Necrotic cell-derived high mobility group box 1 attracts antigen-presenting cells but inhibits hepatocyte growth factor-mediated tropism of mesenchymal stem cells for apoptotic cell death.

Tissue damage due to apoptotic or necrotic cell death typically initiates distinct cellular responses, leading either directly to tissue repair and regeneration or to immunological processes first, to clear the site, for example, of potentially damage-inducing agents. Mesenchymal stem cells (MSC) as well as immature dendritic cells (iDC) and monocytes migrate to injured tissues. MSC have regenerative capacity, whereas monocytes and iDC have a critical role in inflammation and induction of immune responses, including autoimmunity after tissue damage. Here, we investigated the influence of apoptotic and necrotic cell death on recruitment of MSC, monocytes and iDC, and identified hepatocyte growth factor (HGF) and the alarmin high mobility group box 1 (HMGB1) as key factors differentially regulating these migratory responses. MSC, but not monocytes or iDC, were attracted by apoptotic cardiomyocytic and neuronal cells, whereas necrosis induced migration of monocytes and iDC, but not of MSC. Only apoptotic cell death resulted in HGF production and HGF-mediated migration of MSC towards the apoptotic targets. In contrast, HMGB1 was predominantly released by the necrotic cells and mediated recruitment of monocytes and iDC via the receptor of advanced glycation end products. Moreover, necrotic cardiomyocytic and neuronal cells caused an HMGB1/toll-like receptor-4-dependent inhibition of MSC migration towards apoptosis or HGF, while recruitment of monocytes and iDC by necrosis or HMGB1 was not affected by apoptotic cells or HGF. Thus, the type of cell death differentially regulates recruitment of either MSC or monocytes and iDC through HGF and HMGB1, respectively, with a dominant, HMGB1-mediated role of necrosis in determining tropism after tissue injury.

Antimicrobial effects of murine mesenchymal stromal cells directed against Toxoplasma gondii and Neospora caninum: role of immunity-related GTPases (IRGs) and guanylate-binding proteins (GBPs).

Mesenchymal stromal cells (MSCs) have a multilineage differentiation potential and provide immunosuppressive and antimicrobial functions. Murine as well as human MSCs restrict the proliferation of T cells. However, species-specific differences in the underlying molecular mechanisms have been described. Here, we analyzed the antiparasitic effector mechanisms active in murine MSCs. Murine MSCs, in contrast to human MSCs, could not restrict the growth of a highly virulent strain of Toxoplasma gondii (BK) after stimulation with IFN-γ. However, the growth of a type II strain of T. gondii (ME49) was strongly inhibited by IFN-γ-activated murine MSCs. Immunity-related GTPases (IRGs) as well as guanylate-binding proteins (GBPs) contributed to this antiparasitic effect. Further analysis showed that IFN-γ-activated mMSCs also inhibit the growth of Neospora caninum, a parasite belonging to the apicomplexan group as well. Detailed studies with murine IFN-γ-activated MSC indicated an involvement in IRGs like Irga6, Irgb6 and Irgd in the inhibition of N. caninum. Additional data showed that, furthermore, GBPs like mGBP1 and mGBP2 could have played a role in the anti-N. caninum effect of murine MSCs. These data underline that MSCs, in addition to their regenerative and immunosuppressive activity, function as antiparasitic effector cells as well. However, IRGs are not present in the human genome, indicating a species-specific difference in anti-T. gondii and anti-N. caninum effect between human and murine MSCs.

Human but not murine multipotent mesenchymal stromal cells exhibit broad-spectrum antimicrobial effector function mediated by indoleamine 2,3-dioxygenase.

Human multipotent mesenchymal stromal cells (MSCs) exhibit multilineage differentiation potential, support hematopoiesis, and inhibit proliferation and effector function of various immune cells. On the basis of these properties, MSC are currently under clinical investigation in a range of therapeutic applications including tissue repair and immune-mediated disorders such as graft-versus-host-disease refractory to pharmacological immunosuppression. Although initial clinical results appear promising, there are significant concerns that application of MSC might inadvertently suppress antimicrobial immunity with an increased risk of infection. We demonstrate here that on stimulation with inflammatory cytokines human MSC exhibit broad-spectrum antimicrobial effector function directed against a range of clinically relevant bacteria, protozoal parasites and viruses. Moreover, we identify the tryptophan catabolizing enzyme indoleamine 2,3-dioxygenase (IDO) as the underlying molecular mechanism. We furthermore delineate significant differences between human and murine MSC in that murine MSC fail to express IDO and inhibit bacterial growth. Conversely, only murine but not human MSC express inducible nitric oxide synthase on cytokine stimulation thus challenging the validity of murine in vivo models for the preclinical evaluation of human MSC. Collectively, our data identify human MSC as a cellular immunosuppressant that concurrently exhibits potent antimicrobial effector function thus encouraging their further evaluation in clinical trials.

Inhibition of human herpes simplex virus type 2 by interferon gamma and tumor necrosis factor alpha is mediated by indoleamine 2,3-dioxygenase.

Genital herpes simplex virus type 2 (HSV-2) is a significant clinical problem. Infection in pregnancy may result in disseminated infection of the newborn with encephalitis. We analyzed the antiviral effects induced by interferon-gamma (IFN-gamma) and tumor necrosis factor alpha (TNF-alpha) in cervix carcinoma cells (HeLa) and astrocytoma cells (86HG39). We found that replication of HSV-2 in HeLa cells and in 86HG39 cells is inhibited after stimulation of the cells by IFN-gamma and TNF-alpha. The antiviral effect of IFN-gamma is enhanced in the presence of TNF-alpha, while stimulation by TNF-alpha alone did not induce antiviral activity. We found that IFN-gamma induces a strong activation of the tryptophan-degrading enzyme indoleamine 2,3-dioxygenase (IDO) and in addition, that the IFN-gamma-induced IDO activity was enhanced in the presence of TNF-alpha. Furthermore, we found that the induction of IDO activity is responsible for the inhibition of herpes simplex virus replication, since the presence of excess amounts of l-tryptophan abrogates the antiviral effect induced by IFN-gamma and the combination of IFN-gamma and TNF-alpha. We therefore conclude that the antiviral effect against HSV-2 mediated by type II interferon and TNF-alpha are dependent on IDO activation.

Role of indoleamine-2,3-dioxygenase in alpha/beta and gamma interferon-mediated antiviral effects against herpes simplex virus infections.

Gamma interferon (IFN-gamma)-mediated indoleamine-2,3-dioxygenase (IDO) activity in human astrocytoma cells and in native astrocytes was found to be responsible for the inhibition of herpes simplex virus replication. The effect is abolished in the presence of excess amounts of L-tryptophan. Both IFN-alpha and IFN-beta restricted herpes simplex virus replication in both cell types, but (in contrast to the results seen with IFN-gamma) the addition of an excess amount of L-tryptophan did not inhibit the induced antiviral effect.

Restriction of Toxoplasma gondii growth in human brain microvascular endothelial cells by activation of indoleamine 2,3-dioxygenase.

One of the first steps in the development of cerebral toxoplasmosis is the penetration of the blood-brain barrier, which is comprised of microvascular endothelial cells. We examined the capacity of human brain microvascular endothelial cells (HBMEC) to interact with Toxoplasma gondii. We found that stimulation of HBMEC with gamma interferon (IFN-gamma) resulted in the induction of toxoplasmostasis. The capacity of HBMEC to restrict Toxoplasma growth after IFN-gamma stimulation was enhanced in the presence of tumor necrosis factor alpha (TNF-alpha). In addition, we found that IFN-gamma induced a strong induction of indoleamine 2,3-dioxygenase (IDO) activity in HBMEC, and this enzyme activity was enhanced by costimulation with TNF-alpha. The addition of excess amounts of tryptophan to the HBMEC cultures resulted in a complete abrogation of the IFN-gamma-TNF-alpha-mediated toxoplasmostasis. We therefore conclude that IDO induction contributed to the antiparasitic effector mechanism inducible in HBMEC by IFN-gamma and TNF-alpha.

Potential role of human brain microvascular endothelial cells in the pathogenesis of brain abscess: inhibition of Staphylococcus aureus by activation of indoleamine 2,3-dioxygenase.

Cerebral abscess is a rare complication of staphylococcal septicemia in infants associated with high mortality and morbidity. In the pathogenesis of abscess formation, S. aureus, one major causative agent, interacts with endothelial cells of the brain vessels before reaching the central nervous system. This study examined the growth of S. aureus in human brain microvascular endothelial cells (HBMEC) cultures stimulated with cytokines. IFN-gamma inhibited S. aureus replication by the induction of indoleamine 2,3-dioxygenase (IDO) in HBMEC. This activation of IDO in HBMEC could be shown by RT-PCR and by detection of kynurenine in culture supernatants of activated cells. Resupplementation of L-tryptophan abrogated the inhibitory effect of IFN-gamma on the growth of staphylococci, hence confirming the activation of indoleamine 2,3-dioxygenase as being responsible for the induced bacteriostasis. Addition of TNF-alpha enhanced the IFN-gamma mediated antibacterial effects, whereas TNF-alpha alone had no influence on staphylococcal growth. Stimulation of HBMEC with IFN-gamma failed to activate inducible nitric oxide synthase (iNOS) and subsequent production of nitric oxide (NO). Thus, intra- and extracellular depletion of L-tryptophan seems to be an important process in the defense against staphylococcal brain abscesses by means of creating an unfavorable microenvironment.

Inhibition of indoleamine 2,3-dioxygenase in human macrophages inhibits interferon-gamma-induced bacteriostasis but does not abrogate toxoplasmastasis.

Induction of indoleamine 2,3-dioxygenase (IDO) by IFN-gamma results in growth inhibition of Toxoplasma and Chlamydia spp. as well as tumor cells. This is caused by the degradation, and therefore depletion, of L-tryptophan necessary for cell protein synthesis. Human macrophages stimulated with IFN-gamma express IDO and inhibit the growth of intracellular toxoplasma and chlamydia as well as that of extracellular bacteria such as group B streptococci. Here we describe experiments in which the L-tryptophan analog, 6-chloro-DL-tryptophan (CDLT) caused a dose-dependent inhibition in the IFN-gamma-induced IDO-mediated L-tryptophan degradation in monocyte-derived macrophages and glioblastoma cells. An inhibition of IDO activity of up to 80 % was observed at concentrations of CDLT of 750 microM. Expression of IDO at this concentration, as shown by Northern blot analysis, was unimpaired. This inhibition of IDO was coupled in glioblastoma cells by a complete abrogation of the IFN-gamma-induced toxoplasmastasis in these cells. IDO inhibition by CDLT in human macrophages resulted in a complete abrogation of the IFN-gamma-induced growth inhibition of streptococci and staphylococci. In contrast to this, IFN-gamma-induced toxoplasmastasis was not inhibited in human macrophages by CDLT-mediated IDO inhibition.

Interleukin-1 inhibits gamma interferon-induced bacteriostasis in human uroepithelial cells.

The most prominent gamma interferon (IFN-gamma)-induced antimicrobial effector mechanisms are the induction of nitric oxide (NO) synthase (NOS) and of indoleamine 2,3-dioxygenase (IDO) activity. We have recently found that human glioblastoma cells and human macrophages inhibit the growth of group B streptococci after stimulation with IFN-gamma. In this report, we show that in addition, human RT4 (uroepithelial) cells can inhibit the growth of enterococci. Murine macrophages (RAW cells) are unable to inhibit bacterial growth after IFN-gamma stimulation. Stimulation of human glioblastoma cells, macrophages, and RT4 cells with human IFN-gamma results in a strong expression of IDO activity; however, NO production remains undetectable. In strong contrast, murine RAW cells produce large amounts of NO when stimulated with murine IFN-gamma and IDO activity is not detectable. Interleukin-1 (IL-1) induces NO synthase in human RT4 cells when the cells are costimulated with IFN-gamma. We found that IL-1 inhibits IFN-gamma-stimulated IDO activity and antimicrobial effects in RT4 cells, while in human glioblastoma cells, which lack detectable NO synthase activity, neither of these effects was altered by costimulation with IFN-gamma and IL-1. The IL-1-mediated inhibition of IDO activity and of subsequent antibacterial effect is due to the production of NO. This conclusion was supported by evidence that N(G)-monomethyl-L-arginine, a competitive inhibitor of inducible NOS activity, is able to block the inhibitory action of IL-1 on IFN-gamma-induced bacteriostasis. We therefore conclude that NO production does not inhibit the growth of enterococci but might be involved in the regulation of IDO activity in some human cells.

Inducible anti-parasitic effector mechanisms in human uroepithelial cells: tryptophan degradation vs. NO production.

In murine cells the most important effector mechanism directed against the intracellular pathogen Toxoplasma gondii is the production of toxic nitrogen oxides. In contrast the induction of the tryptophan degrading enzyme indolamine 2,3-dioxygenase (IDO) has been described to be the most effective anti-parasitic mechanism in most human cells. In this report we analysed IDO induction and NO production in the human uroepithelial carcinoma cell line RT4. We found that after stimulation with IFN-gamma these cells were able to restrict toxoplasma growth. This was due to an activation of IDO, and the anti-parasitic effect mediated by RT4 cells was abrogated by the addition of L-tryptophan. In addition we found that the costimulation of RT4 cells with IL-1 and IFN-gamma results in the production of nitric oxide, and that in RT4 cells stimulated with both these cytokines, IDO activity and toxoplasmostasis was lower than in cells stimulated with IFN-gamma alone. This IL-1-mediated inhibition of IFN-gamma-induced IDO activity and toxoplasmostasis could be blocked by monomethyl L-arginine, an inhibitor of NO production. We therefore conclude that the induction of indolamine 2,3-dioxygenase activity in human cells is a very important effector mechanism directed against Toxoplasma gondii, and that in human cells the production of NO might be involved in the regulation of IDO activity.

IFN-gamma activated indoleamine 2,3-dioxygenase activity in human cells is an antiparasitic and an antibacterial effector mechanism.

In nearly all human cells IFN-gamma stimulation leads to an activation of indoleamine 2,3-dioxygenase (IDO) activity, which is responsible for anti-toxoplasma and anti-chlamydia effects. We have recently shown that IDO activation is also a defense mechanism against extracellular beta-hemolytic streptococci groups A, B, C and G in human glioblastoma cells, fibroblasts and macrophages. Similar effects were also seen with enterococci and in approximately 65% of staphylococci tested, including multiresistant strains of both species. In addition, we have found that IDO activity is differentially regulated in different cells. For example we have found that TNF-alpha enhances IFN-gamma induced IDO activity and antimicrobial effect in human glioblastoma cells whereas both IFN-gamma mediated effects were blocked by TNF-alpha as well as by IL-1 in a human uroepithelial cell line. We were able to show that the IL-1 and TNF-alpha mediated inhibition of IFN-gamma-induced IDO activity in uroepithelial cells is due to stimulation of inducible nitric oxide synthase. In human astrocytoma cells, IL-1 and TNF-alpha did not inhibit IDO activity and in concordance with this finding these cells did not show a detectable nitric oxide production.

Cytokine mediated regulation of interferon-gamma-induced IDO activation.

Stimulation of human monocyte-derived-macrophages (MDM) with interferon gamma induces the L-tryptophan degrading enzyme indoleamine 2,3-dioxygenase (IDO). It has been well documented that the growth of some intra-cellular parasites such as Chlamydia and Toxoplasma in human fibroblasts and glioblastoma cells is inhibited by IDO mediated L-tryptophan depletion. We have recently shown that IDO induction in cord blood MDM is also responsible for the growth inhibition of extra-cellular group B streptococci and thus for the first time shown an anti-bacterial effect of IDO activation. In view of this immunological function we sought to investigate the regulation, and in particular the downregulation of IDO by the immune system. We describe here the effect of cytokines on IDO activation and in particular the inhibitory function of IL-10, TGF beta and IL-4.

Inhibition of group B streptococcal growth by IFN gamma-activated human glioblastoma cells.

Group B streptococci are the most important bacteria inducing neonatal septicemia and meningitis. The aim of this study was to assess the role of IFNgamma in the induction of anti-microbial effector mechanisms in human brain tumor cells. Different human glioblastoma/astrocytoma cell lines, stimulated with IFNgamma, restricted the growth of group B streptococci. In addition, we found that TNF alpha is able to enhance the IFNgamma-mediated anti-microbial effect. In contrast to group B streptococci, other bacteria which are also capable of inducing meningitis, like E. coli and all but one of the tested Streptococcus pneumoniae strains, were not influenced by the IFNgamma treated cells. We found that the IFNgamma or the IFNgamma/TNF alpha induced activation of indoleamine 2,3-dioxygenase is responsible for the inhibition of streptococcal growth, since the addition of supplemental L-tryptophan completely blocks the IFNgamma induced bacteriostasis.

Interferon-gamma-induced activation of indoleamine 2,3-dioxygenase in cord blood monocyte-derived macrophages inhibits the growth of group B streptococci.

Neonatal sepsis is most often caused by group B streptococci (GBS) and is a major cause of death in the neonatal period. The response of the immune system in the newborn child has received much attention and is thought to be deficient in a number of ways. The effector response of neonatal monocyte-derived macrophages (MDM) was investigated. Interferon-gamma induced the activation of indoleamine 2,3-dioxygenase in MDM and inhibited the growth of GBS. Both effects were enhanced by the addition of tumor necrosis factor-alpha to the culture conditions. The coincident supplementation of L-tryptophan with the bacteria abrogated the bacterial growth inhibition, thus confirming the causative role of L-tryptophan depletion. Control of the extracellular as well as intracellular L-tryptophan levels may thus be one of the effector mechanisms with which the immune system defends the host against GBS dissemination and disease.

Studies of persistent infection by Chlamydia trachomatis serovar K in TPA-differentiated U937 cells and the role of IFN-gamma.

Inoculation of phorbol ester-differentiated U937 cells as a model for human macrophages with Chlamydia trachomatis of the urogenital serovar K resulted in a persistent infection, with maximal growth at day 7, until day 10 post-infection. At these times inclusion bodies were present in 0.5-2% of the cells. Typical inclusion bodies containing elementary bodies and reticulate bodies were observed by electron microscopy. Furthermore, single chlamydial particles resembling atypical elementary or intermediate bodies were identified in the cytoplasm in > 80% of the host cells. IFN-gamma exerts antichlamydial activity in epithelial and fibroblastoid cells, but the infection of U937 cells by C. trachomatis was not affected by IFN-gamma. The activity of the tryptophan-degrading enzyme indoleamine 2,3-dioxygenase (IDO) was not detected in untreated or in IFN-gamma-treated or chlamydiae-infected or mock-infected U937 cells. The presence of atypical persisting chlamydiae and the lack of IDO expression in U937 cells indicates that the development of these atypical bacteria is independent from IFN-gamma-mediated tryptophan deprivation and other IFN-gamma-mediated effects. Evaluation of persistently infected cells revealed that the expression of the chlamydial major outer-membrane protein, heat-shock protein (hsp60) and lipopolysaccharide (LPS) antigens was not significantly altered in the course of the culture. An intense staining of the LPS on the surface of the host cells was demonstrated by immunofluorescence. The data show that phorbol ester-differentiated U937 cells restrict chlamydial growth strongly but not completely through a mechanism distinct from IDO-mediated tryptophan deprivation. The mechanisms of persistence of chlamydiae in monocytes, which differ considerably from those described for other cells, require further investigation.

Cellular immune reactions directed against Toxoplasma gondii with special emphasis on the central nervous system.

Toxoplasma gondii is an obligate intracellular parasite which, after primary infection of humans, is maintained in a dormant state by the host cellular immune system. In the event of an acquired immunosuppression, those parasites surviving as dormant cysts in the host may undergo a change in status, proliferate and cause a life-threatening toxoplasmic encephalitis. Over the last decade much knowledge has accumulated concerning the immune response against T. gondii. This review focuses attention particularly on the anti-parasitic effector mechanisms and the cellular immune reactions in the central nervous system during the course of reactivated toxoplasmic encephalitis.

Protamine enhances the activity of human recombinant interferon-gamma.

Recombinant interferon-gamma (IFN-gamma) is a potent immune regulatory cytokine and is involved in the defense against several intracellular organisms, such as Chlamydia and Toxoplasma. Furthermore IFN-gamma is able to inhibit the growth of human tumor cell lines. The ability to inhibit the growth of intracellular organisms makes the therapeutic use of recombinant human IFN-gamma in certain patient groups, such as those with chronic granulomatous disease, leprosy, and HIV infection, very attractive. We have shown recently that IFN-gamma-mediated effects can be blocked by heparin and that this inhibitory effect can be abrogated by the addition of protamine. In this report, we show that the antagonistic effect of protamine on heparin-mediated inhibition of IFN-gamma activity is mainly due to the capacity of protamine to enhance IFN-gamma activity. We found that protamine enhances the capacity of IFN-gamma to inhibit the growth of different brain tumor cell lines, to induce indolamine 2, 3-dioxygenase activity, to induce toxoplasmostasis, and to induce MHC class II antigen expression in human glioblastoma cells and in human native fibroblasts. We were able to demonstrate that IFN-gamma binds to protamine, and, therefore, we assume that the effect of protamine on IFN-gamma is due to a direct interaction between the two molecules.

Hematopoietic transplant potential of unrelated cord blood: critical issues.

To date, hematopoietic stem and progenitor cells from human umbilical cord blood (CB) have been employed in approximately 90 allogeneic (56 sibling and 34 unrelated) matched and mismatched transplantations worldwide with easy and successful restoration of hematopoiesis. Requests for stem cell preparations from CB will continue to increase. Thus, as a pilot study, the examination and standardization of unrelated cord blood-derived stem cell preparations and banking as well as their biologic characterization were initiated. Up to October 1995, a total of 574 samples [mean volume 79 +/- 26 ml, total nucleated cells (NC) 8.5 +/- 5 x 10(8), BFU-E 9.5 +/- 8.6 x 10(5), CFU-GM 5.7 +/- 6.3 x 10(5), CFU-GEMM 1.6 +/- 1.9 x 10(5)] from cord-derived or placental-derived residual blood have been defined by hematologic, immunologic, and microbiologic criteria. These CB samples were collected from the umbilical cord vein immediately after vaginal full-term delivery (n = 450) or cesarean section (n = 124) and stored frozen in liquid nitrogen. Seven percent of all samples collected could not be considered for potential transplants because of volumes < 40 ml. Only 5.0 ml of a CB sample is required for routine laboratory testing, consisting of HLA class I typing, HLA class II typing by sequence-specific oligonucleotide probes (PCR-SSOP), ABO typing, sterility control, assessment of progenitor and stem cells by colony-forming and LTC-IC assays, and CD34+ status. To assess the potential problem of contaminating maternal cells, a PCR was performed on 7 representative samples. During the initial 6 months of the unrelated CB collection program, a median bacterial contamination rate of 18% (20% skin flora species, 80% perineal flora species) was encountered, which has since been reduced to < 1% through practical experience. With regard to viral infections, maternal sera was tested for HBsAg (0.6% positive), anti-HCV (0%), anti-HAV (IgG 18%, IgM 0%), anti-HIV-1-2 (0%), anti-EBV (IgG 98%, IgM 0%), anti-HTLVI-II (0%), anti-CMV (IgG 43%, IgM 0.4%), toxoplasmosis (46%) and syphilis (0%). In addition, all cord blood samples were tested by PCR for CMV infection. With regard to its clinical relevance, it is important that only 0.3% of all the samples were positive for CMV by this sensitive method. This may represent a critical advantage of CB grafts over bone marrow (BM) since, in contrast, > 40% of the unrelated BM donors have been identified to be positive for CMV. An additional advantage of CB is that since 20% of CB samples were collected from ethnic minorities, it appears possible to balance common HLA types and uncommon HLA types represented in this group. In summary, with the extensive practical experience of the obstetric collection team as well as the stem cell-processing laboratory, it appears feasible to obtain a 90% yield of unrelated CB-derived stem cell preparations for banking, which clearly should meet the medical and regulatory qualification criteria required for clinical transplantation. To test the feasibility of hematopoietic transplant potential of unrelated CB for adult patients, ex vivo expansion of CD34+-enriched stem/progenitor cell populations isolated from fresh or frozen CB was attempted in the presence of rh-IL-3, rh-IL-6, rh-EPO, rh-GM-CSF, and rh-SCF with or without fit 3. At varying time points (days 0, 2, 4, 7, 14, 21), the contents of these cultures were analyzed for the numbers of cells, CFC (BFU-E, CFU-GM, CFU-GEMM), and LTC-IC. In this setting, the increase of cells was 200-fold, that of CFC 70-fold, and most importantly that of LTC-IC was 4.5-fold after 7 days in culture in the presence of flt3. In conclusion, LTC-IC derived from CB can be maintained and considerably expanded ex vivo from highly enriched CD34 + CB cell populations from fresh or frozen CB samples.

Anti-parasitic effector mechanisms in human brain tumor cells: role of interferon-gamma and tumor necrosis factor-alpha.

Toxoplasma gondii, an obligate intracellular parasite, is able to replicate in human brain cells. We recently showed that interferon (IFN)-gamma-activated cells from glioblastoma line 86HG39 were able to restrict Toxoplasma growth. The effector mechanism responsible for this toxoplasmostatic effect was shown by us to be the IFN-gamma-mediated activation of indolamine 2,3-dioxygenase (IDO), resulting in the degradation of the essential amino acid tryptophan. In contrast, glioblastoma 87HG31 was unable to restrict Toxoplasma growth after IFN-gamma activation, and IFN-gamma-mediated IDO activation was weak. We observed that tumor necrosis factor (TNF)-alpha alone is unable to activate IDO or to induce toxoplasmostasis in any glioblastoma cell line tested. Interestingly, we found that TNF-alpha and IFN-gamma were synergistic in the activation of IDO in glioblastoma cells 87HG31, 86HG39 and U373MG and in native astrocytes. This was shown by the measurement of enzyme activity as well as by the detection of IDO mRNA in TNF-alpha + IFN-gamma activated cells. This IDO activity results in a strong toxoplasmostatic effect mediated by glioblastoma cells activated simultaneously by both cytokines. Antibodies directed against TNF-alpha or IFN-gamma were able to inhibit IDO activity as well as the induction of toxoplasmostasis in glioblastoma cells stimulated with both cytokines. Furthermore, it was found that the addition of L-tryptophan to the culture medium completely blocks the antiparasitic effect. We therefore conclude that both TNF-alpha and IFN-gamma may be involved in the defense against cerebral toxoplasmosis by inducing IDO activity as an antiparasitic effector mechanism in brain cells.

Establishment of T-helper type 1- and T-helper type 2-like human Toxoplasma antigen-specific T-cell clones.

As an in vitro model for human cerebral toxoplasmosis, we analysed the interaction between glioblastoma cells, Toxoplasma and Toxoplasma antigen-specific T-helper cells. We established 46 different human CD4+ T-cell clones from four different donors. All T-cell clones responded to Toxoplasma antigen derived from three different Toxoplasma strains. We found that the supernatants of 44 clones induced toxoplasmostasis in glioblastoma cells. The anti-parasitic effector mechanism activated in glioblastoma cells by T-cell supernatants was the induction of the tryptophan-degrading enzyme indolamine 2,3-dioxygenase. Enzyme induction, as well as the anti-parasitic effect, was blocked by a monoclonal antibody directed against interferon-gamma (IFN-gamma), and the addition of L-tryptophan to the cultures completely blocked the anti-parasitic effect induced by T-cell supernatants. The supernatants from two of the 46 established T-cell clones (3A22 and 1A15) were unable to induce indolamine 2,3-dioxygenase activity or, as expected, toxoplasmostasis in glioblastoma cells. We further analysed the supernatants from these two clones, and found that they contained large amounts of IL-4 and no, or only limited amounts of, IFN-gamma. We therefore conclude that Toxoplasma-antigen is able to activate T-helper type 1 (Th1)- and Th2-like human T cells, and only IFN-gamma-producing cells are capable of inducing anti-parasitic effector mechanisms.