The Lymphotoxin β Receptor Is Critically Involved in Controlling Infections with the Intracellular Pathogens Mycobacterium tuberculosis and Listeria monocytogenes¹

Mycobacterium tuberculosis currently is the number one bacterial killer in the world. Eight to 10 million people worldwide newly develop tuberculosis every year and for at least one-third of these patients the disease is lethal (1). These numbers are likely to rise with the emergence of multidrug-resistant strains of M. tuberculosis and the increase of coinfection with HIV. It is thus imperative to define the mechanisms that enable the host to effectively contain M. tuberculosis infection to enlist them in novel therapeutic strategies.

Animal models have been instrumental in elucidating the mechanisms of both protection and pathology in response to M. tuberculosis infection (2). In particular, the use of gene-targeted mice has provided incontrovertible evidence that an efficient cooperation between T cells and macrophages is critical for control of infection (3, 4, 5, 6). This is primarily because macrophages which encounter pathogens capable of surviving intracellularly need to be reprogrammed by specific T cells, essentially via TNF and IFN-γ, to become fully competent in antibacterial functions, a process in which NO and superoxide generation act both as signaling and as effector molecules (7). The concept of the activated macrophage was originally established in the murine model of Listeria monocytogenes infection which has since become a standard experimental system for identifying the roles of newly discovered cells and mediators in response to intracellular pathogens (8, 9).

In addition, antibacterially inefficient host cells may need to be lysed and new effector cells, such as granulocytes and macrophages, must be recruited and activated to engulf and destroy the released microorganisms (10, 11, 12). In the case of persistent mycobacteria, the close interaction of T cells and macrophages is facilitated in an organized microenvironment called granulomas, which at the same time serves as a physical barrier to dissemination of bacteria released from incompetent macrophages (13). Therefore, granulomas are highly dynamic structures which are composed of effector cells that are chronically recruited and activated by stimuli such as chemokines and IFN-γ- and TNF-mediated processes (14, 15).

Lymphotoxin (LT)⁴ α, LTβ, and the recently identified TNF family member LIGHT (homologous to lymphotoxins, exhibits inducible expression, and competes with HSV glycoprotein D for HVEM, a receptor expressed by T lymphocytes) along with the prototype cytokine TNF may be defined as a core group of cytokines clustered within the growing TNF superfamily (16, 17, 18, 19). The specific cell surface receptors for the ligands of the TNF core family are TNFRp55, TNFRp75, LTβR and herpesvirus entry mediator (HVEM) that reveal overlapping, but distinct ligand interactions (16, 20, 21). TNF₃ binds to the TNFRp55 and TNFRp75, whereas LTα₃ engages the TNFRp55, TNFRp75, and HVEM as homotrimer (16, 20, 22). In combination with the membrane-bound LTβ, LTα binds as the LTα₁β₂ heterotrimer to the LTβR (16, 21, 23, 24). TNF₃, LTα₃ in concert with the TNFRp55, were shown to be of critical importance for the host defense against intracellularly replicating bacterial pathogens and for the formation and maintenance of fully differentiated granulomas (25, 26, 27, 28, 29, 30, 31). In contrast, the detailed role of the LTβR and its cognate ligands in antimicrobial resistance and granuloma formation has remained unexplored.

To ascertain the functional relevance of the LTβR for the immune defense against intracellular microorganisms, we investigated the course of infection in two infection models (M. tuberculosis and L. monocytogenes) in mice deficient for the LTβR, LTα, LTβ, or LIGHT. Collectively, the results of this comprehensive study indicate that LTα₁β₂ signaling via the LTβR on macrophages is essential for controlling the infection with both pathogens.

LTβR-knockout (KO) (21), TNFRp55-KO (25), and LTα-KO (32) were N₅ backcross generations to C57BL/6; LTβ-KO (23) were N₁₀ backcrosses to C57BL/6. LIGHT-KO (22) were N₃ backcrosses to C57BL/6 and control mice in experiments involving LIGHT-KO mice were N₃ backcross heterozygous mice. Wild-type mice were C57BL/6 mice purchased from Charles River Breeding Laboratories (Sulzfeld, Germany). LTα-KO and C57BL/6 mice congenic for the CD45.1 Ag were purchased from The Jackson Laboratory (Bar Harbor, ME). All other mice were bred at the GSF National Research Center for Environment and Health (Martinsried, Germany) under specific pathogen-free conditions and were housed in isolator cages under barrier conditions at the BSL 3 facility at the Research Center Borstel (Borstel, Germany). All mice used were between 8 and 16 wk old. In any given experiment, mice were matched for age and sex. All experiments performed were in accordance with the German Animal Protection Law and were approved by the Animal Research Ethics Boards of the regional Ministries of Environment, Nature and Forestation (Kiel, Germany) and the government of Upper Bavaria (Munich, Germany).

Bone marrow (BM) cells were harvested by flushing femurs and tibias of donor mice with cold RPMI 1640 medium (Seromed, Biochrom KG, Berlin, Germany) supplemented with 10% heat-inactivated FCS (Seromed), 2 mM l-glutamine (Seromed) 50 μM 2-ME (Life Technologies, Eggenstein, Germany), 50 μg/ml streptomycin (Seromed), and 100 U/ml penicillin. Recipient wild-type C57BL/6 mice congenic for the CD45.1 (Ly5.1) Ag were irradiated with 10 Gy using a ¹³⁷Cs irradiator (Buchler, Braunschweig, Germany). After washing and counting, 5 × 10⁶ BM cells were injected i.v. into recipients. The degree of chimerism was determined in PBL (B220⁺ B cells and CD3⁺ T cells) 8 wk after reconstitution and in alveolar Mac3⁺ macrophages 21 days postinfection with M. tuberculosis. For bronchoalveolar lavage, M. tuberculosis-infected mice were anesthetized. Cells were collected by performing five intratracheal lavages with 0.8 ml of cold PBS containing 0.5 mM EDTA. Flow cytometric analysis of the differential CD45 Ag expression on donor and recipient BM-derived cells was performed. Eighty-nine percent of peripheral blood T cells and >95% peripheral blood B cells and alveolar Mac3⁺ macrophages were found to be donor derived.

M. tuberculosis (H37Rv) was grown in Middlebrook 7H9 broth (Difco, Detroit, MI) supplemented with Middlebrook OADC enrichment medium (BD Biosciences, Heidelberg, Germany), 0.002% glycerol, and 0.05% Tween 80. Midlog phase cultures were harvested, aliquoted, and frozen at −80°C. After thawing, viable cell counts were determined by plating serial dilutions of the cultures on Middlebrook 7H10 agar plates followed by incubation at 37°C. Before infection of experimental animals, stock solutions of M. tuberculosis were briefly sonicated and diluted in PBS. Pulmonary infection was performed by intranasal instillation. Mice were anesthetized by i.p. injection of 1.25% ketamine (Bayer, Leverkusen, Germany) and 0.025% Rompun (WDT, Garbsen, Germany) to fully suppress swallowing reflexes. Intranasal instillation was performed by applying 20 μl of a suspension containing 1–2.5 × 10⁴ CFU of M. tuberculosis/ml to the nares of anesthetized mice. Actual inoculum size was determined 24 h after infection by determining the bacterial load in the lungs of infected mice and is indicated in the figure legends. Bacterial loads in infected organs were evaluated at different time points after infection with M. tuberculosis to follow the course of infection. Lungs and livers of sacrificed animals (four to five mice per group) were removed and weighed aseptically and homogenized in distilled water. Ten-fold serial dilutions of organ homogenates were plated in duplicates onto Middlebrook 7H10 agar plates containing 10% OADC and incubated at 37°C for 3 wk. Colonies on plates were enumerated and results expressed as log₁₀ CFU per organ.

Infections with Listeria monocytogenes were performed as described previously (26). Briefly, overnight cultures of L. monocytogenes (ATCC strain 43251) were grown in brain-heart infusion (Difco), adjusted to an OD of 0.75, and serially diluted in medium. Titrated numbers of L. monocytogenes were i.p. inoculated in a volume of 350 μl into the mice. The dose of bacteria was checked by plating 10-μl aliquots of a serial 10-fold dilution on Columbia blood agar plates and counting the CFU after overnight incubation at 37°C. Livers and spleens were removed 2 days after infection, organ homogenates prepared, and bacterial counts determined as above.

One lung lobe per mouse was fixed in 4% Formalin-PBS, set in paraffin blocks, and sectioned (2–3 μm). Histology was performed using standard protocols for H&E staining and for acid-fast staining. For immunohistology, tissue sections were deparaffinized, placed in 10 mM sodium citrate buffer (pH 6.0), and pressure-cooked for exactly 1.5 min. After blocking for 20 min in 1% H₂O₂ solution, slides were incubated with appropriately diluted polyclonal rabbit anti-mouse inducible NO synthase (iNOS; Biomol, Hamburg, Germany) in TBS/10% FCS for 30 min in a humid chamber. Appropriately diluted goat anti-rabbit IgG (bridging Ab; Dianova, Hamburg, Germany) and diluted rabbit anti-goat IgG-peroxidase conjugate (tertiary Ab) (Dianova) were used in sequential incubations of 30 min each. For detection of T cells, anti-CD3 mAb (clone CD3.12; Biotrend, Cologne, Germany) was used as the primary Ab, diluted rabbit anti-rat IgG (Dianova) as the secondary, and goat anti-rabbit IgG peroxidase as the tertiary Ab. Development was performed using 3,3′-diaminobenzidine (Sigma-Aldrich, Deisenhofen, Germany) and urea superoxide (Sigma-Aldrich), and hemalum was used to counterstain the slides.

For flow cytometric determination of BM chimerism, Abs directed against CD3 (FITC, clone 500A2; BD Biosciences), B220 (FITC, clone RA3-6B2; BD Biosciences), Mac3 (FITC, clone M3/84; BD Biosciences), CD45.1 (biotin, clone A20; BD Biosciences), and CyChrome-conjugated streptavidin (BD Biosciences) were used.

For detection of LTβR expression after infection with M. tuberculosis, noninfected and infected mice were anesthetized and injected i.p. with 150 U heparin (Ratiopharm, Ulm, Germany). Lungs were perfused through the heart with warm PBS. Once lungs appeared white, they were removed and sectioned. Dissected lung tissue was then incubated in collagenase A (0.7 mg/ml; Roche Diagnostics, Mannheim, Germany) and DNase (30 μg/ml; Sigma-Aldrich) at 37°C for 2 h. Digested lung tissue was gently disrupted by sequential passage through a 23-gauge cannula, a metal sieve, and a 70-μm pore size nylon cell strainer. Recovered lung cells were washed and incubated with FcR-blocking Ab (clone 2.4G2; BD Biosciences). Cells were then incubated for 30 min with either anti-LTβR-mAb (clone 5G11b, rat IgG2a; T. Hehlgans et al., manuscript in preparation) or isotype control (rat IgG2a, clone A110-2; BD Biosciences). Cells were washed again and incubated with biotinylated isotype-specific anti-rat IgG2a mAb (clone RG7/1.30; BD Biosciences). After subsequent washing, cells were stained with streptavidin-CyChrome (BD Biosciences) and FITC-coupled anti-Mac3 (clone M3/84; BD Biosciences). CyChrome fluorescence intensity was analyzed after gating on macrophages identified by side scatter profile and high expression of Mac3.

For detection of LTβR expression on peritoneal macrophages, thioglycolate-induced peritoneal exudate cells were recovered 4 days after i.p. injection of 1 ml of thioglycolate by peritoneal lavage, washed, and incubated with FcR-blocking Ab (clone 2.4G2; BD Biosciences). Cells were then incubated for 30 min with either anti-LTβR-mAb (clone 5G11b, rat IgG2a) or buffer alone. Cells were washed again and stained with allophycocyanin-coupled anti-Mac1α (clone M1/70; BD Biosciences) and FITC-coupled isotype-specific anti-rat IgG1/2a mAb (clone G28-5; BD Biosciences). FITC fluorescence intensity was analyzed after gating on macrophages identified by side scatter profile and high expression of Mac1α (CD11b).

RT-PCR was performed essentially as described previously (33). In brief, weighed lung tissue samples were homogenized in 5 ml of 4 M guanidinium-isothiocynanate buffer and diluted to obtain equalized amounts of milligrams of lung per milliliter of buffer. Total RNA was purified using an RNA isolation kit (High Pure RNA Tissue kit; Roche Diagnostics), and 1 μg RNA was transcribed into cDNA using Moloney murine leukemia virus-RT (Life Technologies) and oligo(dT) (12–18 mer; Sigma-Aldrich) as a primer. PCR was performed on a Light Cycler (Roche Diagnostics) using the proprietary Light-Cycler-DNA Master SYBR Green I kit (Roche Diagnostics) and the following primer sets: β₂-microglobulin: sense, 5′-TGACCGGCTTGTATGCTATC-3′; antisense, 5′-CAGTGTGAGCCAGGATATAG-3′; IFN-γ: sense, 5′-AACGCTACACACTGCATCTTGG-3′; antisense, 5′-GACTTCAAAGAGTCTGAGG-3′; TNF: sense, 5′-GATCTCAAAGACAACCAACTAGTG-3′; antisense, 5′-CTCCAGCTGGAAGACTCCTCCCAG-3′.

After amplification (denaturation at 94°C for 1 s, annealing at 60°C for 5 s, and extension at 72°C for 5 s), melting curve analysis was performed to exclude the presence of confounding primer-dimers. Semiquantitative comparisons of amplified products were made based on the crossing points obtained for each sample compared with a serially diluted, arbitrarily selected standard cDNA run in parallel. In this way, arbitrary units could be assigned to mRNA levels present in each sample. Units were normalized by calculating the mRNA ratios of cytokine/β₂-microglobulin for each sample.

TNF levels in lung homogenates prepared with a proteinase inhibitor mixture (Roche Diagnostics) were analyzed in 3-fold serial dilutions by a sandwich ELISA (OptEia; BD Biosciences) using a modified protocol. Before adding biotinylated Abs, endogenous biotin was blocked by incubating samples with an avidin/biotin block reagent (Vector Laboratories, Peterborough, U.K.). After incubation with HRP coupled to avidin and developing with tetramethylbenzidine substrate reagent, the absorbance was read on a microplate reader (Sunrise; Tecan, Männedorf, Switzerland). Using a test wavelength of 450 nm and a reference wavelength of 630 nm, samples were compared with appropriate recombinant cytokine standards. The detection limit for TNF was 5 pg/ml.

Quantifiable data are expressed as the means of individual determinations and SD. Statistical analysis was performed using the Student’s t test.

LTα is biologically active both in its homotrimeric form and in heterotrimers along with LTβ (LTα₁β₂ or LTα₂β₁) (16, 21, 23, 24). To elucidate whether the homotrimeric and the heterotrimeric forms are involved in early resistance against pulmonary M. tuberculosis infection, bacterial multiplication was assessed in LTα- and LTβ-KO mice 5 wk following intranasal challenge with ∼500 CFU of strain H37Rv (Fig. 1). Both LTα- and LTβ-KO mice had significantly higher CFU counts in their lungs and livers than infected congenic C57BL/6 mice at this time point, suggesting that not only LTα homotrimers, but also LTαβ heterotrimers participate in the early immune response to M. tuberculosis infection.

LTα₁β₂ heterotrimers bind to and signal via the LTβR (16, 21). To directly test the hypothesis whether the LTβR pathway is involved in early defense against M. tuberculosis, the outcome of infection was examined in mice deficient for the LTβR.

First, LTβR-KO and congenic C57BL/6 mice were compared for their capacity to resist an intrapulmonary infection with M. tuberculosis H37Rv. In three independent experiments, all LTβR-KO mice intranasally infected with ∼500 CFU of H37Rv succumbed to infection by day 45, whereas all C57BL/6 mice survived this infecting dose and showed no signs of illness until termination of the experiment on day 60 (Fig. 2).

To investigate the course of infection with M. tuberculosis in LTβR-KO mice, mycobacterial replication in the lungs, livers, and spleens of mice infected intranasally with ∼500 CFU of M. tuberculosis was monitored over time. Although C57BL/6 mice were capable of containing the growth of M. tuberculosis organisms in the lungs after day 14 postinfection, bacterial multiplication progressed almost uncurtailed in LTβR-KO mice in this organ (Fig. 3,A). Simultaneously infected TNFRp55-KO mice were similarly unable to restrict the growth of M. tuberculosis and succumbed to infection around day 30, consistent with previously published data (Fig. 3,A). A reduction in the rate of M. tuberculosis growth was evident in the livers and spleens of immunocompetent mice at 3 wk after infection. In contrast, LTβR-KO mice were permissive for M. tuberculosis growth in these organs up to the fifth week of infection when bacterial replication plateaued at a comparatively high level of ∼5 × 10⁶ CFU (Fig. 3, B and C). However, most LTβR-KO mice sacrificed on day 35 for analysis of bacterial loads already showed severe signs of infection (decreased body weight and scruffy hair). Mice infected with lower inocula (∼250 and 100 CFU of M. tuberculosis) also had significantly increased bacterial burdens in their lungs 5 wk after infection (data not shown). Thus, LTβR-KO mice were invariably more susceptible than wild-type mice to intrapulmonary challenge with M. tuberculosis.

When tissue sections of mice infected intranasally with ∼500 CFU of M. tuberculosis were examined microscopically, C57BL/6 mice were found to develop the typical circumscript granulomatous lesions in their lungs (Fig. 4, A and C). Early during infection (day 21), LTβR-KO mice showed less inflammatory infiltration in both interstitial and alveolar spaces, compared with infected wild-type mice. By day 35 after infection, widespread pulmonary necrosis developed in LTβR-KO mice, eventually leading to destruction of almost the entire lung (Fig. 4, B and D). No signs of necrosis were observed in the livers or spleens of infected LTβR-KO mice (data not shown). Whereas in C57BL/6 mice epithelioid cell differentiation was well advanced on day 35 of infection and acid-fast bacilli were only sparsely detectable (Fig. 4, C and E), LTβR-KO mice showed extensive sheets of dead or degenerating cells surrounding necrotic tissue with fewer signs of epithelioid cell differentiation at this time point (Fig. 4,D). Instead, acid-fast bacilli were abundant within areas of tissue necrosis and numerous within individual macrophages (Fig. 4 F).

It appeared possible that the observed essential role for LTβR in resistance was due to the specific pathogen chosen (M. tuberculosis) or the specific route of infection (intranasal challenge). We therefore infected LTβR-KO mice and wild-type mice with a facultative intracellular pathogen, L. monocytogenes (strain ATCC 43251), and performed infection via the i.p. route (Fig. 5). Wild-type mice had an LD₅₀ of ∼3 × 10⁵ CFU of L. monocytogenes. At this inoculum dose, LTβR-KO mice all succumbed to infection. The LD₅₀ for LTβR-KO mice was determined to be ∼3 × 10³ CFU. Thus, LTβR-KO mice were 100-fold more susceptible to infection.

In additional experiments, Listeria CFU were determined in the livers and spleens of infected mice on days 2 and 4 after infection with 1/10 LD₅₀. At least 10- to 100-fold higher bacterial burdens were found in LTβR-KO mice compared with wild-type mice (data not shown), indicating that in the absence of the LTβR, the replication of L. monocytogenes cannot be sufficiently controlled. Taken together, these data indicate that the LTβR initiates antibacterial protection against prototypic intracellular bacteria independently of the route of infection.

LTα-, LTβ-, and LTβR-KO mice all exhibit anomalies in the development of lymph nodes, albeit to a varying extent (21, 23, 24, 34, 35). To exclude that a defunct lymphoid anatomy was the sole factor responsible for the loss of early resistance to M. tuberculosis infection in these mice, BM chimeras were constructed in lethally irradiated C57BL/6 mice infused with either BM obtained from LTβR-KO mice or from wild-type mice. Reconstitution resulted in ∼90% of donor-derived T cells and >95% donor-derived B cells in the peripheral blood and >95% donor-derived Mac3⁺ macrophages in bronchoalveolar lavage cells (data not shown). Recipients of LTβR-KO BM were highly impaired in resisting a M. tuberculosis intrapulmonary challenge, showing significantly increased bacterial CFU counts in their lungs and livers when directly compared with recipients of wild-type BM at 6 wk postinfection (Fig. 6).

Thus, decreased resistance of the LTβR-KO mice used in this study cannot be attributed solely to lymph node abnormalities, but is directly caused by the lack of signaling via the LTβR expressed on hemopoietic cells during pulmonary M. tuberculosis infection. In the experiments detailed below, we therefore addressed possible defects in LTβR-KO mice which might account for the observed impairment of host defense mechanisms.

To exclude the possibility that the lack of LTβR signaling was associated with decreased TNF or IFN-γ production known to be critical for controlling infection with intracellular bacteria, the expression of mRNAs for TNF and IFN-γ was determined in the lungs of LTβR-KO and wild-type mice 21 and 30 days after intranasal infection with M. tuberculosis. As shown in Fig. 7, mRNA levels for both cytokines were identical or higher in the lungs of infected LTβR-KO mice. We were unable to measure IFN-γ protein in lung homogenates taken at the same time points, but TNF protein levels could be determined and, again, were similar in wild-type and LTβR-KO mice (day 21 after infection: 312 pg/g lung in wild-type and 455 pg/g lung in LTβR-KO mice; day 30 after infection: 1733 pg/g lung in wild-type and 3442 pg/g lung in LTβR-KO mice). These results clearly demonstrate that the deficient antibacterial response in LTβR-KO mice is not caused by a flawed TNF or IFN-γ regulation.

It is generally assumed that the LTβR is expressed solely on stromal (nonhemopoietic) cell types (16), although there is some evidence that it may also be detectable on macrophage-like cells (36). Indeed, the insufficient control of infections with intracellular bacteria by chimeric mice after transfer of LTβR-KO BM into lethally irradiated wild-type mice suggests that the LTβR is expressed in the hemopoietic system. To re-evaluate the expression of LTβR on macrophages, macrophages were obtained by tryptic lung digestion from wild-type and, as a control, LTβR-KO mice, and were labeled with mAbs directed against the LTβR and Mac3. As depicted in Fig. 8, the LTβR was not detected on quiescent, resident macrophages, but became detectable following M. tuberculosis infection on activated wild-type macrophages (days 28 and 42 postinfection), whereas on LTβR-deficient macrophages (day 28 postinfection), no labeling was observed. Thioglycolate-elicited peritoneal macrophages from wild-type, but not LTβR-KO mice, also expressed the LTβR (data not shown). Thus, it appears possible that the LTβR directly induces antibacterial processes in macrophages.

Activation of macrophages for production of NO is a known corollary of effective antimycobacterial immunity in mice. Immunohistological staining with a polyclonal Ab specific for iNOS revealed that LTβR-KO were deficient in timely expression of iNOS protein within developing granulomatous lesions. Twenty-one and 30 days after infection, granulomatous lesions in C57BL/6 mice contained highly activated macrophages as evidenced by marked staining for iNOS protein (Fig. 9,A). In LTβR-KO mice, however, pulmonary lesions present on day 21 postinfection showed little or no material reactive with the Ab (Fig. 9 B), whereas just before death, when large necrotic lesions had developed, some positive staining for iNOS was detected, particularly within smaller lesions or at the circumference of necrotic lesions (data not shown). Thus, expression of this important antibacterial effector protein was significantly delayed in M. tuberculosis-infected LTβR-KO mice and could not be compensated for by TNF or IFN-γ alone.

Recently, it was suggested that defective recruitment of T cells into granulomatous lesions was responsible for the ineffective defense observed in M. tuberculosis-infected LTα-deficient mice (31). However, in the lungs of LTβR-KO mice infected with M. tuberculosis, T cells were readily found to be colocalized with macrophages inside granulomas as evidenced by staining with a monoclonal anti-CD3 Ab (small arrows in Fig. 9, C and D).

The LTβR recognizes not only LTαβ heterotrimers, but also the recently identified ligand LIGHT (20). To define the role of LIGHT in antimycobacterial protection, the bacterial load was determined in the lungs of LIGHT-KO mice (22) infected intranasally with M. tuberculosis. In contrast to mice deficient in LTα, LTβ, or LTβR, LIGHT-KO mice effectively controlled bacterial replication in their lungs (Fig. 10) and exhibited similar inflammatory responses as wild-type mice (data not shown). To examine the possibility that LIGHT may influence the course of infection at a later stage, mice were monitored until day 118 after infection. Again CFU counts in the lungs were similar in control heterozygous (6.4 ± 0.1 log₁₀ CFU) and homozygous LIGHT-KO mice (6.63 ± 0.2 log₁₀ CFU). In conclusion, LTαβ heterotrimers, but not LIGHT, are the critical ligands initiating the protective host immune response against M. tuberculosis.

The data presented here demonstrate that activation of macrophages for antibacterial effector function in the course of infections with intracellular bacteria critically requires intact LTβR signaling. Since LIGHT was found not to be essential, the cognate ligand for LTβR interaction is LTα₁β₂. This conclusion was reached using a number of gene-targeted mouse strains in two different experimental models of intracellular infection. Specifically, LTα-, LTβ-, LTβR-, but not LIGHT-KO mice had significantly increased bacterial loads leading to necrotic lesions following pulmonary infection with M. tuberculosis, and LTβR-KO mice proved dramatically more susceptible to L. monocytogenes infection. Furthermore, the observed defective antibacterial response in LTβR-KO mice could not be attributed to defunct anatomy of secondary lymphoid organs, impaired TNF production, or dysregulated recruitment of CD3⁺ cells into granulomatous lesions, but correlated with the absence or grossly delayed expression of iNOS in mycobacteria-infected macrophages.

It is well known that the core members of the TNF superfamily exhibit intrinsic functions during the organogenesis of the secondary lymphoid organs (21, 23, 24, 34, 35, 37). For example, LTβR-deficient animals lack lymph nodes and Peyer’s patches (21). LTα-KO mice principally share this phenotype; however, within the mesenterium the presence of a lymph node-like structure was described (34, 35). In most LTβ-deficient mice, mesenteric and cervical lymph nodes can be observed, but other lymph nodes and Peyer’s patches are absent (23, 24). Taken together, the LTβR controls the formation of secondary lymphoid organs with the LTα₁β₂ heterotrimer playing a major role. In fact, when the other known ligand of the LTβR, LIGHT (20), was inactivated in the germline of the mouse, no gross anomalies within lymphoid tissues were observed (22).

To investigate whether the increased host susceptibility evident in mice lacking LTα, LTβ, or the LTβR was merely caused by the absence of lymph nodes, a BM transfer approach was chosen. Wild-type hosts were lethally irradiated and LTβR-KO BM or, as a control, wild-type BM was infused. Leukocytes derived from LTβR-deficient and control BM stem cells readily colonized the secondary lymphoid organs and other tissues of the recipient (Ref. 38 and data not shown). In concordance with results obtained in LTβR-KO mice, chimeras reconstituted with LTβR-KO BM were unable to restrain mycobacterial growth, indicating that this defect is intrinsic to the hemopoietic compartment and cannot solely be attributed to anatomical defects in secondary lymphoid organs. However, it cannot be excluded that the absence of lymph nodes may also have contributed to exacerbation of M. tuberculosis infection in the KO mice, particularly in view of the fact that CFU counts in LTβR-KO mice consistently reached higher levels than in irradiated wild-type mice reconstituted with BM derived from LTβR-KO mice. In this regard, it is noteworthy that the induction of experimental autoimmune encephalomyelitis is severely reduced in LTα-KO mice lacking lymph nodes, whereas in chimeric mice reconstituted with LTα-deficient BM, the development of experimental autoimmune encephalomyelitis is similar to that in wild-type mice (39, 40).

Previous studies documented important functions for TNF and for LTα₃ in the defense against intracellular infections (25, 30, 31, 41, 42, 43). In particular, TNFRp55-KO mice succumbed to challenges with L. monocytogenes, Mycobacterium avium, or M. tuberculosis, showing unaltered or only slightly delayed recruitment of inflammatory cells into infectious lesions (26, 27, 29). To dissect the ligands involved, studies were also performed in TNF-KO mice and in LTα-KO mice (31, 44). TNF-KO and TNF/LTα-double-KO mice rapidly died of infection with L. monocytogenes, showing at least 10,000-fold higher bacterial titers in infected organs in comparison to wild-type mice (45). Unfortunately, the histopathology of infected tissues derived from these mice was not reported. Collectively, these studies support a model where TNF and possibly LTα₃, via TNFRp55, effectively prevent intercellular spreading of L. monocytogenes, although the ultimate effector mechanisms remain ill-defined.

Experimental infections with M. tuberculosis in these KO mice have yielded a slightly different picture. TNF-KO mice infected with M. tuberculosis showed structural deficiencies in granuloma formation which were interpreted to be the crucial malfunction underlying their heightened susceptibility to infection (30, 46). Likewise, chimeric mice created by infusing LTα-deficient BM into irradiated wild-type recipients appeared to have no major defect in macrophage activation when compared with wild-type mice, as evidenced by similar amounts of NO reaction products in the serum in the course of M. tuberculosis infection (31). However, these mice showed a delay in chemokine induction and appeared to be unable to recruit T cells into inflammatory lesions, suggesting that a defective organization of the granulomatous response was the critical factor leading to unrestrained mycobacterial growth (30, 46).

The results from experiments detailed in this report indicate that, with respect to initiating antibacterial responses to intracellular infections, the role of the LTβR and its ligand LTα₁β₂ may be distinct from that served by the TNFRp55 and its ligands TNF and LTα₃. A prominent and consistent feature of the inflammatory response in LTβR-KO mice was the grossly delayed expression of iNOS protein within granuloma macrophages. This could not be attributed to a decreased presence of CD3⁺ cells within granulomatous lesions nor to defective regulation of TNF or IFN-γ mRNA expression. Thus, inefficient macrophage activation, rather than abnormal granuloma formation, was the principal factor leading to excessive M. tuberculosis growth in the lungs of LTβR-KO mice. It is, however, extremely difficult to quantitatively assess the kinetics of granulomatous inflammation in the lung, and it remains possible that a slight delay in granuloma formation may also have contributed to decreased resistance in these mice.

To our knowledge, this is the first report implicating LTαβ heterotrimers in the generation of macrophage antibacterial activity. These findings are in contrast to a recent report in which chimeric mice reconstituted with LTβ-deficient BM efficiently controlled infection with M. tuberculosis (31). The reasons for this discrepancy are not clear, but we have consistently observed exacerbation of bacterial loads in LTβ-KO, LTβR-KO, and chimeric mice infused with LTβR-KO BM in several independent experiments. In corroboration of our data, a study in a related model of infection (Mycobacterium bovis bacillus Calmette-Guérin) showed that treatment with a LTβR-Fc fusion protein negatively affected antibacterial protection (47). Since LTβR-Fc fusion proteins antagonize both LTα₁β₂ and LIGHT, thus inhibiting LTβR and HVEM functions simultaneously, a molecular dissection of the relevant interaction partners could not be unambiguously defined in that study.

Taken together, our data support the hypothesis that macrophage activation is incomplete when LTβR signaling is absent. Since we found CD3⁺ T cells, bearing the cognate ligand for the LTβR to be localized inside granulomas even in LTβR-KO mice, the most straightforward way to explain this is to assume an intrinsic defect of the macrophage in the absence of the LTβR. Significantly, we and others found this receptor to be expressed on macrophages (Ref. 36 and this study). From our data, it is evident that some form of macrophage activation is necessary for LTβR expression to become detectable, and this would explain why enhanced mycobacterial growth in LTβR-KO mice is not evident before day 21 postinfection when T cell-dependent macrophage activation emerges. Only at this stage would activated macrophages then be poised to receive costimulatory signals from T cells (which we demonstrated to be present within granulomatous lesions at this point, Fig. 9 B) expressing membrane-bound LTα₁β₂. It would also appear that a deficiency in LTβR signaling cannot be fully compensated by TNF and IFN-γ nor direct T cell-macrophage contact-dependent events. However, a more indirect effect involving adefect in the maturation of APC or flawed priming of the immune response is still possible. In this respect, the LTβR may be required also on other hemopoietic cells to provide coactivating stimuli for full macrophage antibacterial activity.

Ligation of the LTβR leads to activation of NF-κB engaging the “classical” pathway via inhibitor of κB kinase (IKK) β/IKKγ, resulting in RelA containing NF-κB species and an “alternative” pathway where NF-κB complexes composed of NF-κB2/RelB are activated via NFκB-inducing kinase and IKKα (48, 49, 50, 51, 52). Ultimately, the signaling of LTβR appears to be mainly required for the generation of sufficient amounts of type I IFNs which in turn are potent activators of iNOS in vitro and in vivo (53, 54, 55). Impairment of iNOS expression in LTβR-deficient mice may thus be attributed to deficient signaling via type I IFN receptors (54, 55, 56), a hypothesis that will be addressed in future studies. Alternatively, additional, as yet unidentified, mediators of the host defense may be under the control of the LTβR.

Our studies add another receptor-ligand system to the already intricate network orchestrating the immune response to intracellular infections. Since LTs and the LTβR clearly are essential for antibacterial protection, it appears possible that specific deficiencies in these molecules may account for the heightened susceptibility to intracellular infections of some patients. A similar scenario was described for patients suffering from disseminated mycobacterial infections who had deficiencies in the IL-12/IFN-γ pathway of macrophage activation (57). The elucidation of the LTβR-initiated processes that contribute to arming macrophages with antibacterial effector functions may eventually enable the development of novel immunotherapeutic tools for adjuvant treatment of patients with intracellular infections, e.g., those with multidrug-resistant M. tuberculosis.

We thank Svenja Kröger, Alexandra Hölscher, Karin Mink, Evelyne Schaller, and Jenniffer Meinecke for excellent technical help and Jens Würthner and Sandra Beer for critical reading of this manuscript.