In Immature Human Dendritic Cells Gondii Toxoplasma Inhibition of the Growt

Inhibition of the Growth of Toxoplasma gondii in Immature Human Dendritic Cells Is Dependent on the Expression of TNF- α Receptor 2 This information is current as of September 25, 2021. Anja Giese, Silke Stuhlsatz, Walter Däubener and Colin R. MacKenzie J Immunol 2004; 173:3366-3374; ; doi: 10.4049/jimmunol.173.5.3366 http://www.jimmunol.org/content/173/5/3366 Downloaded from

References This article cites 33 articles, 18 of which you can access for free at:

http://www.jimmunol.org/content/173/5/3366.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 25, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2004 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Inhibition of the Growth of Toxoplasma gondii in Immature Human Dendritic Cells Is Dependent on the Expression of TNF-␣ Receptor 21

Anja Giese,*† Silke Stuhlsatz,* Walter Da¨ubener,* and Colin R. MacKenzie2*

An effective immunity to Toxoplasma gondii in humans is dependent on the cellular immune response. Toxoplasma can infect and replicate in almost all nucleated cells, and the most important cytokine regulating the growth in humans is IFN-␥; however, the role of TNF-␣ has to date been largely described to be synergistic. We show that, compared with mature human dendritic cells (mDC), immature human DC (iDC) demonstrate a reduced parasite proliferation when infected with Toxoplasma. This toxoplasmostasis was only present in iDC after 11 days of culture and was not present in DC that had been matured ex vivo using a cytokine mixture (mDC). Spontaneous toxoplasmostatic activity has previously only been described in fresh human monocytes, and the Downloaded from mechanism involved is as yet unclear. We show that, in comparison with an absence of expression in mDC, TNF-R2 is expressed in both iDC and monocytes infected with Toxoplasma, and furthermore, that blocking the TNF-R2 with Abs abrogates the toxoplasmostasis in the iDC. These ﬁndings demonstrate a functional role for TNF-R2 in the newly described spontaneous toxoplasmostasis of iDC. The Journal of Immunology, 2004, 173: 3366–3374.

he obligate intracellular protozoan parasite Toxoplasma more, cells of the mononuclear line have been proposed to play a http://www.jimmunol.org/ gondii causes a chronic asymptomatic infection in immu- part in the dissemination of Toxoplasma by offering the parasites T nocompetent individuals. The ability to evade the immune asylum from the humoral and cellular immune systems, and more- system over decades and then, in the event of severe immunosup- over, by acting as couriers from the infection site to distant organs pression, to reappear in the form of a life-threatening encephalitis where encystation takes place (12). Recently, it has been reported makes Toxoplasma both an important health hazard as well as an that human monocytes and DC are permissive for Toxoplasma interesting organism to study the survival techniques of intracel- proliferation, which would support this hypothesis (13). In direct lular organisms. Furthermore, Toxoplasma is capable of replicat- contrast to this, fresh human monocytes display a spontaneous ing in almost every nucleated cell, and the control of growth and Toxoplasma growth-inhibitory effect (14). suppression of reactivation require IFN-␥ (1). A number of mech- In murine toxoplasmosis, the IFN-␥-mediated effect is depen- by guest on September 25, 2021 anisms have been described to be effective in controlling Toxo- dent on TNF-␣ (15), which exerts its effect exclusively via plasma infection, such as NO (2), oxygen radical production (3), TNF-R1 and not via TNF-R2 (16, 17). In humans, the role of TNF depletion of iron (4), and depletion of tryptophan (5). The mech- receptors is less well described. A recent report described an up- anism involved in human macrophages, however, remains unclear. regulation of TNF-R2 and inhibition of production of TNF-␣ in the Whereas NO as a defense mechanism against Toxoplasma in hu- human monocyte cell line THP-1 after infection with Toxoplasma man cells is controversial (6–8) and tryptophan depletion in hu- (18). TNF-␣ has its major effect through the TNF-R1, and it has man macrophages is not responsible for toxoplasmostasis (9), re- been proposed that TNF-R2, due to its higher affinity for TNF-␣, active oxygen intermediates appear to be at least partly responsible serves to capture TNF-␣ and pass it on to TNF-R1, which is then for an anti-Toxoplasma defense (10). Recently, toxoplasmostatic involved in signal transduction (19). Signal transduction via 3 activity was shown in murine dendritic cells (DC), and this was TNF-R2 has not been described in human cells. described to be neither due to NO nor tryptophan depletion, but to Our aim in this study was to elucidate the underlying mecha- be abrogated by coincubation with oxygen radical scavengers (11). nism or mechanisms for the observed resistance of immature DC DC have typically been ascribed an immune stimulatory and reg- (iDC) to Toxoplasma growth. We set out to determine whether the ulatory function and not necessarily an effector function. Further- innate toxoplasmostatic activity of these iDC was due to a known characteristic of iDC such as apoptosis or IFN-␥ secretion, or perhaps due to some other cytokine-induced mechanism. The cyto- *Institute of Medical Microbiology, Heinrich-Heine-University, Du¨sseldorf, Germa- kine gene expression of iDC, mature DC (mDC), and monocytes † ny; and Department of Zoophysiology, Oldenburg University, Oldenburg, Germany was thus examined, as well as the permissiveness of the DC. Received for publication December 17, 2003. Accepted for publication June 22, 2004. In the present study, we provide evidence that, in addition to The costs of publication of this article were defrayed in part by the payment of page fresh human monocytes, human DC with an immature phenotype charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. (iDC) display a spontaneous toxoplasmostasis. To our knowledge, 1 This work was supported by a grant from the Research Committee, Medical Faculty this is the first report of such activity in cells other than monocytes. of the Heinrich-Heine-University to C.R.M. Furthermore, this activity was restricted to cells that had been in 2 Address correspondence and reprint requests to Dr. Colin R. MacKenzie, Institute of culture for ϳ11 days, which may suggest a selection of a sub- Medical Microbiology, Heinrich-Heine-University, Moorenstr.5, 40225 Du¨sseldorf, population of DC exhibiting this effect. Compared with iDC, DC Germany. E-mail address: [email protected] that were matured ex vivo using the cytokine mixture IL-1, IL-6, 3 Abbreviations used in this paper: DC, dendritic cell; iDC, immature DC; mDC, ␣ mature DC; MOI, multiplicity of infection; mTNF, membrane-bound TNF; YFP, TNF- , and PGE2 (mDC) displayed no inhibition of Toxoplasma yellow fluorescent protein. growth without IFN-␥ stimulation. Both iDC and mDC showed a

strong Toxoplasma growth inhibition after stimulation with IFN-␥. Determination of apoptosis Our experiments demonstrate that in both iDC and monocytes that Annexin FACS analysis. Apoptosis was evaluated using the annexin V- exhibit this toxoplasmostasis, infection with Toxoplasma results in binding assay kit (FACS annexin V-FITC apoptosis detection kit; R&D the expression of TNF-R2. In comparison, mDC and macrophages Systems), according to the manufacturer’s protocol. Briefly, 1 ϫ 105 DC ␮ do not express TNF-R2 after infection. Furthermore, by blocking were washed in PBS and incubated with 0.25 g/ml annexin V FITC in a calcium-containing buffer. Necrotic cells were detected using 0.5 ␮g/ml the TNF-R2 using neutralizing anti-TNF-R2 Abs, we demonstrate propidium iodide. After incubation for 15 min at room temperature, cells that the TNF-R2 has a functional role in the toxoplasmostasis of were washed and analyzed by flow cytometry, as described above. iDC. We know of no other report showing an association between Caspase-3 colorimetric assay. Apoptosis was further evaluated using the TNF-R2 and an effector function in DC. caspase-3 colorimetric assay kit (R&D Systems), according to the manufacturer’s protocol. Briefly, 1 ϫ 106 cells were incubated in lysis buffer and centrifuged (10,000 ϫ g,4°C, 1 min). Supernatants were collected, and the Materials and Methods enzymatic activity was determined photometrically. Differentiation and culture of monocyte-derived DC Array analysis of infected and noninfected iDC and mDC PBMC were isolated from buffy coats of heparinized blood of healthy Six hours after infection with T. gondii, extracellular parasites were re- volunteer donors at the blood bank of the University Hospital (Du¨sseldorf, moved by washing in PBS. Total RNA of mDC and iDC was isolated by Germany) by Ficoll Hypaque (Amersham Biosciences, Freiburg, Ger- density ultracentrifugation (100,000 ϫ g, 16 h), following guanidinium- many) density centrifugation. CD14ϩ cells were separated using immuno- isothiocyanate (Merck, West Point, PA) lysis of the cells. Any remaining magnetic bead selection (MACS; Miltenyi Biotec, Bergisch Gladbach, genomic DNA was degraded by digestion with DNase I (Applied Biosys- Germany), according to the manufacturer’s instructions. The obtained tems, Foster City, CA). RT-PCR was performed with cDNA labeling kit monocytes were shown in FACS analysis to be Ͼ95% pure, and subse- and human cytokine-specific primers (both R&D Systems). During reverse Downloaded from quently, 5 ϫ 106 cells were cultured in 3 ml of IMDM supplemented with transcription, cDNA was labeled by incorporation of 32P-labeled dCTP. 10% FCS (BioWhittaker, Apen, Germany), 75 ng/ml GM-CSF (Novartis Specific activity of labeled cDNA was determined, and equal amounts Pharmaceuticals, Kundl, Austria), and 100 U/ml IL-4 (R&D Systems, of DC cDNA were used to hybridize the cytokine cDNA expression arrays Wiesbaden, Germany). Cytokines were replenished every 3 days by re- (R&D Systems). Before hybridization, arrays were prehybridized with 100 moving 1 ml of supernatant and replacing it with fresh medium containing ␮g/ml sonicated salmon sperm DNA (R&D Systems) for1hat58°C. the same final concentration of cytokines. After a maximum incubation Hybridization was conducted overnight in a water bath at 58°C. The arrays were washed twice with 0.5ϫ standard saline citrate phosphate/EDTA, 1% period of 10 days, cells were harvested and split into two subcultures. http://www.jimmunol.org/ To obtain mDC, one population was supplemented with additional IL-1 SDS, and 0.1ϫ standard saline citrate phosphate/EDTA, 1% SDS. The (1000 U/ml), TNF-␣ (1000 U/ml) (both R&D Systems), IL-6 (1000 cDNA arrays were exposed to a phosphor imager screen (BAS; Fuji Photo ␮ U/ml; PeproTech, Rocky Hill, NJ), and PGE2 (1 g/ml; Sigma-Aldrich, Film, Dusseldorf, Germany) for 24 h. Transcription of candidate genes was St. Louis, MO) for an additional 3 days. A fraction of the DC was checked by RT-PCR using specific primer pairs. submitted to FACS analysis to confirm their maturation status. RT-PCR analysis

Flow cytometry Total RNA of infected and noninfected cells was prepared, as described above. The RT-PCR was performed using the Advantage RT-PCR Kit (BD DC were stained using FITC-conjugated CD86, HLA-DR, and CD80 Clontech, Palo Alto, CA). One microgram of total RNA was reverse tran- mAbs and PE-conjugated CD83, CD40, and CD11b mAbs (BD Bio- scribed in a 20 ␮l final volume containing 1 ␮M oligo(dT) primer. The by guest on September 25, 2021 sciences, San Jose, CA). To block Fc receptors, cells were first incubated reagents for PCR were as follows: 20 nmol each primer; 10 mM each of for 20 min in 1% rabbit serum in PBS, pH 7.3 (Serotec, Serag, Wiessner, dGTP, dCTP, dATP, and dTTP; 1.5 mM MgCl2 and1UofTaq polymer- Germany) with 0.1% sodium azide (Sigma-Aldrich) and washed with PBS ase (all PerkinElmer, Wellesley, MA); and distilled water to a volume of 50 containing 5% FCS and 0.1% sodium azide. The subsequent staining was ␮l. The PCR reagents were subjected to 35 cycles of: denaturing (45 s, performed with PE/FITC-conjugated mAbs for 30 min at 4°C. Controls 95°C), annealing (30 s), and extension (2 min, 72°C). TNF-␣ (product size, were stained with labeled isotype-matched control mAbs (IgG1 PE, BD 414 bp) and GAPDH (product size, 947 bp) primers were from BD Clon- Biosciences; IgG2a FITC, (DakoCytomation, Hamburg, Germany). Flow tech; TNF-R1 (product size, 447 bp) and TNF-R2 (product size, 665 bp) cytometric analysis was performed using the FACSCalibur apparatus and primers were from R&D Systems. CellQuest software from BD Immunocytometry Systems (San Jose, CA). TNF-␣ ELISA ␣ T. gondii culture and proliferation assay TNF- protein was measured in the supernatants of infected and noninfected DC and monocytes using an ELISA from R&D Systems, according T. gondii tachyzoites of the BK strain (a gift from H. M. Seitz and M. to the manufacturer’s instructions. Saathof, Institute for Medical Parasitology, University of Bonn, Bonn, Ger- many) were maintained in L929 mice fibroblasts in IMDM containing 5% Blocking of TNF-R2 with neutralizing mAbs FCS. T. gondii tachyzoites expressing yellow fluorescent protein (YFP) Thirteen-day-old DC (1 ϫ 105 per well) were incubated in a 48-well cul- were maintained in culture under similar conditions. BK-Toxoplasma were ture plate (Costar Corning, New York, NY) with neutralizing anti-human transfected with a YFP plasmid modified from the plasmid, tubYFP-YFP/ TNF-R2 mAb (R&D Systems) at different concentrations from 0.5 to a sagCAT, originally a gift of B. Striepen (Center for Tropical and Emerging maximum of 7 ␮gfor1hat37°C. Subsequently, cells were infected with Global Diseases, University of Georgia, Athens, GA), and described by T. gondii in a ratio of 1:5 (MOI, 5) for an additional 24 h. Proliferation of Gubbels et al. (20). This plasmid has a Toxoplasma protein, designated by Toxoplasma was determined by [3H]uracil incorporation, as described us as B4, replacing the YFP1 locus in the plasmid (a kind gift from A. above. Ailyati, Institute of Medical Microbiology, University of Du¨sseldorf, Du¨s- seldorf, Germany). In experiments determining Toxoplasma proliferation Results in DC, the cells were infected with T. gondii at a multiplicity of infection of 5 tachyzoites per DC (MOI, 5). Growth of Toxoplasma was measured by iDC inhibit the growth of Toxoplasma the [3H]uracil incorporation method (21). Twenty-four hours after infec- Toxoplasma are generally able to proliferate in human nucleated 3 tion, the cells were pulsed with 0.012 MBq of [ H]uracil (Amersham Bio- cells with the single exception of fresh human monocytes that have sciences) and cultured for an additional 24 h. Host cells and Toxoplasma were lysed by freeze thawing, and [3H]uracil incorporation was measured been shown to exhibit a spontaneous toxoplasmostasis. Many other using a beta counter (1205 Betaplate; PerkinElmer, Jugesheim, Germany). cells such as macrophages are only able to inhibit the growth of In experiments aimed at determining the permissiveness of the iDC or Toxoplasma after stimulation with IFN-␥. DC in culture were also mDC, the cells were infected with YFP-Toxoplasma at an MOI of 50 for permissive of Toxoplasma growth; however, after approximately 10 min at 37°C, and thereafter the extracellular Toxoplasma was removed by washing the cells in PBS. DC were stained with PE-labeled CD83 and the eleventh day in culture, this permissiveness changed and Tox- then subjected to FACS analysis, and the proportion of infected (YFP- oplasma proliferation was markedly reduced. Fig. 1A shows the positive) DC was determined. intracellular proliferation of Toxoplasma in DC on days 7, 9, 13, 3368 TOXOPLASMOSTASIS OF iDC Downloaded from

FIGURE 2. FACS analysis of iDC and mDC. Monocyte-derived DC were split at day 10 of culture, and one-half treated additionally with IL-1, ␣ IL-6, TNF- , and PGE2 for an additional 3 days (thin line). (The bold line represents the iDC.) Thereafter, the iDC and mDC were stained with FITC (HLA-DR and CD86)- or PE (CD83, CD40, CD11b)-conjugated mAbs and http://www.jimmunol.org/ analyzed using a FACSCalibur apparatus. The cells were gated according to forward and side scatter of unstained cells. The broken line represents the isotype-matched Abs to irrelevant Ag. The results show a typical analysis of at least 10 different preparations.

iDC and mDC are equally permissive to infection by Toxoplasma

iDC are especially efficient at phagocytosis, an attribute that is lost by guest on September 25, 2021 in the maturation process. In addition, Toxoplasma are able to invade nucleated cells in an active process that ensures entry into the cell within a few minutes. It is therefore possible that differences in the susceptibility or permissiveness of the DC result in the observed difference in Toxoplasma growth in the iDC and mDC, FIGURE 1. iDC inhibit Toxoplasma growth. A,1ϫ 106 DC on days 7, dependent on the mode of entry into the cell. To demonstrate the 9, 13, and 15 of culture, and B,1ϫ 106 13-day-old iDC, macrophages, and presence of such a difference in susceptibility to Toxoplasma in- mDC (day 13) were infected with Toxoplasma at an MOI of 5, incubated fection, we infected DC with YFP-expressing Toxoplasma and an- for 24 h, and then pulsed with 0.012 MBq of [3H]uracil for the final 24 h alyzed the cells in FACS after staining with PE-labeled CD83. In of culture. Toxoplasma proliferation was determined by beta counting. both iDC and mDC, the percentage of infected cells was deter- Ϯ Representative data ( SD of triplicate determinations) of three similar mined by calculating the cells staining for YFP and for the CD83 experiments are shown. molecule. Cells thus negative for CD83 and positive for YFP rep- resented infected iDC and cells double positive, infected mDC. In and 15 of culture. At each of these stages, the maturation status of Fig. 4, the mean of paired results of the percentage of infected iDC the DC was checked by FACS, and the DC were found to have and mDC is shown (iDC, 46.6 Ϯ 17.8%; mDC, 40.0 Ϯ 13.5%). properties of iDC. Fig. 2 shows the phenotype of the iDC and Thus, we were able to show that the resistance to Toxoplasma mDC; the surface markers used in this study to define iDC were growth in iDC is not due to the iDC being less permissive to low HLA-DR, low CD86, low CD80, low CD40, and negative infection than the mDC. This experiment is especially suited to CD83. The figure shows a representative FACS of the iDC from demonstrate any differences in infection in that the iDC and mDC days 7 through 15. No differences in the surface marker expression can be infected with Toxoplasma in a single well, thereby creating of the DC were detected in the DC on days 7, 9, 13, or 15, and in identical experimental conditions for both cell types. These results particular the DC did not mature in this time (see below, Fig. 3). show that the resistance of iDC to Toxoplasma growth is not sec- We were interested to observe whether DC that were submitted to ondary to reduced cell susceptibility to infection. a forced maturation in vitro, by the addition of a cytokine mixture ␣ ␤ Both iDC and mDC demonstrate an IFN-␥-induced of TNF- , IL-1 , IL-6, and PGE2 for 3 days, as described in Materials and Methods, were also able to restrict the growth of toxoplasmostasis Toxoplasma. Fig. 1B shows the Toxoplasma proliferation in mac- To determine whether the DC were capable of inhibiting the rophages, iDC, and mDC (all on day 13 of culture). As observed growth of Toxoplasma after stimulation with IFN-␥ (as are almost before, there is a marked restriction of Toxoplasma growth in iDC all effector cells), we incubated the DC with different concentra- compared with mDC and macrophages (all from the same donor). tions of IFN-␥. After 24 h, the DC were infected with Toxoplasma The Journal of Immunology 3369

FIGURE 4. iDC and mDC are equally susceptible to infection with Tox- oplasma. Toxoplasma-expressing YFP were added to DC (MOI, 50) and

incubated for 10 min at 37°C. Thereafter, the extracellular Toxoplasma Downloaded from FIGURE 3. Both iDC and mDC display IFN-␥-induced toxoplasmosta- were removed by washing in PBS, and the DC were stained with PE- 6 sis. A total of 1 ϫ 10 13-day-old DC (iDC and mDC) in medium with labeled CD83 Ab. The iDC and mDC were then subjected to FACS, and 10% FCS was incubated in the presence of IFN-␥ at varying concentrations the proportion of infected CD83-positive (mDC) and CD83-negative (iDC) or absence of IFN-␥ for 24 h. The cells were then infected with Toxo- was calculated. The results show the mean of 10 experiments ϮSD. Using plasma (MOI, 5) in 96-well microtiter plates and incubated for an addi- paired t test, the pairing correlation coefﬁcient (r) was 0.9307 (p Ͻ 0.0001). tional 24 h, after which the cultures were pulsed with [3H]uracil for 24 h

and the radioactive uracil incorporation was measured in a beta counter. http://www.jimmunol.org/ The values for Toxoplasma growth in both iDC and mDC are normalized to growth in iDC not stimulated with IFN-␥, assigned a value of 1. The results depicted are the means of three values ϮSD and are representative and whether this Toxoplasma-driven process may be in part re- of at least three identical experiments. sponsible for the growth inhibition of the parasite. iDC infected with Toxoplasma for 24 h were analyzed for markers of maturation, and the results are shown in Fig. 5. As can be seen, infection of iDC with Toxoplasma does not induce maturation of the cells. (MOI, 5), and Toxoplasma growth was measured, as described. Fig. 3 shows the results of a typical experiment in which the growth of the Toxoplasma is displayed relative to that in iDC. The by guest on September 25, 2021 parasite proliferation in iDC is approximately one-half that of the mDC, and both the iDC and mDC show an IFN-␥-inducible toxoplasmostasis. It is interesting to note that the toxoplasmostasis of the mDC does not reach levels seen in iDC (or, for that matter, of macrophages). These results also show that the iDC possess both an IFN-␥-independent as well as an IFN-␥-dependent toxoplasmostatic effector mechanism. To attempt to determine which mechanism may be involved, we incubated the iDC with and without IFN-␥ in the presence or absence of different oxygen radical scavengers, such as mannitol, 1,4 diazabicyclo [2,2,2] octane, super- oxide dismutase, and benzoic acid, and were unable to show any effect on either the spontaneous toxoplasmostasis of iDC or the IFN-␥-induced toxoplasmostasis of iDC or mDC. Furthermore, we were unable to detect a production of NO or H2O2 using FACS analysis of dichlorofluorescein-treated iDC and mDC (data not shown). This experiment shows that both the iDC and mDC exhibit an IFN-␥-induced Toxoplasma growth-inhibitory effect. The growth permissiveness of the mDC relative to iDC is preserved even after stimulation with IFN-␥, and the iDC exhibit an IFN-␥- dependent and an IFN-␥-independent toxoplasmostasis. iDC remain immature after infection with T. gondii iDC are well suited to phagocytosis of pathogens, after which they migrate to the regional lymph nodes from the periphery and con- FIGURE 5. iDC do not mature after infection with Toxoplasma. Some comitantly develop an mDC phenotype more suited to Ag presen- iDC (day 12) were infected with Toxoplasma, and after 24 h the infected tation. Many different Ags and pathogens have been shown to (bold line) and uninfected (thin line) cells were analyzed by FACS using influence DC maturation such as LPS, HIV, Mycobacterium tu- FITC anti-CD80, PE anti-CD11b, FITC anti-CD86, FITC anti-CD HLA- berculosis, and Gram-negative bacteria such as Escherichia coli. DR, PE anti-CD83, or PE anti-CD40 mAbs. The broken line represents the This has not been shown for Toxoplasma, and it was of interest to control using isotype-matched Abs to irrelevant Ag. The results of one determine whether Toxoplasma-infected iDC undergo maturation representative experiment of five are shown. 3370 TOXOPLASMOSTASIS OF iDC

Toxoplasma growth inhibition is not due to apoptosis of the infected iDC Apoptosis of an infected cell and subsequent ingestion by other APC is an effective immune mechanism toward eliminating the pathogen and stimulation of the adaptive immune system. If Tox- oplasma-infected iDC undergo apoptosis, parasite growth may be inhibited, and thus a difference in the rate of apoptosis between iDC and mDC could account for the toxoplasmostasis observed in iDC. Apoptosis was measured by determining both the caspase-3 activity and the binding of annexin V to the cell surface of infected and noninfected DC. In both experiments, the iDC and mDC were infected for 24 h, and thereafter, apoptosis was determined. Non- infected cells were treated in the same way as the infected cells, and as a control apoptosis was induced in U937 cells with TNF-␣ plus cycloheximide. In Fig. 6A, the binding of annexin shows that both a so-called early and late apoptosis in the iDC was not increased after infection with Toxoplasma. In both iDC and mDC, the late apoptosis was signiﬁcantly greater; however, infection with Toxoplasma did Downloaded from not result in an increased apoptosis. Interestingly, in the mDC, there was a marked increase in the early apoptosis induced by Toxoplasma infection; however, this was not enough to offset the total Toxoplasma proliferation that presumably occurred in the remaining nonapoptotic cells. These results were conﬁrmed by the caspase-3 activity assay (Fig. 6B), in which a similarly greater http://www.jimmunol.org/ apoptosis was observed in mDC after infection. These results suggest that the observed Toxoplasma growth restriction is not secondary to a selective depletion of the host cell in iDC cultures due to apoptosis induced by Toxoplasma infection.

TNF-R2 gene expression in DC is dependent on maturity and Toxoplasma infection

mDC and iDC were incubated for 6 h with Toxoplasma tachyzoites by guest on September 25, 2021 (MOI, 5); thereafter, the extracellular Toxoplasma was removed by washing the cells, and the total RNA was extracted. Cytokine arrays were hybridized with cDNA from the infected and uninfected iDC and mDC. A comparison of the expression patterns identified the differential expression of the TNF-R2 gene. RT-PCR analysis confirmed this result, as seen in Fig. 7A. A signal for TNF-R2 in iDC is present only after infection, and furthermore, mDC express FIGURE 6. Apoptosis is not responsible for Toxoplasma growth inhi- TNF-R2 neither in the infected nor the uninfected state. In con- bition in iDC. Some DC were infected with Toxoplasma (MOI, 5) for 24 h trast, TNF-R1 is expressed in both iDC and mDC as well as in before determination of apoptosis in both infected and uninfected cells. monocytes, and moreover, the state of infection has no effect on Apoptosis was measured by annexin V expression in FACS (A) in iDC and the expression of this gene. mDC, and as a control in U937 cells treated with cycloheximide and To ensure that the expression of TNF-R2 in mDC did not occur TNF-␣ (C/T). Early apoptosis, as determined by the annexin V-positive at an earlier time point and therefore was not detected, we har- and propidium iodide-negative cells, is expressed as the percentage of pos- vested RNA from DC at 30 min, 1 h, 2 h, 4 h, and 6 h postinfec- itive cells in FACS. Apoptosis was determined by measuring caspase-3 tion, and performed RT-PCR to detect TNF-R2 mRNA. As shown activity, as described in Materials and Methods, under similar conditions Ϯ in Fig. 7B, there is no detectable mRNA for TNF-R2 in mDC, (B). Representative data ( SD of triplicate determinations) of three similar determinations are shown. whereas expression in iDC is detectable at 1 h postinfection, reach- ing a maximum at 4 h. The results demonstrate that TNF-R2 expression in iDC is a result of infection with Toxoplasma, and that the mDC do not express TNF-R2, whereas TNF-R1 is equally secreting TNF-␣ after Toxoplasma infection, we measured TNF-␣ expressed in all cells tested. Furthermore, TNF-R2 is also ex- expression and also protein in the culture supernatants before and pressed in monocytes after infection with Toxoplasma,afinding after infection. Fig. 8A shows the TNF-␣ mRNA in infected and that is interesting in view of the fact that monocytes also exhibit an noninfected iDC, mDC, and monocytes. No obvious difference in innate toxoplasmostatic effect. the expression of TNF-␣ is seen before and after infection and between the different cell types. We detected very small amounts ␣ TNF- expression by iDC, mDC, and monocytes is independent of TNF-␣ protein in the supernatants of all cells (50–400 pg/ml), of Toxoplasma infection although this did increase upon infection of iDC and mDC (Fig. TNF-␣ has been shown in mice to be important in the control of 8B). This is very low in comparison with TNF-␣ levels normally Toxoplasma infection, although its role in human toxoplasmosis is detected in LPS-stimulated cells (usually at ng/ml levels). Further- perhaps secondary to IFN-␥. To determine whether the iDC were more, we incubated iDC with neutralizing anti-TNF-␣ Abs for 30 The Journal of Immunology 3371 Downloaded from http://www.jimmunol.org/

FIGURE 8. TNF-␣ is secreted in small amounts by iDC, mDC, and monocytes. The TNF-␣ secretion by 1 ϫ 105 iDC, mDC, and Mo before and after infection with Toxoplasma (MOI, 5) was measured by determining the TNF-␣ mRNA (A) 6 h after infection and by ELISA of supernatants (B), 6 and 24 h after infection. Representative data (ϮSD of triplicate determinations) of three similar determinations are shown. by guest on September 25, 2021

Toxoplasmostasis is inhibited by neutralizing the TNF-R2 TNF-␣ has been shown to play a supportive role in the defense against Toxoplasma; however, there was no effect in the absence of IFN-␥. Because we were unable to detect IFN-␥ in the supernatants of iDC and equally could not detect mRNA for IFN-␥, we assume that IFN-␥ does not play a part in the described FIGURE 7. Toxoplasma-infected iDC and monocytes express TNF-R2. toxoplasmostasis of iDC (data not shown). It was possible that A, Cell cultures of iDC, mDC, and monocytes (Mo) were split, and half of the expression of TNF-R2 was merely marker or signal for an the cells were infected with Toxoplasma for 6 h (iDCT, mDCT, and MoT), and the other half were left uninfected (iDC, mDC, and Mo). Thereafter, immunological process taking place in the infected cell and in the cells were washed, and RNA was extracted and subjected to RT-PCR itself had no, or little, direct role. We thus used neutralizing using primers specific for TNF-R1 and TNF-R2. B, Different cultures of anti-TNF-R2 mAbs to determine whether TNF-R2 had a func- iDC, mDC, and Mo were infected with Toxoplasma, and after 30 min, 1 h, tional role in the toxoplasmostasis in iDC. Fig. 9A shows the 2 h, 4 h, or 6 h, the cells were washed and RT-PCR was performed on the effect of incubating the iDC with anti-TNF-R2 Abs at varying harvested RNA. In both experiments, amplification of GAPDH is shown as concentrations before infecting the cells with Toxoplasma. The a control for the amount of RNA per reaction. Data are representative of Toxoplasma growth is depicted relative to that of untreated iDC three similar experiments. (given the value 1). Increasing concentrations of neutralizing Ab result in a decreased toxoplasmostatic effect. At 7 ␮g/ml, the toxoplasmostasis is almost completely abrogated. In con- min before infection with Toxoplasma and were unable to dem- trast, the Abs have no effect on mDC, as is shown in Fig. 9B.To onstrate any effect on the toxoplasmostasis in the iDC (data not exclude an unspecific Ab effect, we incubated the iDC and mDC shown). The expression of TNF-␣ and the detection of low with Abs of the same Ig class as that of the anti-TNF-R2 Ab, amounts of TNF-␣ protein in supernatant of all cells, both before which recognize a nonhuman epitope. We observed no effect and after infection, do not support a major role of TNF-␣ in the (data not shown). These results demonstrate a functional role innate toxoplasmostasis of iDC, and the additional failure of neu- for TNF-R2 by showing an abrogation of the toxoplasmo- tralizing TNF-␣ Abs to abrogate this effect at the very least does stasis by use of neutralizing TNF-R2 Abs in a concentration- not increase the likelihood of a possible role for TNF-␣. dependent manner. 3372 TOXOPLASMOSTASIS OF iDC

mouse and human models of Toxoplasma infection, the major stimulus causing a toxoplasmostasis is, regardless of the mechanism induced, IFN-␥. It has been clearly demonstrated in IFN-␥ knockout mice that deficiency of IFN-␥ results in a lethal infection (22, 23) and, in humans, IFN-␥ stimulates many cells such as macrophages (9), endothelial cells (24), and astrocytes (25) to exhibit a toxoplasmostatic effect. To date, a spontaneous growth restriction has only been described in fresh human monocytes (14, 26). Wilson and Remington (14) described the Toxoplasma growth-inhibitory effect of monocytes infected within the first hour of culture. This activity was later further characterized to be present only in the adherent population of monocytes and not to be present in nonadherent monocytes (27). This toxoplasmostatic activity was shown to be IFN-␥ independent and partly due to reac- Ϫ Ϫ tive oxygen species such as OH ,O2 , and H2O2 (10). There was, however, some activity against Toxoplasma that was independent of reactive oxygen species. In the present study, we describe for the first time a spontaneous inhibition of Toxoplasma growth in

iDC. We show that the toxoplasmostasis is not due to the presence Downloaded from of IFN-␥ in the cultures because we were neither able to detect IFN-␥ in the culture supernatants nor the presence of contaminat- ing IFN-␥-producing T cells or NK cells (data not shown). In addition, we show that the growth inhibition is not due to increased apoptosis of the iDC. We also show that the infected cells do not,

after infection, develop a mature phenotype, as measured by http://www.jimmunol.org/ CD83, CD40, CD86, and HLA-DR up-regulation. This is in contrast to Subauste and Wessendarp (28), who demonstrated that the costimulatory molecule markers CD86, CD40, and CD80 were up-regulated after infection with Toxoplasma, but not the maturation marker CD83. This apparent discrepancy may be due to the fact that our iDC are held for longer in culture than the 7-day-old DC used in the study by Subauste and Wessendarp (28). This explanation would be supported by the fact that the cell age appears to be critical for the toxoplasmostatic activity, which by guest on September 25, 2021 suggests a selection of a subpopulation of iDC or a further differentiation into a subtype as yet uncharacterized. Our preliminary investigations in determining an underlying mechanism for the toxoplasmostatic activity of the iDC showed us that the known effector mechanisms described in cells of the innate immune system were not responsible. Using specific inhibitors, we FIGURE 9. TNF-R2 has a functional role in toxoplasmostasis. DC (A, excluded reactive oxygen species, NO, and IDO activity as a iDC; B, mDC) were incubated for 1 h with neutralizing anti-TNF-R2 mAb, source of the growth control of the intracellular Toxoplasma.We and thereafter infected with Toxoplasma (MOI, 5), incubated for an addi- therefore opted for another approach to search for the mechanism tional 24 h, and then pulsed with 0.012 of MBq [3H]uracil for the final 24 h and used cytokine gene expression arrays using cDNA from iDC, of culture. Toxoplasma proliferation was determined by beta counting. mDC, and monocytes both before and after infection with Toxo- Data are shown relative to the Toxoplasma growth in untreated iDC, which plasma. Investigation of the differential cytokine gene expression Ϯ is assigned a value of 1. Representative data ( SD of triplicate determi- led to the observation that after infection both iDC and monocytes nations) of at least three similar experiments are shown. up-regulate their expression of TNF-R2 (p75), which is not detectable in sensitive RT-PCR in the uninfected cells, and in addition, neither detectable in infected nor uninfected mDC or macro- Discussion phages. This was in contrast to the ubiquitous expression of Considering that DC are among the first cells to come into contact TNF-R1 in all four cell types tested in both infected and uninfected with Toxoplasma, that the cellular immune response is essential to states. This study showed a maximal expression of TNF-R2 after the mounting of this response, and that the mechanisms by which infection at4hinboth iDC and monocytes, which correlates well Toxoplasma and many other intracellular pathogens are controlled with the results from Belloni et al. (18), who demonstrated a are not yet clear, we felt it would be interesting and appropriate to TNF-R2 up-regulation in Toxoplasma-infected THP-1 cells, also study the influence of DC maturation on their response to Toxo- 3–4 h after infection with Toxoplasma. This is in contrast to an- plasma. We observed early on that the Toxoplasma proliferation in other report that showed no increased expression of TNF-R2 after DC was reduced after a certain time of culture. The conversion infection with Toxoplasma (29). Both of these studies used THP-1 from a Toxoplasma growth-permitting cell to a growth-restricting cells in contrast to the iDC used in this study. The role of TNF-␣ cell occurred fairly constantly at about day 11 of culture. Further- receptors in Toxoplasma has been best investigated in the mouse more, we show that maturation of the iDC that are able to restrict model, especially using TNF-R1 and TNF-R2 knockout mice. the growth of Toxoplasma using a mixture of TNF-␣, IL 6, IL-1, TNF-R1 knockouts were susceptible to chronic toxoplasmosis, and PGE2 abrogated this toxoplasmostatic effect of the DC. In both whereas they were resistant to the acute infection (17, 16). The The Journal of Immunology 3373

TNF-R2 knockout mice were not more susceptible to acute or References chronic Toxoplasma infection than the wild-type mice, thus putting 1. Suzuki, Y., M. A. Orellana, R. D. Schreiber, and J. S. Remington. 1988. Inter- into question the importance of TNF-R2 for the control of feron-␥: the major mediator of resistance against Toxoplasma gondii. Science 240:516. toxoplasmosis in mice. The expression of TNF-R2 on mouse 2. Adams, L. B., J. B. Hibbs, Jr., R. R. Taintor, and J. L. Krahenbuhl. 1990. Mi- endothelial cells was shown to play a major causative role in the crobiostatic effect of murine-activated macrophages for Toxoplasma gondii: role pathology of cerebral malaria caused by infection of mice with for synthesis of inorganic nitrogen oxides from L-arginine. J. Immunol. 144:2725. 3. Murray, H. W., and Z. A. Cohn. 1979. Macrophage oxygen-dependent antimi- Plasmodium berghei Anka (30, 31). In this study, the TNF-R2- crobial activity. I. Susceptibility of Toxoplasma gondii to oxygen intermediates. deficient mice were not susceptible to cerebral malaria, and the J. Exp. Med. 150:938. 4. Dimier, I. H., and D. T. Bout. 1998. Interferon-␥-activated primary enterocytes authors postulated a role of TNF-R2 in the metabolic homeosta- inhibit Toxoplasma gondii replication: a role for intracellular iron. Immunology sis of the brain endothelial cells, although a specific role of 94:488. TNF-R2 was not determined. A number of other studies have 5. Pfefferkorn, E. R. 1984. Interferon ␥ blocks the growth of Toxoplasma gondii in human fibroblasts by inducing the host cells to degrade tryptophan. Proc. Natl. shown a role for TNF-R1 in the control of other intracellular Acad. Sci. USA 81:908. pathogens such as Listeria momocytogenes (32), Leishmania 6. Murray, H. W., and R. F. Teitelbaum. 1992. L-arginine-dependent reactive ni- major (33), and Mycobacterium tuberculosis (34). In contrast to trogen intermediates and the antimicrobial effect of activated human mononuclear phagocytes. J. Infect. Dis. 165:513. this, the function of TNF-R2 in infection has not been entirely 7. Peterson, P. K., G. Gekker, S. Hu, and C. C. Chao. 1995. Human astrocytes elucidated. The affinity for the two known ligands, TNF-␣ and inhibit intracellular multiplication of Toxoplasma gondii by a nitric oxide-medi- lymphotoxin ␣, is far higher for TNF-R2 than TNF-R1, al- ated mechanism. J. Infect. Dis. 171:516. 8. Schneemann, M., G. Schoedon, S. Hofer, N. Blau, L. Guerrero, and A. Schaffner. though the signaling occurs for the most part through TNF-R1. 1993. Nitric oxide synthase is not a constituent of the antimicrobial armature of Downloaded from The TNF-R2 may have a central role in the passing on of the human mononuclear phagocytes. J. Infect. Dis. 167:1358. 9. MacKenzie, C. R., R. Langen, O. Takikawa, and W. Daübener. 1999. Inhibition ligand to the TNF-R1, as it also has a higher dissociation of indoleamine 2,3-dioxygenase in human macrophages inhibits interferon-␥- constant than does TNF-R1 (19). Studies in human cells are, induced bacteriostasis but does not abrogate toxoplasmastasis. Eur. J. Immunol. however, lacking. To our knowledge, this is the first description 29:3254. 10. Murray, H. W., B. Y. Rubin, S. M. Carriero, A. M. Harris, and E. A. Jaffee. 1985. of a possible role for TNF-R2 in the control of parasite infection Human mononuclear phagocyte antiprotozoal mechanisms: oxygen-dependent vs in humans. Although it is not clear exactly what function oxygen-independent activity against intracellular Toxoplasma gondii. J. Immu- http://www.jimmunol.org/ nol. 134:1982. TNF-R2 has, the expression alone is obviously not merely a 11. Aline, F., D. Bout, and I. Dimier-Poisson. 2002. Dendritic cells as effector cells: signal event for another mechanism ultimately responsible for ␥ interferon activation of murine dendritic cells triggers oxygen-dependent inhi- the control of parasite intracellular growth. This is clearly bition of Toxoplasma gondii replication. Infect. Immun. 70:2368. 12. Remington, J. S., and G. Desmonts. 1976. Toxoplasmosis. In Infectious Diseases shown by the abrogation of the fecundity of Toxoplasma in iDC of the Fetus and Newborn Infant. J. S. Remington and J. O. Klein, eds. when the cells are treated with neutralizing anti-TNF-R2 Ab. W. B. Saunders Co., Philadelphia, pp. 191–332 The obvious question is then what role does TNF-␣ have in the 13. Channon, J. Y., R. M. Seguin, and L. H. Kasper. 2000. Differential infectivity and division of Toxoplasma gondii in human peripheral blood leukocytes. Infect. system? Our data show that the production of TNF-␣ by both Immun. 68:4822.

iDC and mDC is very low, although it does increase in both cell 14. Wilson, C. B., and J. S. Remington. 1979. Activity of human blood leukocytes by guest on September 25, 2021 types after infection. We were unable to show that neutralizing against Toxoplasma gondii. J. Infect. Dis. 140:890. 15. Langermans, J. A., M. E. Van der Hulst, P. H. Nibbering, P. S. Hiemstra, TNF-␣ using Abs had any effect in the control of growth in iDC. L. Fransen, and R. Van Furth. 1992. IFN-␥-induced L-arginine-dependent toxo- It does remain possible that membrane-bound TNF-␣ (mTNF␣) plasmastatic activity in murine peritoneal macrophages is mediated by endoge- ␣ ␣ nous tumor necrosis factor- . J. Immunol. 148:568. or autocrine intracellular TNF- has a key role, and that we are 16. Deckert-Schlu¨ter, M., H. Bluethmann, A. Rang, H. Hof, and D. Schluter. 1998. unable to neutralize this with extracellular Abs. The iDC in Crucial role of TNF receptor type 1 (p55), but not of TNF receptor type 2 (p75), culture do tend to clump together in large aggregates after in murine toxoplasmosis. J. Immunol. 160:3427. ␣ 17. Yap, G. S., T. Scharton-Kersten, H. Charest, and A. Sher. 1998. Decreased re- infection, which implies a cell-cell contact suggesting mTNF sistance of TNF receptor p55- and p75-deficient mice to chronic toxoplasmosis as a possible initiator or regulator of Toxoplasma growth control. despite normal activation of inducible nitric oxide synthase in vivo. J. Immunol. What effect TNF-␣ has, either in terms of autocrine secretion and 160:1340. 18. Belloni, A., I. Villena, J. E. Gomez, H. Pelloux, A. Bonhomme, M. Guenounou, stimulation or in the membrane bound, remains open and requires J. M. Pinon, and D. Aubert. 2003. Regulation of tumor necrosis factor ␣ and its further study. It is interesting to note, however, that a role for specific receptors during Toxoplasma gondii infection in human monocytic cells. ␣ Parasitol. Res. 89:207. TNF- alone in the control of Toxoplasma has never been shown 19. Tartaglia, L. A., D. Pennica, and D. V. Goeddel. 1993. Ligand passing: the 75- in humans, despite the clear synergistic effect it has with IFN-␥ in kDa tumor necrosis factor (TNF) receptor recruits TNF for signaling by the this respect. We were unable to detect any IFN-␥ in the cultures; 55-kDa TNF receptor. J. Biol. Chem. 268:18542. 20. Gubbels, M. J., C. Li, and B. Striepen. 2003. High-throughput growth assay for however, we demonstrate that both the iDC and mDC are IFN-␥ Toxoplasma gondii using yellow fluorescent protein. Antimicrob. Agents Che- responsive and can inhibit the growth of Toxoplasma after mother. 47:309. 21. Pfefferkorn, E. R., and L. C. Pfefferkorn. 1977. Specific labeling of intracellular stimulation. Toxoplasma gondii with uracil. J. Protozool. 24:449. In conclusion, our data show for the first time a spontaneous 22. Deckert-Schluter, M., A. Rang, D. Weiner, S. Huang, O. D. Wiestler, H. Hof, and toxoplasmostasis in human iDC of a certain age, which is matched D. Schluter. 1996. Interferon-␥ receptor-deficiency renders mice highly susceptible to toxoplasmosis by decreased macrophage activation. Lab. Invest. 75:827. by the effect in fresh monocytes, as has been described in the 23. Scharton-Kersten, T. M., T. A. Wynn, E. Y. Denkers, S. Bala, E. Grunvald, literature. We show that TNF-R2 is expressed in the cells exhib- S. Hieny, R. T. Gazzinelli, and A. Sher. 1996. In the absence of endogenous iting the Toxoplasma growth control and not in those cell types IFN-␥, mice develop unimpaired IL-12 responses to Toxoplasma gondii while failing to control acute infection. J. Immunol. 157:4045. tested that do not exhibit toxoplasmostasis, and furthermore, that 24. Nagineni, C. N., K. Pardhasaradhi, M. C. Martins, B. Detrick, and J. J. Hooks. blocking of the TNF-R2 leads to the abrogation of the parasite 1996. Mechanisms of interferon-induced inhibition of Toxoplasma gondii repli- growth control. Further investigations are underway in our labo- cation in human retinal pigment epithelial cells. Infect. Immun. 64:4188. 25. Daübener, W., V. Posdziech, U. Hadding, and C. R. MacKenzie. 1999. Inducible ratory to attempt to elucidate the underlying mechanism behind the anti-parasitic effector mechanisms in human uroepithelial cells: tryptophan deg- expression of the TNF-R2 and the role of soluble TNF-␣ and/or radation vs. NO production. Med. Microbiol. Immunol. 187:143. ␣ 26. McLeod, R., R. Estes, D. G. Mack, and E. G. McLeod. 1983. Effects of human mTNF , and also to try and determine which subpopulation, if alveolar macrophages and peripheral blood monocytes on Toxoplasma gondii. any, of DC is involved. J. Infect. Dis. 147:957. 3374 TOXOPLASMOSTASIS OF iDC

27. Channon, J. Y., and L. H. Kasper. 1996. Toxoplasma gondii-induced immune 31. Piguet, P. F., C. D. Kan, and C. Vesin. 2002. Role of the tumor necrosis factor suppression by human peripheral blood monocytes: role of ␥ interferon. Infect. receptor 2 (TNFR2) in cerebral malaria in mice. Lab. Invest. 82:1155. Immun. 64:1181. 32. Pfeffer, K., T. Matsuyama, T. M. Kundig, A. Wakeham, K. Kishihara, 28. Subauste, C. S., and M. Wessendarp. 2000. Human dendritic cells discriminate A. Shahinian, K. Wiegmann, P. S. Ohashi, M. Kronke, and T. W. Mak. 1993. between viable and killed Toxoplasma gondii tachyzoites: dendritic cell activa- Mice deﬁcient for the 55 kd tumor necrosis factor receptor are resis- tion after infection with viable parasites results in CD28 and CD40 ligand sig- tant to endotoxic shock, yet succumb to L. monocytogenes infection. Cell naling that controls IL-12-dependent and -independent T cell production of 73:457. IFN-␥. J. Immunol. 165:1498. 29. Derouich-Guergour, D., D. Aldebert, I. Vigan, E. Jouvin-Marche, P. N. Marche, 33. Vieira, L. Q., M. Goldschmidt, M. Nashleanas, K. Pfeffer, T. Mak, and P. Scott. D. Aubert, P. Ambroise-Thomas, and H. Pelloux. 2002. Toxoplasma gondii in- 1996. Mice lacking the TNF receptor p55 fail to resolve lesions caused by infection can regulate the expression of tumor necrosis factor-␣ receptors on human fection with Leishmania major, but control parasite replication. J. Immunol. cells in vitro. Parasite Immunol. 24:271. 157:827. 30. Stoelcker, B., T. Hehlgans, K. Weigl, H. Bluethmann, G. E. Grau, and 34. Flynn, J. L., M. M. Goldstein, J. Chan, K. J. Triebold, K. Pfeffer, D. N. Mannel. 2002. Requirement for tumor necrosis factor receptor 2 expression C. J. Lowenstein, R. Schreiber, T. W. Mak, and B. R. Bloom. 1995. Tumor on vascular cells to induce experimental cerebral malaria. Infect. Immun. necrosis factor-␣ is required in the protective immune response against Myco- 70:5857. bacterium tuberculosis in mice. Immunity 2:561. Downloaded from http://www.jimmunol.org/ by guest on September 25, 2021