To evaluate the effect of macrophage SOCS3 on the outcome of experimental TB, we infected LysM^creSOCS3^loxP/loxP mice (deficient for SOCS3 in macrophages) and cre-negative littermates. Bacterial loads were determined at different time-points following aerosol Mtb infection. Compared to SOCS3^loxP/loxP mice, bacterial loads in lungs of LysM^creSOCS3^loxP/loxP mice were significantly enhanced 21 and 25 days after infection (Figure 1A) confirming the high susceptibility of these mice shown by Carow et al. (15).

We next assessed the pathological consequences of macrophage-specific SOCS3 deficiency during experimental TB. After 23 days of infection, the total number of cells infiltrating the lungs of Mtb-infected LysM^creSOCS3^loxP/loxP mice was increased threefold when compared to SOCS3^loxP/loxP mice (data not shown). The main cell types progressively infiltrating the lungs of macrophage-specific SOCS3-deficient mice were predominantly CD11b⁺ macrophages and granulocytes (Figure 1B). Twenty-three days after Mtb infection, their numbers were approximately 3- and 10-fold increased as compared to SOCS3^loxP/loxP mice, respectively (Figure 1C). By contrast, the absolute number of CD4⁺ T cells was comparable in the lungs of both SOCS3^loxP/loxP and LysM^creSOCS3^loxP/loxP mice (Figure 1C). Histopathological analysis of lungs taken from LysM^creSOCS3^loxP/loxP mice confirmed an enhanced inflammatory cell influx (Figure 1D). Organized granuloma formation was absent in cre-positive SOCS3^loxP/loxP mice until day 28 (Figure 1D; H&E staining) and this enhanced inflammatory environment alongside the mycobacterial overgrowth (Figure 1D; ZN staining) led to accelerated death of these mice (Figure 1E).

To examine why mycobacteria growth was not controlled in LysM^creSOCS3^loxP/loxP mice, the induction of pro- and anti-inflammatory cytokines in lung homogenates after Mtb infection was characterized by determining the concentrations of IL-6, tumor necrosis factor (TNF), IL-10, IFN-γ, and IL-12/IL-23p40 in a cytometric bead array (CBA) (Figure 2A). Compared to SOCS3^loxP/loxP mice, the production of IL-6, TNF, and IL-10 was significantly elevated in the lungs of LysM^creSOCS3^loxP/loxP mice after Mtb infection whereas the amount of IFN-γ was comparable in both cre-negative and cre-positive SOCS3^loxP/loxP mice (Figure 2A). In striking contrast, during the course of experimental TB the expression of IL-12/IL-23p40 was significantly reduced in macrophage-specific SOCS3-deficient mice when compared to infected SOCS3^loxP/loxP mice (Figure 2A). Because IL-12 is required to promote a protective Th1 immune response (17), we next investigated the frequency of antigen-specific IFN-γ-producing cells using whole cell suspensions or purified CD4⁺ T cells from lungs of infected animals in an ELISPOT assay (Figure 2B). On day 21 of Mtb infection, the frequency of IFN-γ-producing CD4⁺ T cells was significantly reduced in the lungs of Mtb-infected LysM^creSOCS3^loxP/loxP mice (Figure 2B). Taken together, these data show a dysregulation of pro- and anti-inflammatory cytokine production in LysM^creSOCS3^loxP/loxP mice during Mtb infection.

Our results are in line with previous studies addressing the impact of a macrophage-specific SOCS3 deficiency on the outcome of experimental toxoplasmosis (16) and TB (15) which implicated the reduced IL-12/IL-23 production and subsequent impaired Th1 immune response to be involved in the impaired control of Mtb infection in LysM^creSOCS3^loxP/loxP mice.

Based on in vitro experiments, Carow et al. excluded impaired macrophages effector functions as a cause for the increased susceptibility of Mtb-infected LysM^creSOCS3^loxP/loxP mice (15). The IFN-γ-dependent induction of the effector molecule NOS2 and subsequent production of reactive nitrogen intermediates (RNI) in classically activated macrophages, which are toxic to intracellular Mtb, is required for control of mycobacterial growth (10). Hence, the reduced frequency of Th1 cells in LysM^creSOCS3^loxP/loxP mice might also result in a diminished induction of this enzyme. To evaluate whether a modulated expression of Nos2 accounts for the increased susceptibility of Mtb-infected LysM^creSOCS3^loxP/loxP mice, we determined the expression of Nos2 and the production of RNI in lung homogenates of infected mice (Figure 3A). Quantitative real time (qRT)-PCR revealed that 21 and 28 days after Mtb infection Nos2 was considerably induced in lungs of SOCS3^loxP/loxP mice (Figure 3A). However, Nos2 expression was even higher in lung homogenates of infected LysM^creSOCS3^loxP/loxP mice. To determine the NOS2-dependent production of RNI during experimental TB we quantified the production of nitrate in lung homogenates from SOCS3^loxP/loxP and LysM^creSOCS3^loxP/loxP mice (Figure 3A). In uninfected mice, low amounts of nitrate were detected in lung homogenates. After infection with Mtb, the production of nitrates increased in SOCS3^loxP/loxP mice. In lung homogenates of infected LysM^creSOCS3^loxP/loxP mice, nitrate levels were significantly increased by threefold compared to littermate controls. Therefore, it appears that if the increased susceptibility of LysM^creSOCS3^loxP/loxP mice to experimental TB cannot be attributed to an impaired expression and function of NOS2, other mechanisms may overwrite this normally protective response. Induction of Nos2 can be counter-regulated by a concomitant Arg1 enzyme activity (12, 13). Arg1, a marker of alternative activation, hydrolyzes l-arginine to urea and l-ornithine and has been discussed to regulate RNI production in macrophages through depletion of l-arginine, the substrate for NOS2 (18, 19). Additionally, Arg1-dependent production of polyamines may promote growth of intracellular pathogens (20, 21). Therefore, we speculated that in LysM^creSOCS3^loxP/loxP mice, an increased arginase activity in the lungs modulates protective effector mechanisms against Mtb usually exerted by classically activated macrophages. A kinetic quantification of gene expression in lung homogenates revealed that Arg1 is only moderately induced in lungs of SOCS3^loxP/loxP mice during the course of Mtb infection (Figure 3B). By contrast, Arg1 expression in LysM^creSOCS3^loxP/loxP mice was highly induced early during Mtb infection. Accordingly, arginase activity in lungs of LysM^creSOCS3^loxP/loxP mice was also strikingly elevated when compared to enzyme activity in lungs of infected SOCS3^loxP/loxP mice (Figure 3B). Importantly, high Arg1 gene expression precedes the increased gene expression of Nos2 (Figures 3A,B).

The expression of Nos2 in infected tissue characterizes classical macrophage activation, which is required for protective macrophage effector responses. By contrast, alternatively activated macrophages are defined by tissue expression of Arg1 and other markers, such as Ym1, Fizz1, and the IL-4Rα-subunit especially during pulmonary inflammation (22). To evaluate whether SOCS3 deficiency generally intensify alternative macrophage activation during Mtb infection, gene and surface expression of different markers were quantified by qRT-PCR in lung homogenates and flow cytometry of single-cell suspensions, respectively (Figures 3C,D). Interestingly, LysM^creSOCS3^loxP/loxP mice constitutively expressed elevated levels of Ym1 and Fizz1 in lung homogenates (Figure 3C). Whereas in SOCS3^loxP/loxP mice, Ym1 and Fizz1 were only marginally stimulated after Mtb infection, gene expression of these markers for alternative macrophage activation was highly induced in lung homogenates of Mtb-infected LysM^creSOCS3^loxP/loxP mice. During the course of infection, the expression of these markers was significantly increased in LysM^creSOCS3^loxP/loxP mice compared to SOCS3^loxP/loxP animals. To further evaluate alternative macrophage activation in LysM^creSOCS3^loxP/loxP mice during experimental TB, we also quantified expression of Il4ra by qRT-PCR in lung homogenates and on the single-cell level by flow cytometric analysis of IL-4Rα expression on CD11b⁺ F4/80⁺ lung cells (Figure 3D). Before infection, gene expression was similarly low in both SOCS3^loxP/loxP and LysM^creSOCS3^loxP/loxP mice. After infection with Mtb, only LysM^creSOCS3^loxP/loxP mice showed Il4ra gene expression levels that were significantly different from a basal expression in SOCS3^loxP/loxP mice. On day 23 of experimental TB, surface expression of the IL-4Rα was significantly increased on macrophages in single lung cell suspensions of LysM^creSOCS3^loxP/loxP mice when compared to the surface expression on macrophages from infected SOCS3^loxP/loxP mice. Our results so far indicate that during Mtb infection SOCS3 is involved in regulating early alternative macrophage activation.

Because we found that Nos2 and Arg1 are highly expressed in the absence of macrophage SOCS3 but Arg1 precedes the expression of Nos2; we considered that, in LysM^creSOCS3^loxP/loxP mice, most bacteria might initially and preferentially multiply in Arg1-expressing cells. We, therefore, conducted double staining of mycobacteria and NOS2 or Arg1 in lung sections of LysM^creSOCS3^loxP/loxP mice 1 week after infection with Mtb. In SOCS3^loxP/loxP mice, NOS2- and Arg1-expressing cells were not detectable at this early time point (data not shown) reflecting the low gene expression of both enzymes (see Figures 3A,B). In lungs of LysM^creSOCS3^loxP/loxP mice, 1 week after infection Arg1 precedes Nos2 gene expression (see Figures 3A,B) and accordingly Arg1-positive cells were abundant (Figure 4A), whereas NOS2-positive cells were hardly detectable (Figure 4B). Importantly, at this early time point, mycobacteria were mostly found in Arg1-expressing cells. Together, these results indicate that in Mtb-infected LysM^creSOCS3^loxP/loxP mice an early exacerbated arginase activity favors the initial replication of Mtb particularly in Arg1-expressing cells.

Having shown that in the absence of macrophage SOCS3 Mtb evades in early appearing Arg1-expressing cells, we next analyzed the fate of mycobacteria in different types of macrophages during the course of infection in vivo. A defective recruitment of CD11b^medF4/80^med infiltrating macrophages in the lungs of infected mice is associated with susceptibility to Mtb (23). The examination of resident CD11b^hiF4/80^hi and infiltrating CD11b^medF4/80^med macrophages in lungs of Mtb-infected SOCS3^loxP/loxP and LysM^creSOCS3^loxP/loxP mice (as defined in the representative plot in Figure 5A) revealed a comparable frequency in both strains with a relative proportion of approximately 45–60% infiltrating macrophages and approximately 10% resident macrophages (Figure 5A). Infection of experimental mice with recombinant Mtb expressing mCherry enabled us to track bacteria in these cell types by flow cytometry (representative plot in Figure 5B). However, because Mtb^mCherry was less virulent than wild-type Mtb, lungs were isolated on day 36 after infection. Whereas in lungs of SOCS3^loxP/loxP mice, Mtb was predominantly found in infiltrating macrophages, in LysM^creSOCS3^loxP/loxP mice mycobacteria were detectable not only in infiltrating CD11b^medF4/80^med but also, to the same extent, in resident CD11b^hiF4/80^hi cells (Figure 5B).

We next determined the relative expression of NOS2 and Arg1 by intracellular staining of both enzymes in resident CD11b^hiF4/80^hi and infiltrating CD11b^medF4/80^med macrophages in lungs of Mtb-infected SOCS3^loxP/loxP and LysM^creSOCS3^loxP/loxP mice (Figure 5C). Most of the infiltrating macrophages in lungs of both SOCS3^loxP/loxP and LysM^creSOCS3^loxP/loxP mice were found negative for both, NOS2 and Arg1 or expressed only NOS2. Very few CD11b^medF4/80^med cells expressed Arg1 alone or both, NOS2 and Arg1 (Figure 5C). Approximately 50% of resident CD11b^hiF4/80^hi macrophages in lungs of SOCS3^loxP/loxP animals were found single positive for NOS2, about 20% were double positive for NOS2, and Arg-1 and 25% double negative. A small proportion expressed Arg1 only. By contrast, in lungs of LysM^creSOCS3^loxP/loxP mice the frequencies of Arg1-positive and Arg1/NOS2-positive but also that of Arg1/NOS2-negative CD11b^hiF4/80^hi cells increased at the cost of a decreased relative amount of NOS2-positive cells (Figure 5C).

Together, the absence of SOCS3 is associated with an uncontrolled expression of Arg1 and concomitant enhanced replication of Mtb in lung macrophages during experimental TB.

Suppressor of cytokine signaling 3 is involved in the regulation of gp130-mediated IL-6 signaling and suppresses the induction and maintenance of STAT1- and STAT3-dependent downstream effects (4, 5). We, therefore, determined whether IL-6-dependent signaling was causally involved in the increased susceptibility of LysM^creSOCS3^loxP/loxP mice during Mtb infection in vivo. To this end, we neutralized IL-6 by using a monoclonal anti-IL-6 antibody in vivo. During experimental TB, IL-6 neutralization had no effect on bacterial loads in lungs of C57LB/6 mice (Figure 6A). In contrast, neutralization of IL-6-dependent signals significantly reduced CFU in lungs of LysM^creSOCS3^loxP/loxP mice infected for 21 and 28 days with Mtb. However, bacterial loads in lungs from anti-IL-6-treated cre-positive SOCS3^loxP/loxP mice were still increased compared to infected SOCS3^loxP/loxP mice. IL-6 neutralization also greatly ameliorated pathology in LysM^creSOCS3^loxP/loxP mice infected with Mtb for 28 days (Figure 6B). The reduced expression of Il12b and Ifng in lung homogenates of Mtb-infected LysM^creSOCS3^loxP/loxP mice could only partially be restored by treatment with anti-IL-6 (Figure 6C). Because we show here that macrophage-specific SOCS3 deficiency promotes IL-6-induced classical and alternative macrophage activation during experimental TB in vivo, we next assessed whether IL-6 neutralization affected macrophage activation in the lungs of Mtb-infected mice. Treatment of C57BL/6 mice with anti-IL-6 had neither an effect on classical macrophage activation (measured by Nos2 gene expression and nitrate production in lung homogenates; Figure 6D) nor on alternative macrophage activation (determined by Ym1, Fizz1, Arg1 gene expression, and arginase activity in lung homogenates; Figures 6E,F). IL-6 neutralization reduced classical macrophage activation only to some extent with significantly decreased levels of RNI in lung homogenates of LysM^creSOCS3^loxP/loxP mice 21 after Mtb infection (Figure 6D). In infected SOCS3-deficient mice anti-IL-6 mAb treatment had no effect on gene expression of Ym1 and Fizz1 (Figure 6E) but IL-6 neutralization significantly reduced Arg1 gene expression and arginase activity in lung homogenates of LysM^creSOCS3^loxP/loxP mice that were infected with Mtb for 21 and 28 days (Figure 6F).

Together, these results demonstrate that SOCS3 prevents an IL-6-dependent dysregulated state of macrophage activation in vivo that promotes mycobacterial growth.

All animal experiments performed were in accordance with the German Animal Protection Law and were approved by the Animal Research Ethics Board of the Ministry of Environment, Kiel, Germany.

SOCS3^loxP/loxP mice were backcrossed 10 generations to C57BL/6 under specific-pathogen-free conditions at the Technical University of Munich (Germany) and the University of Erlangen (Germany). Mating with LysM^cre mice on the C57BL/6 genetic background produced LysM^creSOCS3^loxP/loxP and cre-negative SOCS3^loxP/loxP littermates. In addition, in some experiments, C57BL/6 mice (Charles River, Sulzfeld, Germany) were used as controls.

Mycobacterium tuberculosis H37rv or recombinant Mtb^mCherry (39) were grown in Middlebrook 7H9 broth (Difco, Detroit, MI, USA) supplemented with Middlebrook OADC enrichment medium (Life Technologies, Gaithersburg, MI, USA), 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 Mtb were diluted in sterile distilled water to a defined concentration and pulmonary infection was performed using an inhalation exposure system (Glas-Col, Terre-Haute, IN, USA).

Bacterial loads in lungs were evaluated at 3 and 4 weeks of infection. Lungs were removed aseptically, weighed, and transferred into PBS containing a proteinase inhibitor cocktail (Roche Diagnostics, Mannheim, Germany) and homogenized using the FastPrep™ System (MP Biomedicals, Solon, OH, USA). Tenfold serial dilutions of organ homogenates were plated onto Middlebrook 7H10 (Life Technologies, Darmstadt, Germany) agar plates containing 5% glycerine (Applichem, Darmstadt, Germany) and 10% heat-inactivated bovine serum (Biowest, Nuaillé, France) and incubated at 37°C for 21 days.

For histology, one lung lobe per mouse was fixed in 4% formalin-PBS, set in paraffin blocks, and sectioned (2–3 µm). Histopathological analyses were performed using standard protocols for hematoxylin/eosin staining. Acid-fast bacilli were detected using a modified Ziehl–Neelsen protocol. For the immunohistochemical detection of Arg1 and NOS2 (Upstate, Lake Placid, NY, USA) tissue sections were deparaffinized and pressure cooked in 10 mM citrate buffer, pH 6. After peroxidase quenching with 3% H₂O₂/TBS and blocking with 3% BSA, sections were incubated with primary antibodies against Arg1 (Santa Cruz Biotechnology, Santa Cruz, CA, USA) or NOS2 (Upstate) overnight followed by the incubation with HRP-conjugated goat-anti-rabbit secondary antibody (Dianova, Hamburg, Germany). Development was performed by using Elite ABC Kit (Vector, Burlingame, CA, USA) and diaminobenzidine (Vector).

Weighed lung samples before and at different time points of infection with Mtb, were homogenized in 4 M guanidinium-isothiocyanate buffer and total RNA was extracted by acid phenol extraction. cDNA was obtained using murine moloney leukemia virus (MMLV) reverse transcriptase (Superscript II, Invitrogen, Karlsruhe, Germany) and oligo-dT (12–18mer; Sigma) as a primer. Quantitative PCR was performed on a Light Cycler (Roche). Data were analyzed employing the “2nd Derivate Maximum method” and “Standard Curve method” using hypoxanthine-guanine phosphoribosyl transferase (Hprt) as a housekeeping gene to calculate the level of gene expression in relation to Hprt. The following primer and probe sets were employed: Arg1: sense 5′-CCT GAA GGA ACT GAA AGG AAA-3′, antisense 5′-TTG GCA GAT ATG CAG GGA GT-3′, probe 5′-TTC TTC TG-3′; Fizz1: sense 5′-TAT GAA CAGATG GGC CTC CT-3′, antisense 5′-GGC AGT TGC AAG TAT CTC CAC-3′, probe 5′-GGC AGG AG-3′; Hprt: sense 5′-TCC TCC TCA GAC CGC TTT T-3′, antisense 5′-CCT GGT TCA TCA TCG CTA ATC-3′, probe 5′-AGT CCA G-3′; ifng: sense 5′-atc tgg agg aac tgg caa aa-3′, antisense 5′-ttc aag act tca aag agt ctg agg ta-3′, probe 5′-CAGAGCCA-3′; Il4ra: sense 5′-GAG TGG AGT CCT AGC ATC ACG-3′, antisense 5′-CAG TGG AAG GCG CTG TAT C-3′, probe 5′-CTT CCA GC-3′; Il12b: sense 5′-atc gtt ttg ctg gtg tct cc-3′, antisense 5′-gga gtc cag tcc acc tct aca-3′, probe 5′-agc tgg ag-3′; Nos2: sense 5′-CTT TGC CAC GGA CGA GAC-3′, antisense 5′-TCA TTG TAC TCT GAG GGC TGA C-3′, probe 5′-AGG CAG AG-3′; Ym1: sense 5′-GAA CAC TGA GCT AAA AAC TCT CCT G-3′, antisense 5′-GAG ACC ATG GCA CTG AAC G-3′, probe 5′-GGA GGA TG-3′.

The concentrations of cytokines in lung homogenates from uninfected and infected mice were determined by a cytometric bead array (CBA) (BD Biosciences, Heidelberg, Germany) as described (40). IL-6 was analyzed using a CBA mouse-flex-set (BD Biosciences). To determine TNF, IL-12/IL-23p40, IL-10, and IFN-γ, beads were conjugated with purified antibodies (BD Bisosiences) using a functional-bead-conjugation buffer set following the manufacturer’s instructions (BD Biosciences). The quantity of cytokines per lung was calculated based on the ratio of lung to sample weight.

To determine arginase activity in murine tissue, weighed pieces of organs were homogenized in 100 µl of 0.1% Triton X-100 (Sigma) containing a protease inhibitor cocktail (Roche). 50 µl of 10 mM MnCl₂ (Merck, Darmstadt, Germany) and 50 mM Tris–HCl (Merck) were added to all samples and the enzyme was activated by heating for 10 min at 55°C. Arginine hydrolysis was conducted by incubating 25 µl of the activated lysate with 25 µl of 0.5 M l-arginine (Merck) at 37°C for 60 min. The reaction was stopped with 400 µl of H₂SO₄ (96%)/H₃PO₄ (85%)/H₂O (1/3/7, v/v/v). As a degree of arginase activity, the urea concentration was measured at 540 nm after addition of 25 µl α-isonitrosopropiophenone (Sigma; dissolved in 100% ethanol) followed by heating at 95°C for 45 min. One unit of arginase activity is defined as the amount of enzyme that catalyzes the formation of 1 µmol urea/min. To detect RNI in uninfected and infected mice, lung homogenates were collected at different time points. After deproteination of homogenates using Micron YM-30 centrifugal filters (Millipore, Schwalbach, Germany), NO₃ was converted into NO₂ utilizing a commercial nitrate reductase kit (Cayman; Axxora, Lörrach, Germany). After adding Griess reagents, the content of NO₂ was determined by photometric measurement reading the absorbance at 540 nm on a microplate reader (Sunrise; Tecan, Männedorf, Switzerland) as previously described (40).

For fluorescence microscopy, deparaffinized formalin-fixed lung sections were boiled for 45 min in 0.01 M citrate buffer at pH 6.0 for antigen retrieval. Endogenous peroxidase activity was blocked by a 10-min incubation in 1% hydrogen peroxide, unspecific antibody binding was blocked by a 30 min incubation in 3% BSA and endogenous biotin or avidin/streptavidin binding proteins were blocked using the Avidin/Biotin blocking kit (Vector Laboratories). Sections were then incubated overnight at 4°C with goat anti-Arg1 antibody (1:50 dilution, Santa Cruz Biotechnologies) or rabbit anti-NOS2 antibody (1:200 dilution, Merck Millipore). Sections were subsequently incubated with rabbit anti-goat IgG-biotin or goat anti-rabbit IgG-biotin (both Dianova), respectively, followed by an incubation with Streptavidin-Cy5 (Dianova). Finally, sections were co-stained with rabbit anti-Mtb-FITC (Biozol) and DAPI (Roche). Fluorescent stainings were analyzed using a TCS SP5 confocal microscope and LAS AF software (both from Leica, Wetzlar, Germany). Color images were produced by pasting each of the original grayscale images (shown in Figure S2 in Supplementary Material) into the red, the green, or the blue channels. Cross-reactivity of goat anti-Arg1 against Mtb could be excluded as goat anti-Arg1-dependent signals were absent in lung sections of Mtb-infected L-Nil-treated Arg1^flox/floxTie2cre mice, which are deficient for Arg1 in all macrophage populations.

For antigen-specific restimulation and flow cytometric analysis, single-cell suspensions of lungs were prepared from Mtb-infected mice at different time points. Lungs were perfused through the right ventricle 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) at 37°C for 2 h. Digested lung tissue was gently disrupted by subsequent passage through a 100 µm pore size nylon cell strainer. Suspensions were depleted of remaining erythrocytes using hypotonic red cell lysis buffer, containing NH₄Cl and NaHCO₃. Recovered vital lung cells were counted using an automatic cell counter (ViCell^®; Beckman Coulter, Krefeld, Germany), diluted in RPMI 1640 medium (Sigma) supplemented with 10% FCS (Life Technologies), 0.05 mM β-mercaptoethanol (Sigma), and penicillin and streptomycin (100 U/ml and 100 µg/ml; Life Technologies) and used for further experiments.

For flow cytometric analysis of surface markers, single-cell suspensions of lungs were incubated with an anti-FcγRIII/II monoclonal antibody (clone 2.4.G2) and stained with optimal concentrations of the following specific antibodies: anti-CD11b-eFluor450 (Ebioscience, Frankfurt, Germany), anti-CD11b-V500 (clone M1/70; BD Bioscience), anti-F4/80-Alexa-647 (clone BM8; Invitrogen), anti-F4/80-Pacific Blue (clone BM8; Biolegend), anti-Ly6G-PerCp-Cy5.5 (clone1A8; BD Bioscience), and anti-IL-4Rα-PE (clone mIL-4R-M1; BD Biosciences). For intracellular staining of NOS2 and Arg1, single-cell suspensions of lungs were stained for surface markers before permeabilization with Cytofix/Cytoperm^® (BD Bioscience). Staining of NOS2 and Arg1 was performed using optimal concentrations of the following specific antibodies: NOS2-FITC (clone 6/iNOS/NOS type II; BD Bioscience), goat-anti mouse Arg1 (clone V-20; Santa Cruz Biotechnology) and polyclonal donkey-anti-goat IgG-Dylight 650 (Abcam, Cambridge, UK). Appropriate isotype controls were used. Fluorescence intensity was measured using a FACSCanto^® II flowcytometer (BD Biosciences). Analysis was performed utilizing the FCS Express^® program (De Novo Software, Los Angeles, CA, USA).

Detection of antigen-specific IFN-γ-producing cells from infected lungs was conducted using an ELISPOT assay kit (BD Biosciences). To enrich CD4⁺ T cells, single-cell suspensions were incubated with magnetic CD4 microbeads (Miltenyi, Bergisch Gladbach, Germany) and separated from other cells on a MACS separation unit (Miltenyi). Separated CD4⁺ T cells were collected in Iscoves-modified Dulbeccos medium (IMDM; Life Technologies) supplemented with 10% FCS (Life Technologies), 0.05 mM β-mercaptoethanol (Sigma), and penicillin, and streptomycin (100 U/ml and 100 µg/ml; Life Technologies), counted using a cell counter (ViCell^®; Beckman Coulter), diluted in IMDM and used for further experiments. For measuring the antigen-specific IFN-γ response in lungs from infected mice, single-cell suspensions or purified CD4⁺ T cells from lungs were seeded in wells of anti-mouse IFN-γ-coated and blocked 96-well multitestplates at an initial concentration of 1 × 10⁵ cells/well in IMDM. After doubling dilutions of these cells were made, mitomycin-D (Sigma)-inactivated splenocytes from uninfected control mice were used as APCs at a concentration of 1 × 10⁶ cells/well. Cells were stimulated with the Mtb ESAT6_1–20 (Research Center Borstel, Germany) at a concentration of 10 µg/ml in the presence of 10 U/ml recombinant mouse IL-2 (Peprotech, Hamburg, Germany). After 20 h of incubation in 5% CO₂ at 37°C, plates were washed, and biotinylated anti-mouse IFN-γ was used to detect the captured cytokine. Spots were visualized using streptavidin-HRP as substrate. Spots were automatically enumerated using an ELISPOT reader (EliSpot 04 XL; AID, Straßberg, Germany). The frequency of responding cells was determined.

To neutralize endogenous IL-6 during experimental TB, 250 µg of a monoclonal rat IgG1, kappa anti-IL-6 antibody (clone MP5-20F3, InVivo BioTechServives, Henningsdorf, Germany) were injected i.p. at days −1, 2, 6, 9, 13, 16, 20, and 23 after infection with Mtb.

Statistical analysis was performed using GraphPad Prism version 4.03 (GraphPad Software, San Diego, CA, USA). Quantifiable data are expressed as means of individual determinations and SD. After analyzing for Gaussian distribution, unpaired Student’s t-test or the Mann–Whitney test was applied defining different error probabilities (*p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001). ANOVA was performed using the Bonferroni multiple comparison test different error probabilities (*p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001). Statistical survival analysis was performed using the Log-rank test.