Cholesterol Oxidase Is Indispensable in the Pathogenesis of Mycobacterium tuberculosis

Despite considerable research effort, the molecular mechanisms of Mycobacterium tuberculosis (Mtb) virulence remain unclear. Cholesterol oxidase (ChoD), an extracellular enzyme capable of converting cholesterol to its 3-keto-4-ene derivative, cholestenone, has been proposed to play a role in the virulence of Mtb. Here, we verified the hypothesis that ChoD is capable of modifying the bactericidal and pro-inflammatory activity of human macrophages. We also sought to determine the contribution of complement receptor 3 (CR3)- and Toll-like receptor 2 (TLR2)-mediated signaling pathways in the development of macrophage responses to Mtb. We found that intracellular replication of an Mtb mutant lacking a functional choD gene (ΔchoD) was less efficient in macrophages than that of the wild-type strain. Blocking CR3 and TLR2 with monoclonal antibodies enhanced survival of ΔchoD inside macrophages. We also showed that, in contrast to wild-type Mtb, the ΔchoD strain induced nitric oxide production in macrophages, an action that depended on the TLR2, but not the CR3, signaling pathway. Both wild-type and mutant strains inhibited the production of reactive oxygen species (ROS), but the ΔchoD strain did so to a significantly lesser extent. Blocking TLR2-mediated signaling abolished the inhibitory effect of wild-type Mtb on ROS production by macrophages. Wild-type Mtb, but not the ΔchoD strain, decreased phorbol myristate acetate-induced phosphorylation of extracellular signal-regulated kinases 1 and 2 (ERK1/2), which are involved in both TLR2- and CR3-mediated signaling pathways. Our finding also revealed that the production of interleukin 10 by macrophages was significantly lower in ΔchoD-infected macrophages than in wild-type Mtb-infected macrophages. However, tumor necrosis factor-α production by macrophages was the same after infection with mutant or wild-type strains. In summary, we demonstrate here that ChoD is required for Mtb interference with the TLR2-mediated signaling pathway and subsequent intracellular growth and survival of the pathogen in human macrophages.


Introduction
The initial immune response against Mycobacterium tuberculosis (Mtb) starts with recognition and ingestion of mycobacteria by alveolar-resident macrophages. A number of receptors present on the cell surface of macrophages, including the mannose receptor, Toll-like receptors (TLRs) and complement receptors (CR), have been implicated in the recognition and/or uptake of mycobacteria. Phagocytosis of Mtb can involve uptake of tubercle bacilli after opsonization with serum complement as well as non-opsonic ingestion. Recognition of a specific cell wall structure of mycobacteria by TLR2 results in recruitment of adaptor protein MyD88 (myeloid differentiation primary response) to the Toll/ interleukin-1 receptor (TIR) domain of TLR2, followed by recruitment of IL-1 receptor-associated kinase (IRAK)-1 and -4, which in turn leads to the phosphorylation of target signaling proteins, including MAPKs (mitogen-activated protein kinases), PI3K (phosphoinositide 3-kinase), and NF-kB (nuclear factor-kB). Signals initiated by the interaction of Mtb with TLR2 result in the induction of inflammatory and antimicrobial responses of innate immune cells [123].
Complement receptor 3 (CR3), also known as integrin a M b 2 , CD11b/CD18 and Mac-1, is a heterodimeric surface receptor that recognizes mycobacterial cell wall structures (non-opsonic phagocytosis) and Mtb coated with C3b or iC3b (opsonic phagocytosis). After iC3b binding, the b-subunit of CR3 mediates activation of Src-family tyrosine kinases, which subsequently phosphorylate phospholipase C. Non-opsonic binding of Mtb is important during primary infection owing to the limited presence of complement components in the alveolar space. Mtb cell wall structures are recognized by the lectin-like domain of the CR3 asubunit. It has been reported that neither opsonic nor non-opsonic phagocytosis of Mtb via CR3 induces killing of Mtb [4], [5].
Ingestion of Mtb by macrophages induces a variety of intracellular antimicrobial mechanisms, including production of the bactericidal agents, reactive oxygen species (ROS) and reactive nitrogen intermediates (RNIs), as well as cytokines, such as tumor necrosis factor-a (TNF-a) and interleukin 10 (IL-10), that contribute to regulation of immune cell responses [6][7][8].
It is well known that Mtb is able to accumulate, and/or degrade cholesterol and use it as a source of carbon and energy and cholesterol utilization is an important determinant of Mtb survival in macrophages [9][10][11][12][13][14]. The initial step in cholesterol degradation is its oxidation and isomerization to cholestenone. This process is mediated by cholesterol oxidase (ChoD) and/or hydroxysteroid dehydrogenase (HsdD) [11], [15]. 3b-hydroxysteroid:oxygen oxidoreductase, commonly known as Chox, is a flavoenzyme found in a wide range of bacteria. In some bacterial species, ChoD is an extracellular enzyme that appears to be present as both secreted and cell-surface-associated forms [16][17][18][19][20]. We demonstrated previously that the Mtb H37Ra mutant strain lacking the gene encoding ChoD grows slower than wild-type H37Ra in peritoneal macrophages, lungs, and spleens of mice [21]. On the other hand, ChoD does not appear to be essential for cholesterol degradation by mycobacteria [15], [22], [23].
Here, we investigated the functional responses of human macrophages to the wild-type Mtb H37Rv strain and a mutant in which the native choD gene is deleted (DchoD). To this end, we assessed the interaction of the mutant with macrophages, which underlies in vitro recognition by phagocytes and is required for subsequent responses; examined the intracellular growth of bacteria, the production of nitric oxide (NO), ROS, and the cytokines TNF-a and IL-10 by macrophages; and determined the role of TLR2-and CR3-mediated signaling pathways in the response of macrophages to infection with wild-type and mutant strains.

Bacterial Growth Conditions
Wild-type, mutant, and complemented strains were grown in Middlebrook 7H9 broth supplemented with 10% OADC enrichment and 0.05% Tween-80 (in roller bottles) for 4-6 days to reach an optical density value of 1 at 600 nm (OD 600 ). A portion of the bacterial culture was suspended in Middlebrook 7H9 broth (approximately 1610 9 cells/ml) and labeled with 100 mg/ml of FITC for 2 hours at room temperature with gentle agitation in the dark. Bacteria were then washed once with Middlebrook 7H9 supplemented with 4% BSA and twice with Middlebrook 7H9 broth without BSA [24], [25]. Thereafter, unlabeled and FITClabeled bacteria were resuspended in Middlebrook 7H9 broth, divided into equal portions, and stored at 285uC. After 1 week, one portion of unlabeled and one of FITC-labeled bacteria were thawed, and colony-forming assays were used to determine the number of bacteria (CFUs).
Prior to infection of macrophages, bacteria were thawed, washed twice in RPMI-1640 medium, and then opsonized (or not) by incubating with 20% human AB serum in RPMI-1640 medium for 30 minutes at 37uC with gentle agitation. After opsonization, bacteria were washed once with RPMI-1640 medium. Opsonized and non-opsonized bacteria were resuspended in culture medium (CM; see below), and clumps were disrupted by multiple passages through a 25-gauge needle. Serial dilutions of bacteria were prepared in CM.

Cell Culture
In this study, we used the differentiated human monocytemacrophage cell line THP-1 (ACTC TIB-202; USA) as a model for macrophages. The limited number of monocyte-derived macrophages that can be obtained from blood samples is a substantial obstacle to their use in our protocols, which require very large numbers of cells. THP-1 cells were cultured in CM consisting of RPMI-1640 supplemented with 1 mM sodium pyruvate, 10% FBS, 0.05 mM 2-ME, 100 U/ml of penicillin, and 100 mg/ml of streptomycin at 37uC in a humidified 5% CO 2 atmosphere. Monocytes were differentiated into macrophages as described previously [14] by incubating with PMA (20 ng/ml) for 24 hours (37uC/5% CO 2 ). The ability of these macrophages to adhere to plastic dishes (an indicator of monocyte differentiation to macrophages) was examined under a light microscope. The macrophage-like phenotype of cells was also examined by assessing CD14 expression, as we described previously [14]. After incubation with PMA, CM was removed and macrophages were infected with bacteria. Macrophages infected with bacteria were always cultured in CM without antibiotics.

Expression of TLR2 and CR3 on Macrophages
Macrophages were detached from plates using a trypsin-EDTA solution (2-5 minutes, 37uC/5% CO 2 ); trypsin was subsequently neutralized by adding RPMI-1640 medium containing 10% FBS. Cells were then centrifuged (1306 g, 6 minutes) and resuspended in D-PBS supplemented with 1% FBS. The viability of cells was determined by Trypan Blue exclusion and shown to be approximately 95%. Before staining with anti-TLR2 monoclonal antibody (mAb), crystallizable fragment receptors (FcRs) were blocked in D-PBS containing 10% human AB serum for 15 minutes at room temperature to prevent nonspecific antibody binding. Thereafter, cells were washed twice in D-PBS containing 1% FBS and stained with 10 mg/ml of PE-conjugated anti-TLR2 mAb (or 10 mg/ml appropriate IgG2a isotype control) or 20 ml PE-conjugated anti-CR3 mAb (or 20 ml appropriate IgG1 isotype control) for 30 minutes at 4uC. Cells were then washed twice, resuspended in 200 ml of D-PBS containing 1% FBS, 1% paraformaldehyde (PFA) and sodium azide, and stored at 4uC until FACS (fluorescence-activated cell sorting) analysis.
The appropriate concentration of anti-TLR2 and anti-CR3 antibodies that completely blocked the expression of TLR2 and CR3 on cells was determined by adding different mAb concentrations (10, 25, and 35 mg/ml or 25, 35, 45 and 55 mg/ml, respectively) to macrophages and incubating for 1 hour (37uC/5% CO 2 ). Macrophages were then stained with PE-conjugated anti-TLR2 mAb, PE-conjugated anti-CR3 mAb or isotype controls, as described above.
All samples were examined with a FACS LSR II BD flow cytometer (Becton Dickinson, USA) equipped with BD FACSDiva Software. The results are presented as median fluorescence intensity (MFI), which correlates with the surface expression of the target molecule.

THP-1 cells (1610 5 cells/well) were distributed into 8-well
Permanox chamber slides (Nunc, Denmark) and differentiated into macrophages. Cells were then pre-treated with 10 mM IRAK1/4 inhibitor or 35 mg/ml of anti-TLR2 or 55 mg/ml anti-CR3 blocking mAb for 1 hour (37uC/5% CO 2 ) or left untreated (as indicated in figures). Thereafter, macrophages were infected with opsonized or non-opsonized, FITC-labeled, wild-type, mutant (DchoD), or complemented (DchoD-choD) strains at a multiplicity of infection (MOI) of 10 for 2 hours (37uC/5% CO 2 ). Non-ingested bacteria were removed by extensively washing macrophages with warm HBSS. Fluorescence quenching by extracellular bacteria was removed by adding an equal volume of 2 mg/ml Trypan Blue solution. Phagocytes were fixed by incubating with 3% PFA for 15 minutes (37uC/5% CO 2 ) and washed twice with HBSS. The number of infected macrophages and the number of bacteria engulfed per macrophage were determined by fluorescence microscopic examination (Nikon ECLIPSE TE 2000 U). In all cases, 200 macrophages were counted.

Intracellular Growth of Bacteria
THP-1 cells (1610 5 cells/well) were distributed into 24-well plates (Nunc), differentiated into macrophages, and then pretreated with 10 mM IRAK1/4 inhibitor, 35 mg/ml of anti-TLR2 mAb or 55 mg/ml of anti-CR3 blocking mAb, as described above. Macrophages were then infected with opsonized or non-opsonized wild-type, DchoD, or DchoD-choD Mtb strains at an MOI of 1. Noningested bacteria were removed by extensively washing with warm HBSS. Fresh CM and IRAK1/4 inhibitor or antibodies (as indicated) were added, and cells were cultured for 6 days. On day 0 and days 2, 4 and 6 post-infection, macrophages were lysed with 1 ml of 0.1% Triton X-100. Appropriate dilutions of cell lysates were plated onto Middlebrook 7H10 agar supplemented with 10% OADC. After 21 days of culture, CFUs were counted. The data are presented as fold increase in CFUs/ml, calculated as CFUs/ml on day 6 divided by CFUs/ml on day 0.

NO and ROS Production
THP-1 cells (1610 5 cells/well) were distributed into 96-well plates (Nunc) and differentiated into macrophages. After that, macrophages were pre-treated with 10 mM IRAK1/4 inhibitor or 55 mg/ml of anti-CR3 mAb, or were left untreated (see above), as indicated in figures. Cells were then infected for 2 hours with opsonized or non-opsonized wild-type, DchoD, or DchoD-choD strains at an MOI of 10. Extracellular bacteria were removed by extensively washing macrophages with warm HBSS. Macrophages with ingested bacteria were cultured for 24 hours (ROS production) or 48 hours (NO production).
In the case of ROS production, after 24 hours of culturing, the supernatants were harvested and 1 mg/ml of PMA, 40 U of HRP (to initiate ROS production), 1 mM luminol (to enhance chemiluminescence (CL), and HBSS were added to cells. CL was recorded over 4 hours at 5-minute intervals. Data were acquired as relative light units (RLU), and the area under the curve of CL versus assay time (total RLU) was calculated. Data are presented as the percentage inhibition of ROS production, calculated according to the formula: 1-(total RLU for cells infected with bacteria and stimulated with PMA/total RLU for cells stimulated with PMA) 6100.
The presence of nitrite (stable metabolite of NO) in the culture supernatants of macrophages infected with bacteria was detected using the Griess reagent. Nitrite concentration was calculated from a standard curve using sodium nitrite as a reference. OD was determined using a Multiscan RC ELISA reader (Labsystem, Finland).

Western Blot Analysis of ERK1/2
THP-1 cells (5610 6 cells/well) were distributed into 24-well plates (Nunc), differentiated into macrophages, and then infected with bacteria (MOI = 10) as described above. Thereafter, in one set of experiments, macrophages were incubated with 1 mg/ml of PMA for 2 hours or left untreated. In a second set of experiments, macrophages were cultured for 24 hours and then stimulated with PMA. After treatment with PMA, cells were detached, centrifuged (1 minute, 12,0006 g), and lysed in lysis buffer (1% Triton-X 100, 20 mM Tris, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM PMSF) containing 16 Halt protease and phosphatase inhibitor cocktail by incubating for 30 minutes on ice. The protein concentration in each lysate was determined using a DC Protein Assay Kit.
Cell lysates containing equal amounts of protein were run on 10% Mini-Protean TGX Precast Gels along with a molecular weight standard. The proteins were transferred to PVDF membranes using Trans Blot Turbo (Bio-Rad, USA) at 2.5 A (,25 V) for 10 minutes. The membranes were blocked with protein-free (TBS) blocking buffer for 20 minutes and then incubated with primary anti-MAPK Abs (1:1000), anti-phospho-ERK1/2 Abs (1:1000), or anti-b-actin Abs (1:4000) at room temperature for 1 hour. After washing five times in 26 TBS/ Tween-20, membranes were incubated with HRP-conjugated goat anti-rabbit IgG (H+L) (1:4000) or HRP-conjugated goat antimouse IgG (H+L) (1:4000) at room temperature for 1 hour, and then washed. Immunoreactive proteins were visualized using an enhanced chemiluminescence system (Thermo Scientific, USA). Densitometric analyses of blots and analyses of visualized bands were performed using a FluoroChem MultiImage FC Cabinet (Alpha Innotech Corporation, San Leandro, CA, USA) and Alpha Ease FC software 3.1.2. The results are presented as the optical density intensity (ODI) of the area under each band's peak.

TNF-a and IL-10 Production
THP-1 cells (1610 6 cells/well) were distributed into 24-well plates (Nunc), differentiated into macrophages, and then infected with bacteria (MOI = 10), as described above, and cultured for 24 hours. The presence of IL-10 and TNF-a in the culture supernatants was assessed using Quantikine ELISA kits. The sensitivities of IL-10 and TNF-a assays were 3.9 and 1.6 pg/ml, respectively.

Statistical Analysis
Data are presented as means 6 SEMs. Statistical significance was verified using nonparametric Wilcoxon's signed-rank or Mann-Whitney U tests. The Statistica 8.0 (StatSoft, Poland) software package was used for statistical calculations. Statistical significance was defined as p #0.05.

Results
The Surface Expression of TLR2 and CR3 on Macrophages The expression levels of TLR2 and CR3, determined by flow cytometry as MFI values, were 11567 and 340643, respectively. We also determined the concentration of blocking anti-TLR2 and anti-CD3 mAbs sufficient to neutralize surface expression of each receptor. As shown in Figure 1, we found that after pre-incubation of macrophages with 35 or 55 mg/ml of blocking mAbs, surface expression of TLR2 (MFI = 3265) and CR3 (MFI = 3864) was virtually undetectable.

Phagocytosis of Mycobacteria
Three strains of Mtb H37Rv were used in each experiment: wild-type Mtb; a site-directed mutant in which the native choD gene was replaced with a truncated, non-functional choD (DchoD); and a complemented strain carrying an intact choD gene under control of a P hsp promoter introduced into the attB site (DchoD-choD), engineered as described previously (25). We found that the percentage of macrophages ingesting wild-type and DchoD Mtb strains ranged from 25% to 40%, although there was no significant difference in the percentage of macrophages that ingested either strain ( Fig. 2A). We also noted that inhibition of TLR2-or CR3mediated signaling pathways decreased the efficiency of phagocytosis that manifested as a decrease in the percentage of macrophages that took up opsonized or non-opsonized both wild type and DchoD Mtb strains ( Fig. 2B and C). Notably, the anti-TLR2 mAb decreased phagocytosis to a greater degree than the anti-CR3 mAb.
Intracellular Growth of Wild-type, DchoD, and DchoD-choD Strains A survey of bacteria surviving inside macrophages, determined by colony-forming assays and expressed as fold increase in CFUs, was used to test bacterial intracellular growth. In a preliminary study, we tested the survival of wild-type and DchoD strains in macrophages 2, 4 and 6 days after infection, and found that the intracellular growth of Mtb increased significantly with increasing culture duration. We also observed that the growth of wild-type and DchoD strains was similar up to 4 days post-infection, but differed significantly on day 6 post-infection (Fig. 3).
A comparison of the intracellular growth of mutant and wildtype Mtb in macrophages performed 6 days after infection showed that intracellular replication of opsonized and non-opsonized DchoD strain in macrophages was significantly impaired compared to that of wild-type and complemented (DchoD-choD) strains (Fig. 4A). Neither inhibition of the TLR2-mediated signaling pathway with blocking mAb or IRAK1/4 inhibitor nor inhibition of CR3-mediated signaling pathway with blocking mAb had a significant effect on the survival of the wild-type strain in macrophages ( Fig. 4B and C). However, treatment with blocking antibodies or IRAK1/4 inhibitor significantly increased the growth of both opsonized and non-opsonized DchoD in macrophages compared to macrophages not treated with mAbs ( Fig. 4 B  and C), indicating that the TLR2-and CR3-mediated signaling pathways are involved in limiting DchoD replication in macrophages. Dimethyl sulfoxide (DMSO), used as a vehicle to prepare IRAK1/4 inhibitor solutions, at a final concentration of 0.5% had no effect on the growth of Mtb strains in macrophages. Similarly, IRAK1/4 inhibitor at the concentration of 10 mM had no significant effect on the viability of macrophages up to six days (% of viable macrophages with and without IRAK1/4 inhibitor at 6 th day amounted 90% and 89%, respectively).

NO and ROS Production by Macrophages Infected with Wild-type, DchoD, or DchoD-choD Strains
Bacteria-induced NO production by macrophages was determined 48 hours after infection, a time chosen because preliminary experiments showed that the level of nitrite (a stable metabolite of NO) was almost undetectable in 24-hour culture supernatants. We found that both opsonized and non-opsonized DchoD, but not wild- type or complemented strains, induced NO production (Fig. 5A). However, DchoD failed to stimulate NO production when the TLR2-mediated signaling pathway was disrupted by IRAK1/4 inhibitor ( Fig. 5B and C). In contrast, treatment of macrophages with IRAK1/4 inhibitor or anti-CR3 mAb had no effect on NO production after infection with the wild-type strain (Fig. 5 B  We found that neither the mutant strain nor wild-type Mtb influenced ROS production by non-stimulated macrophages 24 hours post-infection. However, the wild-type strain strongly inhibited the ability of macrophages to produce ROS in response to stimulation with PMA. We further found that opsonized and non-opsonized DchoD exhibited a significantly weakened ability to suppress ROS production by macrophages compared to wild-type and complemented strains (Fig. 6A). In subsequent experiments, we treated macrophages with IRAK1/4 inhibitor or anti-CR3 mAb before infection with Mtb. As shown in Figure 6B and C, IRAK1/4 inhibitor significantly reduced the ability of the wildtype strain to impair ROS production by macrophages. In contrast, the decrease in ROS production by macrophages infected with DchoD was significantly greater in the presence of IRAK1/4 inhibitor than in the absence of inhibitor (Fig. 6 B and  C). It was found that in the presence of IRAK1/4 inhibitor, wildtype strain had similar effect on the ROS production as mutant strain without the inhibitor. For opsonized bacteria, the percentage of ROS inhibition observed in wild-type Mtb-infected macrophages treated with IRAK1/4 inhibitor or DchoD-infected macrophages without inhibitor was 49% and 31%, respectively. In the case of infection with the non-opsonized bacteria, both wildtype Mtb in the presence of the inhibitor and mutant strain without inhibitorinhibited ROS production by 31%. Treatment of macrophages with anti-CR3 mAb did not affect ROS inhibition by macrophages infected with the wild-type strain. A trend toward increased inhibition of ROS production that did not reach significance was observed in macrophages infected with the mutant strain in the presence of anti-CR3 mAb. Neither the vehicle for PMA (0.1% ethanol in HBSS) nor 0.5% DMSO in HBSS affected ROS production by macrophages. IRAK1/4 inhibitor at the concentration of 10 mM had no significant effect on the viability of macrophages after 24 hours of culture (% of viable macrophages in the presence and absence of IRAK1/4 inhibitor amounted 91% and 95%, respectively).

Effect of Wild-type and DchoD Strains on PMA-stimulated ERK1/2 Phosphorylation
We found that during a 2-hour phagocytosis experiment, neither wild-type nor mutant strains affected PMA-induced phosphorylation of ERK1/2 in macrophages (data not shown). Because it is known that Mtb displays a very long infection cycle time in macrophages, we tested the impact of Mtb strains on ERK1/2 phosphorylation 24 hours post-infection. These experiments showed that PMA-induced phosphorylation of ERK1/2 was significantly inhibited by opsonized and non-opsonized wildtype strain (Fig. 7A), but not by the DchoD mutant strain (Fig. 7B).

TNF-a and IL-10 Production Response in Macrophages Infected with Wild-type, DchoD, or DchoD-choD Strains
The production of TNF-a by macrophages infected with wildtype or DchoD strains was very similar (Fig. 8A). However, DchoD strain (opsonized and non-opsonized) stimulated significantly lower macrophage production of IL-10 than the wild-type strain and complemented mutant (Dcho-choD) did (Fig. 8B). In the absence of Mtb infection, macrophages released relatively low amounts of TNF-a (10.960.4 pg/ml) and IL-10 (1.360.4 pg/ml).

Discussion
Despite considerable research effort, the molecular mechanisms of Mtb virulence remain unclear. Pathogenic mycobacteria can affect the function of immune cells through secreted extracellular proteins or cell wall components [26]. We have previously shown that the Mtb DchoD mutant is attenuated in mice and peritoneal macrophages [21]. On the other hand, ChoD does not appear to be essential for cholesterol degradation, suggesting that the observed attenuation is not attributable to nutrient limitations during intracellular growth [15]. In the current study, we found that ChoD is required for modification of the antibacterial activity of human macrophages by Mtb. We also observed that inactivation of TLR2-or CR3-mediated signaling (with blocking mAbs) improved the growth of the DchoD mutant in macrophages to the level of the wild-type strain, suggesting that ChoD is required to engage the TLR2-and CR3-mediated signaling that results in prolonged survival of bacilli inside macrophages.
There was little difference in the engulfment of wild-type and DchoD strains by THP-1 cells, and blocking TLR2-or CR3mediated signaling pathways decreased the ingestion of both investigated strains. However, as previously observed in a mouse model [21], the DchoD mutant appeared to be attenuated in THP-1-derived macrophages compared to wild-type Mtb. This observation was verified using the complemented mutant, DchoD-choD, carrying an intact copy of choD introduced into the attB site of chromosomal DNA. Cholesterol transport/degradation mutants have previously been described as attenuated in vivo, suggesting an essential nutrient role of cholesterol in the pathogenic process. For example, Dmce4c, an Mtb mutant defective in cholesterol transport, appears to be less virulent then the wild-type strain in a mouse model [12]. Similarly, a 3-ketosteroid 9a-hydrolase mutant is attenuated in mouse bone marrow macrophages [10],, and the side-chain-degradation mutant, DfadA5, is impaired in the late stage of mouse infection [27]. Additionally, the DkstD strain, which is unable to degrade the cholesterol ring structure, is attenuated in human macrophages differentiated from THP-1 cells [14]. It was also previously shown that the Digr strain, which is defective in degradation of the 26-propionate side chain fragment, is attenuated in mice [28], [29]. The in vivo and in vitro attenuation of these mutants was related to their inability to use cholesterol as a nutrient in the pathogenicity process and/or to their accumulation of toxic degradation intermediates [10]. On the other hand, ChoD probably does not play an essential role in cholesterol degradation [15], [22], [23]. Thus, we hypothesized that the attenuated growth of the DchoD strain was not due to the inhibition of cholesterol degradation but rather was attributable to the enhanced functional response of macrophages.
Upon activation by engulfment of Mtb, phagocytes produce ROS and NO, which participate in bacteria killing and/or growth inhibition [8], [30][31][32]. The NO response of macrophages was effectively suppressed by wild-type and complemented Mtb, but not by DchoD; in macrophages infected with DchoD, NO overproduction remained intact. This induction of NO production by DchoD was blocked by IRAK1/4 inhibitor and anti-TLR2 mAbs, but not by a mAb to CR3, demonstrating its dependence on the TLR2, but not the CR3, pathway. This finding is consistent with a previous report that demonstrated an essential role for TLR2 in inducible nitric oxide synthase (iNOS) expression in Mtbinfected macrophages [33]. Other investigators [34], [35] have reported that Mtb induces ROS production in macrophages 30    minutes post-infection. Here, we observed that PMA-stimulated ROS production was attenuated by 80% in macrophages infected with wild-type Mtb, but only by 20% in DchoD-infected macrophages. Moreover, the inhibition of ROS production observed in the wild-type strain was partially blocked by IRAK1/4 inhibitor, but not by an anti-CR3 mAb. IRAK1 and 24 have been reported to interact with protein kinase C, which phosphorylates NADPH oxidase [36], [37] and can also directly phosphorylate NADPH oxidase to effectively promote the production of ROS [35]. Our data are in accord with a report that the CR3-mediated signaling pathway is not involved in NO or ROS production in Mtb-infected macrophages [38], and collectively suggest that functional ChoD in the wild-type strain acts through the TLR2-mediated signaling pathway to play an essential role in suppressing the bactericidal activity of macrophages.
Consistent test results obtained by us with the use of antibodies anti-TLR2 and IRAK1/4 inhibitor argue in favor for signaling pathways involvement through TLR2. It is known that IRAK1/4 are important mediators in signal transduction of the TLR family, including also TLR1/2/6/5/7/8/9 and they may act to potentiate the downstream signaling [3]. Therefore we cannot completely exclude the participation of other TLRs in the response of macrophages to the Mtb infection.
Activation of TLR2 results in phosphorylation of the MAPK family member ERK1/2 [39]. ERK1/2, in turn, participates in induction of iNOS expression and activity in phagocytes as well as phosphorylation of NADPH oxidase components p47 phox and p67 phox , which are responsible for NO and ROS production [40][41][42]. We found here that wild-type Mtb, but not DchoD, blocked the ability of PMA to induce ERK1/2 phosphorylation in macrophages, confirming that ChoD is required for Mtb to disrupt the TLR2-activated signaling pathway that leads to inhibition of the antibacterial response of macrophages. We also found that ChoD of Nocardia erythropolis significantly decreased PMA-stimulated phosphorylation of ERK1/2 in macrophages (data not shown).
As reported previously [43], the TLR2 pathway can be utilized by Mtb as a survival mechanism (e.g., through inhibition of phagosome maturation); however, the Mtb factor responsible for disrupting TLR2-mediated signaling has remained unknown. Our results suggest the importance of ChoD in this process. In this context, it should be noted that, during the phagocytosis process, TLR2 proteins are recruited to the macrophage phagosome [44], where they may interact with phagocytosed Mtb. Moreover, it was recently reported that Mtb is able to escape the phagosome and gain access to the cytosol of infected host macrophages [45], [46]; thus, ChoD could exert a suppressive effect on signaling proteins in the cytosol. Whether ChoD affects TLR2 signaling proteins directly or inactivates this signaling pathway through interactions with lipid rafts, as has been suggested [47], remains to be clarified.
Another possible impact of Mtb is through direct actions of the bacteria on TLR2 activity [48]. In this study, Madan-Lala and coworkers demonstrated that the cell envelope-associated serine hydrolase, Hip1, is an important Mtb virulence factor that modulates pro-inflammatory responses in Mtb-infected macrophages. The authors concluded that Hip1 limits the magnitude of macrophage responses by suppressing activation of the TLR2 signaling pathway. It has also been demonstrated that direct treatment of cells with ChoD from Pseudomonas fluorescens depletes cholesterol from cell membranes, thereby affecting the conformation and function of chemokine receptors in the plasma membrane [47]. It is accepted that, after penetration into macrophages, tubercle bacilli reside predominantly in a cholesterol-rich region of the cellular plasma membrane. Changes in the cholesterol level in the plasma membrane modulate the activity of proteins and receptors located in lipid rafts [49], which can determine the activity of cytosolic signaling protein, such as NF-kB [50].
The phagocytosis of Mtb initiates the production of various proinflammatory cytokines, including TNF-a and IL-10. IL-10 has been shown to inhibit phagosome maturation and ROS and RNI production in phagocytes [7], [51], [52]. In the current study, we observed an elevation in the level of IL-10 produced by macrophages infected with the wild-type strain compared to that in macrophages infected with the DchoD mutant. The multiplication of both strains was the same during the period of the experiment. This increase in IL-10 level in macrophages in response to infection by the wild-type strain was associated with an absence of ROS and NO production. TNF-a acts in synergy with interferon c (IFN-c) in the killing of Mtb through the induction of NO and ROS production. Moreover, TNF-a is a critical contributor to granuloma formation and is also involved in Mtbinduced apoptosis [7], [53], [54]. However, we found here that wild-type and mutant strains did not differ in the induction of TNF-a production by macrophages, indicating that ChoD affects only a subset of the functional activities of macrophages.
On the basis of the above observations and our experimental data, we hypothesize that ChoD is indispensable for Mtb effects on the TLR2-mediatated signaling pathway in the pathogenesis process. Mtb defective for the synthesis of ChoD stimulated macrophages to produce NO and ROS and limited the production of IL-10, resulting in reduced survival of bacteria inside macrophages. Our findings demonstrate that ChoD of Mtb participates in the virulence of tubercle bacilli and promotes pathogen survival in human macrophages.