Lipoprotein LprI of Mycobacterium tuberculosis Acts as a Lysozyme Inhibitor*

Mycobacterium tuberculosis executes numerous defense strategies for the successful establishment of infection under a diverse array of challenges inside the host. One such strategy that has been delineated in this study is the abrogation of lytic activity of lysozyme by a novel glycosylated and surface-localized lipoprotein, LprI, which is exclusively present in M. tuberculosis complex. The lprI gene co-transcribes with the glbN gene (encoding hemoglobin (HbN)) and both are synchronously up-regulated in M. tuberculosis during macrophage infection. Recombinant LprI, expressed in Escherichia coli, exhibited strong binding (Kd ≤ 2 nm) with lysozyme and abrogated its lytic activity completely, thereby conferring protection to fluorescein-labeled Micrococcus lysodeikticus from lysozyme-mediated hydrolysis. Expression of the lprI gene in Mycobacterium smegmatis (8–10-fold) protected its growth from lysozyme inhibition in vitro and enhanced its phagocytosis and survival during intracellular infection of peritoneal and monocyte-derived macrophages, known to secrete lysozyme, and in the presence of exogenously added lysozyme in secondary cell lines where lysozyme levels are low. In contrast, the presence of HbN enhanced phagocytosis and intracellular survival of M. smegmatis only in the absence of lysozyme but not under lysozyme stress. Interestingly, co-expression of the glbN-lprI gene pair elevated the invasion and survival of M. smegmatis 2–3-fold in secondary cell lines in the presence of lysozyme in comparison with isogenic cells expressing these genes individually. Thus, specific advantage against macrophage-generated lysozyme, conferred by the combination of LprI-HbN during invasion of M. tuberculosis, may have vital implications on the pathogenesis of tuberculosis.

Mycobacterium tuberculosis, the causative agent of tuberculosis, is one of the most dreaded human pathogens that affect millions of lives worldwide (1). The inexplicable success of M. tuberculosis as a human pathogen relies on its ability to utilize a number of defense strategies to adapt its metabolism in response to a plethora of environmental challenges during the course of its pathogenic cycle (2,3). Several of the components of its protein machinery have emerged as virulence factors with vital implications. Among these, lipoproteins play pivotal roles in several functions related to its virulence and host-pathogen interactions (4 -6). Genome analysis of mycobacteria indicated that the number of lipoproteins is much higher in the pathogenic mycobacteria in comparison with their non-pathogenic counterparts, and M. tuberculosis houses the highest number of lipoproteins.
LprI is a novel lipoprotein that is exclusively present in pathogenic mycobacteria belonging to M. tuberculosis complex. The genomic location of the lprI gene is conserved among pathogenic mycobacteria where it has been identified as an operon partner of the glbN gene, which encodes truncated hemoglobin (HbN) 3 (7,8). The lprI gene is positioned adjacent to the glbN gene, separated through an intergenic distance of 58 nucleotides, and their co-transcription has been observed in Mycobacterium bovis (9), indicating that these two proteins may have functional correlation. The HbN of M. tuberculosis carries potent nitric oxide (NO) detoxification ability (8 -10) and is post-translationally modified by a glycan linkage that facilitates adherence and phagocytosis of cells during macrophage infection (11). The functional relevance of the co-occurrence of LprI with HbN in M. tuberculosis is unknown. Similar co-existence of a lipoprotein, LprG, along with Rv1410, which encodes a small molecule transporter, P55 (an operon conserved in M. tuberculosis complex), has also been observed in M. tuberculosis (12) where they function in a cooperative and synchronized manner. Because the HbN of M. tuberculosis plays a vital role in NO detoxification (8 -10) and is also involved in modulating host-pathogen interactions during intracellular infection of M. tuberculosis, it is likely that the glbN-lprI operon may also have some significant implications in the physiology of tubercle bacillus. LprI is an uncharacterized lipoprotein of M. tuberculosis complex; therefore, in the absence of any knowledge on LprI, its physiological function and implications of its co-existence with the HbN in M. tuberculosis are difficult to understand. To investigate the crucial implications of HbN in modulating the host-pathogen interactions and immune system of the host by providing aid in the better survival and sustenance of M. tuberculosis during intracellular infection, it is important to study the functionality of the LprI to understand functional implications of their co-occurrence in M. tuberculosis.
In silico analysis unraveled that LprI carries a lysozyme-binding motif of the membrane-bound lysozyme inhibitors of the * This work was supported in part by the Council of Scientific and Industrial C-type lysozyme (MliC) family of proteins, belonging to a class of lysozyme inhibitors, which have been explored in Gram-negative bacteria only (13). The presence of a lysozyme inhibitor motif in LprI appears very significant as lysozyme constitutes a major fraction of proteins present in the granules of neutrophils and in macrophage secretions (14,15). Furthermore, the level of lysozyme in mammalian tissues increases remarkably in several mycobacterial diseases, and it has been suggested to be a nonspecific diagnostic marker along with antibody levels in tuberculosis (16,17). Recently, it has been identified as one of the proteins with significant mycobactericidal activity from the pool of neutrophil granule proteins (18). Lysozyme inhibitors are important as they are essential for the survival of several pathogenic Gram-negative bacteria during infection in their animal hosts (19,20). In view of the antibacterial activity of lysozyme and because of the protection it endows at the cell surface, lysozyme inhibitors have been referred as the "guards of the great wall" (19) and thus are recognized as attractive targets for antibacterial drug design (21).
Primary studies conducted during initial stages of tuberculosis research mainly focused on studying the effect of lysozyme on mycobacterial strains indicated that slow growing pathogenic mycobacteria are more resistant toward lysozyme than fast growing non-pathogenic strains (22,23). Despite the important role of lysozyme in protection against microbial infection (24) and identification of a range of antibacterial activities of lysozyme in vitro and in vivo against bacillary infection, the mechanism of its defense remains to be elucidated.
The present study demonstrates that LprI of M. tuberculosis is a novel glycosylated lipoprotein that directly interacts with lysozyme and abrogates its hydrolytic activity in vitro and in vivo. A high up-regulation of the lprI and the glbN genes in a temporal fashion, observed during macrophage infection, suggests their functional correlation that may be vital for the pathogenesis of M. tuberculosis.

Bacterial Strains and Growth Conditions-Recombinant
LprI was cloned and expressed in Escherichia coli strains JM109 and BL21(DE3) and the Mc 2 155 strain of Mycobacterium smegmatis. M. tuberculosis strains H37Ra and H37Rv were used for overexpression of LprI in native host. M. tuberculosis (H37Ra) carries a copy of lprI gene identical to that of M. tuberculosis (H37Rv). The cultures were grown in Middlebrook 7H10 agar or 7H9 liquid broth (Difco) supplemented with ovalbumin, 10% bovine serum albumin fraction V, dextrose, sodium chloride, 0.2% glycerol, and 0.05% Tween 80 at 37°C at 200 rpm. Kanamycin (30 mg/ml) and hygromycin (50 mg/ml) were supplemented in the growth medium whenever required.
Cloning, Expression, and Purification of Recombinant LprI-Full-length lprI was amplified from the genomic DNA of M. tuberculosis (H37Ra) using the forward primer 5Ј-GATT-CATATGAGATGGATCGGCGTCCTGGTGACC-3Ј carrying the NdeI site. lprI lacking the signal peptide was amplified using the forward primer 5Ј-GATCCATATGTGCGCAGCA-AACCCTCCGGCTAAC-3Ј with a BamHI site. The same reverse primer carrying an XhoI site was used for both the genes (5Ј-GATCCTCGAGCGAGGTGCGGCAGACGAACG-3Ј). The amplified products were cloned in pET29A at the appropriate sites under the T7 promoter. Similarly, for the cloning and expression of full-length lprI and glbN genes in mycobacteria, the respective genes were cloned in pSC301 vector under the superoxide dismutase promoter. The forward and reverse primer sequences used for lprI are 5Ј-TATACTGCAGATGA-GATGGATCGGC-3Ј and 5Ј-TATAGATATCTCACGAGGT-GCGGCA-3Ј. For glbN, the forward and reverse primer sequences used are 5Ј-TATACTGCAGATGGGACTACTG-TCA-3Ј and 5Ј-TATAGATATCTTAGACTGGTGCCGTGG-TGCT-3Ј. The lprI and glbN gene pair was cloned using the forward primer for glbN and reverse primer for lprI. The cultures of E. coli wild-type and recombinant LprI lacking signal peptide (LprI⌬ss) were grown to an initial A 600 of 0.5 followed by induction with 100 mM isopropyl 1-thio-␤-D-galactopyranoside. Protein expression was checked in various fractions. LprI⌬ss was purified under denaturing conditions in 6 M urea through nickel-nitrilotriacetic acid chromatography and dialyzed against Tris-buffered saline (50 mM Tris, 200 mM NaCl, pH 8.0).
Sequence-and Structure-based Analysis of LprI-The amino acid sequences of LprI of M. tuberculosis and its related pathogenic strains were retrieved through UniProtKB and aligned using the T-Coffee online sequence alignment program. Bioinformatics tools used to study various attributes of LprI include ScanProsite (25), PFAM (26), LipoP (27), and DOLOP (28). For structural analysis studies, a model structure of LprI was obtained by submitting its sequence to the I-TASSER program (29) provided by the University of Michigan. The I-TASSER program generated five tentative models of which the first model (model 1) is suggested to be the most reliable in terms of quality prediction. Validation of the model structure of LprI was carried out using PROCHECK (30), and a Ramachandran plot, which showed that in the model structure of LprI 94.4% of residues are present in favored and allowed regions, was computed. PyMOL (version 0.99) (31) software was used for conducting the structural analysis in the study.
Phase Partitioning Analysis of LprI-M. tuberculosis (H37Ra) was grown to late log phase, and the cells were harvested, washed once, and resuspended in PBS supplemented with protease inhibitors. Cells were lysed by sonication and centrifuged at 20,000 ϫ g to separate cell lysate and cell debris. The lysate (1 ml) was treated with 1% (v/v) Triton X-114, and the aqueous and detergent phases were separated using an established method (33). The phases were equilibrated with PBS and Triton X-114 and analyzed by 12% SDS-PAGE followed by immunoprobing with ␣-LprI polyclonal antibodies generated against recombinant LprI.
Tryptic Digestion of M. smegmatis Surface Proteins-Protease digestion was performed as described elsewhere (34). M. smegmatis recombinants expressing LprI and its partially deglycosylated mutants were grown for 2 days. Cells were harvested by centrifugation at 5000 ϫ g for 5 min and washed with PBS. The wet weight of cells was normalized to 50 mg followed by their resuspension in PBS supplemented with 200 g of trypsin (Sigma-Aldrich), 20 mM CaCl 2 , and 0.5 M D-arabinose. An untreated control sample was also included. All samples were incubated at 37°C for 30 min. Cells were removed, washed once with PBS, and analyzed by SDS-PAGE and immunoblotting.
Agar Diffusion Assay-Two wells were punched in a nutrient agar plate containing M. lysodeikticus cells. In the first well, 50 g of HEWL was added, and in the second well, an equimolar mixture of HEWL and LprI⌬ss was added. The plate was allowed to incubate overnight at 30°C.
Intrinsic Fluorescence Quenching-The tryptophan fluorescence emission spectra of LprI (500 nM) and HEWL (500 nM) as individual proteins and in equimolar combinations (500 nM each) were recorded between 290 and 450 nm after excitation at 280 nm. The degree of fluorescence quenching at 344 nm indicates the interaction between LprI and HEWL.
Fluorescence-based Lysozyme Sensitivity Assay-The EnzChek Lysozyme Assay kit was procured from Molecular Probes (Invitrogen) and used as described previously (35). Standard curves for the activity of HEWL (Sigma L6876, supplied as 58,100 units/mg) on DQ TM lysozyme substrate, fluoresceinconjugated M. lysodeikticus (supplied as a 1-mg vial), were plotted. Briefly, the reaction mixture was prepared in a 100-l volume containing varying concentrations of HEWL (10 -128 units/ml) and HEWL substrate in Tris-buffered saline. The kinetic curves were monitored by measuring the fluorescence for 60 min at 5-min intervals and were found to be linear in nature in the range of 12-64 units/ml HEWL. A concentration of 48 units/ml HEWL was then used for setting up reactions with varying concentrations of LprI to prepare the kinetic curves. The rate of lysis was evaluated from the slope values of the kinetic curves.
Deglycosylation of LprI by ␣-Mannosidase-Analysis of LprI association with glycosyl residues was carried out through digestion of M. tuberculosis lysate with jack bean ␣-mannosidase from Canavalia ensiformis (Sigma-Aldrich). Briefly, LprIexpressing M. tuberculosis (H37Ra) cells were disrupted in 0.05 M sodium acetate buffer, pH 4.5. The cell lysates containing proteins equivalent to 100 g were incubated with 50 g of ␣-mannosidase at 37°C for 8 h. Samples were incubated at 100°C for 10 min and analyzed by 12% SDS-PAGE followed by immunoblotting using ␣-LprI polyclonal antibodies generated against recombinant LprI.
Resazurin Microtiter Plate Assay-A serial 2-fold dilution of HEWL was prepared in 96-well plates, and 100 l of standard mycobacterial cell suspension (A 600 ϭ 0.05) in 7H9 broth was added to each well. The plates were sealed and incubated at 37°C for 24 h and 1 week for M. smegmatis and M. tuberculosis, respectively. Resazurin dye (30 l of a 0.03% solution) was added to each well, and the plates were reincubated for 10 -12 h for color development. The change in color of the resazurin dye from blue to pink indicated the viability of the bacilli.
Macrophage Infection and Determination of Colony-forming Units (cfu)-As described previously, peritoneal macrophages and monocyte-derived macrophages (MDMs) were used as primary cell lines for intracellular infection (36,37), and THP-1 and RAW 264.7 macrophages were used as secondary cell lines. These cell lines were infected with either wild-type or recombinant strains of M. smegmatis expressing LprI, HbN, and LprI-HbN at a multiplicity of infection of 1:5 for 2 h. For the intracellular survival assay in the presence of lysozyme, 100 g/ml HEWL was exogenously supplemented at the time of infection. After 2 h, the extracellular bacteria were removed by washing three times with PBS followed by 1 h of gentamicin treatment to ensure complete removal of non-internalized bacteria. The infected macrophages were incubated for variable time intervals and then solubilized with 0.06% SDS, and the viability of bacteria was determined after appropriate dilutions and plating on Middlebrook agar plates supplemented with ovalbumin, 10% bovine serum albumin fraction V, dextrose, and sodium chloride and then counting the number of colonies.
Ethics Statement-This study was approved by the Institutional Ethics and Biosafety committee and was conducted in accordance with the ethical guidelines of the Central Ethics Committee on Human Research, Indian Council of Medical Research-2000 and those as contained in the Declaration of Helsinki. Animal experiments were approved by the Institutional Animal Ethics Committee and were performed according to National Regulatory Guidelines, Ministry of Environment and Forest, Government of India.
Quantitative Real Time PCR-To determine the expression levels of lprI and glbN in infected macrophages, cells were harvested at different time points, and total RNA was isolated using TRIzol reagent (Invitrogen). The RNA was then converted into cDNA using the iScript cDNA synthesis kit (Bio-Rad). The generated cDNA then served as a template for quantitative PCR analysis using lprI-and glbN-specific primers using the SYBR Green method. The forward primers used for lprI and glbN are 5Ј-ACAAAGCAGCTGGACGATG-3Ј and 5Ј-CCAACTATC-GGCCTTCTTCA-3Ј, respectively, and the reverse primers are 5Ј-TAGAACTGTGCGGTCAATGG-3Ј and 5Ј-GCTGAAGT-GGTGCATGGTAA-3Ј, respectively. The expression levels were normalized in each sample using M. tuberculosis (H37Ra) 16S RNA. 16S RNA of mycobacteria was amplified from the same samples using the primers 5Ј-GTGGCGAACGGGTGA-GTAAC-3Ј and 5Ј-CATCAGGCTTGCGCCCATTG-3Ј.

LprI Is a Novel Lipoprotein of M. tuberculosis Complex-
The open reading frame (ORF) Rv1541c in the genome of M. tuberculosis has been annotated as a lipoprotein, LprI (38), of 197 amino acids with a calculated molecular mass of 21.6 kDa. In silico analysis of LprI (Fig. 1A) indicated that it has a classical lipoprotein character, having a signal peptide (residues 2-16) at the N terminus with a lipobox motif, LSAC, where the presence of a cysteine residue usually acts as a potential lipid attachment site (Fig. 1A). It carries a DUF1311 motif (residues 34 -110) and an MliC motif, which is present in a family of lysozyme inhibitors of certain virulent Gram-negative bacteria (13). The DUF1311 domain, identified on the basis of four conserved cys- teine residues (PFAM, PF07007), spans amino acids 34 -110 with Cys residues present at positions 35, 48, 89, and 98. The functional importance of the DUF1311 motif in proteins is not yet known. Interestingly, LprI, which has the MliC motif, has been identified only in virulence-associated mycobacterial strains, belonging to M. tuberculosis complex where the lprI gene co-exists with the HbN-encoding glbN gene as an operon partner. This genomic organization is conserved only in pathogenic mycobacteria of M. tuberculosis complex (Fig. 1B), suggesting a functional correlation between LprI and HbN.
Structural  Fig. 2A). When the modeled structure of LprI was overlaid upon MliC Pa , the ␤-barrels of LprI and MliC Pa closely overlapped with each other with a root mean square deviation of 3.060 Å (Fig. 2B). In MliC Pa , the two key conserved regions span residues 86 -92 (ISASGAK) and 97-104 (QYIWWTKG). These loops are tightly bound to HEWL through hydrogen bonding and ionic interactions (39). Fig. 2C illustrates the conserved region of MliC Pa (SGSGAKY) that interacts with HEWL and the corresponding similar region of LprI (SGSGARY). Both loops of MliC Pa and LprI are oriented in the same fashion and are entering into the active site cleft of HEWL. High structural resemblance and the conservation of the loop regions in MliC Pa and LprI suggest that LprI may have a similar mechanism for lysozyme binding.
Dimeric LprI Inhibits Hydrolytic Activity of Lysozyme-To experimentally validate the function of LprI, the lprI gene encoding the mature LprI protein (lacking N-terminal residues 1-14) was expressed in E. coli. The mature LprI protein was purified with an apparent molecular mass of ϳ20 kDa (Fig. 3A), and its subunit association properties were analyzed via gel filtration chromatography. Two distinct peaks of almost equal size appeared in the chromatogram, indicating that the protein is present in two oligomeric states. The first peak of the protein corresponded to a high molecular mass of more than 100 kDa, suggesting an aggregated form of LprI, whereas the second peak of the protein exhibited a size of 40 kDa, which indicated the dimeric state of the protein (Fig. 3B). When both the fractions were separated by 12% SDS-PAGE, a single protein band of ϳ20 kDa was obtained.
To check the lysozyme inhibitory activity of LprI, we used M. lysodeikticus, which is used as a standard strain for checking the sensitivity of lysozyme (40). The hydrolytic activity of lysozyme was first tested in the presence of LprI through a lytic zone assay against M. lysodeikticus. A clear lytic zone in the control well that carried only HEWL was visualized, whereas a very thin lytic zone appeared in the other well that carried combination of HEWL and the dimeric fraction of LprI (Fig. 3C). Notably, the protection against lysozyme was seen with the dimeric fraction of the LprI protein only and not with the oligomeric fraction of the protein. These results indicated that dimeric LprI is functional as a lysozyme inhibitor and imparts protection to M. lysodeikticus from HEWL-mediated toxicity. Because LprI was expressed as inclusion bodies in E. coli, oligomers might have resulted during the denaturation/renaturation process used for the protein purification.
LprI Directly Interacts with Lysozyme and Inhibits Its Hydrolytic Activity in Vitro-To check that LprI interacts physically with the lysozyme, we first studied their interactions through the fluorescence spectra of HEWL and LprI recorded individually and in their equimolar combination. Fig. 3D shows a broad peak (ϳ343 nm) along with a long wavelength tail (typical of relatively exposed tryptophan residues) for HEWL as recorded previously in other cases (41), and the peak of LprI alone appeared at 333 nm. Nearly 15% quenching of the fluorescence was measured in the case of the LprI-HEWL mixture with a maximal quenching taking place at 343 nm with respect to the additive spectra of individual proteins. The downward shift in LprI-HEWL spectra indicated intrinsic quenching as a result of the interactions between these two proteins.
The lysozyme inhibitory activity of LprI was further evaluated through a fluorescence-based lysozyme activity assay. This assay relies on the action of lysozyme on the cell walls of M. lysodeikticus conjugated with fluorescein fluorophore such that it relieves the fluorescence quenching of fluorescein. Lysozyme activity levels on fluorescently labeled M. lysodeikticus were determined from the standard curves at different HEWL concentrations. The sensitivity of the assay was verified, and it appeared to detect up to 20 units/ml HEWL. The kinetic curves of HEWL in the range of 24 -64 units/ml showed a linear response over a period of 60 min (data not shown). The activity was then monitored in a similar manner using 57.8 nM (48 units/ml) HEWL and its titration with varying concentrations of LprI. As shown in Fig. 4A, the HEWL activity decreases with increasing concentrations of LprI. From the kinetic curves, the rate of lysis for each of the LprI titrations was determined. The concentration of LprI, which decreases the initial HEWL activity to that of 27.9 nM HEWL (see Fig. 4B, inset), was determined from the plots of lytic activity versus LprI concentration (Fig.  4B), and the dissociation constant (K d Յ 2 nM) was obtained for LprI and HEWL binding. At equimolar concentrations, LprI completely inhibits the lytic activity of HEWL, thus confirming a 1:1 stoichiometry of their binding.
LprI Is a Glycosylated Lipoprotein-To understand the physiological function(s) of LprI in its native host, the lprI gene was overexpressed in its native host M. tuberculosis, and M. smeg-matis, which lacks lprI, served as an lprI knock-out model of mycobacteria. The expression of the lprI gene in these mycobacterial hosts was examined through immunoblotting, which indicated that LprI expressed in mycobacteria migrates slowly on SDS-PAGE and that its molecular mass corresponds to ϳ30 kDa, which is much higher than the LprI expressed in E. coli (Fig. 5A), indicating the possibility of post-translational modification of LprI in mycobacteria. Interestingly, LprI from M. tuberculosis migrated more slowly on SDS-PAGE as compared with the LprI expressed in M. smegmatis (Fig. 5A). The reason for this difference in size may be attributed to certain differences in the glycosylation machinery of these two mycobacterial strains (42). Analysis of LprI protein through the NetOglyc program (43) indicated that it may have multiple gly-cosylation sites and might exist as a highly glycosylated protein (Table 1). Based on the general convention that the presence of a proline residue at Ϫ1 and/or ϩ3 position from a threonine residue strongly favors glycosylation (44), we checked potential glycosylation sites of LprI manually, and three threonine residues, Thr-24, Thr-28, and Thr-117, were identified as potential sites for the glycan linkage. To validate these sites, LprI mutants carrying single and combined mutations at the predicted glycosylation sites (Thr-24, Thr-28, and Thr-117) were expressed in M. smegmatis and analyzed by SDS-PAGE. In comparison with the LprI protein expressed in M. smegmatis, all mutants of LprI altered at single glycosylation sites displayed a faster migration on SDS-PAGE (Fig. 5B), suggesting deglycosylation of LprI at the respective sites. However, the LprI mutant altered at all An analogous loop of LprI (red) is shown to extend into HEWL in a similar orientation and interacts with its active site residues. The interactions between MliC and HEWL take place through hydrogen and ionic bonds (as shown in the inset). All the molecular graphics were generated using PyMOL software (version 0.99) software. FEBRUARY 5, 2016 • VOLUME 291 • NUMBER 6 JOURNAL OF BIOLOGICAL CHEMISTRY 2943 three positions did not show complete deglycosylation of the protein or a larger difference in migration in comparison with the LprI mutant mutated at a single position. These results suggest that LprI may have other sites for the glycan linkages or that an alternative glycan linkage site(s) becomes exposed due to conformational changes after combined mutations at the three glycosylation sites of LprI. To validate that the changes in the molecular mass of the LprI are due to the glycan linkage, we attempted to remove it by deglycosylation. Because O-mannosylation has been detected in the majority of M. tuberculosis proteins (44), we treated whole cell lysate of LprI-expressing M. tuberculosis with ␣-D-mannosidase and subjected it to SDS-PAGE followed by Western blotting with ␣-LprI antibodies (Fig. 5C). A clear reduction in the molecular mass of LprI was Whole cells of E. coli expressing LprI⌬ss (lacking the signal peptide) were fractionated into pellet and supernatant and analyzed by 12% SDS-PAGE. LprI⌬ss appeared in the pellet fraction as inclusion bodies; this fraction was solubilized in 6 M urea and purified using nickel-nitrilotriacetic acid chromatography. B, subunit association of LprI. The gel filtration profile of recombinant LprI purified from E. coli is shown (the inset shows the calibration curve using the marker proteins albumin, ovalbumin, chymotrypsinogen, and ribonuclease A with molecular masses of 67, 43, 25, and 13.7 kDa). C, effect of LprI on M. lysodeikticus cells. An agar-based diffusion assay shows that, in the presence of LprI, the lytic activity of HEWL is inhibited. Two wells were punched in a nutrient agar plate supplemented with M. lysodeikticus cells. In the first well, 50 g of HEWL was added, and in the second well, equal amounts (50 g) of HEWL and LprI⌬ss were added, and the plate was allowed to incubate overnight at 30°C. The distinct zone of lysis visible in the periphery of the well containing HEWL alone is absent when LprI⌬ss is added along with HEWL. D, LprI-HEWL interaction probed through intrinsic fluorescence quenching. The fluorescence emission spectra of 500 nM HEWL (dashed) and 500 nM recombinant LprI (solid curve) were obtained separately. The dashed-dotted curve shows the experimental spectrum obtained for a 1:1 mixture of HEWL and LprI (total protein concentration, 1 M). The dotted curve is the expected emission spectrum for the mixture obtained by adding the spectra of the individual proteins. The difference between the dashed-dotted and dotted curves is attributed to the interaction between the two proteins. mAU, milliabsorbance units; arb., arbitrary.

LprI, a Novel Lysozyme-binding Lipoprotein
observed, and the magnitude of loss of mass appeared to make it equivalent to that of unglycosylated LprI produced in E. coli.
Glycosylation Modulates Cell Surface Association of LprI-Lipoproteins are usually localized at the cell membrane, anchored through their invariant cysteine residue, or appear at the cell surface where they are engaged in a diverse array of functions (4,12). Therefore, we checked cellular localization of the LprI in M. tuberculosis and M. smegmatis after cell fractionation and immunoblotting. LprI was detected only in the membrane and the cell wall fractions of M. tuberculosis and M. smegmatis and remained absent in the cytoplasmic fractions of both strains (Fig. 5D). The localization of LprI was further analyzed through Triton X-114 phase partitioning, which is an established tool to verify lipoproteins. Here, LprI was detected only in the detergent phase, indicating its selective partitioning with this phase and not with the aqueous phase (Fig. 5E), which suggests its association with the membrane lipids. To investigate the biological significance of protein glycosylation in subcellular localization of proteins (45) and the appearance of LprI on the cell wall in addition to its membrane association, we checked whether glycosylation has any role in surface association of LprI in M. tuberculosis. Interestingly, in the modeled structure of LprI, glycosylation sites Thr-24, Thr-28, and Thr-117 are arranged on the protein surface in the same line, and among these, the Thr-24 and Thr-28 sites are close to the protein anchorage site (Cys-16) that might contribute to its surface exposure (Fig. 5F). Therefore we checked the cellular localization of wild-type and partially deglycosylated mutants of LprI altered at Thr-24, Thr-28, and Thr-117 residues. Our results indicated that the LprI mutants are mainly localized on the cell membrane, showing a weak signal on the cell wall (results not shown). To rule out the possibility of background signal for LprI due to membrane contamination with the cell wall fractions, we used a trypsin sensitivity test, which is widely used to check the cell surface association of proteins (11,34). Trypsin-digested whole cells of M. smegmatis expressing wild-type LprI displayed complete digestion of the protein and did not show any signal after Western blotting. In comparison, all three glycosylation-deficient mutants of LprI exhibited retention of LprI inside the cells as observed in the case of the control cytoplasmic protein, green fluorescent protein (GFP), indicating defective exposure of LprI mutants on the cell surface (Fig. 5G). These results indicated that glycosylation of LprI at three surface-exposed Thr residues may be contributing to its cell surface association.
LprI Confers Protection from Lysozyme-mediated Growth Inhibition-We further checked whether LprI is able to provide any growth protection to mycobacteria from the hydrolytic activities of lysozyme by monitoring the influence of lysozyme on growth properties of LprI-overexpressing M. tuberculosis and M. smegmatis. A significant increase in the tolerance of both strains of mycobacteria toward lysozyme was observed in the presence of LprI. M. smegmatis expressing LprI had a 6-fold higher resistance against lysozyme (500 g/ml HEWL) than isogenic control cells (75 g/ml HEWL) (Fig. 6A, upper panel). Further evaluation of its protective efficacy against HEWL was tested in its native host, M. tuberculosis (H37Rv) after exposing it to increasing concentrations of HEWL. The pattern observed in Fig. 6A (lower panel) indicates that, whereas the survival of wild-type cells decreases continually with the increase in HEWL concentration, the presence of LprI sustains the growth of M. tuberculosis even at 500 g/ml HEWL. These results were further verified by resazurin microtiter plate assay, which has been used successfully to detect the susceptibility of mycobacterial strains against various antibiotics (46). From Fig. 6B  (upper panel), it is evident that wild-type M. smegmatis can resist up to 12.5 g/ml HEWL only, whereas the tolerance of lprI-expressing M. smegmatis is elevated up to 25 g/ml. In comparison, native M. tuberculosis survived up to 62.5 g/ml HEWL, which is relatively higher than the tolerance level of M. smegmatis. The corresponding LprI-overexpressing strains of M. tuberculosis continued to survive until 500 g/ml HEWL (Fig. 6B, lower panel). These complementary approaches con-  FEBRUARY 5, 2016 • VOLUME 291 • NUMBER 6 firmed that LprI confers enhanced resistance to its host against the hydrolytic activities of lysozyme.

LprI, a Novel Lysozyme-binding Lipoprotein
LprI Facilitates Phagocytosis and Cell Survival of M. smegmatis during Intracellular Infection-Our repeated attempt to generate an LprI knock-out mutant of M. tuberculosis remained unsuccessful. Therefore, to understand the physiological function of LprI, we utilized M. smegmatis as an alternative model primarily due to the fact that it lacks lprI gene and that it has been used as an alternative host to understand the function(s) of several M. tuberculosis proteins (47,48). Because high concentrations of lysozyme are secreted as a part of the host defense system in various organs including mucosal surfaces and macrophages (49,50) and recombinant LprI effectively relieved the toxicity of lysozyme in vitro, we checked the implications of LprI during intracellular infection of primary cell lines, viz. MDMs and peritoneal macrophages (Fig. 7, A and  B), that are known to secrete significant levels of lysozyme (51,52). Enhanced phagocytosis and subsequently a significant increase in the survival postinfection were observed in lprIexpressing cells in comparison with wild-type cells in both the cell lines used. At the zero time point, nearly 8-and 10-fold increases in cfu of lprI-expressing M. smegmatis occurred during infection of MDMs and peritoneal macrophages, respectively. Furthermore, the bacterial replication inside the macrophages again showed an explicit increase up to 2-6-fold for lprI-expressing M. smegmatis relative to its isogenic control cells, thereby suggesting vital implications of LprI in the pathogenesis of tuberculosis when levels of lysozyme are up-regulated after bacillary infection (17).
LprI Is an Operon Partner of the HbN in M. tuberculosis-Because the LprI-and HbN-encoding genes share the same operon in M. tuberculosis, we checked their expression level in aerobically growing cells. Expression of LprI in M. tuberculosis remained low at early exponential phase but increased significantly at stationary phase. Although the expression of the HbN appeared relatively low during early growth phase, it followed a similar pattern, showing a significant increase during stationary phase (Fig. 8A). Interestingly, RT-PCR analysis revealed that the transcript levels of both lprI and glbN are elevated synchronously during macrophage infection with the passage of time, exhibiting a nearly 50-fold increase by 72 h after infection of  FEBRUARY 5, 2016 • VOLUME 291 • NUMBER 6 JOURNAL OF BIOLOGICAL CHEMISTRY 2947 mouse peritoneal macrophages by M. tuberculosis (Fig. 8B). The transcripts of both glbN and lprI were normalized to that of 16S RNA from the same samples and expressed as -fold change (Fig. 8B). The synchronized increment in the expression levels of the glbN and lprI genes indicated the requirement of both proteins, LprI and the HbN, in the tubercle bacilli during infection.

LprI, a Novel Lysozyme-binding Lipoprotein
LprI-HbN Pair Alleviates Lysozyme Stress during Macrophage Infection-Due to the absence of LprI and the presence of HbN, which lacks NO dioxygenase activity and glycosylation, M. smegmatis appeared to be a suitable knock-out system for LprI and the LprI-HbN pair. Therefore, we checked the implications of these two proteins in intracellular infection and survival of M. smegmatis after expressing the lprI and glbN genes  individually and as a glbN-lprI gene pair that mimics the physiological scenario present in M. tuberculosis. Co-expression of LprI and HbN in M. smegmatis was validated via Western blotting (Fig. 8C). These strains were used to study the implications of LprI and the HbN alone and in combination in intracellular survival of M. smegmatis using two different cell lines, RAW 264.7 and THP-1 macrophages. The viable cell counts, recovered at different time intervals postinfection in the THP-1 cell line, are shown in Fig. 8D. Although the expression of the glbN gene alone and in combination with lprI exhibited enhanced phagocytosis as shown by the increase in the bacterial counts at 0 h of infection of the cells and a substantial increase in the cell survival of M. smegmatis, the expression of lprI alone in M. smegmatis did not show any distinct effect on the cfu counts in comparison with isogenic control cells. Similar results were obtained using the RAW 264.7 cell line (data not shown). The protective effect of LprI was not visible, possibly due to a very low level of lysozyme present in the laboratory-grown cell lines; however, these observations re-established our previous finding (11) that the presence of HbN causes a significant increase in bacterial infectivity and cell survival. The presence of LprI and the HbN as a single unit also behaved in a similar manner; nevertheless, it can be deduced that in this case the effect may be directly related to the presence of the HbN.
To verify that the absence of any beneficial effect of LprI on the survival of M. smegmatis is due to a low concentration of lysozyme in secondary cell lines, we performed the intracellular survival assay in the presence of exogenously supplemented lysozyme (100 g/5 ϫ 10 5 cells). Viable cell counts were determined at time points appropriate to assess the infectivity and survival of the M. smegmatis wild type and recombinants (Fig.  8E). In concord with our expectation, lysozyme led to a rapid and significant killing of wild-type cells in comparison with LprI-overexpressing M. smegmatis cells, which were able to infect and thrive uninhibited during macrophage infection by virtue of their lysozyme inhibitory activity. In contrast, viable cell counts for the HbN-carrying M. smegmatis dropped significantly in the presence of lysozyme and were comparable with that of wild-type control cells, exhibiting similar infectivity and cell survival. Interestingly, the presence of LprI and HbN together in M. smegmatis conferred a much enhanced increase in macrophage infectivity and cell survival in the presence of lysozyme with respect to cells carrying LprI alone. The beneficial response of the LprI-HbN unit during macrophage infection in the presence of lysozyme thus clearly demonstrates the physiological relevance of the co-existence of LprI and HbN in M. tuberculosis and their role during intracellular infection and pathogenesis.
Proposed Model for the Physiological Function of LprI-HbN Pair-Based on our experimental results, a model for the function of LprI and HbN has been proposed ( Fig. 9) that may provide mechanistic insight into the functional correlation of their role in pathogenic mycobacteria. Both LprI and HbN are expressed together in M. tuberculosis as surface-exposed glycosylated proteins. Because HbN-expressing cells exhibit increased phagocytosis/invasion into the macrophages and might be able to adhere better on the macrophage surface due to the presence of mannose-linked HbN on the cell surface (11), these cells may also need protection from the lysozyme secreted by the macrophages on the cell surface. The presence of LprI and HbN together on the cell surface of M. tuberculosis thus may provide a specific advantage by protecting macrophageadhered M. tuberculosis from the lytic activity of the macrophage-secreted lysozyme and may facilitate phagocytosis. Such a mechanism thus justifies the presence of LprI and HbN as a unit in pathogenic mycobacteria and their absence in nonpathogenic mycobacteria.

Discussion
This study demonstrates that LprI is a novel lysozyme-binding lipoprotein of M. tuberculosis that confers protection to its host from the hydrolyzing activity of lysozyme. Because lysozyme is produced abundantly in epithelial secretions and lysosomal compartments of phagocytic cells (17,50), the cell surface association and lysozyme binding properties of LprI may be crucial for maintaining the structural integrity of M. tuberculosis during intracellular infection. Lysozyme resistance may be especially important for pathogenic mycobacteria as they need to protect the thick peptidoglycan layer of their cell wall from the hydrolyzing activities of lysozyme produced by their host during microbial infection. The exclusive presence of LprI in members of M. tuberculosis complex suggests its functional relevance for pathogenesis.
The efficacy of LprI against lysozyme was substantiated in vitro through a fluorescence-based assay wherein purified LprI rendered significant diminution on the rate of lysozyme-mediated lysis of fluorescein-conjugated strain of M. lysodeikticus. Furthermore, LprI displayed a strong binding affinity toward HEWL with a dissociation constant of ϳ2 nM. The affinity of this binding is comparable with that observed for other complexes of lysozyme inhibitors with lysozyme (53, 54) Moreover, the LprI expression in M. smegmatis imparted a significant growth advantage over the isogenic control cells in the presence of inhibitory levels of lysozyme. Our results demonstrated that LprI abrogates the hydrolytic activities of lysozyme via proteinprotein interactions as supported by physical interactions of LprI with lysozyme in vitro and the structural model of LprI in complex with HEWL exhibiting orientation of loop regions of its MliC motif toward the active site cleft of HEWL. These results substantiated that LprI is functional as a lysozyme inhibitor under in vitro and in vivo conditions. A remarkable difference in the tolerance level of M. tuberculosis and M. smegmatis toward lysozyme exists as M. tuberculosis is able to tolerate nearly 3-fold higher levels of lysozyme. Expression of the lprI gene in M. smegmatis raises its tolerance level toward lysozyme more or less similar to that of M. tuberculosis, indicating that the lack of lprI gene might be the reason for its higher lysozyme sensitivity. Elevated expression of lprI gene in M. tuberculosis also enhanced its tolerance level toward HEWL, suggesting that the increase in LprI content may allow M. tuberculosis to cope with higher levels of lysozyme secretions in certain cellular compartments like neutrophils and phagocytic and epithelial cells (17,55). This is in line with our observations that transcriptional activity of the lprI gene is elevated in M. tuberculosis up to 50-fold within 72 h of macrophage infection. Increased biosynthesis of LprI has also been observed in M. tuberculosis during mouse lung infection (56). A severalfold increase in phagocytosis and survival of LprI-expressing M. smegmatis during infection of peritoneal macrophages and MDMs reinforced the significance of LprI in pathogenesis. The first step during macrophage infection is the adhesion of bacteria on the macrophage surface and then invasion. Because LprI-expressing M. smegmatis cells can abrogate the lytic activity of macrophage-secreted lysozyme on the surface, they are able to survive better during adhesion and are taken up better during phagocytosis in comparison with the control cells. The protective effect of LprI was found to be lysozyme-specific as phagocytosis of LprI-expressing M. smegmatis was not affected during infection of laboratory-grown THP-1 and RAW264.7 cell lines where lysozyme secretion is very low and ranges from 0.01 to 28 g/10 6 cells (57, 58), whereas the physiological levels of lysozyme secreted by phagocytic cells and neutrophils are high, ranging from 50 to 500 g/ml (59). Addition of lysozyme exogenously during infection of secondary cell lines THP-1 and RAW264.7 in a concentration equivalent to that of the physiological environment validated this reasoning where a beneficial effect of LprI was apparent in enhancing the infectivity and survival in contrast to the wild-type M. smegmatis cells that were rapidly killed due to lysozyme-mediated toxicity. From these results, it can be inferred that the LprI-expressing cells are protected from the macrophage-secreted lysozyme during invasion and are internalized better than the control cells. The observed role of LprI in abrogating the lytic activity of lysozyme may have crucial implications inside the human host where infection sites are rich in lysozyme, especially due to the fact that up-regulation of lysozyme expression occurs during pulmonary diseases (16,60).
LprI appears to be a highly glycosylated lipoprotein, carrying mannose as a glycan linkage, and remains localized on the cell membrane and the cell surface of M. tuberculosis. Thus, it can easily interact with the lysozyme secreted by the host defense machinery. Contrary to our expectation, mutations at three identified glycosylation sites (Thr-24, Thr-28, and Thr-117) of LprI did not fully restore the size of unglycosylated recombinant LprI expressed in E. coli, suggesting the presence of additional sites for the glycan linkage on LprI. The role of glycosylation appears to be distinct from its lysozyme inhibitor function as recombinant LprI produced by E. coli directly interacts with lysozyme in 1:1 stoichiometry and efficiently abrogates its hydrolytic activity. It is thus likely that the post-translational modification of LprI by mannose linkage in M. tuberculosis is required for other functions. Site-directed mutagenesis at the three glycosylation sites of LprI led to a drastic reduction in the In the absence of LprI, lysozyme is able to disrupt the integrity of the bacillary cell wall, leading to its killing and rapid clearance (lower side). The presence of LprI allows the bacilli to abrogate the hydrolytic activity of lysozyme and invade its host successfully. B, M. tuberculosis represents an analogous situation where the infection of host macrophages by M. tuberculosis is established with ease. LprI, expressed as a cell surface protein, binds and inhibits lysozyme activity threatening the integrity of bacilli. The bacilli are thus protected from lysozyme action and are now able to invade the macrophages where the invasion is facilitated by the presence glycosylated HbN, which also protects the bacilli from the macrophage-generated toxic NO. Both proteins remain surface-exposed and together constitute a well evolved arsenal to disarm the two important host defenses, NO and lysozyme.
surface exposure of LprI in M. tuberculosis, indicating their contribution in escorting the protein on the cell surface as has been observed in some membrane-anchored proteins where glycan linkage contributes to apical delivery or surface transport of proteins (61,62). Because Thr-24 and Thr-28 are located very close to the membrane anchorage site (Cys-16), oriented outward in the same plane along with Thr-117 in the modeled structure of LprI, glycan linkage at these sites may modulate the surface localization and positioning of LprI that may be important for its function during M. tuberculosis infection. However, further studies on glycosylation of LprI and its role in protein function are required.
LprI coexists as an operon partner of HbN in M. tuberculosis, and co-occurrence of these two proteins is preserved only in related pathogenic mycobacteria. It is surprising that the biosynthesis of these two functionally unrelated proteins is co-regulated in M. tuberculosis. A synchronized up-regulation of lprI and glbN genes, observed during macrophage infection, supports the assumption that both the proteins are required during infection and may be vital for virulence. M. smegmatis provided an excellent model system to understand the functional relevance of these two proteins due to the fact that it lacks the combination of LprI-HbN and is more sensitive to lysozyme than M. tuberculosis. Additionally, HbN produced by M. smegmatis is not functional as an NO scavenger and lacks posttranslational modifications identified in the HbN of M. tuberculosis (10). HbN is glycosylated at its C terminus via its association with mannosyl residues that possibly facilitate phagocytosis of M. tuberculosis (11). Interestingly, phagocytosis of M. smegmatis expressing the HbN-LprI pair was much higher than that of the isogenic cells expressing HbN or LprI alone, and these cells survived better during intracellular infection. It is likely that co-occurrence of LprI with HbN at the cell surface is required to protect the bacillus from the macrophagesecreted lysozyme. This is supported by the observation that the macrophage infectivity of M. smegmatis cells carrying the HbN alone is reduced drastically in the presence of lysozyme, but the presence of the LprI-HbN pair confers resistance to these cells against lysozyme toxicity and facilitates their infectivity and survival, suggesting the importance of LprI in pairing with HbN as a protective partner. Because HbN-carrying cells are able to infect better only in the absence of lysozyme but not in the presence of lysozyme, we can hypothesize that the co-expression of LprI and HbN together might be functionally relevant during infection as these cells are protected from the lysozyme on the macrophage surface and internalized better. The presence of HbN then protects these cells from macrophage-produced reactive nitrogen due to its potent nitric oxide dioxygenase activity (8,9).
Taken together, the present study provides the first report on a novel lysozyme-binding lipoprotein of M. tuberculosis and suggests that, besides protective roles of LprI and HbN in detoxification of lysozyme and NO, respectively, their combination may particularly be crucial in protecting the structural integrity of the cells during phagocytosis, which increases in HbN-expressing cells of M. tuberculosis. Because NO and lysozyme are two major host defense factors against microbial infection, a detailed understanding of the functional coordina-tion of the LprI-HbN pair may allow development of new antimycobacterial strategies for therapeutic intervention.