The JI PBL Intereron Source
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kubota, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kubota, K.
The Journal of Immunology, 2006, 176: 7576-7588.
Copyright © 2006 by The American Association of Immunologists

A Novel Functional T Cell Hybridoma Recognizes Macrophage Cell Death Induced by Bacteria: A Possible Role for Innate Lymphocytes in Bacterial Infection

Koichi Kubota1

Department of Microbiology, Kitasato University School of Medicine, Sagamihara, Kanagawa, Japan


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
We have established a novel TCR{alpha}beta (TCRVbeta6)+CD4–CD8– T cell hybridoma designated B6HO3. When the B6HO3 cells were cocultured with bacterial-infected J774 macrophage-like cells, IFN-{gamma} production by B6HO3 cells was triggered through direct cell-cell contact with dying J774 cells infected with Listeria monocytogenes (LM), Shigella flexneri, or Salmonella typhimurium that expressed the type III secretion system, but not with intact J774 cells infected with heat-killed LM, nonhemolytic lysteriolysin O-deficient (Hly–) LM, plasmid-cured Shigella, or stationary-phase Salmonella. However, the triggering of B6HO3 cells for IFN-{gamma} production involved neither dying hepatoma cells infected with LM nor dying J774 cells caused by gliotoxin treatment or freeze thawing. Cycloheximide and Abs to H-2Kd, H-2Dd, Iad, CD1d, TCRVbeta6, and IL-12 did not inhibit the contact-dependent IFN-{gamma} response, indicating that this IFN-{gamma} response did not require de novo protein synthesis in bacterial-infected J774 cells and was TCR and IL-12 independent. Thus, in an as yet undefined way, B6HO3 hybridoma recognizes a specialized form of macrophage cell death resulting from bacterial infection and consequently produces IFN-{gamma}. Moreover, contact-dependent interaction of minor subsets of splenic {alpha}beta T cells, including NKT cells with dying LM-infected J774 and bone marrow-derived macrophage (BMM) cells, proved to provide an IFN-{gamma}-productive stimulus for these minor T cell populations, to which the parental T cell of the B6HO3 hybridoma appeared to belong. Unexpectedly, subsets of {gamma}{delta} T and NK cells similarly responded to dying LM-infected macrophage cells. These results propose that innate lymphocytes may possess a recognition system sensing macrophage cell "danger" resulting from bacterial infection.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Macrophages are well known for their scavenger function. They are able to ingest and digest a variety of particulate bodies, such as macromolecular complexes, aged erythrocytes, apoptotic cells, and microbial pathogens (1). Upon infection, in addition to ingesting microbial pathogens, macrophages recognize pathogen-associated molecular patterns of bacteria, such as LPS or bacterial CpG DNA, via pattern recognition receptors that include several members of the TLR family (2), and consequently produce a variety of soluble mediators, such as cytokines and chemokines that recruit immune cells and induce inflammation (1). Among other things, IL-12 produced by macrophages (3), in conjunction with other cytokines like IL-18 (4), induces NK cells (5, 6) and some TCR{alpha}beta-bearing memory-type CD8+ T cells (7, 8, 9) to produce IFN-{gamma}, which in turn activates macrophages to ultimately kill internalized bacteria via generation of reactive nitrogen and oxygen intermediates (10). The macrophages, therefore, constitute both sentinels and the first line of defense against bacterial pathogens, and IFN-{gamma} is the most important key factor of early host resistance against bacterial infection.

Microbes, on the other hand, have evolved a variety of countermeasures to macrophage defense mechanisms (1, 11). Intracellular bacterial pathogens evade the macrophage killing mechanisms that eliminate intracellular pathogens, and thus they can survive and replicate inside macrophages (12, 13, 14, 15). In addition, a variety of bacterial pathogens are known to induce cell death in macrophages (15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26). However, the biological significance of the pathogen-induced cell death in macrophages is not fully defined with respect to the host-parasite relationship and it may differ with individual pathogens; viz, the macrophage cell death may be one evasion mechanisms of bacteria from killing by macrophages, or it may constitute a host defense response to the incoming pathogen to halt its survival within macrophages and/or to present Ags to the adaptive immune system through bystander dendritic cells (27, 28), or it may be the result of a bacterial strategy to promote disease through activation of caspase-1 which can cleave the pro-forms of the inflammatory cytokines IL-1beta and IL-18 (18, 19, 20).

In previous studies, we developed a novel T cell hybridization system, referred to as a functional T cell hybridoma, which comprised production of growth-arrested hybrids between the YACUT T cell lymphoma (29) and normal T cells and subsequent spontaneous cell transformation of the hybrids (30, 31). Compared with the ordinary T cell hybridomas using the BW5147 T cell lymphoma (32), our hybridoma system is unique in that the hybridomas dictate most of the terminally differentiated phenotypes of normal parental T cells, thus preserving the effector functions of T lymphocytes, such as T cell contact helper activity (30) and cytotoxic activity (31). Thus, we were interested in the feasibility of constructing functional T cell hybridomas with hitherto unknown phenotypes by using this hybridization system (32), as T lymphocytes were becoming recognized as consisting of functionally more heterogeneous populations than previously anticipated (33, 34, 35, 36). In a series of attempts to produce functional T cell hybridomas derived from minor T cell populations, we obtained a novel T cell hybridoma designated B6HO3. We focused in this study on the characterization of this hybridoma and found that the B6HO3 T cell hybridoma recognizes and responds to a specialized form of cell death in bacterial-infected J774 macrophage cells with the production of IFN-{gamma}, a crucial cytokine for microbicidal functions of macrophages. This finding predicted that a minor subset of T cells should have the ability to recognize cell death in macrophages resulting from bacterial infection. In fact, we have shown in this study that not only minor subsets of spleen {alpha}beta T cells but also subsets of {gamma}{delta} T and NK cells produce IFN-{gamma} in response to direct cell-cell contact with dying Listeria monocytogenes (LM)2-infected macrophage cells. Thus, this study suggests that the innate immunity may involve a lymphocyte recognition system sensing macrophage cell "danger" resulting from bacterial infection.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosures
 References
 
Mice and cell lines

We derived TCR{alpha}beta-deficient T cell line YC11 from the hypoxanthine-guanine phosphoribosyltransferase-negative YACUT T cell line (29). YC11 cells and J774 macrophage-like cells were maintained in DMEM medium containing 10% FCS. Mouse hepatoma cell line Hepa 1-6 (37) and mouse fibroblast cell line L929 were obtained from Riken Cell Bank and maintained in DMEM supplemented with 10% FCS and 4.5 g/ml glucose. Bone marrow-derived macrophages (BMM) were grown from marrow cells from C57BL/6 mice according to the method described elsewhere (38). C3H/He, DBA/2, and C57BL/6 mice were purchased from Japan SLC and Oriental Yeast. Mice were used in accordance with the institutional guideline.

Bacteria

Virulent LM (EGD strain; Hly+) and its nonhemolytic strain (ATCC15316; Hly–) were obtained from Dr. T. Fujimura (Department of Dermatology, Kitasato University School of Medicine, Sagamihara, Japan). A mutant strain of LM, DP-L4048 (39), which produces a mutant listeriolysin O (LLO) protein that is not subjected to phosphorylation, was obtained from Dr. D. A. Portnoy (Department of Molecular and Cell Biology, University of California, Berkeley, CA). LM strains were grown in tryptic soy broth overnight. Escherichia coli (ATCC25922), Salmonella typhimurium (ST; a stock strain of our laboratory), and Staphylococcus aureus (SA; a stock strain of our laboratory) were grown in brain-heart infusion broth overnight. The bacteria were washed twice with PBS by centrifugation and stored at –80°C until use. Heat-killed LM (HKLM) was prepared by incubating suspension of bacteria at 70°C for 1 h. S. typhimurium expressing type III secretion system (TTSS) was prepared according to the method described elsewhere (21). Briefly, overnight cultures of S. typhimurium were diluted to an OD measured at 600 nm of 0.1 in L-broth containing 0.3 M sodium chloride, incubated at 37°C for 3 h with shaking, and used immediately after washing twice with PBS by centrifugation. Shigella flexneri 2a strain YSH6000 and its 230-kbp plasmid-cured mutant YSH6200, obtained from Dr. C. Sasakawa (Institute of Medical Science, University of Tokyo, Tokyo, Japan) (22, 40), were grown in tryptic soy broth at 30°C overnight, then diluted 1/50 with brain-heart infusion broth, incubated at 37°C for 2.5 h on a horizontal shaker, and used immediately after washing twice with PBS.

mAbs and reagents

The following mAbs were used: FITC-conjugated, PE-conjugated, or biotinylated mAbs specific for TCR{alpha}beta (H-57-59), CD4 (GK1.5), CD8 (53-6.7), CD11c (HL-3), CD69 (H1.2F3), CD45R/B220 (RA3-6B2), DX5, TCRVbeta6 (44-22-1), CD25 (3C7), CD3 (145-2C11), NK1.1 (PK136), FcRII/III (2.4G2), and mouse IFN-{gamma} (XMG1.2) (all purchased from BD Pharmingen). Anti-CD1d (1B1) mAb and PE-Cy5-conjugated streptavidin were purchased from eBioscience. PE-conjugated anti-TCR{gamma}{delta} (GL3), PE-conjugated hamster IgG and PE-conjugated anti-mouse IL-4 (BVD6-24G2) were purchased from Caltag Laboratories. Anti-CD244 mAb (C9.1) was produced and fluorescein conjugated in our laboratory (41). Biotinylated anti-CD122 mAb (TM-beta1) was obtained from Dr. T. Tanaka (Osaka University, Suita, Japan). Anti-mouse IL-12 polyclonal Abs and its control goat Igs were purchased from PeproTech. Biotinylated anti-mouse TCR{gamma}{delta} (GL3) was purchased from Immunotech. Anti-H-2Kd, anti-H-2Dd, and anti-Iad mAbs were purchased from Meiji Institute of Health Science. Fab of anti-Vbeta6 mAb was produced by Takara Bio. PE-conjugated streptavidin was purchased form DakoCytomation. Gliotoxin and cycloheximide (CHX) were purchase from Sigma-Aldrich.

Cell fusion and hybrid selection

YC11 cells (10 x 106) were hybridized with 10 x 106 secondary C3H/He anti-DBA/2 MLC cells as described previously with slight modifications (31). Briefly, after fusion, cells were suspended in culture medium (DMEM supplemented with 10% FCS, 10 mM HEPES, 5 x 10–5 M 2-ME, and 1 mM nonessential amino acids) containing hypoxanthine/aminopterin/thymidine, distributed into the wells of a 24-well microculture plate and cultured for 18 days to eliminate unhybridized YC11 cells. Eighteen days after fusion, unhybridized MLC cells were eliminated by the panning method as described previously (30). The adherent cells on the panning plate were recovered by vigorous pipetting with culture medium, mixed with 6 x 106 irradiated DBA/2 mouse spleen cells and then pelleted by centrifugation. The pelleted cells were suspended in culture medium and distributed into the wells of a 24-well microculture plate and then incubated at 37°C in a CO2 incubator. On the next day, human rIL-2 (25 U/ml) was added to the culture. After 5 days, growing cells were collected and distributed at one cell per well into 96-well microculture plates containing irradiated DBA/2 spleen cells, and IL-2 was added at day 2. After 5 days, growing cells in each well were passaged into a 24-well microculture plate containing irradiated DBA/2 spleen cells followed by addition of IL-2 at day 2. Cells in most wells proliferated vigorously for 4–6 days and then died over time. However, cells in some wells ceased to proliferate without dying out and were observed to be gathering round macrophages which had been used as feeder cells. These survived cells were passaged into a 24-well microculture plate containing irradiated C3H/He spleen cells and IL-2. They proliferated for several days and then ceased to proliferate. After this passaging of the hybrid cells was repeated five times with intervals of 2 wk, autonomously proliferating cells appeared in one of the wells. This hybridoma cell line was designated B6HO3 and used in the present study.

Flow cytometric analysis of surface Ags

This method was described previously (30).

Ab-mediated cross-linking of the TCR-CD3 complex

Anti-CD3 mAb (25 µg/ml) diluted in Tris buffer (0.05 M Tris-HCl, pH 9.5) was incubated in a 96-well microculture plate at 4°C overnight. After unbound mAb was removed, 50 x 103 of B6HO3 cells or YACUT cells were added to each well and cultured for 16 h.

RNA extraction and RT-PCR

Total RNA was extracted from B6HO3 hybridoma and YACUT lymphoma by using the High Pure RNA isolation kit (Boehringer Mannheim). RT-PCR was performed by using the Access RT-PCR system (Promega) with two primer sets, beta2-microglobulin-specific primers as an endogenous standard (42) and cytokine-specific primers, in the same tube. Primers used for PCR are shown in Table I. They were synthesized by SIGM Genosys and designed with Primer3 software (Broad Institute, Cambridge, MA).


View this table:
[in this window]
[in a new window]
  		Table I. Primer sequences used for RT-PCR and expected sizes of the PCR products

 
Cytokine assay

Mouse IFN-{gamma}, TNF-{alpha}, and GM-CSF levels in culture supernatants were determined by ELISA using commercial ELISA kits (BioSource International). For early T lymphocyte activation protein 1 (Eta-1), a mouse osteopontin determination kit (ImmunoBiologicals) was used. For intracellular cytokine staining, cells pretreated with the blocking anti-Fc{gamma}R mAb (2.4G2) were stained with mAbs against surface markers, and then the cells were permeabilized using the Cytofix/Cytoperm Plus kit (BD Pharmingen) and stained using either FITC-conjugated anti-IFN-{gamma} or PE-conjugated anti-IL-4 mAb. The data were acquired using FACScan and analyzed using CellQuest software (BD Biosciences). For measuring intracellular IFN-{gamma} production by B6HO3 cells, B6HO3 cells were separated from J774 cells using magnetic cell sorting. Briefly, cells in cocultures of B6HO3 with LM-infected or noninfected J774 cells were collected in microcentrifuge tubes and incubated with biotinylated anti-TCR{gamma}{delta} followed by addition of streptavidin-conjugated microbeads (Miltenyi Biotec), since this mAb nonspecifically bound to J774 cells, but not to B6HO3 cells. Magnetically labeled J774 cells were removed by placing the tube in the magnetic field of a MACS separator and then the rest of the cells were assessed for intracellular IFN-{gamma}. For measuring intracellular IFN-{gamma} production by NWNA spleen cells, dying LM-infected J774 cells were excluded on the basis of forward and side light scatter.

In vitro stimulation of B6HO3 hybridoma and NWNA spleen cells by bacterial-infected J774 cells

J774 cells (1 x 105 cells/well) or BMM cells (5 x 104 cells/well) suspended in culture medium without antibiotics were seeded into a 96-well microculture plate and cultured at 37°C overnight in a CO2 incubator. After each well was washed with DMEM without antibiotics, bacteria were inoculated into each well (100 µl/well) of the microculture plate at various multiplicity of infection(MOI) and the plate was centrifuged at 800 x g for 1 min and then incubated at 37°C in a CO2 incubator. After a 1-h infection period, 50 x 103 B6HO3 cells or 8 x 105 NWNA spleen cells per 0.2 ml of culture medium containing gentamicin (200 µg/ml) were added to each well. The microculture plate was incubated at 37°C for 24 h in a CO2 incubator. For B6HO3 cells, culture supernatants were harvested and assessed for IFN-{gamma} levels by ELISA. For NWNA cells, intracellular IFN-{gamma} staining was conducted and analyzed on FACScan.

Macrophage cytotoxicity assay

Cytotoxcity was assessed by measuring the release of cytosolic lactate dehydrogenase (LDH) into the supernatants. J774 cells were incubated with bacteria for a 1-h infection period and then cultured in medium containing 200 µg/ml gentamicin. At various time points, the culture supernatants were harvested and their LDH levels were measured using the colorimetric Cytotox 96 kit (Promega). The relative LDH release was calculated as 100 x (experimental release – spontaneous release)/(total release – spontaneous release), where spontaneous release is the amount of LDH activity in the supernatant of uninfected J774 cells and total release is the activity in cell lysates.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosures
 References
 
Surface membrane phenotype of B6HO3 hybridoma

The B6HO3 hybridoma and its parental YC11 lymphoma were characterized by immunofluorescence for surface staining of Abs specific for a variety of T cell surface Ags. Fig. 1 mainly illustrates FACS profiles of the Ags that were specifically expressed by the B6HO3 cells. The B6HO3 cells expressed the H-2Dk of the C3H/He parental lymphocytes and H-2Dd of the YC11 lymphoma (Fig. 1, A–D), which, along with the fact that the mean chromosome numbers of B6HO3 were 80 (data not shown), indicates that the B6HO3 cells were hybrids. B6HO3 cells expressed 2B4 (CD244), CD11c, IL-2R{alpha} (CD25), IL-2Rbeta (CD122) and TCR{alpha}beta utilizing the Vbeta6 gene segment (Fig. 1, I–T), but did not express CD4 and CD8 (Fig. 1, E–H), whereas none of these Ags were expressed on the parental YC11 cells. CD244 and CD11c, which are primarily expressed by NK cells (41) and dendritic cells, respectively, are known to be expressed on activated/memory-type T cells (43, 44). Other activation markers for T cells, such as CD69, CD45RBhigh, and CD44high (45), were also detected on the B6HO3 cells as well as the parental YC11 cells (data not shown). These results demonstrate that the B6HO3 is the CD4–CD8–TCR{alpha}beta+ T cell hybridoma exhibiting the activated/memory phenotype. Since some double-negative T cells have been reported to express surface markers such as B220, CD16/CD32 (46), or DX5 (35), we also examined the expression of those Ags on B6HO3 cells and found that B6HO3 cells were negative for the Ags (data not shown).


Figure 1
View larger version (58K):
[in this window]
[in a new window]
  		FIGURE 1. Fluorescence flow cytometric analysis of surface membrane phenotypes. YC11 lymphoma cells and B6HO3 hybridoma cells were stained with (shaded curve) or without (open curve) each of mAbs indicated on the bottom of the panel and analyzed using FACScan. a.u., Arbitrary units.

 
Cytokine production by B6HO3 cells upon Ab-mediated cross-linking of the TCR-CD3 complex

B6HO3 cells were next studied for their ability to produce cytokines following activation of the cells by cross-linking of the TCR-CD3 complex with solid-phase anti-CD3 mAb. Fig. 2, A and B, show the analysis of RT-PCR products from GM-CSF, IFN-{gamma}, IL-2, IL-4, and TNF-{alpha} mRNAs extracted from the B6HO3 hybridoma and the YACUT lymphoma before and after cross-linking of the CD3-TCR complex. The CD3-TCR ligation induced B6HO3 cells to express GM-CSF and IFN-{gamma} mRNA, but not IL-2 and IL-4 mRNA, while parental YACUT lymphoma cells transcribed only IL-4 mRNA in response to CD3-TCR ligation, whose responsiveness was thus suppressed in the B6HO3 hybridoma. TNF-{alpha} mRNA was constitutively expressed in both YACUT and B6HO3 cells, and CD3-TCR ligation did not enhance the levels of its expression in both of the cells. Since some double-negative T cells have been reported to produce MCP-1(CCL2), Eta-1 (47), or IL-10 (34), investigation of their mRNAs expression was included in this experiment. As shown in Fig. 2C, although YACUT cells did not express Eta-1 mRNA before and after CD3-TCR ligation, B6HO3 cells constitutively expressed Eta-1 mRNA, and the expression level was enhanced by CD3-TCR ligation. MCP-1 and IL-10 mRNAs were not detectable in both B6HO3 and YACUT cells before and after CD3-TCR ligation (data not shown).


Figure 2
View larger version (47K):
[in this window]
[in a new window]
  		FIGURE 2. Cytokine expression in B6HO3 hybridoma and its parental YACUT lymphoma upon Ab-mediated cross-linking of the TCR-CD3 complex. B6HO3 cells (A) and YACUT cells (B) were cultured in the presence (+) or absence (–) of solid-phase anti-CD3 mAb for 16 h, and total RNA was extracted, subjected to RT-PCR with two primer sets, beta2-microglobulin-specific primers as an endogenous standard and cytokine specific primers, in the same tube, and then electrophoresed on 3.5% agarose gel. Cytokines examined were GM-CSF, IFN-{gamma}, IL-2, IL-4, and TNF-{alpha}, whose PCR primer pairs and product sizes are shown in Table I. The product sizes are also shown in parentheses. The band corresponding to 301 bp in each lane is a PCR product specific for beta2-microglobulin. Lane M, Molecular size marker. C, Eta-1 mRNA expression by B6HO3 and YACUT cells was examined as in A and B. D, B6HO3 and YACUT cells were cultured in the presence (+) or absence (–) of solid-phase CD3 mAb for 24 h, and culture supernatants were assessed for the production of cytokines by ELISA.

 
To see production of cytokines at a translational level, secretion of GM-CSF, IFN-{gamma}, TNF-{alpha}, and Eta-1 into culture supernatants following activation of B6HO3 and YACUT cells by Ab-mediated cross-linking of the TCR-CD3 complex was assessed by ELISA (Fig. 2D). As expected from the results obtained in the RT-PCR experiment (Fig. 2, A and B), B6HO3 cells secreted both GM-CSF and IFN-{gamma} into culture supernatants upon CD3-TCR ligation. Unexpectedly, TNF-{alpha} secretion was observed only in B6HO3 cells upon CD3-TCR ligation although TNF-{alpha} mRNA was constitutively expressed in both B6HO3 and YACUT cells without its expression levels being enhanced by CD3-TCR ligation (Fig. 2, A and B). Similarly, although Eta-1 mRNA was constitutively expressed in B6HO3 cells (Fig. 2C), Eta-1 secretion by B6HO3 cells was strongly induced upon activation of the cells by CD3-TCR ligation (Fig. 2D).

B6HO3 hybridoma produces IFN-{gamma} when cocultured with dying LM-infected J774 macrophage cells

In the course of the cell fusion experiments, we noticed that the IL-2-dependent hybrid cells, precursor cells of the B6HO3 hybridoma before its transformation, were gathering around the macrophages that had been used for feeder cells. Thus, we hypothesized that the hybridoma cells might functionally interact with macrophages. To test this possibility, we cocultured B6HO3 cells with the J774 macrophage-like cell line alone or J774 cells infected with each one of the following bacteria: LM, HKLM, LLO-deficient (Hly–) LM, E. coli, ST, and SA, and then supernatants harvested at 24 h of coculture were assessed for IFN-{gamma} levels by ELISA as illustrated in Fig. 3A. High amounts of IFN-{gamma} were detectable only in cocultures of B6HO3 cells with LM-infected J774 cells. J774 cells cultured with HKLM and J774 cells infected with Hly– LM, Escherichia, Salmonella, or Staphylococcus displayed no significant ability of stimulating B6HO3 cells to produce IFN-{gamma}. Interestingly, among the bacterial-infected J774 cells examined, wild type LM-infected J774 cells were the only cells that were microscopically observed to undergo cell death.


Figure 3
View larger version (23K):
[in this window]
[in a new window]
  		FIGURE 3. IFN-{gamma} production by B6HO3 hybridoma cocultured with LM-infected J774 cells. A, Supernatants were collected from 24-h cultures of the following combinations: J774 cells alone, B6HO3 cells alone, B6HO3 and J774 cells, B6HO3 and LM, LM-infected J774 cells, B6HO3 and LM-infected J774 cells, Hly-deficient(Hly–) LM-infected J774 cells, B6HO3 and Hly– LM-infected J774 cells, HKLM and J774 cells, B6HO3 and HKLM-inoculated J774 cells, E. coli-infected J774 cells, B6HO3 and E. coli-infected J774 cells, ST-infected J774 cells, B6HO3 and ST-infected J774 cells, SA-infected J774 cells, or B6HO3 and SA-infected J774 cells. Infection was performed at MOI of 50:1. IFN-{gamma} levels of supernatants were measured by ELISA. B, IFN-{gamma} production by B6HO3 cells was assessed by intracellular cytokine staining. B6HO3 cells were cultured alone (open curve) or cocultured with J774 cells (lower panel, shaded curve) or LM-infected J774 cells (upper panel, shaded curve) for 24 h. Monensin was added for the last 10 h of culture. B6HO3 cells were separated from J774 cells by magnetic cell sorting, stained for intracellular IFN-{gamma}, and analyzed using FACScan. Depletion of J774 cells was confirmed by the fact that no cells were detected in the forward and light scatter area gated for B6HO3 after magnetic cell sorting of control LM-infected J774 cells. Representative data from one of two experiments are shown.

 
To verify that B6HO3 cells themselves produce IFN-{gamma}, flow cytometric analysis of IFN-{gamma} production was performed by using intracellular cytokine staining (Fig. 3B). Approximately 56% of the B6HO3 cells cocultured with LM-infected J774 cells were positive for intracytoplasmic IFN-{gamma} (Fig. 3B, upper panel), whereas no IFN-{gamma} was detectable in B6HO3 cells cocultured with noninfected J774 cells (Fig. 3B, lower panel). These results indicate that the B6HO3 hybridoma produces IFN-{gamma} in response to dying LM-infected J774 cells.

Cell-cell contact is necessary for activation of B6HO3 cells by dying LM-infected J774 cells, but the TCR is not involved in its recognition mechanism

We next addressed a question of whether soluble mediators secreted by LM-infected J774 cells were responsible for the IFN-{gamma} production by B6HO3 cells. To this end, we compared the ability of LM-infected J774 cells to induce IFN-{gamma} production by B6HO3 cells with that of LM-infected J774 supernatants collected at 24 h postinfection (Fig. 4A). Since IL-12 plays a pivotal role in the production of IFN-{gamma} by T and NK cells (3), we also examined whether IL-12 played a role in this response by coculturing B6HO3 cells with LM-infected J774 cells in the presence of neutralizing Ab specific for IL-12. Fig. 4A shows IFN-{gamma} levels in supernatants collected from those 24-h cocultures. A culture supernatant from LM-infected J774 cells was incapable of inducing IFN-{gamma} by B6HO3 and addition of anti-IL-12 Ab to the coculture did not inhibit the production of IFN-{gamma}. These results indicate that the IFN-{gamma} response is IL-12 independent and not triggered by soluble mediators.


Figure 4
View larger version (31K):
[in this window]
[in a new window]
  		FIGURE 4. Requirement of direct cell-cell contact between B6HO3 cells and LM-infected J774 cells for IFN-{gamma} production by B6HO3 hybridoma, but no requirement of the TCR and de novo protein synthesis for B6HO3 hybridoma recognition of LM-infected J774 cells. A, B6HO3 cells were cultured for 24 h with a 24-h culture supernatant (*1) from LM-infected J774 cells or cocultured with LM-infected J774 cells in the presence or absence of neutralizing Ab to IL-12 (2 µg/ml), and then culture supernatants were assessed for IFN-{gamma} levels by ELISA. B, B6HO3 cells were cocultured for 24 h with LM-infected J774 cells in either a 24-well microculture plate or a Transwell (*2). C, Effect of Abs on production of IFN-{gamma} by B6HO3 hybridoma. B6HO3 cells were cocultured for 24 h with LM-infected J774 cells in the presence or absence of the mAb indicated. Data are expressed as the percentage of IFN-{gamma} production in the presence vs absence of the mAb indicated. D, J774 cells were treated or untreated with 10 µg/ml CHX for 10 min, infected with LM, and B6HO3 cells were added to the cultures at 6 h postinfection without removing CHX. After 4 h of coculture, total RNA was extracted from each culture indicated and IFN-{gamma} mRNA expression was assessed by RT-PCR as in Fig. 2. E, Culture supernatants from LM-infected J774 cells treated or untreated with 10 µg/ml CHX for 10 h were assessed for TFN-{alpha} levels by ELISA. Representative data from one of two experiments are shown.

 
To determine whether cell-cell contact between B6HO3 cells and dying LM-infected J774 cells is required for the production of IFN-{gamma}, B6HO3 cells were cocultured with LM-infected J774 cells in a Transwell to separate both cells with a permeable membrane, which inhibited cell-cell contact (Fig. 4B). Separation of B6HO3 cells from dying LM-infected J774 cells failed to induce IFN-{gamma}, indicating that cell-cell contact between B6HO3 cells and LM-infected J774 cells was required for the activation of B6HO3 cells.

To evaluate the contribution of the TCR expressed on B6HO3 cells in the cognitive interaction between B6HO3 cells and dying LM-infected J774 cells, we examined the effect of Fab of anti-TCRVbeta6 mAb and mAbs to H-2Kd, H-2Dd, Iad, and CD1d, potential ligands for the TCR, on the IFN-{gamma} response by adding each one of these mAbs to cocultures of B6HO3 cells with LM-infected J774 cells (Fig. 4C). Fab of anti-TCRVbeta6 mAb as well as anti-H-2Dd, anti-Iad, and anti-CD1d mAbs did not inhibit IFN-{gamma} production by B6HO3 cells. Anti-H-2Kd mAb and a mixture of anti-H-2Kd and anti-H-2Dd mAbs slightly enhanced the IFN-{gamma} production for an unknown reason. These results indicate that the TCR expressed on B6HO3 cells is not involved in the mechanism underlying the cognitive interaction between B6HO3 cells and dying LM-infected J774 cells, and thus the IFN-{gamma} response is TCR independent.

De novo protein synthesis is not required for B6HO3 hybridoma cell recognition of dying LM-infected J774 cells

Since a variety of proteins, such as cytokines and heat shock proteins, are induced to synthesize in bacterial-infected macrophages, we next asked whether de novo protein synthesis in LM-infected J774 cells was required for the B6HO3 cell recognition of dying LM-infected J774 cells. To this end, J774 cells that had been treated with CHX, an inhibitor of protein synthesis, were infected with LM, and at 6 h postinfection the infected cells were cocultured with B6HO3 cells without removing CHX. After 4 h of coculture, total RNA was extracted from the cultures and assessed for the expression of IFN-{gamma} mRNA by RT-PCR (Fig. 4D). Although no IFN-{gamma} mRNA was detectable in cultures of J774 cells, LM-infected J774 cells, and B6HO3 cells alone, IFN-{gamma} mRNA expression was detected in cocultures of B6HO3 cells with LM-infected J774 cells, and this IFN-{gamma} mRNA expression was not inhibited by the treatment of J774 cells with CHX. The inhibitory effect of CHX on protein synthesis was confirmed by the parallel experiment showing that TNF-{alpha} production by LM-infected-J774 cells was strongly inhibited in the presence of CHX (Fig. 4E). These results indicate that the B6HO3 cell recognition of dying LM-infected J774 cells does not require de novo protein synthesis in LM-infected J774 cells.

IFN-{gamma} production by B6HO3 hybridoma is also induced through direct cell-cell contact with dying J774 cells infected with S. flexneri or ST that expressed TTSS

Since S. flexneri and ST had been reported to induce cell death in macrophages (16, 19, 20, 21, 22, 23, 24, 25, 26), we next determined whether cell death in J774 cells caused by Shigella or Salmonella infection could also induce B6HO3 cells to produce IFN-{gamma}. Fig. 5A shows the time course of cell destruction following infection of J774 cells with LM or Shigella as assessed by LDH release into culture supernatants. Infection of J774 cells with LM resulted in cell destruction beginning at 8 h postinfection, reaching a maximum of 60% death within 20 h. Infection of J774 cells with wild-type Shigella YSH6000 resulted in almost complete destruction of the cells within 16 h, while plasmid-cured Shigella-infected J774 cells did not undergo cell death. Fig. 5B shows IFN-{gamma} levels in supernatants from 24-h cocultures of B6HO3 cells with wild-type Shigella- or plasmid-cured Shigella-infected J774 cells. Like dying LM-infected J774 cells, dying wild-type Shigella-infected J774 cells induced IFN-{gamma} production by B6HO3 cells, whereas intact J774 cells infected with plasmid-cured Shigella YSH6200 did not significantly induce B6HO3 cells to produce IFN-{gamma}. Fig. 5C shows the result of an experiment using a Transwell to separate B6HO3 cells from dying wild-type Shigella-infected J774 cells, and the result of an experiment in which the ability of a supernatant collected from a 24-h culture of wild-type Shigella-infected J774 cells to induce IFN-{gamma} production by B6HO3 cells was examined. Separation of wild-type Shigella-infected J774 cells and B6HO3 cells by a permeable membrane completely abrogated the IFN-{gamma} production by B6HO3 cells, and a culture supernatant from wild-type Shigella-infected J774 cells had no ability to stimulate B6HO3 cells to produce IFN-{gamma}. These results indicate that direct cell-cell contact between B6HO3 cells and dying Shigella-infected J774 cells is necessary for the IFN-{gamma} production by B6HO3 cells.


Figure 5
View larger version (30K):
[in this window]
[in a new window]
  		FIGURE 5. IFN-{gamma} production by B6HO3 cells cocultured with and time course of LDH release from LM-, Shigella- and Salmonella-infected J774 cells. A and D, J774 cells were infected with LM ({blacksquare}) (MOI, 50:1), wild-type Shigella (YSH6000) (•) (MOI, 4:1), plasmid-cured Shigella (YSH6200) ({blacktriangleup}) (MOI, 20:1), ST culture grown under conditions for TTSS expression (*) (MOI, 10:1), or ST in stationary phase ({blacktriangledown}) (MOI, 50:1). At the indicated time points, LDH contents in culture supernatants were determined and relative LDH release was calculated. B and E, Supernatants from 24-h cocultures of B6HO3 cells with LM (MOI, 50:1)-, wild-type Shigella (YSH-6000)(MOI, 4:1)-, plasmid-cured Shigella (YSH6200)(MOI, 20:1)-, TTSS-expressing ST (MOI, 10:1)-, or stationary-phase ST (MOI, 50:1)-infected J774 cells were assessed for IFN-{gamma} levels by ELISA. C and F, B6HO3 cells were cocultured for 24 h with wild-type Shigella (YSH-6000)(MOI, 4:1)-infected J774 cells or with J774 cells infected with ST grown under conditions for TTSS expression in a 24-well microculture plate or in a Transwell, and culture supernatants were assessed for IFN-{gamma} production by ELISA. *1 and *5, B6HO3 cells were added to an upper chamber of a Transwell. *2, A 24-h culture supernatant from wild-type Shigella-infected J774 cells. *3, ST grown under conditions for TTSS expression. *4, Stationary-phase culture of ST. *6, A 24-h culture supernatant from J774 cells infected with ST grown under conditions for TTSS expression. Data are representative from one of two experiments.

 
The same results were obtained with ST as shown in Fig. 5, D–F. J774 cells infected with Salmonella grown under conditions for TTSS expression underwent rapid cell death (Fig. 5D) and simultaneously stimulated B6HO3 cells to produce IFN-{gamma} (Fig. 5E). However, J774 cells infected with Salmonella in stationary phase neither underwent cell death (Fig. 5D) nor induced IFN-{gamma} production by B6HO3 cells (Fig. 5E). In the same way as Shigella, the necessity of cell-cell contact between B6HO3 cells and dying Salmonella-infected J774 cells for the induction of IFN-{gamma} was verified by both the cell separation experiment and the experiment using a supernatant collected from 24-h cultures of J774 cells infected with Salmonella grown under conditions for TTSS expression (Fig. 5F).

These results indicate that cell death in bacterial-infected J774 cells is closely associated with the ability of the infected cells to stimulate B6HO3 cells to produce IFN-{gamma} in a direct cell-cell contact-dependent manner.

Cell death in J774 cells induced by gliotoxin or freeze thawing and cell death in hepatoma cells caused by LM infection do not trigger B6HO3 hybridoma to produce IFN-{gamma}

To determine whether B6HO3 cells recognize macrophage cell death per se, we examined whether cell death in J774 cells induced by the treatment of gliotoxin (48) or freeze thawing, which induce apoptosis and necrosis, respectively, in macrophages, could also induce B6HO3 to produce IFN-{gamma}. Fig. 6A shows the time course of cell destruction following a 3-h treatment of J774 cells with gliotoxin as assessed by LDH release into culture supernatants. The gliotoxin treatment of J774 cells resulted in cell destruction with similar kinetics to cell death in LM-infected J774 cells. However, the dying gliotoxin-treated J774 cells, when cocultured with B6HO3 cells, did not induce IFN-{gamma} production by B6HO3 cells as assessed by IFN-{gamma} levels in culture supernatants (Fig. 6B). Fig. 6B also shows that dead J774 cells subjected to freeze and thaw did not stimulate B6HO3 cells to produce IFN-{gamma}. These results indicate that the B6HO3 hybridoma discriminates between typical apoptotic and necrotic cell death in macrophages and macrophage cell death caused by bacterial infection.


Figure 6
View larger version (56K):
[in this window]
[in a new window]
  		FIGURE 6. Time course of LDH release from gliotoxin-treated J774 cells, and B6HO3 hybridoma cell recognition of dying J774 cells caused by gliotoxin treatment or freeze thawing and of dying Hepa 1-6 cells infected with mutant LM strain DP-L4048. A, J774 cells were treated with gliotoxin (10 µM) for 3 h, washed extensively, and at the indicated time point, culture supernatants were collected and LDH contents were determined. B, J774 cells infected with LM, or treated with gliotoxin as in A, or subjected to freeze thawing, were cocultured with B6HO3 cells for 24 h and supernatants were assessed for IFN-{gamma} production. C–F, phase-contrasted microscopic observations of control Hepa 1-6 cells (C) and J774 cells (E), and those of dying Hepa 1-6 hepatoma cells (D) and J774 cells (F) infected at MOI of 4:1 with mutant LM strain DP-L4048 at 6 h postinfection. G, B6HO3 cells were cocultured with dying DP-L4048-infected Hepa 1-6 cells or dying DP-L4048-infected J774 cells for 24 h, and IFN-{gamma} levels in supernatants were assessed by ELISA. Gentamicin was added to cultures at 6 h postinfection. Data are representative from one of two experiments.

 
We next addressed a question of whether B6HO3 cells could produce IFN-{gamma} in response to cell death in LM-infected cells other than macrophages. To this end, we infected Hepa 1-6 hepatoma cells with mutant LM strain DP-L4048, because infection of Hepa 1-6 cells with wild-type LM did not result in cell death in Hepa 1-6 cells (data not shown). The mutant strain DP-L4048 expresses mutant LLO protein, whose potential phosphate acceptor residues in the PEST-like sequence were all changed to a residue that cannot accept phosphate, and thus the LLO proteins are not subjected to degradation in the cytosol, being capable of inducing rapid membrane damage (39). Fig. 6, C–F, show phase-contrasted microscopic observations of normal Hepa 1-6 cells and J774 cells (Fig. 6, C and E) and those infected with the mutant LM stain DP-L4048 Fig. 6, D and F) in the absence of gentamicin at 6 h postinfection. As has been reported (39), J774 cells infected with the mutant strain underwent rapid cell death within 6 h (Fig. 6F). Similarly, the mutant strain induced rapid cell death in Hepa 1-6 cells (Fig. 6D). To see whether both of the dying cells have the ability to induce B6HO3 cells to produce IFN-{gamma}, the DP-L4048-infected J774 and Hepa 1-6 cells were cocultured with B6HO3 cells for 24 h, and IFN-{gamma} levels in the supernatants were assessed by ELISA. Fig. 6G shows that dying DP-L4048-infected J774 cells induced B6HO3 to produce IFN-{gamma}, but dying DP-L4048-infected Hepa 1-6 cells did not. These results indicate that the IFN-{gamma} response of B6HO3 is a bacterial-infected macrophage death-specific phenomenon.

A minor subset of {alpha}beta T cells and subsets of {gamma}{delta} T and NK cells respond to dying LM-infected J774 cells with the production of IFN-{gamma} in a cell-cell contact-dependent manner

The above-mentioned functional phenotype of the T cell hybridoma B6HO3 predicted that a minor subset of T cells should recognize dying LM-infected J774 cells. To address this question, wild-type LM- or Hly– LM-infected J774 cells were cocultured with C3H/He mouse nylon wool nonadherent (NWNA) spleen cells in ordinary wells or Transwells in the presence of either neutralizing Ab specific for IL-12 or control Ab, and 24 h after coculture the NWNA spleen cells were assessed for IFN-{gamma} production by intracellular cytokine staining in combination with surface staining of TCR{alpha}beta and NK cell marker DX5. A representative result is shown in Fig. 7, A–K. Although coculture of NWNA spleen cells with Hly– LM-infected J774 cells, which had not undergone cell death, did not result in the production of IFN-{gamma} by NWNA spleen cells (Fig. 7B), dying wild-type LM-infected J774 cells indeed induced 2.3 ± 0.5% of {alpha}beta T cells to produce IFN-{gamma} (Fig. 7D), of which the proportion was decreased to 0.7 ± 0.1% by addition of anti-IL-12 Ab (Fig. 7G). It is noteworthy that the IFN-{gamma}-producing {alpha}beta T cells expressed intermediate levels of the TCR{alpha}beta. Unexpectedly, dying LM-infected J774 cells also induced a small percentage of non-{alpha}beta T cells (6.1 ± 0.6% of total NWNA spleen cells) to produce IFN-{gamma} (Fig. 7D, lower right quadrant), and those cells were decreased to 3.0 ± 0.4% in the presence of anti-IL-12 Ab (Fig. 7G, lower right quadrant). When the dot plots were gated on IFN-{gamma}-producing non-{alpha}beta T cells and assessed for DX5 expression, 89.2 ± 4.2% and 92.1 ± 2.7% of the IFN-{gamma}-producing non-{alpha}beta T cells were DX5+ in the absence and presence of anti-IL-12 Ab, respectively (Fig. 7, E and H), indicating that most of the IFN-{gamma}-producing non-{alpha}beta T cells were NK cells. Of total NK cells ({alpha}betaTCR–DX5+ cells), 41.4 ± 5.0% were IFN-{gamma}+ cells, which were reduced to 26.3 ± 2.6% in the presence of anti-IL-12 Ab (data not shown). Since ~10% of IFN-{gamma}-producing non-{alpha}betaT cells were not NK cells, we next determined whether {gamma}{delta}T cells also produced IFN-{gamma}. As shown in Fig. 7, I–K, when NWNA spleen cells cocultured with dying LM-infected J774 cells in the presence of anti-IL-12 Ab (Fig. 7J) or control Ab (Fig. 7I) were stained with PE-conjugated mAb specific for TCR{gamma}{delta} (Fig. 7, I and J) or control PE-conjugated hamster IgG (Fig. 7K) followed by intracellular IFN-{gamma} staining, 21 ± 0.2% of {gamma}{delta} T cells, which comprised 2.4 ± 0.3% of the NWNA spleen cell population, produced IFN-{gamma} in the absence of anti-IL-12 Ab and 6.4 ± 1.6% of {gamma}{delta} T cells produced IFN-{gamma} in the presence of anti-IL-12 Ab. Fig. 7, C and F, show that separation of NWNA spleen cells from dying LM-infected J774 through permeable membrane failed to induce NWNA spleen cells to produce IFN-{gamma}, indicating that IFN-{gamma} production by a minor subset of {alpha}beta T cells and subsets of {gamma}{delta} T and NK cells is completely dependent on direct cell-cell contact with dying LM-infected J774 cells. To ensure that LM-infected J774 culture supernatants do not include soluble mediators that can induce T and NK cells to produce IFN-{gamma}, NWNA spleen cells were cultured for 24 h with (Fig. 7M) or without (Fig. 7L) a supernatant collected from 24-h cultures of LM-infected J774 cells, and the production of IFN-{gamma} by NWNA spleen cells was assessed by intracellular cytokine staining. As shown in Fig. 7M, a culture supernatant from LM-infected J774 cells was incapable of inducing IFN-{gamma} by NWNA spleen cells.


Figure 7
View larger version (44K):
[in this window]
[in a new window]
  		FIGURE 7. Subsets of {alpha}beta T cells, {gamma}{delta} T cells, and NK cells produce IFN-{gamma} in response to dying LM-infected J774 cells. C3H/He mouse NWNA spleen cells were cocultured for 24 h with J774 cells alone (A), Hly– LM-infected J774 cells (B), or LM-infected J774 cells in an ordinary 24-well plate (D and G) or a Transwell (C and F) in the presence (2 µg/ml) of anti-IL-12 Ab (F and G) or control Ab (C and D). Monensin was added for the last 6 h of culture. NWNA spleen cells were assessed for TCR{alpha}beta and DX5 expression by double-labeled surface staining and for IFN-{gamma} production by intracellular cytokine staining. The histograms represent TCR{alpha}beta–IFN-{gamma}+ gated cells stained for DX5. I–K, C3H/He NWNA spleen cells cocultured with dying LM-infected J774 cells in the presence of anti-IL-12 Ab (J) or control Igs (I) were assessed for TCR{gamma}{delta} expression and intracellular IFN-{gamma} production. Control staining was conducted by using PE-conjugated hamster IgG (K). L and M, C3H/He mouse spleen cells were cultured for 24 h with (M) or without (L) a 24-h culture supernatant collected from dying LM-infected J774 cells and were assessed for TCR{alpha}beta expression and intracellular IFN-{gamma} production. N and O, Effect of various concentrations of anti-IL-12 Ab on the IFN-{gamma} production by {alpha}beta T and non-{alpha}betaT cells was examined. IFN-{gamma}+ cells were expressed as the percentage of IFN-{gamma}+ cells in the presence of anti-IL-12 Ab vs control Ab. The numbers shown in the upper right quadrants represent mean percentages of IFN-{gamma}+ cells in {alpha}beta T cells or {gamma}{delta} T cells, and those in the lower right quadrants represent mean percentages of IFN-{gamma}+ non-{alpha}beta T cells in total spleen NWNA cells from four experiments.

 
To evaluate the role of IL-12, various amounts of anti-IL-12 Ab were added to the cocultures and their effects on the IFN-{gamma} production by NWNA spleen cells were examined (Fig. 7, N and O). Addition of a range of concentrations between 0.01 and 0.3 µg/ml anti-IL-12 Ab resulted in a decrease in the numbers of IFN-{gamma}-producing {alpha}beta T and non-{alpha}beta T cells by ~25 and ~50%, respectively, but higher concentrations than 1 µg/ml anti-IL-12 Ab no longer reduced the number of IFN-{gamma}-producing cells. These results indicate that 1 µg/ml anti-IL-12 Ab was sufficient to neutralize IL-12 and that ~25% of the IFN-{gamma}-producing {alpha}beta T cells and ~50% of IFN-{gamma}-producing non-{alpha}beta T cells did not require IL-12 for the production of IFN-{gamma}.

Overall, these results indicate that a small percentage of {alpha}beta T cells recognize and respond to dying LM-infected J774 cells by producing IFN-{gamma} in a similar way as the B6HO3 T cell hybridoma. Furthermore, subsets of {gamma}{delta} T and NK cells proved to have the same ability of responding to dying LM-infected J774 cells as this minor subset of {alpha}beta T cells.

Surface membrane phenotype of IFN-{gamma}-producing {alpha}beta T cells

We further examined CD4 and CD8 expression on the IFN-{gamma}-producing {alpha}beta T cells responding to dying LM-infected J774 cells by using three-color flow cytometric analysis. C3H/He mouse NWNA spleen cells that had been cocultured with LM-infected J774 cells in the presence of anti-IL-12 Ab or control Ab were stained for TCR{alpha}beta, intracellular IFN-{gamma}, and CD4 or CD8, subjected to flow cytometry, and then dot plots were gated on IFN-{gamma}-producing {alpha}beta T cells and analyzed for the expression of CD4 and CD8. A representative result is shown in Fig. 8A. Of the IFN-{gamma}-producing {alpha}beta T cells, 5.2 ± 2.3% were CD8+ and 55.8 ± 9.2% were CD4+ in the absence of anti-IL-12 Ab, and 7.3 ± 2.4% were CD8+ and 43.9 ± 9.4% were CD4+ in the presence of anti-IL-12 Ab. These results indicate that the IFN-{gamma}-producing {alpha}beta T cell population mainly consists of CD4+ (~50%) and CD4–CD8– (~44%) {alpha}beta T cells. It was noteworthy that the level of CD4 expression on the surface of the IFN-{gamma}-producing {alpha}beta T cells (Fig. 8, middle and bottom panels) was lower that that of ordinary T cells (Fig. 8, top panel).


Figure 8
View larger version (44K):
[in this window]
[in a new window]
  		FIGURE 8. Phenotypic characterization of IFN-{gamma}-producing {alpha}beta T cells by means of three-color flow cytometric analysis. A, C3H/He NWNA spleen cells that had been cocultured with dying LM-infected J774 cells for 24 h in the presence (2 mg/ml) of anti-IL-12 Ab or control Ab were assessed for the expression of TCR{alpha}beta, intracellular IFN-{gamma}, and CD4 or CD8. A representative result from four experiments is shown, and the numbers within the histograms represent the mean percentages of CD8+ or CD4+ cells in gated IFN-{gamma}+ {alpha}beta T cells. B, C57BL/6 NWNA spleen cells were assessed for the expression of TCR{alpha}beta, NK1.1, and intracellular IFN-{gamma} or IL-4. The numbers within the histograms represent the mean percentages of NK1.1+ cells in gated IFN-{gamma}+ {alpha}beta T and non-{alpha}betaT cells from three experiments. C, The percentages of IFN-{gamma}+ cells in NKT cells were determined from three experiments and shown in the histograms. D, C57BL/6 NWNA spleen cells cultured as in B in the presence or absence of blocking mAb specific for CD1d were subjected to three-color flow cytometric analysis, and the numbers of IFN-{gamma}+ NKT cells per 105 NWNA spleen cells were determined.

 
IFN-{gamma}-producing {alpha}beta T cells involve NKT cells

Because the surface membrane phenotype of the IFN-{gamma}-producing {alpha}beta T cells responding to dying LM-infected J774 cells was similar to that of NKT cells (35), we next determined whether the IFN-{gamma}-producing {alpha}beta T cells were NKT cells or not. To this end, NWNA spleen cells from a C57BL/6 mouse instead of a C3H/He mouse were cocultured with dying LM-infected J774 cells in the presence of anti-IL-12 Ab or control Ab, stained with mAbs specific for TCR{alpha}beta and NK1.1 followed by intracellular IFN-{gamma} and IL-4 staining, and then three-color flow cytometric analysis was performed (Fig. 8B). Dying LM-infected J774 cells did not induce NWNA spleen cells to produce IL-4, but induced 2.9 ± 0.8% of {alpha}beta T cells to produce IFN-{gamma}, of which the proportion was decreased by 0.6 ± 0.2% in the presence of anti-12 Ab. Of the IFN-{gamma}-producing {alpha}beta T cells, 27.2 ± 2.3% were NK1.1+ in the absence of anti-IL-12 Ab, and 24.8 ± 1.8% were NK1.1+ in the presence of anti-IL-12 Ab. The NK1.1+ T cells comprised 1.5 ± 0.2% of the NWNA spleen cell population and this proportion remained unchanged before and after coculture with LM-infected J774 cells (data not shown). These results indicate that IFN-{gamma}-producing {alpha}beta T cells consist of NK1.1+ NKT cells (~25%) and non-NKT cells (~75%). When the opposite analysis was done by gating on NK1.1+ T cells and evaluating the percentage of IFN-{gamma}-producing cells in this NKT cell population, we found that 60 ± 5.5% and 10 ± 2.7% of the NK1.1+ NKT cells were IFN-{gamma}+ in the absence and presence of anti-IL-12 Ab, respectively (Fig. 8C).

Because most of NK1.1+ NKT cells carry an invariant TCR restricted to interactions with the class I-like molecule CD1d, we determined whether CD1d was involved in the NKT cell recognition of dying LM-infected J774 cells. To this end, blocking anti-CD1d mAb was added to cocultures of C57BL/6 mouse NWNA spleen cells with LM-infected J774 cells in the presence of anti-IL-12 Ab or control Ab, and the numbers of IFN-{gamma}+NK1.1+ NKT cells per 105 NWNA spleen cells were determined by three-color flow cytometric analysis (Fig. 8D). Addition of anti-CD1d mAb did not reduce the numbers of both the IL-12-dependent and -independent IFN-{gamma}+NK1.1+ NKT cells that responded to dying LM-infected J774 cells. This result indicates that the V{alpha}14 invariant TCR expressed on NKT cells is not involved in the mechanism by which NKT cells recognize dying LM-infected J774 cells. Also this result is consistent with the conclusion that the contact-dependent interaction between B6HO3 T cell hybridoma and dying LM-infected J774 cells is not mediated by the TCR of B6HO3 cells (Fig. 4C).

Cell death in LM-infected BMM cells is also associated with the IFN-{gamma} production by subsets of {alpha}beta T cells, {gamma}{delta} T cells, and NK cells

Since LM had been reported to induce nonapoptotic cell death in BMM (17), we reiterated the same experiments using BMM instead of the J774 macrophage cell line to corroborate that dying LM-infected macrophages have the ability to stimulate subsets of lymphocytes to produce IFN-{gamma} in a cell-cell contact-dependent manner. C57BL/6 mouse BMM cells were infected with LM and cocultured with syngeneic C57BL/6 mouse NWNA spleen cells in ordinary wells or Transwells in the presence of either neutralizing anti-IL-12 Ab or control Ab, and 20 h after coculture the NWNA spleen cells were assessed for intracellular IFN-{gamma} production by three-color flow cytometric analysis (Fig. 9). LM induced cell death in BMM (Fig. 9B), as has been reported, and coculture of dying LM-infected BMM cells with NWNA spleen cells induced 1.6 ± 0.6% of {alpha}beta T cells, part of non-{alpha}beta T cells (2.0 ± 0.3% of total NWNA spleen cells; Fig. 9D), and 15 ± 2% of {gamma}{delta} T cells (Fig. 9I) to produce IFN-{gamma}. When anti-IL-12 Ab was added to the coculture, the proportion of IFN-{gamma}-producing {alpha}beta T cells, non-{alpha}beta T cells, and {gamma}{delta} T cells decreased to 0.4 ± 0.2%, 1.1 ± 0.2%, and 3.9 ± 0.6%, respectively. Addition of higher concentrations than 2 µg/ml anti-IL-12 no longer reduced the number of IFN-{gamma}-producing cells (data not shown). When the dot plots were gated on TCR{alpha}beta+NK1.1+ cells (NKT cells), we found that 50.9 ± 7.0% and 16 ± 5.2% were IFN-{gamma}+ in the absence or presence of anti-IL-12 Ab, respectively (Fig. 9, G and H). Of TCR{alpha}beta–NK1.1+ cells (NK cells), 46 ± 9% were IFN-{gamma}+, of which the proportion was decreased to 27 ± 5% in the presence of anti-IL-12 Ab (data not shown). Separation of NWNA spleen cells from dying LM-infected BMM cells in a Transwell failed to induce IFN-{gamma} production (Fig. 9F). These results indicate that in addition to a minor subpopulation of {alpha}beta T cells, subsets of {gamma}{delta} T cells, NK cells, and NKT cells produced IFN-{gamma} in response to direct cell-cell contact with dying LM-infected BMM.


Figure 9
View larger version (77K):
[in this window]
[in a new window]
  		FIGURE 9. IFN-{gamma} production by subsets of {alpha}beta T cells, {gamma}{delta} T cells, and NK cells in response to dying LM-infected BMM. C57BL/6 mouse BMM cells were infected with wild-type LM (MOI, 50:1) and cocultured for 24 h with C57BL/6 mouse NWNA spleen cells in an ordinary 24-well plate (C–E and G–K) or a Transwell (F) in the presence (2 mg/ml) of anti-IL-12 Ab (E and H) or control Ab (D). The expression of TCR{alpha}beta, NK1.1, and IFN-{gamma} was analyzed using FACScan. The histograms in D and E represent TCR{alpha}beta– IFN-{gamma}+ gated cells stained for NK1.1, and the histograms in G and H represent TCR{alpha}beta+NK1.1+ gated cells (NKT cells) stained for IFN-{gamma}. I–K, The same NWNA spleen cells in the presence of anti-IL-12 Ab (J) or control Ab (I) were assessed for TCR{gamma}{delta} expression and intracellular IFN-{gamma} production. Control staining was conducted by using PE-conjugated hamster IgG (K). Data are representative from one of three experiments. The numbers shown in the upper right quadrants represent mean percentages of IFN-{gamma}+ cells in {alpha}beta T cells or {gamma}{delta} T cells.

 
Mutant LM strain DP-L4048 infection has been reported to induce cell death in macrophages in a manner that is inversely correlated to the presence of gentamicin in culture medium (39); accordingly, we examined whether the IFN-{gamma} production by NWNA spleen cells was also inversely correlated to the presence of gentamicin. As shown in Fig. 10, A and B, when gentamicin was added 45 min after infection, DP-L4048-infected BMM did not undergo cell death, probably because bacteria that escaped from host phagosomes were killed by the antibiotic that was allowed to flow into the cytoplasm through the plasma membrane permeabilized by the mutant LLO (39), whereas in the absence of gentamicin DP-L4048-infected BMM underwent cell death within 10 h. When NWNA spleen cells were cocultured with these BMM cells, dying DP-4048-infected BMM cells, but not intact DP-L4048-infected BMM cells, induced IFN-{gamma} production by NWNA spleen cells (Fig. 10, C and D). The same results were obtained with B6HO3 hybridoma. When B6HO3 cells were cocultured with the dying or intact DP-L4048-infected BMM cells, only the dying DP-L4048-infected BMM cells induced B6HO3 cells to produce IFN-{gamma} (Fig. 10E).


Figure 10
View larger version (78K):
[in this window]
[in a new window]
  		FIGURE 10. Correlation of cell death in BMM caused by mutant LM strain DP-L4048 with the production of IFN-{gamma} by NWNA spleen cells and by B6HO3 hybridoma cells. A and B, Phase-contrasted microscopic observations of DP-L4048-infected C57BL/6 mouse BMM cells at 20 h postinfection. Gentamicin (100 µg/ml) was added 45 min (A) or 10 h (B) after bacterial internalization. C and D, DP-L4048-infected C57BL/6 BMM cells were cocultured with C57BL/6 mouse NWNA spleen cells for 24 h and NWNA spleen cells were assessed for TCR{alpha}beta expression and intracellular IFN-{gamma} production by flow cytometric analysis. Gentamicin was added at 45 min (C) or at 10 h (D) postinfection. E, DP-L4048-infected C57BL/6 mouse BMM cells were cocultured with B6HO3 hybridoma cells for 24 h and culture supernatants were assessed for IFN-{gamma} levels by ELISA. Gentamicin was added 45 min (*1) or 10 h (*2) after bacterial internalization.

 
These results indicate that cell death in BMM caused by LM infection is closely associated with its capability of stimulating subsets of lymphocytes and the B6HO3 hybridoma to produce IFN-{gamma}, and thus this phenomenon is not limited to the J774 macrophage cell line.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosures
 References
 
We have established in the present study a novel double-negative T cell hybridoma line designated B6HO3, which expresses activated/memory markers such as CD25, CD69, CD44high, and CD45RBhigh and produces IFN-{gamma}, TNF-{alpha}, GM-CSF, and Eta-1, but not IL-2 and IL-4 upon CD3-TCR ligation. We have also shown that the B6HO3 hybridoma recognizes a specialized form of macrophage cell death caused by bacterial infection in a TCR-independent manner and consequently produces IFN-{gamma} in an IL-12-independent manner.

A variety of bacterial pathogens induce cell death in macrophages (16), but the mechanisms by which cell death is induced seem to be manifold and there exists considerable controversy; i.e., combinations of different types of macrophage cells and bacteria have been reported to induce different modes of cell death in bacterial-infected macrophages, with some exhibiting apoptotic or autophagic cell death and others exhibiting necrosis or oncosis (15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26). The mode of cell death in LM-infected J774 cells has been shown to be nonapoptotic (Ref. 17 and my unpublished data), but the exact mechanism is yet undefined. Mutant LM DP-4048, which produces mutant LLO protein that is not subjected to intracellular degradation, has been reported to induce cell death in J774 cells by rapid pore formation in the plasma membrane (39). Shigella was first shown to induce apoptosis in J774 cells in a caspase-1-dependent manner (16, 18), but recent reports have shown that Shigella induces necrotic cell death (22, 23); i.e., the Shigella YSH 6000 strain induces necrosis in J774 cells by forming pores in the plasma membrane (22). Salmonella-induced macrophage death involves several mechanisms depending on the particular stage of infection (15, 19, 20, 21, 25). Salmonella grown under conditions for TTSS expression has been reported to induce rapid cell death that is caspase-1 dependent and characterized by necrosis-like features that include an increase in the cell membrane permeability (21). Thus, since the modes of cell death in the bacterial-infected J774 cells examined in the present study are all reportedly nonapoptotic, the IFN-{gamma} response of B6HO3 appears to be particularly associated with necrosis-like cell death in macrophages resulting from bacterial infection. The results presented here have shown that the IFN-{gamma} response of B6HO3 is macrophage cell death specific (Fig. 6) and does not require de novo protein synthesis in bacterial-infected macrophages (Fig. 4D). Therefore, some intracellular entity that exists intrinsically inside macrophages could be exposed or released outside the surface membrane due to the specialized form of macrophage cell death caused by bacterial infection, and this exposed novel entity could be recognized by B6HO3 cells. The precise mode of the pathogen-induced macrophage cell death that B6HO3 cells can recognize and the recognition mechanism thereof remain to be further explored.

Since our hybridoma system dictates most of the terminally differentiation programs of parental T cells and thus preserves the effector functions of T cells (30, 31), the unique functional phenotype of the B6HO3 hybridoma predicted existence of a hitherto unknown minor T cell population that should respond to pathogen-induced macrophage cell death by producing IFN-{gamma}. In fact, the present study has revealed that in response to direct cell-cell contact with dying LM-infected J774 cells ~1.6 and ~0.7% of splenic {alpha}beta T cells produce IFN-{gamma} in an IL-12-dependent and -independent manner, respectively. These two IFN-{gamma}-producing {alpha}beta T populations consisted of CD4+ (~50%), CD4–CD8– (