Tumor necrosis factor alpha and prostaglandin E2 production by human monocyte-derived macrophages infected with spotted fever group rickettsiae

Significant Growth by Rickettsia Species within Human Macrophage-Like Cells Is a Phenotype Correlated with the Ability to Cause Disease in Mammals

Rickettsia are significant sources of tick-borne diseases in humans worldwide. In North America, two species in the spotted fever group of Rickettsia have been conclusively associated with disease of humans: Rickettsia rickettsii, the causative agent of Rocky Mountain spotted fever, and Rickettsia parkeri, the cause of R. parkeri rickettsiosis. Previous work in our lab demonstrated non-endothelial parasitism by another pathogenic SFG Rickettsia species, Rickettsia conorii, within THP-1-derived macrophages, and we have hypothesized that this growth characteristic may be an underappreciated aspect of rickettsial pathogenesis in mammalian hosts. In this work, we demonstrated that multiple other recognized human pathogenic species of Rickettsia, including R. rickettsii, R. parkeri, Rickettsia africae, and Rickettsiaakari can grow within target endothelial cells as well as within PMA-differentiated THP-1 cells. In contrast, Rickettsia bellii, a Rickettsia species not associated with disease of humans, and R. rickettsii strain Iowa, an avirulent derivative of pathogenic R. rickettsii, could invade both cell types but proliferate only within endothelial cells. Further analysis revealed that similar to previous studies on R. conorii, other recognized pathogenic Rickettsia species could grow within the cytosol of THP-1-derived macrophages and avoided localization with two different markers of lysosomal compartments; LAMP-2 and cathepsin D. R. bellii, on the other hand, demonstrated significant co-localization with lysosomal compartments. Collectively, these findings suggest that the ability of pathogenic rickettsial species to establish a niche within macrophage-like cells could be an important factor in their ability to cause disease in mammals. These findings also suggest that analysis of growth within mammalian phagocytic cells may be useful to predict the pathogenic potential of newly isolated and identified Rickettsia species.

Characteristics of in vitro infection of human monocytes, by Rickettsia helvetica

Eighteen species of rickettsiae are reported to cause infections in humans. One of these is Rickettsia helvetica, which is endemic in European and Asian countries and transmitted by the tick Ixodes ricinus. Besides fever, it has been demonstrated to cause meningitis and is also associated with perimyocarditis. One of the initial targets for rickettsiae after inoculation by ticks is the macrophage/monocyte. How rickettsiae remain in the macrophages/monocytes before establishing their infection in vascular endothelial cells remains poorly understood. The main aim of the present study was to investigate the impact on and survival of R. helvetica in a human leukemic monocytic cell line, THP-1. Our results show that R. helvetica survives and propagates in the THP-1 cells. The infection in monocytes was followed for seven days by qPCR and for 30 days by TEM, where invasion of the nucleus was also observed as well as double membrane vacuoles containing rickettsiae, a finding suggesting that R. helvetica might induce autophagy at the early stage of infection. Infected monocytes induced TNF-α which may be important in host defence against rickettsial infections and promote cell survival and inhibiting cell death by apoptosis. The present findings illustrate the importance of monocytes to the pathogenesis of rickettsial disease.

Enhanced Expression of T-Cell Immunoglobulin and Mucin Domain Protein 3 in Endothelial Cells Facilitates Intracellular Killing of Rickettsia heilongjiangensis

Rickettsia heilongjiangensis is the pathogen of Far eastern spotted fever, and T-cell immunoglobulin and mucin domain protein 3 (Tim-3) is expressed in human vascular endothelial cells, the major target cells of rickettsiae. In the present study, we investigated the effects of altered Tim-3 expression in vivo in mice and in vitro in human endothelial cells, on day 3 after R. heilongjiangensis infection. Compared with corresponding controls, rickettsial burdens both in vivo and in vitro were significantly higher with blocked Tim-3 signaling or silenced Tim-3 and significantly lower with overexpressed Tim-3. Additionally, the expression of inducible nitric oxide synthase and interferon γ in endothelial cells with blocked Tim-3 signaling or silenced Tim-3 was significantly lower, while the expression of inducible nitric oxide synthase, interferon γ, and tumor necrosis factor α in transgenic mice with Tim-3 overexpression was significantly higher. These results reveal that enhanced Tim-3 expression facilitates intracellular rickettsial killing in a nitric oxide-dependent manner in endothelial cells during the early phase of rickettsial infection.

Orientia tsutsugamushi suppresses the production of inflammatory cytokines induced by its own heat-stable component in murine macrophages

Orientia tsutsugamushi is a Gram-negative obligate intracellular bacterium, which causes scrub typhus. To understand the pathogenesis of scrub typhus, we have investigated the induction of tumor necrosis alpha (TNF-alpha) and interleukin-6 (IL-6) by O. tsutsugamushi in two murine macrophage cell lines. Both live and heat-killed orientia stimulated the production of cytokines in J774A.1 cells. Polymyxin B does not affect the secretion of cytokines. These together with the fact that the immature macrophage cell line, P388D1, did not produce TNF-alpha when induced by either live or heat-killed O. tsutsugamushi strongly argue against any roles of lipopolysaccharide (LPS) in cytokine production. Furthermore, the result that the cytokine responses were more brisk when macrophage cell lines had been induced by heat-killed O. tsutsugamushi than by live organisms strongly suggest that a heat-stable molecule might be responsible for the induction of cytokine production and O. tsutsugamushi might have mechanisms suppressing the production of inflammatory cytokines induced by its own heat-stable molecule.

Plasma levels of tumor necrosis factor-alpha and interferon-gamma in Sicilian children with Mediterranean spotted fever

The plasma levels of tumor necrosis factor-alpha and interferon-gamma were measured in 53 consecutive children with serologically confirmed Mediterranean spotted fever and were found to be increased during the acute phase compared with the convalescent phase (tumor necrosis factor-alpha mean 32.17 vs. 4.12 pg/ml, P < 0.0001; interferon-gamma mean 84.17 vs. 2.65 pg/ml, P = 0.0006). Plasma levels of both cytokines were higher in patients with a typical exanthema rather than those with a very mild or no exanthema; tumor necrosis factor-alpha levels were significantly lower in the latter (tumor necrosis factor-alpha 32.17 vs. 9.85 pg/ml, P < 0.0001; interferon-gamma 84.17 vs. 38.14 pg/ml, P = 0.35). Tumor necrosis factor-alpha and interferon-gamma may be harmful or beneficial to the infected host, depending upon the amounts produced and whether they are circulating or confined locally to the site of inflammation.

Absence of tumor necrosis factor alpha, interleukin-6 (IL-6), and granulocyte-macrophage colony-stimulating factor expression but presence of IL-1beta, IL-8, and IL-10 expression in human monocytes exposed to viable or killed Ehrlichia chaffeensis

Ehrlichia chaffeensis is a recently isolated minute gram-negative obligatory intracellular bacterium of monocytes/macrophages and is the etiologic agent of human monocytic ehrlichiosis. It is not known how macrophages respond when they encounter ehrlichiae in terms of cytokine production. In this study, we examined cytokine mRNA expression by incubating E. chaffeensis with THP-1 cells and performing competitive reverse transcription-PCR (RT-PCR). At 2 h postinfection, the levels of interleukin-1beta (IL-1beta), IL-8, and IL-10 mRNAs were significant but lower than those following Escherichia coli lipopolysaccharide (LPS) stimulation. Unlike the situation with E. coli LPS stimulation, however, IL-6, granulocyte-macrophage colony-stimulating factor (GM-CSF), and tumor necrosis factor alpha (TNF-alpha) mRNAs were not induced. Time course and dose-response studies confirmed the absence of IL-6, GM-CSF, and TNF-alpha mRNA induction with E. chaffeensis. Viable E. chaffeensis organisms were not required for IL-1beta IO, IL-8, and IL-10 mRNA induction, since heat-killed E. chaffeensis induced identical time course responses. IL-1beta, IL-8, and IL-10 mRNAs were detected for up to 21, 21, and 24 h postexposure with E. chaffeensis, respectively, which were shut off more rapidly than with LPS stimulation. Although heat treatment of E. chaffeensis had no effect, periodate treatment completely abolished the ability of E. chaffeensis to induce IL-1beta, IL-8, and IL-10 mRNAs. The capture enzyme-linked immunosorbent assay result corresponded with the RT-PCR results, showing that viable and heat-killed E. chaffeensis produced and secreted the same levels of IL-1beta and IL-8. IL-10 production was significantly reduced by heat treatment. Periodate-treated ehrlichiae did not induce production of any of the cytokines tested. Anti-CD14 monoclonal antibody and polymyxin B did not inhibit IL-1beta mRNA expression upon exposure to E. chaffeensis. The absence of TNF-alpha, IL-6, and GM-CSF mRNA induction may delay the development of a protective immune response, thereby allowing E. chaffeensis to set up residence in macrophages.

Release of soluble tumor necrosis factor receptors in Mediterranean spotted fever rickettsiosis

Tumor necrosis factor alpha (TNF) is a key cytokine in the defense against many intracellular pathogens, including Rickettsia conorii, the causative agent of Mediterranean spotted fever (MSF). The levels of two soluble TNF receptors (sTNFR-p55 and sTNFR-p75), the extracellular domains of the two cell surface receptors for TNF, were elevated in the acute-stage plasma samples from 20 patients with serologically confirmed MSF. The median values were 3.1 and 7.8 ng/ml for sTNFR-p55 and sTNFR-p75, respectively. sTNFR values correlated significantly with plasma TNF concentrations. Patients with severe MSF had higher values for both receptor fragments than patients with nonsevere disease. The differences were statistically significant for sTNFR-p55 (median, 5.8 versus 2.0 ng/ml; P = 0.008). Given the proportionately higher values for both TNF and sTNFR-p55 in patients with severe MSF, the sTNFR-p55/TNF ratios for the two patient subgroups did not differ (P = 0.5), while the sTNFR-p75/TNF ratios were significantly different (P = 0.01), with disproportionately lower values in patients with severe disease.

IL-6 and IL-8 production from cultured human endothelial cells stimulated by infection with Rickettsia conorii via a cell-associated IL-1 alpha-dependent pathway

Mediterranean spotted fever due to infection by Rickettsia conorii, is characterized by a general vasculitis. This vasculitis is thought to be due to a direct injury to endothelial cells induced by R. conorii. However, production and activity of cytokines on endothelial cells is an important pathway in inflammation, and part of the underlying mechanism of vasculitis. In the present studies, human umbilical vein endothelial cells (HUVEC) infected with R. conorii actively secrete high levels of IL-8 and IL-6 (P < 0.002, and P < 0.03, respectively, compared with uninfected cells). IL-1alpha, IL-1beta, or TNFalpha were not detected in the culture supernates. Nevertheless, IL-6 and IL-8 production was due, in a large part, to a cell-associated form of IL-1 alpha expressed on R. conorii-infected HUVEC, since production of these cytokines was suppressed by 80% (P = 0.0001) and 85% (P < 0.04) by the addition of IL-1 receptor antagonist, or anti-IL-1alpha antibodies (60% inhibition, P < 0.01 and 65% inhibition, P < 0.05, respectively) and IL-1alpha was measured after lysis of R. conorii-infected HUVEC but not in uninfected cells (P < 0.01). Rickettsial lipopolysaccharide does not seem to be involved, since polymyxin B did not reduce cytokine secretion. On the contrary, infection by intracellular R. conorii appears to be necessary to induce IL-1alpha and subsequently IL-8, since formalin-fixed R. conorii did not induce cytokine production. These observations demonstrate that R. conorii-infected HUVEC secrete IL-6 and IL-8 via the induction of cell-associated IL-1alpha, providing a possible mechanism for the vasculitis observed in Mediterranean spotted fever.

The role of tumor necrosis factor in host defense against scrub typhus rickettsiae. II. Differential induction of tumor necrosis factor-alpha production by Rickettsia tsutsugamushi and Rickettsia conorii

The present study was undertaken to investigate the ability of members of two different groups of Rickettsia to stimulate macrophages or immune lymphocytes to produce TNF. It was found that R. conorii, a spotted fever group rickettsia, readily induced murine peritoneal macrophages or the macrophage-like cell line P388D1 to produce relatively high levels of TNF. The interaction of macrophages with viable organisms or heat-killed organisms resulted in TNF production. In contrast, viable or killed R. tsutsugamushi did not stimulate the production of detectable TNF even though viable organisms grew to high numbers in both cell types. It was found that the appropriate immune spleen cells stimulated with heat-killed R. tsutsugamushi or R. conorii produced TNF, and TNF activity was found in the sera of immune mice after injection with rickettsial antigen. Infection of naive mice with viable R. tsutsugamushi resulted in high TNF levels in ascites, but TNF was not found in ascites obtained from infected athymic (nu/nu) mice. These data support the suggestion that spotted fever group rickettsiae, such as R. conorii, possess components perhaps on the surface that interact with macrophages to induce TNF production and this component is lacking in R. tsutsugamushi. Antigens of R. tsutsugamushi and R. conorii will stimulate immune cells to produce TNF activity. These data are compatible with the suggestion that the TH-1 subset of T cells is predominant in immunity to R. tsutsugamushi.

The role of tumor necrosis factor in host defense against scrub typhus rickettsiae. I. Inhibition of growth of Rickettsia tsutsugamushi, Karp strain, in cultured murine embryonic cells and macrophages by recombinant tumor necrosis factor-alpha

Recombinant murine tumor necrosis factor-alpha (TNF-alpha) inhibited intracellular growth of Rickettsia tsutsugamushi, Karp strain, in the mouse embryo cell line C3H/10T1/2 clone 8 at doses of 100 to 10 U/ml. The growth inhibitory effect of TNF-alpha was also evident when peritoneal exudate macrophages or bone marrow-derived macrophages were used as the host cell for rickettsial growth. Interferon-gamma (IFN-gamma), at doses up to 1,000 U/ml, did not affect the growth of this strain of rickettsiae in the mouse embryo cell line but, as expected, profoundly inhibited rickettsial growth in peritoneal exudate macrophages and bone marrow-derived macrophages. The effect of TNF-alpha on rickettsial growth in the mouse embryo cell line was not reproducibly enhanced by IFN-gamma. Treatment of the cell line with TNF-alpha delayed rickettsial cytopathic effects, but the rickettsiae ultimately grew to high numbers in the cells and caused cell death. These findings show that, at least in our system, R. tsutsugamushi is resistant to IFN-gamma-mediated antirickettsial effects in cells other than macrophages. The results of this study support the suggestion that TNF-alpha may inhibit rickettsial growth in cells other than macrophages.

Relationship of tumor necrosis factor alpha, the nitric oxide synthase pathway, and lipopolysaccharide to the killing of gamma interferon-treated macrophage-like RAW264.7 cells by Rickettsia prowazekii

Macrophage-like RAW264.7 cells are killed by the combination of gamma interferon (IFN-gamma) treatment and infection with Rickettsia prowazekii. The roles of tumor necrosis factor alpha (TNF-alpha), the nitric oxide synthase pathway, and lipopolysaccharide (LPS) in this killing were investigated. R. prowazekii, both the Breinl and Madrid E strains, induced RAW264.7 cells to produce TNF-alpha. However, dead rickettsiae (which cannot kill the IFN-gamma-treated RAW264.7 cells) induced the production of as much TNF-alpha as viable rickettsiae. Inhibition of the production of TNF-alpha (by the addition of actinomycin D or emetine during the rickettsial infection) or neutralization of TNF-alpha (by the addition of polyclonal rabbit anti-mouse TNF-alpha serum both during the IFN-gamma treatment and during the rickettsial infection) did not inhibit the killing of the RAW264.7 cells. Addition of polymyxin B (which inhibits many effects of LPS) during the IFN-gamma treatment did not inhibit the ability of IFN-gamma to prepare the RAW264.7 cells to be killed by R. prowazekii. Suppression of nitrite production by addition of the nitric oxide synthase inhibitor aminoguanidine both during the IFN-gamma treatment and during the rickettsial infection also did not inhibit the killing of the RAW264.7 cells. R. prowazekii-mediated killing of the RAW264.7 cells was dramatically suppressed in cultures treated with IFN-gamma plus LPS compared with that in cultures treated with IFN-gamma alone, and inhibition of nitric oxide synthase restored the rickettsia-induced killing of the RAW264.7 cells in cultures treated with IFN-gamma plus LPS. These data indicate that (i) TNF-alpha, LPS, and the nitric oxide synthase pathway are not required in order for IFN-gamma to prepare RAW264.7 cells to be killed by R. prowazekii; (ii) neither TNF-alpha nor the nitric oxide synthase pathway is responsible for the killing of the IFN-gamma-treated RAW264.7 cells by R. prowazekii; and (iii) in cultures treated with IFN-gamma plus LPS and then incubated with rickettsiae, a nitric oxide synthase pathway-dependent mechanism inhibits the killing of the RAW264.7 cells.

Depletion of gamma interferon and tumor necrosis factor alpha in mice with Rickettsia conorii-infected endothelium: impairment of rickettsicidal nitric oxide production resulting in fatal, overwhelming rickettsial disease

C3H/HeN mice infected intravenously with a dose of Rickettsia conorii (Malish 7 strain) that is sublethal for immunocompetent animals (1.1 x 10(3) PFU) developed disseminated infection of endothelial cells of the brain, lungs, heart, liver, kidney, testis, and testicular adnexa. In R. conorii-infected mice depleted of gamma interferon (IFN-gamma) and/or tumor necrosis factor alpha (TNF-alpha) by intravenous administration of neutralizing monoclonal antibodies on days 0, 2, and 4, the mortality rate was 100%. Death of the cytokine-depleted animals on days 5 and 6 was associated with overwhelming rickettsial infection documented by titration of rickettsial content in the brain and liver and by immunohistologic demonstration of massive quantities of R. conorii in endothelial cells of all organs examined, in macrophages of the liver and spleen, and in hepatocytes. Nondepleted, immunocompetent animals showed markedly reduced rickettsial content in the tissues on day 6, with rickettsial destruction in phagolysosomes not only in macrophages but also in endothelial cells and hepatocytes. All nondepleted, infected mice recovered and appeared completely healthy by day 9. Assay of liver infiltrated by lymphocytes and macrophages revealed mRNA of IFN-gamma and TNF-alpha, indicating that the host defenses were activated at the site of infection. Treatment of mice with an analog of L-arginine reduced the synthesis of nitric oxide and impaired rickettsial killing. Nitric oxide production was also impaired in cytokine-depleted infected mice. These observations support the hypothesis that IFN-gamma secreted by T lymphocytes and natural killer cells and TNF-alpha secreted by macrophages act in a synergistic, paracrine fashion on adjacent rickettsia-infected endothelial cells, hepatocytes, and macrophages to stimulate synthesis of nitric oxide, which kills intracellular R. conorii.

Tumor necrosis factor alpha, interleukin-1 alpha, interleukin-6, and prostaglandin E2 production in murine peritoneal macrophages infected with Ehrlichia risticii

Ehrlichia risticii is a gram-negative obligate intracellular bacterium which primarily infects macrophages and crypt epithelial cells in the intestinal wall and is the etiologic agent of Potomac horse fever. To understand the pathogenesis of the disease, we tested whether E. risticii induces inflammation-associated products in thioglycolate-induced mouse peritoneal macrophages. Mouse peritoneal macrophages produced larger amounts of interleukin-1 alpha (IL-1 alpha) but lower levels of tumor necrosis factor alpha (TNF-alpha), IL-6, and prostaglandin E2 (PGE2) when exposed to live or killed E. risticii than when exposed to Escherichia coli lipopolysaccharide (LPS). Preincubation of macrophages with live or killed E. risticii suppressed TNF-alpha, IL-6, and PGE2 generation but not IL-1 alpha production in response to LPS. Murine gamma interferon treatment of macrophages did not influence TNF-alpha, IL-1 alpha, IL-6, or PGE2 production regardless of exposure to E. risticii. Intracellular cyclic AMP was significantly greater in E. risticii-infected macrophages than in uninfected macrophages. These results suggest that increased levels of IL-1 alpha but not TNF-alpha or PGE2 production by macrophages may be primarily involved in the pathogenesis of the disease caused by E. risticii. Increased intracellular concentration of cyclic AMP in infected macrophages may be chiefly responsible for the high level of IL-1 alpha and inhibition of TNF-alpha production in response to LPS.

Rickettsial stimulation of endothelial platelet-activating factor synthesis

Endothelial cells (EC) synthesize platelet-activating factor (PAF) when activated by agents such as ATP or thrombin, and PAF production occurs as a consequence of endothelial phospholipase A activity. Because interactions between Rickettsia prowazekii and a variety of host cells result in the expression of phospholipase A activity, we assessed the relative abilities of uninfected and rickettsia-infected EC to synthesize PAF. Endothelial cells were infected with rickettsiae and examined at 24-h intervals for rickettsial multiplication, EC viability, and PAF synthesis. By 24 h postinfection, 80% of the EC were infected with an average of 10.6 rickettsiae per cell; by 72 h, the rickettsiae were too numerous to count and the numbers of viable EC began to decrease. Both rickettsia-infected and sham-treated EC synthesized PAF when stimulated with either thrombin or ATP, but rickettsia-infected EC synthesized about three times as much PAF in response to cell activation as did their uninfected counterparts. Additionally, unlike their uninfected counterparts, rickettsia-infected EC synthesized significant amounts of PAF in the absence of cell activation; rickettsia-infected EC synthesized as much PAF in the absence of activation as did uninfected EC in response to ATP. In each case, essentially all of the newly synthesized PAF remained with the cell pellet. Finally, EC incubated with high numbers of rickettsiae (1,000 rickettsiae per EC) for 30 min synthesized more PAF when activated with ATP than did their sham-treated activated counterparts.