Immune Mechanisms Underlying Susceptibility to Endotoxin Shock in Aged Hosts: Implication in Age-Augmented Generalized Shwartzman Reaction.

In recent decades, the elderly population has been rapidly increasing in many countries. Such patients are susceptible to Gram-negative septic shock, namely endotoxin shock. Mortality due to endotoxin shock remains high despite recent advances in medical care. The generalized Shwartzman reaction is well recognized as an experimental endotoxin shock. Aged mice are similarly susceptible to the generalized Shwartzman reaction and show an increased mortality accompanied by the enhanced production of tumor necrosis factor (TNF). Consistent with the findings in the murine model, the in vitro Shwartzman reaction-like response is also age-dependently augmented in human peripheral blood mononuclear cells, as assessed by enhanced TNF production. Interestingly, age-dependently increased innate lymphocytes with T cell receptor-that intermediate expression, such as that of CD8+CD122+T cells in mice and CD57+T cells in humans, may collaborate with macrophages and induce the exacerbation of the Shwartzman reaction in elderly individuals. However, endotoxin tolerance in mice, which resembles a mirror phenomenon of the generalized Shwartzman reaction, drastically reduces the TNF production of macrophages while strongly activating their bactericidal activity in infection. Importantly, this effect can be induced in aged mice. The safe induction of endotoxin tolerance may be a potential therapeutic strategy for refractory septic shock in elderly patients.

Brief Report: Endothelial-Specific X-Box Binding Protein 1 Deficiency Limits Tumor Necrosis Factor-Induced Leukocyte Recruitment and Vasculitis.

OBJECTIVE: Endothelial cell activation by tumor necrosis factor (TNF) and associated leukocyte infiltration are hallmarks of vasculitis. The aim of this study was to investigate the potential role of the cellular stress-associated endothelial X-box binding protein 1 (XBP-1) transcription factor in TNF-induced endothelial cell inflammation and vasculitis. METHODS: Mice with an endothelial cell-specific XBP-1 deficiency were used in a modified local Shwartzman reaction (LSR) model of TNF-induced small vessel vasculitis. To address the contribution of XBP-1 to the TNF-mediated inflammatory response in endothelial cells, we examined the activation of XBP-1 expression by TNF as well as the effect of XBP-1 knockdown in endothelial cells on TNF-induced signaling, proinflammatory gene expression, and leukocyte-endothelial cell adhesion. RESULTS: The active spliced form of XBP-1 in endothelial cells was triggered by TNF. In addition, endothelial XBP-1 contributed to the sustained TNF-triggered NF-κB-dependent transcriptional activation of proinflammatory molecules, which was associated with leukocyte-endothelial cell adhesion. In the LSR model, endothelial cell-specific XBP-1-deficient mice displayed significantly less vascular damage, accompanied by reduced perivascular neutrophil infiltration, as compared with wild-type mice. CONCLUSION: Endothelial XBP-1 is activated by TNF and regulates leukocyte-endothelial cell adhesion in vitro as well as neutrophil infiltration and vascular damage in murine vasculitis.

The effect of synthetic C-reactive protein on the in vitro immune response of human PBMCs stimulated with bacterial reagents.

Synthetic C-reactive protein (CRP) rescues mice from lethal endotoxin shock or bacterial infection by suppressing tumor necrosis factor (TNF-α), but in turn, enhances Kupffer cell phagocytic activity. We herein assessed the influence of CRP in human peripheral blood mononuclear cells (PBMCs). When human PBMCs were stimulated in vitro with penicillin-treated Streptococcus pyogenes, bacterial DNA motifs and lipopolysaccharide with or without synthetic CRP, CRP suppressed the production of TNF-α and IL-12, but not that of IFN-γ. This was also the case for the in vitro Shwartzman reaction induced in PBMCs. CRP also decreased high-mobility group box 1 production from macrophages, which is crucial in the later phase of endotoxin/septic shock. However, CRP upregulated the perforin expression by CD56(+) NK cells and increased their antitumor cytotoxicity. CRP may thus be a potent immunomodulatory factor in the human immune system, suggesting its therapeutic potential for use against human septic shock.

Requirement for Vav proteins in post-recruitment neutrophil cytotoxicity in IgG but not complement C3-dependent injury.

The signals linking neutrophil opsonic receptors, FcgammaRs and complement receptor 3 (Mac-1) to cellular cytotoxic responses are poorly understood. Furthermore, because a deficiency in activating FcgammaRs reduces both IgG-mediated neutrophil recruitment and tissue injury, the role of FcgammaRs specifically in mediating neutrophil cytotoxicity in vivo remains unclear. In this study, we demonstrate that neutrophil Vav 1 and 3, guanine exchange factors for Rac GTPases, are required for IgG/FcgammaR-mediated hemorrhage and edema in the reverse passive Arthus in the lung and skin. Rac GTPases are also required for development of the reverse passive Arthus reaction. A deficiency in Vav 1 and 3 does not affect neutrophil accumulation at the site of immune complex deposition, thus uncoupling neutrophil recruitment and tissue injury. Surprisingly, Vav and Rac proteins are dispensable for the development of the local Shwartzman reaction in vivo and phagocytosis of complement-opsonized RBC in vitro, processes strictly dependent on Mac-1 and complement C3. Thus, FcgammaR signaling through the Vav and Rac proteins in neutrophils is critical for stimulating immune complex disease while Vav- and Rac-independent pathways promote Mac-1/complement C3-dependent functions.

Mice deficient in Vbeta8+NKT cells are resistant to experimental hepatitis but are partially susceptible to generalised Shwartzman reaction.

NKT cells are responsible for hepatitis induced either by concanavalin A (Con-A) or alpha-galactosylceramide (alpha-GalCer), and they are also profoundly involved in the generalised Shwartzman reaction (GSR) induced by consecutive injections of interleukin (IL)-12 and lipopolysaccharide (LPS). In the present study, using NC/Nga (NC) mice and SJL mice lacking the Vbeta(+)8 gene, we examined the role of Vbeta(+)8+NKT cells in hepatitis models and in the GSR. The absence of Vbeta(+)8+NKT cells in the liver mononuclear cells (MNC) was confirmed by the alpha-GalCer/CD1d/Ig dimer. Unexpectedly, other dimer+NKT cells including Vbeta7(+)NKT cells in these mice were found to decrease in comparison to that of C57BL/6 mice. No significant hepatocyte injury was observed after alpha-GalCer or Con-A administration in either mice. The serum interferon (IFN)-gamma, IL-4 and tumour necrosis factor (TNF) levels did not increase in these mice after alpha-GalCer injection, however these cytokines substantially increased after Con-A administration, thus suggesting that the roles of NKT cells differ between the two hepatitis models. However, in GSR, although neither mice showed lower IFN-gamma levels after a priming IL-12 injection, they showed TNF levels comparable to those in normal mice after LPS injection, and thus resulted in a decreased but substantial mortality. Although liver MNC from IL-12-injected SJL mice showed an impaired antitumour cytotoxicity, liver MNC of NC mice exhibited a greater antitumour cytotoxicity than that of C57BL/6 mice because liver NK cells proportionally increased in NC mice. These results confirm the critical role that Vbeta8(+)NKT cells play in both liver and multi-organ injury.

An in vitro Shwartzman reaction-like response is augmented age-dependently in human peripheral blood mononuclear cells.

A lethal human septic shock model, mouse generalized Shwartzman reaction (GSR), was elicited by two consecutive lippolysaccharide (LPS) injections (24 h apart) in which interferon-gamma (IFN-gamma) induced by interleukin (IL)-12 played a critical role in the priming phase, and tumor necrosis factor (TNF) was an important effector molecule in the second phase. We recently reported IL-12/LPS-induced mouse GSR age-dependently enhanced. We herein demonstrate that human peripheral blood mononuclear cells (PBMC) from healthy adults/elderly, cultured with IL-12 for 24 h and with LPS for an additional 24 h, produced a much larger amount of TNF (which increased age-dependently) than did PBMC without IL-12 priming. Whereas macrophages mainly produced TNF following LPS stimulation, macrophages and lymphocytes were necessary for a sufficient TNF production. IL-12-induced IFN-gamma up-regulated Toll-like receptor 4 (TLR-4) on macrophages of adults. Although the PBMC from children produced a substantial amount of IFN-gamma after IL-12 priming, the GSR response, with augmented TNF production and an up-regulated TLR-4 expression of macrophages, was not elicited by LPS stimulation. CD56+natural killer cells, CD56+T cells, and CD57+T cells (NK-T cells), which age-dependently increased in PBMC, produced much larger amounts of IFN-gamma after IL-12 priming than that of conventional CD56-CD57-T cells and also induced cocultured macrophages to produce TNF by subsequent LPS stimulation. The elder septic patients were consistently more susceptible to lethal shock with enhanced serum TNF levels than the adult patients. The NK cells, NK-T cells, and macrophages, which change proportionally or functionally with aging, might be involved in the enhanced GSR response/septic shock observed in elderly patients.

Cholinergic stimulation blocks endothelial cell activation and leukocyte recruitment during inflammation.

Endothelial cell activation plays a critical role in regulating leukocyte recruitment during inflammation and infection. Based on recent studies showing that acetylcholine and other cholinergic mediators suppress the production of proinflammatory cytokines via the alpha7 nicotinic acetylcholine receptor (alpha7 nAChR) expressed by macrophages and our observations that human microvascular endothelial cells express the alpha7 nAChR, we examined the effect of cholinergic stimulation on endothelial cell activation in vitro and in vivo. Using the Shwartzman reaction, we observed that nicotine (2 mg/kg) and the novel cholinergic agent CAP55 (12 mg/kg) inhibit endothelial cell adhesion molecule expression. Using endothelial cell cultures, we observed the direct inhibitory effects of acetylcholine and cholinergic agents on tumor necrosis factor (TNF)-induced endothelial cell activation. Mecamylamine, an nAChR antagonist, reversed the inhibition of endothelial cell activation by both cholinergic agonists, confirming the antiinflammatory role of the nAChR cholinergic pathway. In vitro mechanistic studies revealed that nicotine blocked TNF-induced nuclear factor-kappaB nuclear entry in an inhibitor kappaB (IkappaB)alpha- and IkappaBepsilon-dependent manner. Finally, with the carrageenan air pouch model, both vagus nerve stimulation and cholinergic agonists significantly blocked leukocyte migration in vivo. These findings identify the endothelium, a key regulator of leukocyte trafficking during inflammation, as a target of anti-inflammatory cholinergic mediators.

Critical role of the liver CD8+ CD122+ T cells in the generalized Shwartzman reaction of mice.

We examined the role of mouse CD8+ CD122+ T cells, which increase in number with age, in the generalized Shwartzman reaction. This reaction was induced by IL-12 priming and subsequent LPS challenge (after 24 h) in mice of various ages (4-50 weeks of age). Although most young mice (4 or 6 weeks of age) survived, mortality essentially increased with increasing age of the mice, and all mice of 20 weeks of age or older died within 48 h. Serum TNF-alpha levels after LPS challenge also increased age dependently. The neutralization of either IL-12-induced IFN-gamma or LPS-induced TNF-alpha improved the survival of middle-aged (25-week-old) mice. Both IFN-gamma production after IL-12 priming and TNF-alpha production from the liver mononuclear cells after LPS challenge were also prominent in the middle-aged mice. CD8+CD122+ T cells cultured with IL-12 produced a much larger amount of IFN-gamma than CD8+CD122- T cells. Although the depletion of NK/NK T cells did not decrease the IFN-gamma or TNF-alpha production in the Shwartzman reaction of the middle-aged mice, an additional depletion of CD8+CD122+ T cells did decrease such production and also improved mouse survival. Furthermore, young mice transferred with CD8+CD122+ T cells from aged B6 nude mice showed an enhanced Shwartzman reaction.

Resistance of natural killer T cell-deficient mice to systemic Shwartzman reaction.

The generalized Shwartzman reaction in mice which had been primed and challenged with lipopolysaccharide (LPS) depends on interleukin (IL)-12-induced interferon (IFN)-gamma production at the priming stage. We examined the involvement in the priming mechanism of the unique population of Valpha14 natural killer T (NKT) cells because they promptly produce IFN-gamma after IL-12 stimulation. We report here that LPS- or IL-12-primed NKT cell genetically deficient mice were found to be resistant to LPS-elicited mortality. This outcome can be attributed to the reduction of IFN-gamma production, because injection of recombinant mouse IFN-gamma, but not injection of IL-12, effectively primed the NKT cell-deficient mice. However, priming with high doses of LPS caused mortality of severe combined immunodeficiency, NKT cell-deficient, and CD1-deficient mice, indicating a major contribution of NKT cells to the Shwartzman reaction elicited by low doses of LPS, whereas at higher doses of LPS NK cells play a prominent role. These results suggest that the numerically small NKT cell population of normal mice apparently plays a mandatory role in the priming stage of the generalized Shwartzman reaction.

The liver as a crucial organ in the first line of host defense: the roles of Kupffer cells, natural killer (NK) cells and NK1.1 Ag+ T cells in T helper 1 immune responses.

The liver remains a hematopoietic organ after birth and can produce all leukocyte lineages from resident hematopoietic stem cells. Hepatocytes produce acute phase proteins and complement in bacterial infections. Liver Kupffer cells are activated by various bacterial stimuli, including bacterial lipopolysaccharide (LPS) and bacterial superantigens, and produce interleukin (IL)-12. IL-12 and other monokines (IL- 18 etc.) produced by Kupffer cells activate liver natural killer (NK) cells and NK1.1 Ag+ T cells to produce interferon-gamma and thereby acquire cytotoxicity against tumors and microbe-infected cells. These liver leukocytes and the T helper 1 immune responses induced by them thus play a crucial role in the first line of defense against bacterial infections and hematogenous tumor metastases. However, if this defense system is inadequately activated, shock associated with multiple organ failure takes place. Activated liver NK1.1 Ag+ T cells and NK cells also cause hepatocyte injury. NK1.1 Ag+ T cells and another T-cell subset with an intermediate T-cell receptor, CD 122+CD8+ T cells, can develop independently of thymic epithelial cells. Liver NK cells and NK1.1 Ag+ T cells physiologically develop in situ from their precursors, presumably due to bacterial antigens brought from the intestine via the portal vein. NK cells activated by bacterial superantigens or LPS are also probably involved in the vascular endothelial injury in Kawasaki disease.

Expression of Fas and Fas ligand on mouse renal tubular epithelial cells in the generalized shwartzman reaction and its relationship to apoptosis.

Previously we reported that the consecutive injection of lipopolysaccharide (LPS) into LPS-sensitized mice for the generalized Shwartzman reaction (GSR) appeared to induce the injury of renal tubular epithelial cells via apoptosis. The aim of this study was to characterize the mechanism of renal tubular epithelial cell injury in GSR. The expression of Fas and Fas ligand was immunohistochemically detected on renal tubular epithelial cells from GSR-induced mice, although neither Fas nor Fas ligand was found in cells from untreated control mice or in cells from mice receiving a single injection of LPS. GSR-induced renal tubular epithelial cell injury was produced in neither Fas-negative MRL-lpr/lpr mice nor Fas ligand-negative MRL-gld/gld mice. The administration of anti-gamma interferon antibody together with a preparative injection of LPS prevented the expression of Fas and Fas ligand and the apoptosis of renal tubular epithelial cells. A provocative injection of tumor necrosis factor alpha into LPS-sensitized mice augmented Fas and Fas ligand expression and the apoptosis of renal tubular epithelial cells. The administration of tumor necrosis factor alpha to interleukin-12-sensitized mice resulted in Fas and Fas ligand expression and the apoptosis. Sensitization with interleukin-12 together with anti-gamma interferon antibody did not cause the apoptosis of renal tubular epithelial cells. It was suggested that the Fas/Fas ligand system probably plays a critical role in the development of renal tubular epithelial cell injury through apoptotic cell death.

Involvement of NK1+ T cells and their IFN-gamma production in the generalized Shwartzman reaction.

IL-12 (or LPS) priming and subsequent challenge by LPS produces the generalized Shwartzman reaction. IFN-gamma induced by IL-12 is a crucial cytokine in the priming phase. In vivo depletion of both NK cells and NK1+ alphabeta T cells of mice by anti-NK1.1 Ab greatly reduced the elevation of serum IFN-gamma induced by IL-12 and significantly reduced mortality after subsequent injection of LPS, whereas depletion of NK cells alone by anti-asialo GM1 Ab only partially decreased serum IFN-gamma, and lethality was not changed. Cell sorting and culture experiments confirmed that liver NK1+ alphabeta T cells of IL-12-injected mice produced greater amounts of IFN-gamma than did liver NK cells. MHC class I-deficient mice of C57BL/6 background, which lack a majority of NK1+ alphabeta T cells, produced low amounts of IFN-gamma by IL-12; no mortality was observed after the LPS challenge. However, production of TNF-alpha in the second phase (after LPS challenge) was not inhibited by depletion of NK cells alone or both subsets. IL-12 and subsequent LPS challenge activated NK1+ alphabeta T cells in the liver and induced strong cytotoxicity of these cells not only against tumor cells (including Fas-negative tumors) but also against a syngeneic hepatocyte cell line. Our findings show that IFN-gamma produced by NK1+ alphabeta T cells is essential for the IL-12 priming of the Shwartzman reaction, and the autoreactivity of NK1+ alphabeta T cells in the liver is involved in the hepatic disorders that are sometimes caused by IL-12, LPS, or the generalized Shwartzman reaction.

Involvement of tumor necrosis factor-alpha, interleukin-1 beta, interleukin-8, and interleukin-1 receptor antagonist in acute lung injury caused by local Shwartzman reaction.

A local Shwartzman reaction (LSR) was prepared in rabbit lung as a model of acute lung injury. To induce LSR, intratracheal injection of lipopolysaccharide (LPS) 10 micrograms into the lower lobe of the right lung, followed 24 h later by i.v. injection of LPS (10 micrograms/kg). In the lung with the LSR, myeloperoxidase activity, representing neutrophil accumulation, peaked at 1-2 h and was sustained for 48 h after challenge with i.v. LPS. The lung water content peaked at 12 h, and decreased gradually. Histological findings showed diffuse interstitial widening, intra-alveolar leukocyte infiltration with hemorrhage, and alveolar exudate formation. The production of tumor necrosis factor-alpha (TNF-alpha), interleukin-1 beta (IL-1 beta), interleukin-8 (IL-8), and IL-1 receptor antagonist (IL-1 Ra) in the lung was analyzed. TNF-alpha first elevated and peaked at 0.5 h (66.5 +/- 16.7 ng/g.lung), subsequently, IL-1 beta and IL-8 increased and peaked at 2 h (17.8 +/- 3.4 ng/g.lung and 336.9 +/- 49.6 ng/g.lung, respectively). IL-1Ra was present even before the challenge, and the production increased to show a dual peak (0.5 h, 1.5 +/- 0.2 micrograms/g.lung; and 2 h, 1.6 +/- 0.1 micrograms/g.lung), and a large concentration of IL-1Ra was sustained for 48 h. Immunohistochemistry showed that the cellular source of these cytokines was alveolar macrophages and infiltrating neutrophils. Thus, disclosing the kinetics of the generation of cytokines led to a better understanding of their roles, namely TNF-alpha as an initiator, IL-1 and IL-8 as amplifier and effector, and IL-1Ra as regulator of the intensity of acute inflammation.

Immunomodulation of LPS ability to induce the local Shwartzman reaction.

Immunologically, the septic shock is a natural model of immunomediated vascular pathology where the interaction between cytokines and the endothelium mediates the syndrome and lethality. Tumour necrosis factor (TNF), a non-species-specific cytokine, has outstanding pleiotropic activities as an important mediator of the septic shock syndrome. In rabbits, passive immunization with anti-lipopolysaccharide (LPS) polyclonal antibodies prior to the intravenous (i.v.) injection of LPS inhibits the haemorrhagic necrotic lesion characteristic of the local Shwartzman reaction (an excellent localized in vivo correlate of the septic shock). Paradoxically, tested in an ex vivo assay (short-term whole human blood culture, stimulated with LPS), these antibodies mediated an increase in TNF production by mononuclear phagocytes and, in the rabbit model, they induced an increase in body temperature, as compared with the pre-immune reagent. Although anchoring of immune complexes containing LPS to receptors (Fc or C4b-C3b) on circulating monocytes may facilitate the access of LPS to these cells, access to localized, LPS-sensitized macrophages may be impaired. Consequently inhibition of the local Shwartzman reaction and increased TNF production in the ex vivo system were observed. Concordantly, the higher temperature in the passively immunized animals may be a consequence of a higher, immune complex-induced, systemic TNF production. These experimental results suggest that the use of anti-LPS immunoglobulins, as a potential immunotherapy for septic shock syndrome in vertebrates, may lead to increased TNF production, with adverse effects such as the pyrogenic.

Effect of soluble P55 tumour-necrosis factor binding fusion protein on the local Shwartzman and Arthus reactions.

1. In this study, the effects of a protein synthesis inhibitor, cycloheximide, and a soluble tumour necrosis factor (TNF) binding/IgG fusion protein, p55-sf2, on the priming and challenge stages of the local Shwartzman reaction (LSR) were assessed and compared with their effects on the acute inflammatory response induced by recombinant human tumour necrosis factor-alpha (rhTNF), lipopolysaccharide (LPS) and a reversed passive Arthus (RPA) reaction in rabbit skin. 2. The LSR was induced in skin by giving an intradermal (i.d.) priming injection of LPS followed by two i.v. challenge injections 20 h and 22 h later. Accumulation of 51-Cr-labelled red blood cells and [125I]-albumin were measured at 24 h as markers of haemorrhage and oedema formation, respectively. 3. The RPA reaction was induced in the rabbit by giving i.d. injections of Arthus anti-serum (anti-bovine-gamma-globulin, BGG) followed 5 min later by an i.v. injection of the antigen (BGG). Oedema formation and the accumulation of 111In-labelled neutrophils produced in the RPA reaction and in response to i.d. injection of rhTNF and LPS were measured over the 4 h period after inducing the responses. 4. A single local injection of cycloheximide (10 micrograms/site) did not inhibit neutrophil accumulation or oedema formation produced by 100% Arthus anti-sera. Although LPS injected i.d. induced a marked dose-dependent neutrophil accumulation, there was little associated plasma leakage. Cycloheximide (10 micrograms/site) did not significantly inhibit the neutrophil accumulation induced by LPS (0.1 microgram/site). In the LSR, priming i.d. injections of LPS caused a dose-dependent increase in haemorrhage and plasma leakage at skin sites after challenge with LPS (two injections of 100 micrograms, i.v.). Co-injection of a single dose of cycloheximide (10 micrograms/site) with LPS (30 micrograms/site) caused a marked reduction in the amount of haemorrhage. Local cycloheximide (10 micrograms/site) administered immediately before LSR challenge did not affect the responses produced in the LSR. 5. Neutrophil accumulation induced by TNF (0.17 micrograms/site) was abolished by co-administration of p55-sf2 (3 micrograms/site) whereas neutrophil accumulation induced by i.d. LPS and produced in the RPA reaction was not affected. In the LSR, haemorrhage and oedema formation were inhibited by p55-sf2 (3 micrograms/site) when it was administered i.d. with the LPS priming injection, but not when given i.d. immediately before LSR challenge. 6. These data suggest that the acute neutrophil accumulation produced in the RPA reaction and in response to i.d. LPS may not be dependent on local protein synthesis or TNF production. On the other hand, haemorrhage appears to be dependent on local protein synthesis during the priming phase but not during the challenge stage of the LSR. Importantly, haemorrhage and plasma leakage appear to be dependent on local TNF generation during the priming phase but not during the challenge stage of the LSR. Thus TNF appears to play a key role in the LSR in rabbit skin.

Essential role for natural killer cells in the lethal lipopolysaccharide-induced Shwartzman-like reaction in mice.

Observations in our laboratory have provided evidence that interferon-gamma (IFN-gamma) is a key regulator of inflammatory responses to bacterial lipopolysaccharide (LPS) (Heremans et al., J. Exp. Med. 1990. 171: 1853): treatment of mice with neutralizing monoclonal antibody against IFN-gamma was found to completely prevent lethal shock reactions, in particular the generalized Shwartzman reaction, whereas treatment with IFN-gamma sensitized the mice to the development of such reactions. Since activated T cells and natural killer (NK) cells are the main if not the only potential source of LPS-induced IFN-gamma, we investigated the relative importance of these cells in the development of the generalized Shwartzman-like reaction in mice by depleting them selectively with relevant monoclonal antibodies. Treatment with antibodies directed against the CD4+ T cells subset was not effective in protecting mice. Anti-CD8 antibody did attenuate the reaction to some extent. However, markedly reduced mortality was seen in mice which were depleted of NK cells by systemic administration of polyclonal anti-asialo GM1 or monoclonal anti-NK1.1 antibodies. Failure of T cells to promote the Shwartzman reaction was also evidenced by the observation that thymus-less nude mice, which are deficient in T cells, were more rather than less sensitive to the reaction. Approximately 20 times less LPS was needed to induce the lethal reaction in these mice than in NMRI mice and 58 times more anti-IFN-gamma antibody was required to block mortality. Nu/nu mice reportedly have an over-active NK cell compartiment. IFN-gamma production by these cells in LPS-treated mice may account for the augmented sensitivity. Our data suggest that NK cells may be the most important source of endogenous IFN-gamma which mediates the LPS-induced lethal reactions in mice.

Possible evidence for a Shwartzman reaction in pseudomembranous colitis.

Pseudomembranous colitis is a potentially fatal disorder associated with gastrointestinal surgery and the use of antibiotics. The aetiological agent has been shown to be Clostridium difficile but the pathogenesis of the disease is unknown. It has been suggested that the lesions produced are due to a local Shwartzman effect. Such an effect may be accompanied by activation of the serum complement system and we therefore looked for evidence of complement consumption in 4 patients with the diagnostic clinical and histological features of pseudomembranous colitis. Positive evidence was obtained suggesting that a Shwartzman phenomenon may be involved.