ReviewEndotoxin and Mammalian Host Responses During Experimental Disease
Introduction
The term endotoxin, credited to Pfeiffer 1904 (Westphal, 1984) was introduced to distinguish heat-stable toxin released during bacterial lysis from secreted, heat-labile bacterial exotoxins. In pure chemical form endotoxin is a lipopolysaccharide (LPS) (Rietschel et al., 1990), which has a widely variable composition (or chemotype) and exerts a wide range of often opposing influences on host–pathogen interactions. It is an integral component of the outer membrane of Gram-negative bacteria, essential for viability (Poxton and Edmond, 1995); it should be mentioned, however, that a slow-growing mutant of Neisseria meningitidis lacking any detectable LPS was reported by Steeghs et al. (1998). LPS is also a major bacterial virulence factor (Burns and Hull, 1998; Pollack, 1999) and a prime example of highly conserved bacterial molecular patterns, including lipoteichoic acids and flagellin, that occur primarily on the surface of pathogens (Pugin et al., 1994; Krug et al., 2001; Seya et al., 2001) and engage with the innate immune system via pattern recognition receptors (Janssens and Beyaert, 2003). Variation in LPS structure can also contribute to the avoidance of immune detection through molecular mimicry of host structures. Examples of this include mimicry of gangliosides by Campylobacter jejuni and of Lewis antigens by Helicobacter pylori (Moran and Prendergast, 2001). Pattern recognition receptors (PRRs) of the mammalian innate immune system include Toll-like receptors (TLRs) (Beutler and Poltorak, 2000; Bowie and O’Neill, 2000; Brightbill and Modlin, 2000; Means et al., 2000; Modlin, 2002), a novel group of transmembrane receptors that are the mammalian equivalent of Toll receptors described originally for the fruit fly (Rock et al., 1998). Bacterial molecular patterns have been considered mainly in the context of recognized pathogens—hence the common collective term “pathogen-associated molecular patterns” (PAMPs); however, they also occur in commensal bacteria, in which they have been termed CAMPs (Cario et al., 2002). Because the boundaries between pathogenic organisms and those considered harmless become indistinct in the light of disease caused by environmental bacteria (see below), the more general term microbial-associated molecular patterns (MAMPs) will be used in this review. At least 10 mammalian TLRs have been identified to date, including a receptor for LPS (TLR4), and an understanding of the associated binding and signalling processes may lead to better means of preventing or controlling bacterial disease and avoiding autoimmune type diseases. This review, after a brief summary of the well-documented structure and overall biological effects of endotoxin, will focus on (1) the consequences of experimental exposure of a range of hosts to endotoxin, (2) the pathological outcomes, and (3) recent developments in our understanding of the mechanisms.
Section snippets
Composition, Structure and Function of Endotoxin
The various chemotypes of LPS form a collection of heterogeneous glycoconjugates (Nowotny, 1990) consisting of O-antigen, outer and inner core polysaccharides, and covalently bound lipid (“lipid A”) (Rietschel et al., 1990). There are many authoritative reviews of the composition and structure of endotoxin (Rietschel et al., 1994, Rietschel et al., 1993, Rietschel et al., 1991; Raetz, 1993; Seydel et al., 1993; Moran, 1995; Schromm et al., 1998; Morrison et al., 1999; El Samalouti et al., 2000;
Mechanisms of Host–Endotoxin Interaction
In addition to its interaction with TLRs (see below), endotoxin interacts with the host via a range of receptors and intermediates, including soluble ligands such as high-density lipoprotein, surface-bound and soluble CD14, LBP (Kitchens, 2005), bactericidal/permeability-increasing protein (BPI), antimicrobial peptides (Werling et al., 1996; Vreugdenhil et al., 1999; Wu et al., 2004; Hari-Dass et al., 2005) and the leucocyte integrin CD18 (Paape et al., 1996).
The effects of exposure of host
Outcome of Exposure of the Host to Endotoxin
Reactions to endotoxin administered in crude or purified form have been well-described, although not necessarily backed up by serum analyses to enable endotoxin concentrations to be correlated with observed effects. There are many reports of the effects of LPS on rodents but few (Khan et al., 1998) of naturally occurring LPS-related disease in these animals.
Administration of endotoxin in vivo has been used commonly in the study of cellular and molecular host responses, to increase understanding
Measurement of Endotoxin in Biological Fluids
A common method of endotoxin measurement is based on the initiation of the coagulation cascade in LAL by LPS (Tanaka and Iwanaga, 1993). However, it is notoriously difficult to obtain a reliable and repeatable estimate of endotoxin in plasma or serum samples; this is because the reaction of endotoxin with LAL is susceptible to either inhibition or enhancement due to contaminants such as fungal beta glucans (Hausmann et al., 2000) or factors present in blood. It was observed (Hodgson,
Opposing Effects of Endotoxin
The sequence of reactions initiated by macrophage–LPS interactions may be detrimental or beneficial by contributing to the pathogenesis of disease or to the priming of host defence mechanisms, respectively. Detrimental effects include fever, hypotension, multiple organ failure, and death (Vincent, 1996); beneficial effects include polyclonal antibody induction, adjuvancy, normal development of lymphoid organs, and macrophage activation (Gollahon et al., 1983; Vogel and Hogan, 1990). Much
Endotoxin-related Diseases
Endotoxin is a key aetiological agent in a range of important animal and human diseases, including systemic colibacillosis in sheep and cattle (Hodgson, 1994), salmonellosis in a range of animals (Fierer and Guiney, 2001), systemic pasteurellosis in sheep caused by Pasteurella trehalosi (Hodgson et al., 1993) and septic shock in man (Barron, 1993). Additionally, endotoxin is implicated in bowel oedema in pigs (Beers-Schreurs et al., 1992) and Reye's syndrome in children (Cooperstock et al., 1975
Other Models of Shock and What we have Learnt of Mechanisms
The range of animal models used to study sepsis highlights the difficulty of devising an acceptable experimental approach. Ideally, a model should be (1) characterized by biochemical and physiological changes similar to those in the natural disease, (2) reproducible, (3) as humane as possible, and (4) an accurate predictor of the results of clinical trials of new therapeutic agents. Associated requirements for such a model include use of bacterial phenotypes that evade host defences and doses
Variation in Sensitivity to LPS Shown at Different Stages of Physiological Development and by Different Animal Species
Pulmonary injury is induced by minute amounts (a few micrograms or nanograms) of bacterial endotoxin in sheep, calves, pigs and cats, but not in laboratory animals or dogs (Winkler, 1988). The relatively low sensitivity of mice to LPS varies with their genetic make-up; thus, for example, C3H/HeJ mice, in which the gene controlling LPS responsiveness is defective, are resistant to the effects of endotoxin (Heumann et al., 1996). Hormonal responses, for example growth hormone response, appear to
Concluding Remarks
Just as in the past the discovery of the role of cytokines, chemokines and eicosanoids as mediators of the effects of LPS re-focussed endotoxin research, so now does the current interest in TLRs. The demonstration of different TLRs for different PAMPs (Krug et al., 2001) should enable us to target specific pathogens or pathogenic effects and avoid “blunderbuss” approaches with the attendant danger of side-effects or indiscriminate damage to host tissues and processes. Ultimately, the concept of
Acknowledgments
Thanks are due to Ms H Simm for help with Fig. 3, Fig. 4, Fig. 5a–c and to Drs K. Angus, D. Buxton and M. P. Dagleish for help and advice with Fig. 5. Research on this topic is supported by the Scottish Executive Environment Rural Affairs Department.
References (232)
- et al.
Release of tumor necrosis factor-alpha from bovine alveolar macrophages stimulated with bovine respiratory viruses and bacterial endotoxins
Veterinary Immunology and Immunopathology
(1992) - et al.
Commensal-associated molecular patterns induce selective toll-like receptor-trafficking from apical membrane to cytoplasmic compartments in polarized intestinal epithelium
American Journal of Pathology
(2002) - et al.
Response of newborn and adult sheep to pyrogens: relation between fever and brain eicosanoid changes
Brain Research
(1995) - et al.
Differential regulation of human blood monocyte and alveolar macrophage inflammatory cytokine production by nitric oxide
Annals of Allergy, Asthma and Immunology
(1999) - et al.
Immunity of specific pathogen-free lambs to challenge with an aerosol of Pasteurella haemolytica biotype A serotype 2. Pulmonary antibody and cell responses to primary and secondary infections
Veterinary Immunology and Immunopathology
(1986) - et al.
Experimental induction of pneumonic pasteurellosis in calves by intratracheal infection with
Pasteurella multocida biotype A:3 Research in Veterinary Science
(2002) - et al.
Effects of smooth and rough Pasteurella haemolytica lipopolysaccharides on plasma cyclic-nucleotides and free fatty acids in calves
Veterinary Microbiology
(1987) - et al.
Structure and function of lipopolysaccharides
Microbes and Infection
(2002) - et al.
Laboratory models of sepsis and septic shock
Journal of Surgical Research
(1990) - et al.
The genetic bases for the variation in the lipo-oligosaccharide of the mucosal pathogen, Campylobacter jejuni. Biosynthesis of sialylated ganglioside mimics in the core oligosaccharide
Journal of Biological Chemistry
(2002)