Mycobacterium bovis Requires P27 (LprG) To Arrest Phagosome Maturation and Replicate within Bovine Macrophages

ABSTRACT Mycobacterium bovis causes tuberculosis in a wide variety of mammals, with strong tropism for cattle and eventually humans. P27, also called LprG, is among the proteins involved in the mechanisms of the virulence and persistence of M. bovis and Mycobacterium tuberculosis. Here, we describe a novel function of P27 in the interaction of M. bovis with its natural host cell, the bovine macrophage. We found that a deletion in the p27-p55 operon impairs the replication of M. bovis in bovine macrophages. Importantly, we show for the first time that M. bovis arrests phagosome maturation in a process that depends on P27. This effect is P27 specific since complementation with wild-type p27 but not p55 fully restored the wild-type phenotype of the mutant strain; this indicates that P55 plays no important role during the early events of M. bovis infection. In addition, we also showed that the presence of P27 from M. smegmatis decreases the association of LAMP-3 with bead phagosomes, indicating that P27 itself blocks phagosome-lysosome fusion by modulating the traffic machinery in the cell host.

M ycobacterium bovis causes tuberculosis in a wide variety of mammals with strong tropism for cattle and eventually humans. This mycobacterial species forms, together with M. tuberculosis and many other pathogenic mycobacterial species, the M. tuberculosis complex (MTC). The members of the MTC are genetically related and may have evolved from a common ancestor via successive DNA deletions/insertions, resulting in the present Mycobacterium speciation and their differences in pathogenicity (1). In recent years, considerable advances have been made in the understanding of the molecular bases of pathogenicity, virulence, and persistence of mycobacteria. One significant contribution has been the identification and characterization of essential mycobacterial virulence genes. Most of these studies have been performed in the human bacillus M. tuberculosis because of its worldwide distribution and its impact on global health. Although the bovine macrophage is the preferred site of infection for M. bovis, little is known about the host-pathogen interactions between M. bovis and the bovine macrophage.
P27 is a secreted surface-expressed glycolipoprotein antigen that was first described in M. bovis and is conserved in many species of the Mycobacterium genus (2). The gene that encodes P27 constitutes an operon, together with the Mb1445c gene, usually known as p55 (2). The mutation of the p27-p55 operon in both M. bovis and M. tuberculosis significantly reduces the replication of the bacilli in mice and in macrophage cell cultures; this indicates that both P27 and P55 are relevant for mycobacterial virulence (3,4). The first functional studies of the p27-p55 operon demonstrated that LprG (P27) participates in host-pathogen adhesion through binding to dendritic cellspecific intercellular adhesion molecule-3-grabbing nonintegrin (DC-SIGN) receptor (5).
DC-SIGN receptor, like mannose receptor, is a C-type lectin receptor present in dendritic cells and macrophages. These findings suggest that interactions between these cells and P27 occurred via its mannosylated residues (5). P27 also binds Toll-like receptor 2 (TLR2) (6,7), an important component of the innate immune response against M. tuberculosis (8) and virulent M. bovis (9). The deletion of the p27-p55 operon has shown to reduce the capacity of M. bovis to induce an adequate Th1 response in cattle (10), underlining the immunogenic properties of P27. Later on, two studies reported that the proteins encoded in this operon are involved in the preservation of the cell wall and in the transport of toxic compounds away from the cells (4,11). Recent studies have proposed that the role of the proteins encoded in the p27(lprG)-p55 operon is to facilitate the surface localization of mannose-capped lipoarabinomannan (ManLAM) (12,13). The ManLAM exposed on the bacterial surface binds the mannose receptor on macrophages, allowing the phagocytosis of the mycobacteria and subsequently the survival in arrested phagosomes (14). In this model, the lack of P27 (LprG) would impair the inhibitory mechanism of phagosome-lysosome fusion by reducing the expression of ManLAM in the bacterial surface. Although these previous studies have undoubtedly demonstrated a role of the lprG-p55 operon in the ManLAM-mediated cell-host interactions, the contribution of each individual gene of the lprG-p55 operon to these interactions is still undefined. To better understand the mechanism by which P27 and P55 contribute to the virulence of pathogenic mycobacteria, we studied the role of these proteins in the infection by M. bovis of its natural host, the bovine macrophage.

RESULTS
Mutation of the p27-p55 operon impairs M. bovis replication inside bovine macrophages. We have previously reported that P27 and P55 are essential proteins for full M. bovis virulence in BALB/c mice and M. bovis replication in a murine macrophagic cell line (15). To determine whether this role is conserved in the natural host of M. bovis, the bovine macrophages, we compared the intracellular replication of the mutant to the wild type and the complemented strains in bovine monocyte-derived macrophages (BMDMs). Macrophages were infected at a multiplicity of infection (MOI) of 1, and the bacterial burden was calculated by the recovery of intracellular bacteria at different time points postinfection. The number of intracellular bacteria was similar for wild-type and mutant strains at 3 h postinfection, indicating a similar rate of infection for these strains (Fig. 1). Whereas the bacterial burden in cells infected with either the wild type or the fully complemented strain increased after 48 and 72 h, the replication rate of the MbΔp27 mutant was steady and significantly lower ( Fig. 1). This finding demonstrates that p27-p55 mutation impaired M. bovis replication in bovine macrophages. Notably, the complementation with p55 alone failed to restore the wild-type phenotype. In contrast, the complementation with p27 alone enabled the mutant to replicate at levels equivalent to those of a doublecomplemented or the wild-type strain at 72 h postinfection.
The p27-p55 operon is required for M. bovis phagosome maturation arrest in bovine macrophages. To further understand the mechanism by which P27-P55 counteracts the microbicidal actions of the bovine macrophages, we evaluated the maturation stage of mycobacterial phagosomes by using immunofluorescence and confocal microscopy. M. bovis strains were used to infect bovine macrophages during 1 h of uptake and 2 h of chase, as described in Materials and Methods. M. bovis wild-type association with the late endocytic marker lysosomal-associated membrane protein 3 (LAMP-3) was relatively low. In contrast, the fraction of MbΔp27 associated with LAMP-3 was significantly higher (P Յ 0.05) than that of the wild-type strain ( Fig. 2A and B). These results indicate that the p27-p55 operon participates in the phagosomal arrest induced by intracellular M. bovis to replicate inside macrophages. To assess the contribution of each protein of the p27-p55 operon to the phagosome arrest, we evaluated the association of LAMP-3 with phagosomes containing MbΔp27 complemented with the p27 or p55 gene of the operon. After infection, the association of LAMP-3 with MbΔp27::p55 was not significantly different from the MbΔp27 strain. Con-versely, the association of LAMP-3 with MbΔp27::p27 was lower than that for infections with MbΔp27 and equivalent to those for infections with either with the doublecomplemented MbΔp27::p27/p55 strain or the wild-type strain. To confirm these results, we additionally analyzed the association of the lysosomal enzyme cathepsin D. Similarly, the association fraction of cathepsin D with mycobacteria was higher in bovine macrophages infected with MbΔp27 than in those infected with either the wild-type or the MbΔp27::p27 strain ( Fig. 2C and D). Altogether, we concluded that the M. bovisinduced phagosomal arrest in bovine macrophages required P27.
P27 blocks polystyrene bead-phagosome (BP) maturation. The results described above indicate that only P27, and not P55, seems to be necessary for limiting phagosome-lysosome fusion, suggesting that the function of the P27-P55 system is not involved in the phagosome arresting mechanism. In order to test whether P27 itself influences phagosome maturation, we determined the maturation stage of phagosomecontaining P27 in a bacterium-free context. It has been reported that P27 captures in its hydrophobic pocket and in other protein domains, LAM, lipomannan (LM) and other glycolipids, particularly phosphatidyl-myoinositol mannosides (PIM) (6). Mannosecapped LAM (ManLAM) is a key component of the interaction between pathogenic mycobacterium and host cells (16), and this complex glycolipid is not produced by M. smegmatis. Instead, M. smegmatis produces phospho-myoinositol-capped LAM (PI-LAM) and PIM (17). To determine whether the role of P27 in host cell interaction is dependent on ManLAM, P27 was purified from M. smegmatis as a recombinant hexahistidine-tagged form. Recombinant P27 was absorbed onto polystyrene beads, and the presence of the protein on coated beads was verified by Western blotting (see Fig. S1 in the supplemental material). As a control for the phagocytosis of coated beads, we used bovine serum proteins, recombinant HspX (a heat shock protein not associated with phagosome maturation [purified from Escherichia coli]), and a pool of M. smegmatis protein debris (ProtPool) that binds nonspecifically to the nickel-resin (see Materials and Methods). Coated beads were incubated with BMDMs and, upon phagocytosis, the trafficking of bead-phagosomes (BPs) was analyzed using LAMP-3 and LysoTracker (LTR), a fluorescent dye used for labeling acidic bead-phagosomes (18), and then analyzed by confocal microscopy (Fig. 3). Our results showed that the presence of

Role of P27 in M. bovis Macrophage Infection
Infection and Immunity P27 from M. smegmatis decreased the association of LAMP-3 with bead-phagosomes ( Fig. 3A and B). In contrast, the LAMP-3 association with ProtPool-coated beads markedly increased, indicating that P27 is sufficient to arrest phagosome maturation of latex bead-phagosomes. Confirming this result, the association of LTR with P27-coated BPs significantly decreased in bovine macrophages (Fig. 3B).

P27 increases the survival of M. bovis in bovine macrophages and HeLa cells.
To determine the importance of P27 in the mycobacterial intracellular persistence, we infected bovine macrophages with M. smegmatis wild type or M. smegmatis strains overexpressing P27. After 4 and 24 h of infection, we observed a significant increase in  or HpsX protein for 1 h of uptake. Next, the cells were washed, followed by incubation for 1 h of chase with the addition of the dye LysoTracker Red (50 nM) to detect acidic compartments. The cells were fixed and analyzed by confocal microscopy. The fluorescence intensity association of LysoTracker to the BPs was quantified. Data represent the means Ϯ the SEM of three independent experiments. In all panels, the asterisks indicate significance: *, P Յ 0.05; **, P Յ 0.01; and ***, P Յ 0.001. The data were analyzed using a two-tailed Student t test.
Mutation of the p27-p55 operon produces low expression of iNOS. In addition to phagosome maturation, nitric oxide (NO) production is a mechanism that controls mycobacterial replication. To explore the capacity of the mutant MbΔp27 to induce proinflammatory response in bovine macrophages, we evaluated the expression of cytokines and inducible nitric oxide synthase (iNOS) upon infection. The MbΔp27 strain elicited significantly lower transcript levels of iNOS at 16 h postinfection than the wild-type strain ( Fig. 5A and B). However, infection with the complemented strain MbΔp27::p27/p55 produced a partial recovery of iNOS transcription. Despite the upregulation of iNOS upon M. bovis infection, the production of NO in culture supernatants of infected bovine macrophages was undetectable (data not shown). This result may be because of the lack of interferon gamma stimulation during the cell culturing, which is a well-known enhancer of NO production (19,20).
Although not significant, the expression of the proinflammatory cytokines interleukin-1␤ (IL-1␤), IL-6, and IL-12p35 showed a reduction trend in macrophages infected with the mutant strain compared to those infected with the wild-type strain (Fig. 5C, D, and E).

DISCUSSION
A hallmark of pathogenic mycobacteria is the lack of classical virulence factors that are present in many other bacterial pathogens. In nonpathogenic mycobacteria in particular, many virulence genes, such as P27/LprG, are conserved. These findings suggest that pathogenic genomes have adapted their free lifestyle to the intracellular environment with minimal acquisition of new or exclusive virulence genes. Many of the known virulence factors are needed for the survival of bacilli inside macrophages. These particular virulence factors act by modulating or blocking the defense mechanisms that the host cell displays to eliminate mycobacteria. For instance, the production of effectors, including those involved in responses to nitro-oxidative stress and cell death program or apoptosis, is a clear example of these mechanisms.
An effective strategy used by M. tuberculosis to counteract the action of macrophages is the subversion of the normal progression of phagosome maturation to prevent the formation of an active phagolysosome. This modulation of intracellular endosomal trafficking allows the bacteria to remain in a replicative niche and thus avoid immune detection. Although the process of phagosomal arrest induced by M. tuberculosis is well described, this cell-autonomous mechanism of defense is less clear , or IL-12p35 (E) was calculated using the 2 ϪΔΔCT method with E correction, using GAPDH mRNA expression as the reference gene and uninfected cells as the calibrator. Data were analyzed using a paired Student t test. ns, not statistically significant.

Role of P27 in M. bovis Macrophage Infection
Infection and Immunity in M. bovis. Moreover, at longer time points after infection (e.g., 7 days), M. bovis is able to translocate into the cytosol of human cells (21), although whether M. bovis is localized in the cytosol of bovine macrophages remains to be determined. Two previous studies (12,13) proposed that LprG-P55 serves as a carrier to facilitate assembly or trafficking of ManLAM to the M. tuberculosis cell wall, which in turn binds to the macrophage-mannose receptor and facilitates bacterial entry and the inhibition of phagosome-lysosome fusion. Furthermore, ManLAM induces iNOS through the NF-B signaling pathway (22). Therefore, it is possible that the prominent localization of the mutant MbΔp27-p55 in phagolysosomes, as well as the significant decrease in the levels of iNOS induced by this mutant compared to the wild-type strain, is a consequence of lower ManLAM levels on the cell surface, leading to a reduction in the survival of the mutant in macrophages. This study highlights that P55 is dispensable for the inhibition of phagosomelysosome fusion since the M. bovis mutant expressing P27 alone restored the wild-type intracellular persistence at 72 h postinfection and was capable of inducing phagosome arrest at the same level as that of the wild-type strain. These results suggest that P27 does not require P55 to facilitate ManLAM localization on the bacterial surface, that P27 inhibits phagosome-lysosome fusion by direct binding to the mannose receptor via its mannose residues, or both. The fact that recombinant P27 purified from M. smegmatis inhibited the maturation of bead-containing phagosomes to phagolysosomes suggests that P27 exerts a direct interaction with the host cells, which does not preclude its previously described function in the cell wall assembly. Moreover, the localization of P27 in the extracellular space or in association with the bacterial cell wall (15) also supports a model in which P27 can trigger the blockage of phagosome-lysosome fusion by direct binding to the mannose receptor. We hypothesize that this binding to the mannose receptor is via its associated glycolipids, likely PIM 6 (17).
In summary, we characterized here for the first time the trafficking of M. bovis phagosomes in a primary culture of bovine macrophages. This study demonstrates that virulent M. bovis arrests phagosome maturation in bovine macrophages and that P27 plays a key role in this mechanism. Importantly, although both P27 and P55 are necessary for M. bovis replication inside both bovine and mouse macrophages (15), P27 alone has an additional role in the mycobacterial survival inside bovine macrophages and in nonprofessional phagocytes. Therefore, our results, together with previous reports (12,13,15), indicate that the p27-p55 operon is involved in multiple virulence mechanisms and functions and that P27 plays a major role in the early events of M. bovis infection (5,6). The precise mechanism by which P27 modulates M. bovis trafficking is still unknown, as well as the host cell factors subverted by this virulence factor. Our findings do demonstrate that this protein directly participates in one important survival mechanism displayed by pathogenic mycobacteria. It remains unclear, however, whether there is any cooperative action between LprG/P27 and P55 in the transfer of ManLAM and glycolipids to the bacterial cell envelope.

MATERIALS AND METHODS
Bacterial strains and culture media. The M. bovis strains were grown in Middlebrook 7H9 or Middlebrook 7H10 medium (Difco Laboratories, USA) supplemented with 0.5% albumin, 0.4% dextrose, and 0.5% pyruvate. Middlebrook 7H9 was used with or without 0.05% Tween 80. When necessary, either hygromycin at 50 g/ml or kanamycin at 20 g/ml (Sigma, USA) was added to the media.
PBMC isolation and BMDM differentiation. The animals used in this study were selected from the Instituto Nacional de Tecnologías Agropecuarias (INTA) experimental herd and tested negative for bovine tuberculosis infection for the single intradermal tuberculin test and interferon gamma release test. Portions (60 ml) of blood were taken from each animal under sterile conditions according to the instructions of the Committee for Institutional Care and Use of Animal Experimentation (CICUAE-CICVyA) of INTA. Peripheral blood mononuclear cells (PBMCs) were separated from heparinized blood by centrifugation over Histopaque 1077 (Sigma) according to the manufacturer's protocol. To derive monocytes, we cultured PBMCs on 12-mm glass coverslips, T25 flasks, or 24-well plates containing RPMI 1640 complete medium (Invitrogen, Carlsbad, CA) supplemented with 10% autologous plasma for 16 h at 37°C and 5% CO 2 . Nonadherent cells were removed by extensive washing with phosphate-buffered saline (PBS). Only adherent cells were maintained in culture for 5 days at 37°C and 5% CO 2 to obtain BMDMs. Cell viability was confirmed by trypan blue exclusion assay. were coupled with 20-g/ml concentrations of the following proteins: bovine serum proteins (serum), a pool of proteins of M. smegmatis (ProtPool), or M. smegmatis P27 protein (P27). The coupling was performed using 0.1 M borate buffer (pH 8.5) at room temperature overnight on a rotating wheel. The beads were washed with BSA at 10 mg/ml in borate buffer and finally resuspended in PBS (pH 7.4) containing BSA at 10 mg/ml and stored at 4°C until use.

SDS-PAGE and Western blot analysis.
To detect the adsorption of P27 to the polystyrene beads, beads coated with bovine serum proteins (serum), a pool of proteins of M. smegmatis (ProtPool), or the P27 protein of M. smegmatis (P27) as described above were resuspended in sample buffer containing 1% 2-mercaptoethanol, and the samples were maintained at 4°C. For the Western blot analyses, the samples were subjected to electrophoresis in a 12% SDS-PAGE gel, transferred to a nitrocellulose membrane, and blocked with PBS supplemented with 0.1% (vol/vol) Tween 20 and 5% milk. The nitrocellulose membrane was incubated for 2 h with a polyclonal anti-P27 antibody, washed, and incubated with a secondary alkaline phosphatase-conjugated goat anti-rabbit antibody (Sigma, USA) at a 1:10,000 dilution.
Internalization of polystyrene beads. For internalization, coated beads were diluted 1:500 in complete medium and applied to PBMC-derived bovine macrophages seeded in a 24-well plate with a final concentration of ϳ10 beads per cell. After the indicated times of uptake and chase, the cells were washed with PBS and fixed for 20 min with 4% PFA.
Image analysis. A Leica SP5 AOBS laser scanning confocal microscope was used (Leica Microsystems, Germany). For image acquisitions in fixed samples, a single focal plane was monitored over time (xyt scanning mode) using a 63ϫ/1.4 HCX-PLAPO oil objective lens, an argon laser (488 nm), and a DPSS laser (561 nm), when applicable; a scanner frequency of 200 to 400 Hz; and line averaging 6, using PMT detectors at a scanning resolution of 1,024 by 1,024 pixels or 512 by 512 pixels (zoom of 2.5). The same settings for laser powers, gain, and offset were maintained for the different experiments. However, they were not obtained close together in time. The microscope software (LAS AF) permits saving the exact settings to acquire the images used in the experiments. Therefore, with every experiment the corresponding settings were loaded.
Analyses of all the images were performed using ImageJ (National Institutes of Health) and Fiji. Fiji is a distribution of ImageJ available at http://fiji.sc. Iterative versions of ImageJ used for this work are 1.41m through 1.46a.
To measure the association of a marker (e.g., primary antibodies against LAMP-3 or cathepsin D) with bacterial particles in the cell, the RGB image was split into individual channels, and the red channel (bacteria) was subjected to a pixel threshold. The "wand-tracing tool" was used to select all the bacteria per cell, and then the "analyzed-measure" function of Fiji was used to measure the fluorescence intensity of the marker of interest (corresponding to the secondary antibody used to detect the primary antibody) associated with the bacterial phagosome by redirecting the measurements to the channel of interest in "set measurements" in the analysis function of Fiji.
To quantify the fluorescence association with the bead phagosomes, the corresponding fluorescentchannel movies were loaded into Fiji. A circle enclosing the phagosome in the bright-field image was drawn by using the "elliptical selection" tool. In "set measurements," only "area" and "integrated density" were selected. Subsequently, the fluorescence intensity of the marker of interest (corresponding to primary antibodies against LAMP-3 or LysoTracker) associated with the bead phagosome was measured by redirecting the measurements to the channel of interest in "set measurements" in the analysis function of Fiji (18). All of the images were acquired with the same zoom, and the results of the different experiments were combined. The fluorescence intensity values were plotted and analyzed using Microsoft Excel 2011 (Microsoft) and GraphPad Prism 5 (GraphPad Software, Inc., USA).