, inflammatory, and vascular systems, information derived from our studies has broad implications for all areas of biomedical research and for understanding how other types of microbial pathogens cause human disease.

A major part of our effort is to define the membrane receptor and intracellular signaling pathways used by LPS during septicemia. Through studies of LPS-binding protein (LBP) and CD14, our research has provided an entirely new basis for understanding how components of microbial pathogens initiate host responses. LBP is a plasma protein; CD14 is a protein found both on the membranes of cells of the myeloid lineage and soluble in plasma.

LPS INTERACTIONS WITH LBP AND CD14

The biochemical basis for targeting CD14 to intercede in the development of endotoxic shock is derived from our understanding of the mechanism by which CD14, together with LBP, enables cells to respond to LPS. LBP binds to LPS whether the LPS is in the outer membranes of gram-negative bacteria or free in solution. The key activity of LBP is that it promotes the interaction of LPS with CD14.

If the CD14 is present as an extracellular membrane protein of myeloid cells, LBP enhances phagocytosis of intact gram-negative bacteria and enables cellular responses to concentrations of LPS that would not otherwise initiate the activation of these inflammatory cells. If the CD14 is the soluble form, the LPS-CD14 complexes may be recognized by receptors on a wide variety of cells, including endothelial, epithelial, dendritic, and smooth muscle cells, and the recognition initiates the inflammatory responses of the affected cells. Normally, these inflammatory responses are local and result in resolution of an infection before the infection can spread to be life threatening. However, the response to LPS can also, paradoxically, become life threatening when it is uncontrolled; it spreads throughout the body and causes endotoxic shock.

Interestingly, LBP and CD14 also participate in antiinflammatory responses to LPS and gram-negative bacteria. For example, when LPS binds with LBP to CD14 on the monocyte membrane, the LPS partitions into 2 pathways, 1 that activates inflammatory responses and 1 that causes internalization of the LPS without initiating an inflammatory response. How the partitioning of LPS into these 2 pathways is controlled is not understood and is a focus of current studies. In addition to this more basic focus, we have obtained data from studies in animal models that support our contention that blockade of LPS-CD14 interactions provides a novel target for therapy in septic shock.

Using a rabbit model of endotoxic shock, we showed that a monoclonal antibody to rabbit CD14 blocks LPS-CD14 binding and protects against organ injury and death even when the antibody is administered after initial exposure to LPS. These results support the concept that treatment with antibodies to CD14 provides a new therapeutic window for the prevention of pathophysiologic changes caused by cumulative exposures to LPS and septic shock in humans.

In similar studies, we found that prophylactic or delayed treatment with a monoclonal antibody to rabbit TNF did not provide the degree of protection observed with the antibody to rabbit CD14. Prophylactic treatment with the antibody to TNF neutralized TNF and conferred a survival benefit but gave only partial protection against renal cortical damage and pulmonary injury. Delayed treatment with the antibody at 4 hours after exposure to LPS did not have any significant survival benefit and did not completely block the renal cortical necrosis and pulmonary injury.

The severe physiologic injury and organ failure that often accompany gram-negative sepsis are most likely associated with repeated exposure to small amounts of LPS. Recognition of LPS by the innate immune system results in the release of a multiplicity of mediators. Although some mediators appear to play a more prominent role than do others in producing injury, disappointing results from numerous clinical trials have shown that therapeutic agents that target single mediators such as TNF do not protect subjects from LPS-mediated injury. Any benefit provided by such therapies is usually confined to treatments given either prophylactically or in the early stage of sepsis. Unfortunately, treatment of sepsis in patients with signs and symptoms predictive of shock and multiorgan failure often does not occur until some pathophysiologic manifestations of sepsis are apparent. In this regard, results with our model suggest that a therapeutic window for treatment may exist, because we have shown clear benefit with respect to organ injury and survival in animals given delayed treatment with antibodies to CD14.

REGULATION OF GENE EXPRESSION DURING ENDOTOXEMIA

Recent studies at the molecular level have provided new insights into the mechanisms of innate immunity that are operative during infection. Recognition of bacterial LPS causes multiple host responses, including activation of cells of the innate immune system. One consequence of this activation is new gene expression and de novo synthesis of proteins such as proinflammatory cytokines involved in the elimination of pathogens. During septicemia, exposure to LPS occurs repeatedly, making strict regulation of gene expression necessary. Tolerance to LPS is characterized by diminished production of TNF during prolonged exposure to LPS and is an essential feature in maintaining physiologic control during sepsis.

In studies with mice with genetic deletions of the proteins of the NF-B complex, we showed that increased expression of the p50 subunit of the complex provides a signal involved in the downregulation of the TNF response to LPS. This conclusion is based on the following observations: (1) Tolerance to LPS is not induced in macrophages from p50^--/-- mice. (2) Ectopic overexpression of p50 reduces transcriptional activation of the murine TNF promoter. (3) Analysis of the 4 B sites from the murine TNF promoter during different stages of LPS responsiveness indicated that binding of p50 homodimers to the positively acting B3 element of the mouse TNF promoter is associated with development of the LPS-tolerant phenotype.

CELLULAR ACTIVATION PATHWAYS INVOLVING PROTEIN KINASES

Transfer of information from the extracellular environment to the nucleus often proceeds through intracellular kinase cascades initiated by binding of a ligand to the ligand's membrane receptor. Members of the mitogen-activated protein (MAP) kinase family of serine/threonine kinases play a central role in many cellular responses to extracellular stimuli. These kinases are activated by phosphorylation of adjacent threonine and tyrosine residues. The sequence motif threonine-Xaa-tyrosine defines subgroups of MAP kinases; the Xaa may be glutamate, glycine, or proline.

We discovered 1 member of this family, p38, and have detected 3 additional isoforms of this enzyme. These isoforms have distinct biochemical and biological properties. A variety of experimental studies link activation of the p38 enzyme group to transcriptional and posttranscriptional events associated with production of important mediators of inflammation, and this group of enzymes is a potentially important target for anti-inflammatory drugs.

The events that couple activation of MAP kinases caused by reactions at the cell surface require the participation of intracellular kinase cascades. One group of sequentially activated protein kinases essential for regulation of many important cellular genes regulates a family of transcription factors known as NF-B. Activation of NF-B requires the phosphorylation of an inhibitory protein, IB, by components present in a multiprotein IB kinase complex. MAP kinase kinase kinase-1 (MEKK1) is thought to be important in this process.

We showed that an inducible zinc finger protein, A20, associates with MEKK1 and specifically affects NF-B activity by inhibiting diverse signaling pathways that converge at the level of MEKK1. These data underscore the importance of MEKK1 in NF-B activation and together with previous studies provide evidence that A20 participates in signal-dependent NF-B autoregulation. We noted a correlation of the kinetics of LPS- or TNF-induced expression of mRNA for IB and A20 in different cell lines and in primary macrophages. In peritoneal macrophages, induction of IB and A20 was first detected 15--30 minutes after exposure to LPS, peaked at 1--1.5 hours, and was barely detectable after 2--4 hours. These data are consistent with TNF- or IL-1--mediated induction of A20 in endothelial cells and with the kinetics of TNF or IL-1 stimulation of the IB kinases in HeLa cells. Expression of A20 may inhibit its own transcription, as does expression of Ia.

Our data suggest that A20 and IB serve as keys in the autoregulation of NF-B. We propose a model in which NF-B--inducible expression of IB inhibits translocation of NF-B proteins into the nucleus, and at the same time, expression of A20 attenuates the kinase cascades associated with the MEKK1 pathways. Such stringent control of NF-B activation ensures the rapid but transient pattern of its action, particularly in the setting of repeated stimulation such as infection with microbial pathogens. Finally, the data suggest new approaches for the control of multiple signal transduction pathways leading to the activation of NF-B.

PROTEIN KINASES IN THE REGULATION OF CELL PROLIFERATION AND THE CELL CYCLE

Intracellular kinase cascades also play an essential role in regulating cell growth and thus are important in human diseases that involve dysregulation of cellular growth. Epidermal growth factor (EGF) induces proliferation of a variety of cell types through a prototypic transmembrane tyrosine kinase receptor. Ligation of this receptor by EGF is known to activate ERK1 and ERK2, members of the MAP kinase family, through a Ras- and Raf-dependent signal transduction pathway. Despite a detailed understanding of these events, the complete mechanism by which EGF signals cells to proliferate is not fully understood.

Big MAP kinase 1 (BMK1), also known as ERK5, is a recently identified member of the MAP kinase family that is activated in cells in response to oxidative stress, hyperosmolarity, and treatment with serum. We showed that EGF is a potent activator of BMK1 and that this signaling event requires the tyrosine kinase activity of the EGF receptor. However, in contrast to findings with ERK1/2, EGF-mediated activation of BMK1 occurs independent of Ras and requires the MAP kinase kinase MEK5. Expression of a dominant negative form of BMK1 blocked EGF-induced cell proliferation and prevented cells from progressing through the G₁ phase of the cell cycle. These results indicate that BMK1 is part of a distinct MAP kinase signaling pathway required for EGF-induced cell proliferation and cell-cycle progression.

PUBLICATIONS

Dziarski, R., Tapping, R.I., Tobias, P.S. Binding of bacterial peptidoglycan to CD14. J. Biol. Chem. 273:8680, 1998.

Elass-Rochard, E., Legrand, D., Salmon, V., Roseanu, A., Trif, M., Tobias, P.S., Mazurier, J., Spik, G. Lactoferrin inhibits the endotoxin interaction with CD14 by competition with the lipopolysaccharide-binding protein. Infect. Immun. 66:486, 1998.

Fearns, C., Ulevitch, R.J. Effect of recombinant interleukin-1ß on murine CD14 gene expression in vivo. Shock 9:157, 1998.

Kato, Y., Kravchenko, V.V., Tapping, R.T., Han, J., Ulevitch, R.J., Lee, J.-D. BMK1/ERK5 regulates serum-induced early gene expression through transcription factor MEFC. EMBO J. 16:7054, 1997.

Pugin, J., Kravchenko, V.V., Lee, J.-D., Kline, L., Ulevitch, R.J., Tobias P. Cell activation mediated by glycosylphosphatidylinositol-anchored or transmembrane forms of CD14. Infect. Immun. 66:1174, 1998.

Sanna, G., Duckett, C.S., Richter, B.W.M., Thompson, C.B., Ulevitch, R.J. Selective activation of JNK1 is necessary for the anti-apoptotic activity of hILP. Proc. Natl. Acad. Sci. U.S.A. 95:6015, 1998.

Schiff, D.E., Kline, L., Soldau, K., Lee, J.-D., Pugin, J., Tobias, P.S., Ulevitch, R.J. Phagocytosis of gram-negative bacteria by a unique CD14-dependent mechanism. J. Leukoc. Biol. 62:786, 1997.

Sellati, T.J., Bouis, D.A., Kitchens, R.L., Darveau, R.P., Pugin, J., Ulevitch, R.J., Gangloff, S.C., Goyert, S.M., Norgard, M.V., Radolf, J.D. Treponema pallidum and Borrelia burgdorferi lipoproteins and synthetic lipopeptides activate monocytic cells via a CD14-dependent pathway distinct from that used by lipopolysaccharide. J. Immunol. 160:5455, 1998.