Search

Brucella is a facultative intracellular pathogen and the etiological agent of brucellosis. In some cases, human brucellosis results in a persistent infection that may reactivate years after the initial exposure. The mechanisms by which the parasite evades clearance by the immune response to chronically infect its host are unknown. We recently demonstrated that dendritic cells (DCs), which are critical components of adaptive immunity, are highly susceptible to Brucella infection and are a preferential niche for the development of the bacteria. Here, we report that in contrast to several intracellular bacteria, Brucella prevented the infected DCs from engaging in their maturation process and impaired their capacities to present antigen to naïve T cells and to secrete interleukin-12. Moreover, Brucella-infected DCs failed to release tumor necrosis factor alpha (TNF-alpha), a defect involving the bacterial protein Omp25. Exogenous TNF-alpha addition to Brucella-infected DCs restored cell maturation and allowed them to present antigens. Two avirulent mutants of B. suis, B. suis bvrR and B. suis omp25 mutants, which do not express the Omp25 protein, triggered TNF-alpha production upon DC invasion. Cells infected with these mutants subsequently matured and acquired the ability to present antigens, two properties which were dramatically impaired by addition of anti-TNF-alpha antibodies. In light of these data, we propose a model in which virulent Brucella alters the maturation and functions of DCs through Omp25-dependent control of TNF-alpha production. This model defines a specific evasion strategy of the bacteria by which they can escape the immune response to chronically infect their host.

In a previous study, we analysed the synthesis and properties of a series of imidazo[1,2-alpha]quinoxalines designed in our laboratory as possible imiquimod analogues. We found that these imidazo[1,2-alpha]quinoxalines were in fact potent inhibitors of phosphodiesterase 4 enzymes (PDE4). PDE4 inhibition normally results in an increase in intracellular cAMP which, in PBMC, induces the suppression of TNF-alpha mRNA transcription and thus cytokine synthesis. Such an effect is antagonistic to that of imiquimod. Furthermore, some TNF-alpha-induced activity, such as cell apoptosis which is dependent on the intracellular cAMP levels might also be affected. Therefore, by counteracting the properties of TNF-alpha and/or its production, the imidazo[1,2-alpha]quinoxalines could be considered as potential anti-inflammatory drugs. The present study was performed to confirm or refute this hypothesis. For this, we characterized the effects of imidazo[1,2-alpha]quinoxalines both on TNF-alpha activity and synthesis in regard to their ability to act as inhibitors of PDE4 (IPDE4). We found that the imidazo[1,2-alpha]quinoxalines dose-dependently prevented the TNF-alpha-triggered death of L929 cells, with the 8-series (-NHCH3 in R4) being the most potent. Moreover, when the effect of the 8-series on TNF-alpha production was investigated using gamma9delta2 T cells, it was observed that these compounds impaired the TCR:CD3-triggered TNF-alpha production. Structure-activity analysis revealed that these properties of the drugs did not coincide with their IPDE4 properties. This prompted further exploration into other signalling mechanisms possibly involved in TNF-alpha action and production, notably the p38 MAPK and the PI3K pathway. We demonstrate here that the imidazo[1,2-alpha]quinoxalines targeted these pathways in a different way: they activated the p38 MAPK pathway whilst inhibiting the PI3K pathway. Such effects on cell signalling could account for the imidazo[1,2-alpha]quinoxalines effects on 1) action and 2) production of TNF-alpha, which define these drugs as potential anti-inflammatory agents.

Brucella spp. can establish themselves and cause disease in humans and animals. The mechanisms by which Brucella spp. evade the antibacterial defenses of their host, however, remain largely unknown. We have previously reported that live brucellae failed to induce tumor necrosis factor alpha (TNF-alpha) production upon human macrophage infection. This inhibition is associated with a nonidentified protein that is released into culture medium. Outer membrane proteins (OMPs) of gram-negative bacteria have been shown to modulate macrophage functions, including cytokine production. Thus, we have analyzed the effects of two major OMPs (Omp25 and Omp31) of Brucella suis 1330 (wild-type [WT] B. suis) on TNF-alpha production. For this purpose, omp25 and omp31 null mutants of B. suis (Deltaomp25 B. suis and Deltaomp31 B. suis, respectively) were constructed and analyzed for the ability to activate human macrophages to secrete TNF-alpha. We showed that, in contrast to WT B. suis or Deltaomp31 B. suis, Deltaomp25 B. suis induced TNF-alpha production when phagocytosed by human macrophages. The complementation of Deltaomp25 B. suis with WT omp25 (Deltaomp25-omp25 B. suis mutant) significantly reversed this effect: Deltaomp25-omp25 B. suis-infected macrophages secreted significantly less TNF-alpha than did macrophages infected with the Deltaomp25 B. suis mutant. Furthermore, pretreatment of WT B. suis with an anti-Omp25 monoclonal antibody directed against an epitope exposed at the surface of the bacteria resulted in substancial TNF-alpha production during macrophage infection. These observations demonstrated that Omp25 of B. suis is involved in the negative regulation of TNF-alpha production upon infection of human macrophages.

We studied, using organotypic hippocampal slices in culture, the role of pro-inflammatory cytokines, oxygen radicals and nitric oxide in neuronal death induced either by endotoxic insult [interferon (IFN) gamma, 24 h followed by lipopolysaccharide, 24 h] or by glutamate receptor-mediated excitotoxic insult. We demonstrated that neuronal death induced by endotoxic insult was absolutely dependent on the synthesis of tumour necrosis factor alpha (TNF-alpha). Indeed, TNF-alpha antibodies and SB203580, an inhibitor of p38 stress kinase known to block TNF-alpha and other cytokine synthesis, completely protected neurons from the endotoxic insult. Inhibiting oxygen radical and nitric oxide productions also reduced the endotoxic shock. We also showed that after priming the cultures with IFN-gamma, TNF-alpha was unable to induce neuronal death unless oxygen-free radicals were exogenously provided. In contrast, although glutamate receptor-induced excitotoxicity was associated with a low TNF-alpha synthesis and a modest activation of p38 stress kinase, neither TNF-alpha antibodies nor SB203580 were able to decrease excitotoxic neuronal insult. We did not reduce glutamate receptor-induced neuronal death with superoxide dismutase plus catalase. In conclusion, although inflammation follows glutamate receptor-mediated neurotoxicity, the mechanisms by which an endotoxic insult triggers neuronal death are different from those involved in excitotoxicity.

From undetectable basal levels, TNFalpha is produced in the hypothalamus of rats challenged with a systemic low profile endotoxin (LPS) at least 30 min before its release may be detected in the plasma. The cytotoxic activity of this hypothalamic TNFalpha correlates with its immunoreactivity. Injection of BB-1101, a matrix metalloprotease inhibitor, immediately after the LPS totally inhibits the plasma LPS-induced TNFalpha release without affecting the hypothalamic TNFalpha response. Likewise, L-NAME pretreatment, a nitric oxide inhibitor which reportedly crosses the blood-brain barrier, blunts the plasma TNFalpha response by almost 66%, but leaves the hypothalamic TNFalpha response unchanged. The present study thus describes the early occurrence of hypothalamic TNFalpha in response to systemic LPS injection, which is not subjected to the same regulations as plasma TNFalpha.

We studied the production of tumor necrosis factor alpha (TNF alpha) and interleukin-6 (IL-6) 2 and 7 days following status epilepticus (SE), induced in rats by intra-amygdala injection of kainate. At day 2 the release of both cytokines by hippocampal slices prepared from epileptic animals was increased compared to controls, whereas at day 7 only TNF alpha secretion was enhanced. Since SE-induced neuronal damage is probably due to excitotoxicity, we investigated the effects of agonists of glutamate receptors on TNF alpha release in organotypic hippocampal cultures. A correlation was found between the damage intensity and the release of TNF alpha, suggesting production of this cytokine by macrophagic microglia. We propose a role for TNF alpha and IL-6 in the adaptive phenomena which follow severe limbic seizures.

Brucella species can establish themselves and cause disease in humans, but the mechanisms by which brucellae evade the antibacterial defenses of their host remain largely unknown. We have previously reported that, unlike Escherichia coli K12, intracellular pathogens from the genus Brucella survive and multiply within U937-derived phagocytes, and live Brucella organisms failed to induce TNF-alpha release upon infection. Moreover, exogenously added TNF-alpha restricted intracellular growth of Brucella species. Herein, we demonstrate that Brucella-infected U937 cells are activated to express IL-1 beta and IL-6 at both the mRNA and protein levels, while they cannot accumulate TNF-alpha mRNA. When physically separated from macrophages, live brucellae impaired TNF-alpha production in E. coli-infected cells. Moreover, in agonist-activated macrophages, supernatants from Brucella cultures promoted an inhibition of the induction of both TNF-alpha expression and release, without affecting IL-1 beta or IL-6 induction. These phenomena, observed whatever the Brucella strain assayed, show that brucellae release some high m.w. factor(s) that specifically inhibits TNF-alpha expression in activated human macrophages. The proteic nature of the factor(s) was demonstrated by its heat and protease sensitiveness, and this could explain why U937-derived macrophages did release TNF-alpha when infected with chloramphenicol-treated brucellae. We also found that the Brucella factor(s) specifically acts on human macrophagic cells, but not on murine macrophage-like cells. Our findings provide direct evidence that a secreted Brucella virulence factor(s) inhibiting TNF-alpha expression might contribute to the evasion of Brucella organisms from human antimicrobial defenses.

Tumor necrosis factor alpha (TNF-alpha) plays a central role in activation of first-line defenses of a host against foreign organisms. To determine whether Brucella infection modulated TNF-alpha production, we measured the biological activity of this cytokine in supernatants of U937 cell-derived macrophages and of fresh human monocytes infected with Brucella spp. Neither the smooth nor rough Brucella strains used induced any measurable TNF-alpha excretion upon infection. On the contrary, as reported before for other gram-negative bacteria, phagocytosis of nonpathogenic Escherichia coli was followed by a rapid and transient induction of TNF-alpha release, suggesting an involvement of this cytokine in some autocrine process. As expected, the Brucella strains tested survived and/or multiplied within U937-derived macrophages, whereas E. coli was rapidly eliminated after phagocytosis. Immunoglobulin G opsonization of E. coli strains enhanced their intracellular killing and strongly potentiated TNF-alpha secretion. Immunoglobulin G opsonization of Brucella strains, in contrast, did not lead to TNF-alpha production, although their rate of intracellular multiplication was reduced. Killed brucellae, however, promoted a significant excretion of TNF-alpha from U937-derived macrophages into cell culture supernatants. We finally demonstrated that pretreatment of U937-derived macrophages with exogenous TNF-alpha significantly inhibited intracellular multiplication of Brucella spp. These results and experiments performed on fresh human monocytes or with isolated lipopolysaccharide (LPS) showed that (i) differences in TNF-alpha production observed during macrophage infection by Brucella spp. and E. coli were not due to differences in LPS structure but resulted from active inhibition of TNF-alpha production by a specific process linked to Brucella spp. and (ii) the capacity of Brucella spp. to use pathways avoiding TNF-alpha production during infection may be considered a major attribute of virulence.

We have previously described a synergism between the two physiological hormones, retinoic acid (RA) and 1 alpha,25-dihydroxyvitamin D3 (VD) in the induction of U937 cell differentiation towards a more mature state. Herein, we investigated the regulation of cytokine production during RA and/or VD treatment of U937 cells. Cell differentiation was followed by measurement of their capacity to give oxidative responses, and interleukin-1 beta (IL-1 beta), tumour necrosis factor-alpha (TNF-alpha) and IL-6 gene and protein expression were determined in RA/VD-treated cells, activated or not with lipopolysaccharide (LPS). The undifferentiated and RA-treated U937 cells were unable to produce monokines even when they were stimulated by LPS. VD induced the monokine mRNA expression in U937 cells but failed to induce protein release. However, unlike RA, it primed the cells to secrete monokines upon endotoxin stimulation. A large enhancement of the production of the monokines both at mRNA and protein levels was observed in the U937 cells exposed to the combination of RA + VD. Nevertheless, protein release required a further step of activation of the RA + VD-primed cells. The co-inducer effect of RA and VD was not observed in HL-60 or THP-1 cells and seems to be restricted to U937 cells. These results on cytokine expression support our previous finding that a combination of RA and VD brings the U937 cells to a high stage of myeloid differentiation with major characteristics of monocytes/macrophages.