Interaction between resident luminal bacteria and the host: can a healthy relationship turn sour?

Abstract

INTRODUCTION The mucosal surfaces and cavities of the gastrointestinal (GI) tract in humans and animals are populated by a complex mixture of non-pathogenic microorganisms. These enteric microorganisms, often referred to as the indigenous or normal microflora, are composed of more than 400 species having spatial differences in population size and relative species predominance along the GI tract. The host has evolved various mechanisms of tolerance to these organisms which allows a peaceful and productive coexistence with resident bacterial flora while allowing it to remain responsive to enteropathogenic bacterial species. This discriminatory ability of the intestine toward its luminal microflora is a pivotal feature of efficient tolerance and homeostatic mechanisms. However, under specific conditions, including genetic susceptibility and certain environmental factors, the host may mount an inappropriate response to its own resident luminal bacteria with devastating consequences if not promptly restrained. Indeed, an emerging hypothesis proposes that human inflammatory bowel disease (IBD) is caused by a loss of tolerance toward the host's indigenous luminal content. Thus, the concept that only pathogenic bacteria initiate inflammatory host responses in the intestine should be reconsidered. Although an extensive literature details the molecular mechanisms of host responses to pathogenic bacteria and the ensuing inflammatory disorders, little is known about the intestinal mucosal responses to the non-pathogenic intestinal microflora. This review covers the state of knowledge of the host responses to the indigenous commensal microflora of the intestine, especially the responses of the intestinal epithelial cells (IEC). It also provides an overview of the potential mechanisms involved in the control of intestinal inflammation and the maintenance of homeostasis with special emphasis on the implications for IBD. PHYSIOLOGICAL AND PATHOPHYSIOLOGICAL IMPACT OF HOST/MICROBIAL INTERACTION IN THE INTESTINE Physiological Impact of Host/Microbial Interaction Savage and Dubos defined the term indigenous microflora as including both microorganisms naturally colonizing a particular habitat (autochthonous flora) and microorganisms ubiquitously present in a given environment (allochthonous flora) but unable to colonize the host except under abnormal circumstances (1–4). The normal intestinal microflora does not establish itself all at the same time and is not uniform throughout the GI tract. At various times after the host's birth, certain microorganisms colonize particular regions of the GI tract with luminal habitats that suit their specific needs. The tremendous impact of the endogenous luminal flora on the development and maturation of gut homeostasis is eloquently illustrated by comparative studies of germ-free and conventionalized animals. Studies in gnotobiotic animals have shown that experimentally colonizing the intestine of germ-free rodents with single species selected from the indigenous flora has profound impacts on the anatomical, physiological, and immunological development of the host, including effects on epithelial cell functions and the composition of the gut-associated lymphoid tissue (GALT) (5–7). It is clear that the immunological responses of conventional animals differ greatly from those of germ-free animals. The number of intraepithelial lymphocytes (IEL) is reduced in germ-free animals, particularly αβ T-cell receptor (α, βTCR)-bearing T cells, which also have reduced cytolytic activity. Lamina propria lymphocytes (LPL) are reduced in number and in reactivity to mitogens (8–10). Lymphoid aggregates such as Payer's patches are small and poorly developed in the intestine of germ-free animals compared with those in the intestine of animals harboring a normal indigenous microflora (11). The importance of the indigenous microflora in establishing mucosal lymphocyte populations has been shown in mice with severe-combined immunodeficiency (SCID) who have been reconstituted with mature thymus-derived T-cells (12). It has been found that the number and function of lymphocyte populations is greatly reduced in T-cell reconstituted SCID mice with a reduced microbial flora consisting of only two non-pathogenic clostridial species. This finding suggests that homing and/or expansion of thymus-derived T cell populations in the intestine may be best driven by a complex mixture of luminal bacterial antigens. MacDonald and Carter have shown that normal bacterial flora must be present in order for mice to mount a delayed-hypersensitivity (DTH) reaction, a finding suggesting that enteric bacteria influence peripheral T-cell function as well as local intestinal T-cell function (13). Pathological Impact of Host/Microbial Interaction. A characteristic feature of the mucosal immune system in the normal host is that that the protective cell-mediated and humoral immune responses against enteropathogenic organisms are allowed to proceed while responses to microorganisms of the indigenous flora are prevented. This complex homeostasis between acquisition of tolerance (unresponsiveness) to the indigenous microflora and protective immune responses to enteropathogens presents an intriguing immunological paradox. Under conditions of chronic intestinal inflammation such as IBD, this homeostasis appears to be disrupted (14). Ulcerative colitis (UC) and Crohn disease (CD), two discrete pathologies of IBD, are spontaneously relapsing, immunologically-mediated disorders of the GI tract (15). Microbial agents of the normal microflora are invoked in many of the current etiologic theories of these disorders (16). Previous studies have shown a greater association of luminal enteric bacteria with the intestinal epithelium in patients with IBD than in normal controls (17). Duchmann et al. have demonstrated that immunologic tolerance to the endogenous microflora exists in normal controls but does not exist in patients with active IBD (18,19). These findings are consistent with the clinical observation that there is a lack of inflammation in bypassed distal ileal or colonic segments in IBD patients after proximal diversion of the fecal stream, but a rapid reactivation of immune responsiveness and inflammation when the proximal effluent is introduced into the bypassed ileum (20–22). In addition, a mechanistic role for the enteric non-pathogenic bacterial flora in the pathogenesis of chronic mucosal inflammation has been shown in various animal models of experimental colitis (23,24). Selective intestinal colonization by defined bacterial species in germ-free rodents with experimental colitis triggers disease, although not all enteric bacteria are equal in their capacity to induce chronic mucosal inflammation. For example, reconstitution studies of gnotobiotic HLA-B27 transgenic rats (25,26) and studies in guinea pigs with carrageenen induced colites (27) implicate Bacteroides vulgatus as particularly important to the induction of colitis but have not found any pathological response after colonization with Escherichia coli. Conversely B. vulgatus appears to protect against the development of E. coli-induced experimental colitis in IL-2−/− mice (28), suggesting that different bacterial species have differing effects in initiating and perpetuating intestinal inflammation in different genetically susceptible hosts. The absence of experimental colitis and pathological immune responses to any of the bacterial species in wild-type animals demonstrates the non-pathogenic nature of normal microflora and, most importantly, suggests that the normal host develops immunosuppressive mechanisms to control the constant challenge to the immune system by antigens from commensal microorganisms. Protective Effects of Host/Microbial Interaction. It appears that bacteria of the endogenous microflora not only affect the induction of immune maturation and defense mechanisms of the naïve gut but also induce protective mechanisms in the host with considerable therapeutic relevance for experimental colitis and IBD. For example, Madsen et al. have shown that specific-pathogen free (SPF) IL-10−/− mice colonized at 2 weeks of age display changes in bacterial colonization with simultaneous increases in aerobic adherent and translocated bacteria and reduced levels of Lactobacillus reuteri. Rectal administration of endogenous L. reuteri enhances mucosal barrier function and attenuates the development of colitis in SPF IL-10−/− mice at 4 weeks of age (29). Similar protective effects of lactobacilli have been demonstrated in a rat model of methotrexate-induced enterocolitis. These rats have decreased weight loss, decreased intestinal permeability, and decreased myeloperoxidase levels after oral administration of L. plantarum compared to untreated controls (30). Evidence of the protective activity of certain lactic acid bacteria in human IBD is provided by studies from Gionchetti et al. The authors show that a combination of 8 different lactic acid bacteria, including lactobacilli, bifidobacteria, and streptococci (VSL#3), inhibits relapse of chronic pouchitis (15% recurrence rate in the VSL#3 treated group vs. 100% in the placebo group), inhibits mucosal TNFα and up-regulates IL-10 production in treated pouches (31–33). Non-pathogenic gram-negative E. coli have also been shown to have protective effects against inflammation in patients with ulcerative colitis (34). The mechanisms ensuring hyporesponsiveness to indigenous microflora in normal hosts or the mechanisms by which this microflora protects the host during chronic intestinal inflammation and the role of the intestinal epithelium in regulating mucosal immune responses are virtually unknown. Reaching an understanding of the interplay between host and the luminal endogenous bacteria and the molecular mechanisms controlling intestinal tolerance promises to improve our ability to treat intestinal inflammation. The inflammatory process observed in the intestine is the consequence of the dysregulated activation of mucosal cells such as IECs, mesenchymal cells, macrophages, neutrophils, dendritic cells, and lymphocytes. Although the individual contribution of these cell types to the inflammatory process are different, IECs are the likely primary target of the proinflammatory constituents in the intestinal lumen. BACTERIAL CROSS-TALK TO IECs The host is isolated and protected from the potentially toxic intestinal luminal content by a single layer of IECs. These cells are the first line of defense against noxious luminal agents and, as has been shown recently, are important players in intestinal responses induced by microbial products (35,36). There is accumulating evidence that the collaboration between enteric bacteria and the intestinal epithelium helps create a complex ecosystem by modifying host epithelial cell biology. For example, Bacteroides thetaiotoomicron a prominent species of the normal anaerobic, gram-negative microflora of the distal intestinal tract provides nutrients and binding receptors for other bacterial species by inducing the production of specific fucosylated glycoconjugates by the host (37). Furthermore, the colonization of germ-free mice with B. thetaiotoomicron induces a characteristic profile of genes in epithelial cells, which enhances mucosal barrier function, nutrient absorption, xenobiotic metabolism, differentiation, defense, and angiogenesis (38–40). Remarkably, the single bacterial strain B. thataiotoomicron and, to a lesser extent, E. coli or Bacteroides infantis can recapitulate changes in the phenotype and function of the gut epithelium normally induced by the complete mixed conventional microflora. These findings support the concept that there may be bacteria-specific effects in the crosstalk between organisms and host epithelium. Adherent and invasive pathogenic bacteria, bacterial cell wall components, and cytokines all can stimulate IECs to release proinflammatory products including chemokines, cytokines, and adhesion molecules (35,36). The NF-κB transcription factor plays a key role in the induction of many of the cytokines, chemokines, and adhesion molecules involved in inflammatory disorders, including IBD (41–44). The production and accumulation of these proinflammatory molecules in the vicinity of the mucosa have a dual effect on the inflammatory process. They lead to the activation of adjacent lamina propria mononuclear cells (macrophages, dendritic cells, and myofibroblasts), and contribute to the recruitment of peripheral mononuclear and polymorphonuclear cells which participate in the inflammatory process (Fig. 1). Among the potential stimuli of IEC, bacteria and bacterial products are the most important given their high content in the intestinal lumen. Although the duodenum and jejunum contain only 102-103 bacteria/g of luminal contents and the ileum only 105-109 bacteria/g, the total number of microorganisms explodes in the colon to reach 109-1012 bacteria/g of colonic contents.FIG. 1.: Bacterial interaction with the intestinal epithelium and host response. Intestinal epithelial cells are in close proximity to luminal pathogenic and non-pathogenic microorganisms. Protective host responses to the pathogenic threat include production of pro-inflammatory/immunosuppressive cytokines (e.g. IL-1β, TNF, IL-12, IFN-γ, IL-4, TGF-β, IL-10) and chemokines (e.g. IL-8, IP-10, GRO, MIP) as well as recruitment and activation of polymorphonuclear cells, macrophages, dendritic cells, mast cells, T and B lymphocytes, which orchestrate innate and acquired immune responses in the intestinal mucosa.Recent findings indicate that non-pathogenic enteric bacteria mediate proinflammatory processes in the gut epithelium, which with the appropriate compensatory inhibitory control mechanisms, are not detrimental to the host and are thus referred to as “physiological inflammation.” For example, the non-pathogenic bacterial strain B. vulgatus induces RelA phosphorylation, NF-κB transcriptional activation, and proinflammatory gene expression in primary and IEC lines through the TLR4 signaling cascade (45,46). Immuno-staining of tissue sections of B. vulgatus mono-colonized rats indicates that the induction of RelA phosphorylation is confined to the epithelium with no induction in the underlying lamina propria immune cells. This finding suggests that activation of NF-κB in the gut mucosa is compartmentalized and possibly a transient phenomenon. Hornef et al. have shown that E. coli-derived LPS is internalized by murine epithelial cells to stimulate the IκB/NF-κB system via intracellular TLR-4 (47), supporting the concept that non-pathogenic bacteria have the potential to activate inflammatory signaling processes in the gut epithelium. Kojima et al have found that gram-negative enteric bacteria evoke the sustained expression of hsp25 and hsp72 in the IEC of normal mice, thereby helping to maintain the integrity of the intestinal barrier (48). Chronic metronidazole treatment perturbs the normal colonic flora and decreases expression of mucosal hsp25 and hsp72, rendering the mucosa more susceptible to the deleterious effects of Clostridium difficile toxin A (48). Other studies have shown that the crosstalk between IECs and their adjacent immune cells modulates the activity of bacteria-stimulated NF-κB and proinflammatory cytokine expression in vitro (46,49). Thus, a complex interaction between non-pathogenic bacteria, epithelial cells, and immune cells in the mucosa are prerequisite for the development of mature immune function and defense mechanisms in the gut. PATTERN RECOGNITION RECEPTORS AND INNATE IMMUNE RESPONSE A set of well conserved pattern recognition receptors (PRR) named toll-like receptors (TLR) located on the extracellular membrane, and a family of intracellular sensors called nucleotide-binding oligomerization domain/caspase recruitment domain (NOD/CARD) proteins initiate the key process of bacteria-induced innate host response (50–52). The combined action of both sets of receptors play a pivotal role in the detection of various microbial molecular signatures and in the transmission of various signaling cascades that finally lead to the induction of a complex innate gene program aimed at restoring homeostasis of the host. To date over 10 different TLR and more than 20 NOD proteins have been identified, but only a handful have been assigned a specific ligand (50,53,54). For example, LPS uses the pattern recognition receptor (PRR), TLR4, whereas gram-positive bacterial products (lipoteichoic acid, peptidoglycan, etc.), bacterial flagellin, unmethylated CpG DNA, and muramyl dipeptide utilize TLR2, TLR5, TLR9 and NOD2 respectively to initiate signal transduction in mammalian cells (55–57). The primary role of PRRs is the immunosurveillance of the host and of tissues susceptible to invasion by bacteria including the lung, the gastrointestinal tract, and hematopoeitic-derived cells. TLR4 Signaling to the NF-κB Transcriptional System TLR signals to multiple down-stream target effector systems, including the mitogen-activated protein kinase (MAPK), the extracellular activated kinase (ERK), p38, c-jun NH2-terminal kinase (JNK) pathways, and the IκB/NF-κB transcriptional system, the latter a prototypical proinflammatory cascade (54,58). Among the various TLRs, TRL4 is highly relevant to many inflammatory disorders caused by gram-negative bacteria. The cytoplasmic portion of TLR4 is similar in molecular structure and organization to IL-1R and is thus named the toll/IL-1R (TIR) domain (59). The extracellular portion of TLR4 is characterized by a leucine-rich repeat domain located at the C-terminal portion of the protein. Studies using transient transfection of wild-type and dominant-negative forms of various signaling molecules have provided a working map for LPS signaling to NF-κB (Fig. 2A). From these studies, it appears that the TIR domain of TLR4 promotes the homophilic interaction between the TIR domain-containing adapter protein (TIRAP) (60) or the MyD88-adapter-like protein (Mal) (61) and the myeloid differentiation protein 88 (MyD88) (62,63). This interaction is followed by the recruitment of the IL-1 receptor-associated kinase (IRAK-1, 2 or 4) (64–69). The LPS/TLR4 cascade also launches a MyD88-independent signaling cascade associated mainly with induction of IFN-β gene expression (70). This pathway involves recruitment of the TIR domain containing adaptor-inducing IFN-β (TRIF) and the TRIF related adaptor molecule (TRAM) (71,72). Assemblage of this signaling complex leads to the late induction of NF-κB activity and of proinflammatory gene expression as well as expression of IFN-β (Fig. 2A). After assembly of the TLR-induced membrane proximal signaling complex, IRAK is phosphorylated; it then dissociates from the complex and recruits both the TNF receptor-associated factor-6 (TRAF-6) (59,66) and the transforming growth factor-β activated kinase 1 (TAK1) (73,74).FIG. 2.: TLR4 signaling to the NF-κB transcriptional system. A) The extracellular portion of TLR4 characterized by a leucine-rich repeat (LRR) domain and co-receptor MD-2 mediate LPS binding to the receptor. LPS-binding protein and CD14 facilitate interaction of LPS with TLR4. The cytoplasmic TIR domain of TLR4 contains binding sites for adaptor proteins including TIRAP (TIR-domain containing adaptor protein), Mal (MyD88-adapter-like) and the myeloid differentiation protein (MyD) 88. The gram-negative bacterial compound lipopolysaccharide (LPS) activates TLR4 signal transduction through MyD88-dependent mechanisms including IRAK (IL-1 receptor associated kinase) phosphorylation and recruitment of TRAF 6 (TNF receptor-associated factor 6) and TAK1 (Transforming growth factor-b activated kinase 1). MyD88-independent mechanisms induce IFN-β gene expression through the adaptor proteins TRAM (TIR domain containing adaptor-inducing IFN-β) and TRIF (TRIF related adaptor protein). B) Induction of the IκB/NF-κB system through the TLR4 signaling cascade activates the proximal kinase NIK (NF-κB inducible kinase), which associates with the IKK (IκB kinase) complex to induce IκB phosphorylation and ubiquitination. Upon proteasome-dependent degradation of IκB, NF-κB homo or heterodimer are released from the IκB/NF-κB complex and translocate to the nucleus to induce κB-dependent gene expression.The signal coming from the TRAF6/TAK1 is then transmitted to the core signaling complex of the NF-κB pathway (Fig. 2B) which includes the NF-κB-inducing kinase (NIK), which in turn associates/activates the IκB kinase (IKK) complex (58,75–79). This complex is controlled by the structural regulatory protein IKKγ, also known as NF-κB essential modifier (NEMO). IKKγ directs the activation of the catalytic IKKα and IKKβ subunits (80,81), which then phosphorylates IκBα at serine residue 32 and 36 (82). This phosphorylation is followed by the activation of a complex enzymatic system (E1, E2, E3) that adds multiple ubiquitin proteins at lysine residues 21 and 22 of phosphorylated IκBα (83). Ubiquitinated IκBα is then selectively and rapidly degraded via the non-lysosomal, ATP-dependent 26S proteolytic complex composed of a 700 kDa proteasome (83). Destruction of IκBα liberates NF-κB from inhibition and allows nuclear transmigration of the transcription factor, binding to κB-promoter elements, and induction of gene transcription. The enzyme responsible for ubiquitin conjugation of phosphorylated IκBα is the E3 receptor subunit of IκB (E3RSIκB)(83–85). Interestingly, non-virulent Salmonella strains inhibit NF-κB activity by preventing IκB ubiquitination, possibly through inhibition of E3RSIκB (86). This inhibition of NF-κB suggests that some intestinal bacteria have evolved sophisticated mechanisms to down-regulate the host innate immune response by targeting regulatory elements of the NF-κB pathway. TLR4 Signaling and Host Response TLR4 is associated with many inflammatory diseases. Of these, septic shock syndrome is the paradigm of the deleterious effect of LPS/TLR4 on host integrity. LPS released from gram-negative bacteria is the most frequent cause of septic shock, which affects approximately 400,000 patients/yr in the United States with a mortality of more than 100,000/yr (87,88). Of therapeutic relevance is the finding that blocking NF-κB activity with IκBα enhances survival in an animal model of septic shock (89). The LPS/TRL4 pathway is also linked to heart, lung, and liver tissue injuries. For example, compared to wild-type mice, C3H/HeJ mice (TLR4 natural mutant) show strongly reduced LPS-induced myocardial dysfunction, TNF and IL-1β mRNA expression, and NF-κB activation (90,91). Infection of C57BL/10ScNCr mice (TLR4 deficient) with respiratory syncytial virus causes the mice to have deficient natural killer cell function, decreased IL-12 secretion, and impaired viral clearance compared to wild-type mice (92). Interestingly, mutation in the human gene encoding for IRAK-4, a kinase involved in TLR signal transduction, impairs innate response and increases susceptibility to bacterial infection (93). Finally, C3H/HeJ mice have reduced alcohol-induced liver injury and TNF mRNA expression compared to control mice (94). These findings suggest that TLR4 plays a role in driving LPS-induced inflammation in vivo and that manipulating the pathway may prevent the deleterious consequences of constant activation. Although evidence suggests a potential role for TLR4 signal transduction in intestinal inflammation, the deleterious or beneficial impact of TLR4 signal transduction on specific disease states is not established. For example, expression of TLR4 and MD-2, an LPS co-receptor, is regulated by immune cell-derived signals with increased expression under conditions of chronic intestinal inflammation (68,95–97). However, the TLR4 mutant mice C3H/HeJ are more sensitive to dextran sodium sulphate-induced colitis than wild-type mice (98,99). SPF colonized TLR4−/− but not wild-type mice fail to express peroxisome proliferator-activated receptor (PPAR)-γ in IEC (100), a nuclear receptor shown to inhibit NF-κB activity and to regulate chronic inflammation (101). Whether up-regulation of TLR4 expression and signal transduction contribute to the inflammatory process or instead represent an attempt to reestablish host homeostasis remains to be determined. Of note, selective ablation of NF-κB activity in enterocytes through deletion of IKKβ sensitizes these animals to ischemia-reperfusion induced enterocyte apoptosis, which is associated with loss of a mucosal integrity (102). This local intestinal tissue injury is likely due to the failure of IKK to activate an NF-κB-dependent protective gene program sheilding IECs from the deleterious effects of intestinal ischemia-reperfusion. Thus, while LPS-induced TLR4 signaling clearly contributes to various extra-intestinal inflammatory disorders, its activation in the intestine by endogenous non-pathologic bacteria may be part of a protective mechanism to maintain homeostasis. However, improper regulation of TLR4 signaling caused by either absence or over-activation of the pathway may turn a physiological response into a pathological one (Fig. 3).FIG. 3.: Pattern recognition receptors and host response. A) Commensal bacteria influence the anatomical, physiological and immunological development of the host, including the induction of TLR signaling and activation of the IκB/NF-κB system in intestinal epithelial cells. Protective mechanisms ensure hyporesponsiveness to antigens of the endogenous microflora in the normal host. B) Pathogenic microorganisms (e.g. Shigella) trigger innate host responses through various mechanisms and signaling cascades including the TLR and the CARD/NOD system. Induction of acute inflammatory processes and subsequent activation of acquired immune responses will help to clear the pathogens in the normal host. C) Under conditions of chronic intestinal inflammation in the genetically susceptible host, commensal enteric bacteria perpetuate chronic immune-mediated inflammation. Dashed arrows indicate proposed activation mechanisms.Interestingly, NOD2/CARD15 may share a similar biological effect in the intestine. This PRR is the first innate receptor linked to an increased likelihood of developing intestinal inflammation in a subset of CD patients (103,104). Indeed, mutations in the LRR domain of NOD2/CARD15 disable the signaling function of the protein and prevent MDP-induced NF-κB activity. The loss of function may impair NOD2/CARD15 induced protective genes that shield the intestine against the deleterious impact of some bacteria. Alternatively and not mutually exclusively, a loss of NOD2/CARD15 signaling may prevent the host from mounting an appropriate innate response, leading to a dysregulated adaptive response (105,106). Interestingly, NOD2−/− mice failed to develop intestinal inflammation suggesting that this PRR is not essential for the maintenance of intestinal homeostasis and that compensatory mechanisms likely substitute for the loss of NOD2 (107). The reader is directed to a series of excellent recent reviews on NOD2/CARD15 (108–110). A common theme in innate immunity is that Pathogen Associated Molecular Patterns (PAMP) bind to TLR to induce a host response. However, extracellular constituent or Commensal Associated Molecular Patterns (CAMP) derived from non-pathogenic bacteria also have the ability to trigger an innate host response through these receptors (45,46,111). For example, while non-pathogenic E. coli and LPS induce IL-10 production in peripheral blood monocytes, enteric gram-positive Lactobacillus strains induce IL-12 production in mononuclear cell populations and IFN-γ in NK cells (112–115). Similar to LPS, lipoteichoic acid isolated from various lactobacilli induce NF-κB activity through a TLR dependent pathway (116). Of considerable importance and in contrast to enteropathogens, the indigenous microflora and/or its derived CAMP induce no histopathology in the normal gut mucosa, suggesting that sophisticated mechanisms tightly regulate proinflammatory signaling in the intestine and help maintain homeostasis. NEGATIVE REGULATORS OF TLR SIGNAL TRANSDUCTION Although the purpose of the TLR signaling pathway is to alert and protect the host against the intrusion of pathogenic microorganisms, improper activation of this signaling system, either due to the lack of negative immunoregulatory mechanisms and/or persistent stimulation, would have a negative impact on the host. Indeed, IEC and intestinal mononuclear cells live in a hostile environment containing a high load of bacteria and bacterial products; therefore, a low threshold response to these products would likely be detrimental to the host. Conversely, peripheral monocytes and polymorphonuclear cells (PMN) display a high sensitivity to LPS (<10ng/ml), which activates various signaling systems and lead to the production of numerous inflammatory mediators (58,117). The fact that monocytes and PMNs are so sensitive to LPS likely reflects their biological role of monitoring the blood stream by detecting traces of infectious agents and initiating immune responses. This biological activity is critical for the appropriate host immune response to invading agents. However, the lower gastrointestinal tract harbors the highest bacterial load, and the intestine must establish tolerogenic mechanisms which allow only minimal, physiological responses. Presence of sophisticated mechanisms assuring hyporesponsiveness of the intestine to its own microflora is vital for the maintenance of intestinal homeostasis. Disruption in the function of these mechanisms would open the floodgates preventing response to luminal bacteria and compromise the delicate balance between pro and anti-inflammatory signals in the intestine, especially in a genetically susceptible host. These regulatory mechan