Mycobacterial PknG Targets the Rab7l1 Signaling Pathway To Inhibit Phagosome–Lysosome Fusion

Mycobacterium tuberculosis is an intracellular pathogen that infects alveolar macrophages through aerosols (1). Soon after entry of M. tuberculosis through phagocytosis, during early-to-late endosomal transition, maturation of phagosomes containing M. tuberculosis are arrested, and the phagosomes fail to fuse with lysosomes (2–4). Inhibition of phagosome–lysosome (P–L) fusion is critical for successful M. tuberculosis infection and its subsequent intracellular survival (5). It has been noted that the bacilli either inhibit phagosomal maturation during early-to-late endosomal transition (5–8) or have developed mechanisms to reside inside the acidified lysosomes (9).

The mycobacterial protein kinase G (PknG), a serine/threonine kinase (10), has significant structural similarity with eukaryotic serine/threonine kinases, mainly the PKCα (11). PknG is known to be secreted and released into the cytoplasm of bacteria-infected macrophages and is found to be involved in the inhibition of P–L fusion (12, 13). Several reports have shown that PknG confers a survival advantage to mycobacteria and infection of macrophages with M. tuberculosis and M. bovis Bacillus Calmette-Guerin (BCG) mutants lacking PknG, (M. tuberculosis∆PknG and M. bovis BCG∆PknG) that allows maturation of mycobacterial phagosomes, leading to degradation of the infecting bacilli (11–14). In contrast, expression of M. tuberculosis PknG in M. smegmatis, which lacks PknG expression, resulted in inhibition of P–L fusion and enhanced survival of bacilli inside the host macrophages (12, 15). Interestingly, the kinase activity of PknG was found to be critical for inhibition of P–L fusion (12, 13, 15, 16). However, the exact mechanism by which PknG inhibits P–L fusion during mycobacterial infection remains poorly understood. It is possible that PknG directly interacts with key host proteins involved in the phagosomal maturation process and inhibits P–L fusion. To validate our hypothesis, in the current study, we carried out protein–protein interaction studies using a yeast two-hybrid (Y2H) system in which a leukocyte library was screened using M. tuberculosis PknG as bait, and we were able to confirm that a host RabGTPase protein Rab7l1 (Rab29 in mouse) is an interacting partner of PknG. Further, we were able to demonstrate a role of Rab7l1 in the maturation process of phagosomes and found that PknG manipulates the Rab7l1 signaling pathway to block P–L fusion. To the best of our knowledge, for the first time, we demonstrate a role of Rab7l1 in phagosomal maturation that is hijacked by PknG to favor intracellular survival of the bacilli. This information may be helpful in understanding host–pathogen interaction, the mycobacterial virulence mechanism, and in designing novel therapeutics against mycobacteria.

Details of cloning vectors, primers, and Abs used in this study were described in Supplemental Table I.

Y2H screening was carried out as described earlier (17). Briefly, the PknG gene was cloned into Y2H vector pGBKT7 in frame with GAL4-BD, whereas the leukocyte cDNA library, obtained from Takara Clontech (Clontech Laboratories), was cloned into pACT2 vector in frame with GAL4-AD. The cloned prey library and the PknG bait plasmids were then transformed into Y187 (Mat⁻) and AH109 (Mat⁺) yeast strains, respectively. The transformed AH109 and Y187 yeast strains were mixed and incubated together for mating and then plated on quadruple drop out (QDO) plates (SD/-Trp, -Leu, -Ade, -His) for selection. Putative positive colonies were then substreaked onto QDO plates five times to dilute the nonspecific plasmids, and the prey plasmids were isolated from the positive colonies. Prey genes were then amplified by PCR using GAL4-AD–specific primers containing cDNA insert. The sequences of the inserts were determined by sequencing in an automated DNA sequencer, and the sequences were identified by online PSI-BLASTX searches available at the National Center for Biotechnology Information (NCBI) Web site. One-to-one interaction studies in Y2H were carried out by cotransformation of pGADT7-PknG bait plasmid with various putative genes cloned in pGBKT7 vector. Transformants were first selected in SD/-Trp and -Leu plates and further streaked onto QDO plates to examine Ade and His reporter gene activation.

For β-galactosidase (β-gal) assay (Roche Diagnostics), cell lysates (50 μg) were added to a 96-well plate coated with anti–β-Gal Ab and incubated for 2 h. The plate was washed and incubated with anti–β-gal-digoxigenin Ab for 1 h. After washing the plates, anti–digoxigenin-peroxidase was added and further incubated for 1 h at 37°C. The plates were again washed and followed by an addition of 200 μl of ABTS (diammonium salt) and incubated at 25°C for 30 min or until the color was developed. Absorbance was measured at 405 nm using a microplate reader.

Kinase-defective PknG (PknG-K181M), Rab7l1Q67L (Rab7l1-GTP), and Rab7l1T21N (Rab7l1-GDP) were generated by site-directed mutagenesis using PCR-driven overlap extension methods. Internal primers to introduce the mutation and flanking primers with specific restriction enzyme sites were used to amplify the gene with specific mutations (Supplemental Table I) that were cloned in Y2H vectors (Supplemental Table I) for Y2H one-to-one screening, in a mammalian expression vector for pull-down experiments, or in bacterial expression vectors for purification of recombinant protein.

Human monocyte/macrophage cell line THP-1 and human embryonic kidney cell line HEK-293 cells were obtained from the National Centre for Cell Science (NCCS) (Pune, India). The suspension cell line THP-1 was maintained in RPMI 1640 medium (Hyclone, UT) supplemented with 10% (v/v) heat-inactivated FBS (Hyclone), antibiotic-antimycotic (1×, containing Penicillin G, Streptomycin, and Amphotericin B), 2 mM l-glutamine, and 10 mM HEPES (all from Life Technologies, Carlsbad, CA) and maintained in a humidified incubator at 37°C with 5% CO₂ in air. The HEK-293 cells were maintained in DMEM supplemented with 10% FBS, 1× antibiotic-antimycotic, 2 mM l-glutamine, and 10 mM HEPES and maintained in a humidified incubator at 37°C with 5% CO₂ in air. THP-1 cells were differentiated into macrophages by incubating cells with 5 ng/ml PMA (Sigma-Aldrich, St. Louis, CA) for 12 h, followed by a resting period of 24 h.

Rab7l1 knock-down (Rab7l1-KD) stable cells in THP-1 monocytes were generated using lentiviruses carrying short hairpin RNAs (shRNAs) specific to Rab7l1 obtained from Sigma-Aldrich (Sequence 1: 5′-TTCCGGCCACAGAAGATATCATGTCTTCTCGAGAAGACATGATATCTTCTGTGGTTTG-3′; Sequence 2: 5′-CCGGCGGTTCAGTAAAGAGAACGGTCTCGAGACCGTTCTCTTTACTGAACCGTTTTTG-3′; and Sequence 3: 5′-CCGGAGTTAGTGCCTTCGGTGTAAGCTCGAGCTTACACCGAAGGCACTAACTTTTTTG-3′). As a control, nontargeted scrambled shRNA (SHC016V-1EA) was used (Sigma-Aldrich). The shRNAs were cloned in PLKO.1 vector carrying a puromycin-resistant gene. THP-1 cells were infected with the lentivirus particles carrying Rab7l1-specific shRNA and nontargeted shRNA and the recombinant viruses had titers around 1 × 10⁶ PFU/ml. About 0.1 million cells were infected with 100 μl of virus at one multiplicity of infection (MOI) in 12-well plates for 48 h and were selected with 2 μg/ml puromycin. Single-cell originated lines were checked for expression levels of Rab7l1 by semiquantitative RT-PCR. The cells were maintained with additional 2 μg/ml puromycin but kept without antibiotic during infection with M. smegmatis and M. bovis BCG.

PknG and its mutants were cloned in pCDNA3.1c with an N-terminal 3×-FLAG tag. Rab7l1 was cloned with an N-terminal GFP tag in pEGFP-C1 vector. PknG and its various domains were transfected in HEK-293 cells to check their interactions with endogenously expressing Rab7l1 by coimmunoprecipitation (Co-IP) studies and also to examine the colocalization of PknG and its various domains with trans-Golgi marker trans-Golgi network protein 2 (TGOLN2). For GTPase assay and GTP binding assay, GFP-Rab7l1 alone or along with 3×-FLAG PknG or PknG-K181M was transfected to HEK-293 cells. Briefly, HEK-293 cells were seeded in DMEM with 10% FBS, HEPES, and glutamine and grown overnight. Next day, respective plasmids and Lipofectamine-2000 plasmid transfection reagent (Invitrogen) were diluted in Opti-MEM and then mixed gently to prepare 1 μg of plasmid and 3 μl of Lipofectamine-2000 mix. The mixture was then incubated at room temperature for 45 min and then added to HEK-293 cells kept in Opti-MEM medium. Transfected cells were incubated at 37°C for 6 h in DMEM complete medium. All downstream experiments were performed after 24 h of transfection.

GarA, PknG-WT, and PknG-K181M cloned in pQE-2 were kind gifts from Dr. V. Nandikoori, National Institute of Immunology (New Delhi, India). These recombinant plasmids were transformed into Escherichia coli DH5α strain whereas the pGEX-4T-1-Rab7l1 was transformed into E. coli BL21 (DE3) strain. An overnight-grown primary culture was used to inoculate 250 ml super optimal broth medium for pQE-2 constructs, and terrific broth medium was used for pGEX-4T-1 constructs. The transformed cells were grown at 37°C and 250 rpm till the absorbance at 600 nm reached 0.6. The cultures were then shifted to 22°C for 30 min, and induction of expression of the genes was carried out by addition of isopropyl-β-d-galactopyranoside (IPTG) (Sigma-Aldrich). For pQE-2-PknG-WT and pQE-2-PknG-K181M constructs, 0.1 mM IPTG was used to induce expression of the recombinant proteins and was further incubated at 22°C for 12–16 h. For pQE-2-GarA and pGEX-4T-1-Rab7l1, 1 mM IPTG was used for induction followed by incubation at 37°C for 5 h. For purification of His-tagged proteins, cells were harvested by centrifugation at 5000 rpm for 10 min, and the pellet was washed with PBS followed by resuspension in sonication buffer [20 mM Sodium phosphate (pH 7.4), 50 mM NaCl, 5 mM imidazole, and 1 mM 2-ME]. The cells were lysed by sonication on ice with five to eight pulses for 1 min. The lysate was centrifuged at 15,000 rpm, and the supernatant was added to pre-equilibrated Ni-agarose beads (Qiagen, Germany) followed by gentle agitation at 4°C for 4 h. The protein-bound beads were loaded onto an empty column and washed with 10 column volume of wash buffer (20 mM Sodium phosphate, 200 mM NaCl, and 10 mM imidazole), and the bound protein was eluted in five column volume elution buffer (20 mM Sodium phosphate, 50 mM NaCl, and 150 mM imidazole). Fractions were collected, and aliquots were separated on a 12% SDS-PAGE. The proteins were visualized by Coomassie Brilliant Blue staining, and the fractions containing protein were pooled and dialyzed overnight against dialysis buffer [10 mM Tris-HCl (pH 7.5), 20 mM NaCl, and 20% glycerol]. For purifying GST-Rab7l1 protein, cells were harvested by centrifugation at 10,000 × g for 10 min and resuspended in sonication buffer [50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1 mM EDTA, 10% glycerol, and protease inhibitors]. Cells were sonicated on ice with five to eight pulses for 1 min, and Triton X-100 was added to a final concentration of 0.1%. The cell lysate was centrifuged at 15,000 rpm, and the clear cell supernatant was incubated with glutathione Sepharose 4B beads (GE Healthcare) for 3 h at 4°C. Proteins bound to beads were loaded onto a column and washed with five column volumes of PBS and eluted with 20 mM glutathione in Tris-HCl (pH 8). Fractions containing proteins were pooled and dialyzed in PBS containing 20% glycerol. All the purified proteins were snap frozen in liquid nitrogen and stored in −70°C till further use.

M. smegmatis bacteria were transformed with pVV16 vector (Msmeg-pVV16), pVV16-PknG-WT (Msmeg-PknG-WT), or pVV16-PknG-K181M (Msmeg-PknG-K181M) following the protocol as described earlier. The transformants were maintained in 7H9 broth with antibiotic selection of kanamycin (50 μg/ml) and hygromycin (50 μg/ml) and were grown till the absorbance (A₆₆₀) reached 0.5 to 0.6. Expression of PknG was confirmed by immunoblotting using rabbit anti-PknG Ab (1:1000 dilution) and anti-rabbit Ab coupled to HRP conjugate (Sigma-Aldrich). M. bovis BCG and M. bovis BCG∆PknG were also grown in a similar way in the presence of kanamycin (50 μg/ml). These strains were kind gifts from Dr. J. Pieters (University of Basel, Switzerland). PMA-differentiated THP-1 macrophages were used for infection with Msmeg-pVV16, Msmeg-PknG-WT, Msmeg-PknG-K181M, M. bovis BCG, or M. bovis BCG∆PknG. Before infection, single-cell suspension of the bacterial cultures was prepared by passing the cultures through a 22-gauge syringe three to four times. Number of bacteria in the culture was counted by measuring the absorbance at 640 nm in which absorbance of 0.3 was considered to be equivalent of ∼7 × 10⁷ CFU/ml. MOI was calculated as the ratio of the number of bacteria used for infection to the total number of cells. Cells were synchronized with bacteria by incubation at 4°C for 1 h followed by incubation at 37°C for infection. The infected cells were harvested at various time points for different experiments as indicated in the figures and legends.

PMA-differentiated THP-1 macrophages were infected with GFP-Msmeg-pVV16, GFP-Msmeg-PknG-WT, or GFP-Msmeg-PknG-K181M at 10 MOI for 1 h at 4°C and then incubated for another 1 h at 37°C followed by addition of Lysotracker Red DND-99 (Invitrogen), a lysosomal marker (100 nM), for 15 min. The slides were mounted using VECTASHIELD Mounting Medium (Vector Laboratories, Burlingame, UK) containing DAPI as a nuclear staining marker. PknG and its various mutants were transiently cotransfected with mCherry TGOLN2 plasmid into HEK-293 cells and were examined for their colocalization using confocal microscopy. After 24 h of transfection, the cells were fixed with 0.4% formaldehyde solution for 15 min and then permeabilized with 0.1% Triton X-100. The cells were washed and kept in blocking solution (5% BSA) for 1 h. Next, the cells were incubated with mouse anti-FLAG Ab (1:500 dilution) for 1 h at room temperature and washed extensively with PBS three to five times. Alexa Fluor 488–conjugated goat anti-mouse (green) secondary Ab was used to probe the specific primary (Anti-FLAG) Ab. Colocalization of FLAG-expressing PknG and its truncated mutants with mCherry TGOLN2 was monitored by confocal microscopy. Colocalization of GFP–M. smegmatis harboring either empty vector pVV16 (GFP-Msmeg-pVV16), pVV16-PknG-WT (GFP-Msmeg-PknG-WT), pVV16-PknG-K181M (GFP-Msmeg-PknG-K181M) with Rab7l1, or GFP-Msmeg-pVV16 with Rab5; early endosome autoantigen 1 (EEA1); Rab7; and LAMP2 was monitored by confocal microscopy using appropriate combinations of primary and secondary Abs. All confocal imaging was carried out using the LSM700 confocal microscope, and images were captured at an original magnification of ×64. Randomly selected fields were chosen for colocalization analyses using either LysoTracker Red or various P–L markers stained by Alexa Fluor 594 with GFP-expressing bacteria. Colocalization analyses were performed on a pixel by pixel basis using the AIM/ZEN software supplied along with the LSM700 confocal microscope setup following the methodologies described on the following Web site: https://www.zeiss.com/content/dam/Microscopy/Downloads/Pdf/FAQs/zen-aim_colocalization.pdf.

Percentage of bacterial colocalization was calculated from various fields using the following formula: % of bacterial colocalization = number of bacteria colocalized × 100 / total number of bacteria observed.

HEK-293 cells were transiently transfected with either 3×-FLAG–tagged PknG-WT or its truncated mutants, and at 24 h posttransfection, cells were harvested and lysed with lysis buffer [PBS (pH 7.4), 0.1% NP-40] and protease inhibitor mixture (Roche, Switzerland). About 1000 μg of the lysate was then mixed with 10 μg of anti-FLAG Ab and incubated overnight at 4°C on a rotor. To capture the immune complexes, Sepharose A/G beads (25 μl) were added and incubated further for 2 h at 4°C. The immune complexes were then collected by centrifugation, washed three times with PBS, and boiled in 6× SDS sample buffer. The eluted and denatured immune complexes were separated on a 12% SDS-PAGE and transferred to a nitrocellulose membrane. The membrane was blocked for 1 h at 37°C in 5% fat-free milk in PBS and incubated overnight at 4°C with rabbit anti-Rab7l1 Ab (1:200 dilution). The membrane was next washed and incubated with anti-rabbit IgG coupled to HRP (Sigma-Aldrich). Bands were detected by chemiluminescence following the manufacturer’s protocol (GE Healthcare, Little Chalfont, U.K.).

GST-Rab7l1, His-PknG-WT, or His-PknG-K181M was purified using either glutathione beads for GST-tagged proteins or Ni-NTA beads for His-tagged proteins. His-PknG-WT– or His-PknG-K181M–bound Ni-NTA beads were incubated with GST-Rab7l1–expressing E. coli BL21 (DE3) strain lysate at 4°C for 4 h. These beads were washed with lysis buffer three to five times in a rotating shaker at 4°C for 10 min. Pull-down products in both the cases were eluted, boiled with 6× SDS dye, and run on a 12% SDS-PAGE for immunoblotting using rabbit anti-Rab7l1 Ab (1:200 dilution) and anti-rabbit HRP conjugate.

In vitro kinase assays were performed in a 40-μl reaction containing 25 mM HEPES-NaOH (pH 7.4), 20 mM Mg-acetate, 20 μM MnCl₂, 1 mM DTT, 200 μM Sodium orthovanadate, 100 μM cold ATP, and 10 μCi (γ-[³²P]) ATP. To this reaction buffer, 5 pmol of recombinantly purified His-GarA was incubated along with 1 pmol of His-PknG-WT or His-PknG-K181M for 30 min at 30°C. The reactions were stopped by adding 15 μl of 6× SDS sample buffer and denatured by heating at 95°C for 15 min. Reactions were resolved on 12% SDS-PAGE. The gel was dried at 80°C for 1 h, and bands were visualized using a phosphorimager (Typhoon FLA 9500).

GTPase activity of Rab7l1 was measured by using GTPase assay kit as per the manufacturer’s instruction (Innova Biosciences, U.K.). Along with pEGFP-Rab7l1, 3×-FLAG–tagged PknG-WT and PknG-K181M were transiently transfected in HEK-293 cells, and Rab7l1 was immunoprecipitated by using anti-GFP Ab. GTPase activity of Rab7l1 was measured by using GTPase assay kit as per the manufacturer’s instruction (Innova Biosciences). Briefly, 1000 μg of cell lysate was mixed with 10 μg of anti-GFP Ab and incubated overnight at 4°C followed by addition of Sepharose A/G beads and incubated for another 2 h with gentle agitation. The immune complex was precipitated by centrifugation and washed. Next, 100 μl of substrate buffer along with 10 μl phosphate-free GTP was incubated with the immunoprecipitated GFP-Rab7l1 for 15 to 30 min at room temperature, and the reaction was stopped by adding 50 μl of PiColorLock mix. After 2 min, 20 μl of stabilizer was added to the reaction, mixed thoroughly, and incubated at 37°C for 30 min. The absorbance was measured at 660 nm. In another experiment, Rab7l1 was immunoprecipitated from THP-1 macrophages (1000 μg of lysate) infected with either Msmeg-pVV16, Msmeg-PknG-WT, Msmeg-PknG-K181M, M. bovis BCG, or M. bovis BCG∆PknG using anti-Rab7l1 Ab (10 μg). GTPase activity assay was carried out in the immunoprecipitated complexes using GTPase assay kit.

GTP-bound Rab7l1 was pulled down using 50 μl GTP-agarose beads (Innova Biosciences) either from HEK-293 cells coexpressing GFP-Rab7l1 with 3×-FLAG empty vector, FLAG-PknG-WT, and FLAG-PknG-K181M or from PMA-differentiated THP-1 macrophages infected either with Msmeg-pVV16, Msmeg-PknG-WT, or Msmeg-PknG-K181M or with M. bovis BCG or M. bovis BCG∆PknG. Level of GTP-bound Rab7l1 was determined by immunoblotting using either anti-GFP Ab (1:500 dilution) or anti-Rab7l1 Ab (1:200 dilution). GTP-bound Rab7l1 was also detected in postnuclear supernatant (PNS), phagosome, and Golgi fractions isolated from the infected macrophages.

Phagosomes were isolated using a method as described elsewhere (2) with some modification. Approximately, 200 million PMA-differentiated THP-1 macrophages were infected with various mycobacterial strains. The cells were kept at 4°C for 1 h to avoid discontinuous endocytosis and were then shifted to 37°C for different time points. For phagosome isolation, cells were resuspended in homogenization buffer [250 mM Sucrose, 3 mM imidazole in 10 mM Tris-HCl (pH 7.4) with protease inhibitors and benzonase] and lysed. The unlysed and nuclear part of the cells was removed by centrifugation at 5000 rpm for 10 min to obtain PNS. A sucrose step gradient (1 ml 62%, 4 ml 55%, and 6 ml 37%) was used to separate the phagosomes from other organelles by loading 1 ml of PNS on top of the sucrose gradient cushion and centrifugation at 100,000 × g for 2 h.

Golgi fractions were isolated as previously described (18) with some modifications. Approximately 200 million PMA-differentiated THP-1 macrophages were infected with various mycobacterial strains. Cells were washed and resuspended in homogenization buffer [10 mM Tris-HCl (pH 7.4), 250 mM sucrose, 20 mM EDTA (pH 8), and protease inhibitor]. The lysate was centrifuged at 5000 rpm for 10 min to isolate the PNS. PNS was mixed with 62% sucrose solution to a final concentration of 37% w/v of sucrose and loaded onto the ultracentrifuge tubes. Sucrose solutions of 35 and 29% concentrations were carefully layered onto the PNS to make a discontinuous gradient of 29, 35, and 37% (containing the PNS) of sucrose and centrifuged at 95,000 × g for 2.5 h. The Golgi fractions were isolated from the band appearing between the interface of 35 and 29% sucrose gradient.

Statistical analysis was performed with GraphPad Prism (version 5). Individual statistics of unpaired samples were performed by the Student t test. For CFU counts, two-way ANOVA followed by Bonferroni posttests were used to check statistical significance. *p < 0.05, **p < 0.01, and ***p < 0.001 were considered to be statistically significant. NS indicates a nonsignificant relationship.

To identify the interacting host partner(s) of M. tuberculosis PknG, we carried out Y2H screening using full-length M. tuberculosis PknG (PknG-WT) cloned into plasmid pGBKT7 in frame with GAL4-BD as bait against a human leukocyte’s cDNA library cloned into the pACT2 vector (Clontech Laboratories) in frame with GAL4-AD as prey. In Y2H screening, prey plasmids were rescued from ∼237 positively selected colonies and were sequenced using an automated DNA sequencer. The obtained sequences were searched against a human genomic plus transcript sequence database available online at NCBI using BLASTN. The cDNA sequences from five prey plasmids contained a complete open reading frame that had 100% sequence similarity with a host protein Rab7l1 (NCBI reference sequence: NM_001135662.1) (Supplemental Fig. 1A). The rest of the prey plasmids contained nonspecific sequences, which did not return any significant match. Plasmid obtained from the positive colony was transformed with bait pGBKT7-PknG and selected on a QDO selection plate with various positive and negative controls. Once again, the interaction was found to be positive (Fig. 1A). The specific interaction was again confirmed by exchanging the bait and prey plasmids in which PknG-WT was cloned in frame with GAL4-AD in pGADT7 vector and Rab7l1 was cloned in frame with GAL4-BD in pGBKT7 vector in Y2H assay (Supplemental Fig. 1B). To further confirm the Y2H results, the PknG was cloned in pCDNA3.1c 3×-FLAG vector in frame with the FLAG tag and expressed in HEK-293 cells and Co-IP assay was carried out using anti-FLAG Ab for immunoprecipitation and anti-Rab7l1 Ab for detecting Rab7l1 in these captured complexes. The results indicate that PknG interacts with endogenous Rab7l1 present in HEK-293 cells (Fig. 1B). A direct physical interaction of PknG with Rab7l1 was also confirmed by pull-down assay using purified His-tagged PknG and GST-Rab7l1 protein. His-tagged PknG was immobilized to Ni-NTA agarose beads and incubated with E. coli lysate–expressing GST-Rab7l1. The interacting protein complexes were eluted and probed with anti-Rab7l1 Ab, and it was further confirmed that PknG interacted with Rab7l1 (Fig. 1C).

Because PknG inhibits P–L fusion and several other host proteins are known to play a critical role in phagosomal maturation (12, 14–16), we investigated one-to-one interaction of PknG, if any, using Y2H with Rab5, Rab7, Rabex5, Rabaptin5, PI3K3, Rab11, Rab4, Mon1a, and Mon1b and found that none of these proteins interacted with PknG, although they were well expressed in the system (Supplemental Fig. 1C, 1D). We also carried out Co-IP assay using Flag-tagged PknG, and the pulled-down complexes were probed with specific Abs for EEA1, Rab5, Rab7, Rab14, Rab22a, and LAMP2 (Supplemental Fig. 1E). However, none of these proteins were found to interact with PknG.

PknG is known to possess three distinct domains: N-terminal trx domain, middle trf (kinase) domain, and C-terminal egx domain (10, 11). To delineate the regions of PknG involved in the interaction with Rab7l1, we next constructed various truncated mutants of PknG encompassing these domains. The truncated mutant PknG-(1–146) contained the N-terminal domain, PknG-(147–406) contained the kinase domain, PknG-(407–750) contained the C-terminal domain, PknG-(1–406) contained both the N-terminal domain and the kinase domain, and PknG-(147–750) contained both the kinase domain and the C-terminal domain. In addition, PknG-K181M, a PknG mutant carrying a point mutation in which the lysine residue at position 181 is substituted with methionine, was also used. The K181M mutation in the kinase domain of PknG makes it functionally defective for its kinase activity (12, 13, 15, 16) but is shown to be secreted from the M. tuberculosis cell wall and detected in the host cytosol (12). All these PknG mutants were cloned into pGADT7 vector with GAL4-AD, and we examined their interactions with Rab7l1 cloned in pGBKT7 vector with GAL4-BD in Y2H assays. It was found that Rab7l1 could interact with PknG mutants containing the kinase domain but not with mutants expressing either the N-terminal [PknG-(1–146)] or the C-terminal region [PknG-(407–750)] (Fig. 2A). Interestingly, PknG-K181M was also found to interact with Rab7l1, suggesting that kinase activity of PknG was not essential for its interaction with Rab7l1 (Fig. 2A). In addition, β-gal assay was also carried out to measure the strength of these interactions among various groups. It was found that PknG mutant containing the kinase domain alone [PknG-(147–406)] had significantly weaker interaction as compared with PknG-(1–406) or PknG-(147–750), indicating a role of N-terminal or C-terminal domains in stabilizing the interaction between PknG and Rab7l1 (Fig. 2B). To further confirm the Y2H results, the PknG truncated mutants were also cloned in pCDNA3.1c 3×-FLAG vector in frame with the FLAG tag and expressed in HEK-293 cells, and Co-IP assay was carried out. Anti-FLAG Ab was used for capturing the FLAG-tagged mutant PknG-bound protein complexes, and anti-Rab7l1 Ab was used for detecting Rab7l1 in these captured complexes. It was again found that mutants lacking the kinase domain failed to interact with Rab7l1, but PknG-K181M was able to interact (Fig. 2C).

Because we observed that PknG could physically interact with Rab7l1, we speculated that Rab7l1 probably plays a role in P–L fusion and PknG targets the Rab7l1 to inhibit P–L fusion. Therefore, we generated a Rab7l1-KD stable THP-1 cell line using Rab7l1-specific shRNA. Expression of Rab7l1 mRNA and protein was found to be reduced by more than 90% in Rab7l1-KD cells compared with control THP-1 cells, which received a scrambled shRNA (Supplemental Fig. 2A, 2B). Viability of Rab7l1-KD or control THP-1 cells was not found to be affected upon stimulation with bacterial LPS (Supplemental Fig. 2C). The Rab7l1-KD or control THP-1 cells were differentiated into macrophages using PMA and infected with GFP-expressing M. smegmatis (which lacks PknG expression) carrying either PknG-WT (Msmeg-PknG-WT) or PknG-K181M (Msmeg-PknG-K181M) or harboring the backbone vector alone (Msmeg-pVV16). Colocalization of bacteria-associated GFP fluorescence was monitored to a marker of acidic compartment (LysoTracker Red DND-99) after 1 h of infection. Expectedly, green fluorescence associated with Msmeg-pVV16 and Msmeg-PknG-K181M was found to be predominantly colocalized with LysoTracker Red in control THP-1 macrophages (Fig. 3Ai, 3Aiii). However, a significant reduction in colocalization with LysoTracker Red was observed in case of Msmeg-PknG-WT in these cells (Fig. 3Aii), indicating a role of PknG in avoiding P–L fusion by the infecting bacteria as reported by others (12, 15). Interestingly, none of these strains were found to be significantly colocalized with Lysotracker Red in Rab7l1-KD THP-1 macrophages (Fig. 3Aiv–3Avi). The inability of the nonpathogenic M. smegmatis strains to be localized in lysosomes, whether carrying PknG or not, in Rab7l1-KD cells indicates that Rab7l1 may play a role in P–L fusion and trafficking of mycobacteria into the lysosomes.

Because inhibition in P–L fusion was found to be directly correlated with intracellular survival of mycobacteria (4), we next examined whether M. smegmatis or PknG-deficient M. bovis BCG (M. bovis BCG∆PknG), which are known to be eliminated efficiently by the activated macrophages, could survive better in Rab7l1-KD macrophages. Therefore, PMA-differentiated control or Rab7l1-KD THP-1 macrophages were infected with either Msmeg-PknG-WT, Msmeg-PknG-K181M, Msmeg-pVV16, M. bovis BCG, or M. bovis BCG∆PknG at an MOI of 10, and intracellular survival of these bacilli was monitored by CFU counting at different time points postinfection. As expected, Msmeg-PknG-WT was found to survive better than Msmeg-pVV16 in control THP-1 macrophages (12, 15). However, in case of Msmeg-pVV16 and Msmeg-PknG-K181M, an increased CFU count was observed in Rab7l1-KD macrophages as compared with control THP-1 macrophages (Fig. 3B). The enhanced survival of Msmeg-pVV16 bacilli in Rab7l1-KD over control THP-1 macrophages or Msmeg-PknG-WT in control macrophages over Msmeg-pVV16 was not due to change in phagocytic activity as no significant differences in CFU counts were observed at an early time point (Supplemental Fig. 2D). Also, expectedly, M. bovis BCG∆PknG had significantly lower CFU counts as compared with M. bovis BCG strain in control THP-1 macrophages, highlighting a role of PknG in regulating intracellular survival of bacteria (Fig. 3C). However, in Rab7l1-KD macrophages, both these strains had no significant differences in survival, and also, the CFU counts of M. bovis BCGΔPknG were significantly higher when compared with control THP-1 macrophages (Fig. 3C). Taken together, these observations suggest that mycobacterial PknG confers a survival advantage to the bacteria inside the macrophages possibly by regulating the Rab7l1 function.

Because we observed the role of Rab7l1 in P–L fusion, the probable mechanism by which PknG manipulates Rab7l1 signaling pathway to block P–L fusion was investigated next. First, we examined whether PknG inhibited the endogenous expression of Rab7l1 to inhibit P–L fusion. Accordingly, PMA-differentiated THP-1 macrophages were infected with Msmeg-pVV16, Msmeg-PknG-WT, or Msmeg-PknG-K181M for 1 h and cells were harvested at various time points postinfection, lysed, and probed with anti-Rab7l1 Ab in Western blotting. Presence or absence of PknG did not affect endogenous levels of Rab7l1 (Fig. 4A, Supplemental Fig. 3A). Similarly, THP-1 macrophages infected with either M. bovis BCG or M. bovis BCG∆PknG had no effect on the levels of endogenous Rab7l1 up to 48 h postinfection (Fig. 4B, Supplemental Fig. 3B). Therefore, PknG-mediated inhibition of P–L fusion is not due to reduction in the cellular levels of Rab7l1.

Because PknG is a serine/threonine kinase, it was speculated that PknG could phosphorylate Rab7l1 to modulate its downstream signaling cascades. To test this hypothesis, in vitro kinase assay was carried out using purified recombinant Rab7l1 and PknG-WT or PknG-K181M proteins. It was found that PknG-WT failed to directly phosphorylate Rab7l1 in vitro, although GarA, a known substrate of PknG (19), was phosphorylated (Supplemental Fig. 3C, 3D). As Rab7l1 is a known RabGTPase, we examined whether PknG prevented Rab7l1 function by inhibiting its GTPase activity. Accordingly, THP-1 macrophages were infected with either Msmeg-pVV16, Msmeg-PknG-WT, or Msmeg-PknG-K181M, and GTPase assay was carried out. Interestingly, Msmeg-PknG-WT–infected cells had significantly reduced levels of GTPase activity as compared with those infected with Msmeg-pVV16 or Msmeg-K181M (Fig. 4C). When Rab7l1 was pulled down using GTP-agarose beads from these infected cells, the level of GTP-bound Rab7l1 was found to be reduced in cells infected with Msmeg-PknG-WT as compared with cells either expressing PknG-K181M or the backbone vector alone (Fig. 4D). Similarly, when THP-1 macrophages were infected with M. bovis BCG or M. bovis BCG∆PknG for 2 and 4 h, GTPase activity was significantly decreased in cells infected with M. bovis BCG as compared with cells infected with M. bovis BCG∆PknG at both time points (Fig. 4E). Decreased Rab7l1 GTPase activity was associated with reduced levels of GTP-bound Rab7l1 in M. bovis–infected cells (Fig. 4F). These results indicated that GTPase activity of Rab7l1 was inhibited by PknG and the kinase function of PknG was essential to inhibit Rab7l1 GTPase activity. To further confirm these findings, GFP-tagged Rab7l1 was coexpressed in HEK-293 cells along with 3×-FLAG–tagged PknG-WT or PknG-K181M (Fig. 4G), immunoprecipitated using anti-GFP Ab, and GTPase activity was measured. It was again found that expression of PknG-WT significantly reduced GTPase activity as compared with cells transfected with the vector alone or PknG-K181M, although transient expression of PknG-WT/PknG-K181M did not affect expression levels of GFP-Rab7l1 in these cells (Fig. 4G, 4H). When GTP-bound GFP-Rab7l1 was pulled down from the transfected HEK-293 cells expressing FLAG-PknG-WT or FLAG-PknG-K181M using GTP-agarose beads and probed with anti-GFP Ab to detect GFP-Rab7l1, the level of GTP-bound GFP-Rab7l1 was found to be reduced in cells expressing FLAG-PknG-WT as compared with cells either expressing FLAG-PknG-K181M or the backbone vector alone (Fig. 4I). Together, these data hint a role of PknG-WT in blocking conversion of inactive Rab7l1-GDP to active Rab7l1-GTP and thus explain that an observed lower GTPase activity of Rab7l1 in the presence of active PknG and kinase activity of PknG is required. Therefore, we speculated that PknG probably interacts with the Rab7l1-GDP to interfere its conversion to GTP-bound form of Rab7l1. To further confirm this hypothesis, constitutively active (Q67L, remain locked in the GTP-bound conformation, deficient in GTPase activity) and inactive (T21N, unable to bind GTP, remain locked in the GDP-bound state) mutants of Rab7l1 were generated (18), and interaction of PknG-WT and PknG-K181M with these mutants was examined by Y2H one-to-one interaction assay. Interestingly, specific interaction of PknG-WT and PknG-K181M was observed with GDP-bound Rab7l1T21N but not with GTP-bound Rab7l1Q67L (Fig. 4J).

Earlier studies have demonstrated that Rab7l1-GDP is present in the trans-Golgi network (TGN), whereas Rab7l1-GTP is primarily localized in the cytosol (20, 21). Because PknG appears to preferentially interact with Rab7l1T21N, it is possible that PknG is translocated to the TGN to interact with Rab7l1-GDP. Therefore, HEK-293 cells were transiently transfected with FLAG-tagged full-length PknG-WT and its various truncated mutants along with mCherry TGOLN2, encoding a TGN marker TGOLN2, and their colocalization was examined by confocal microscopy. We found that PknG-WT as well as the fragments that contain kinase domain–like PknG-(147–406), PknG-(1–406), and PknG-(147–750) were colocalized with TGOLN2. However, mutants like PknG-(1–146) and PknG-(407–750), which lack the kinase domain, failed to be colocalized with TGOLN2 (Fig. 5A), indicating that the kinase domain is critical for recruitment of PknG to the TGN. Interestingly, PknG-K181M mutant was also found to be colocalized with TGOLN2 (Fig. 5A), indicating that presence of the kinase domain is important for localization but not kinase function of PknG. The degree of colocalization was estimated by measuring Mander’s overlap coefficient (Fig. 5B). These data, together with data presented in the previous section, suggest that PknG is translocated to Golgi and interacts with Rab7l1 to inhibit its GDP/GTP transition. To confirm translocation of PknG to Golgi during infection, Golgi fractions were isolated from THP-1 macrophages infected with either M. bovis BCG or M. bovis BCGΔPknG, and presence of PknG as well as Rab7l1 was examined by immunoblotting. The results suggest that PknG was present in the Golgi extract where Rab7l1 and EEA1 were also localized (Fig. 5C). Also, we were able to pull down the PknG-Rab7l1 complex from the enriched Golgi fraction of THP-1 macrophages infected with M. bovis BCG (Fig. 5D), suggesting that PknG is not only able to be recruited in the Golgi but can also interact with the Golgi resident Rab7l1 during infection.

Because we found a role of Rab7l1 in P–L fusion, we next examined whether Rab7l1 gets recruited to phagosomes. Therefore, we isolated phagosomes from THP-1 macrophages infected with either M. bovis BCG or M. bovis BCG∆PknG, and recruitment of Rab7l1 was examined in the isolated phagosomes from infected macrophages by immunoblotting using anti-Rab7l1 Ab. It was found that a higher amount of Rab7l1 was accumulated in phagosomes isolated from macrophages infected with M. bovis BCG∆PknG as compared with those isolated from macrophages infected with M. bovis BCG, although total levels of Rab7l1 in the PNS of both groups remained unchanged (Fig. 6A). Recruitment of Rab7l1 to phagosomes was also observed in control THP-1 macrophages infected with Msmeg-pVV16 that do not contain PknG but expectedly not in Rab7l1-KD macrophages, which have reduced expression of Rab7l1 (Fig. 6B). When THP-1 macrophages were infected with GFP-expressing Msmeg-pVV16, Msmeg-PknG-WT, or Msmeg-PknG-K181M and colocalization of GFP fluorescence with Rab7l1 was monitored by confocal microscopy, it was observed that Msmeg-pVV16 and Msmeg-PknG-K181M were colocalized with Rab7l1 but not in the case of Msmeg-PknG-WT (Fig. 6C, 6D). These data suggest that Rab7l1 gets recruited to the bacilli-containing phagosomes, which are inhibited by PknG.

Next, we compared the levels of GTP-bound Rab7l1 and its GTPase activity in phagosome and Golgi fractions isolated from Msmeg-pVV16–infected THP-1 macrophages. We observed an enrichment of GTP-bound Rab7l1 in phagosomes compared with the Golgi body (Fig. 6E), and the levels were found to be directly correlated with the GTPase activity of Rab7l1 (Fig. 6F), suggesting that inactive Rab7l1-GDP is present predominantly in the Golgi fraction, and during infection, Rab7l1-GTP gets recruited to the bacteria-containing phagosomes. Because PknG was found to inhibit Rab7l1-GDP/GTP transition, we next examined the distribution of GTP-bound Rab7l1 in the phagosome and Golgi fractions isolated from THP-1 macrophages infected with M. bovis BCG and M. bovis BCG∆PknG. Interestingly, a higher amount of GTP-bound Rab7l1 was observed in phagosomes from M. bovis BCG∆PknG–infected THP-1 macrophages when compared with phagosomes from M. bovis–infected THP-1 macrophages (Fig. 6G). As expected, in M. bovis BCGΔPknG–infected macrophages, Rab7l1 GTPase activity was also higher in phagosome fractions as compared with Golgi fractions (Fig. 6H). Also, as speculated, a low level of Rab7l1-GTP was present in the Golgi as compared with phagosomes of M. bovis BCG∆PknG–infected macrophages (Fig. 6G). These results suggest that GTP-bound Rab7l1 was predominantly recruited to bacilli-containing phagosomes, and PknG inhibited the conversion of inactive Rab7l1-GDP to active Rab7l1-GTP and thereby interfered with phagosomal recruitment and activity of Rab7l1 during infection.

We next examined whether Rab7l1 is important for the recruitment of other phago-lysosomal markers to promote P–L fusion, which was inhibited by PknG (12). Therefore, we next compared colocalization of GFP-expressing Msmeg-pVV16 with different phagosomal and lysosomal markers in Rab7l1-KD macrophages and compared them with control THP-1 macrophages. Interestingly, we observed that in Rab7l1-KD macrophages, Msmeg-pVV16 did not exhibit significant colocalization with EEA1 (early endosomal marker), Rab7 (late endosomal marker), and LAMP2 (lysosomal marker) as compared with control THP-1 macrophages, although colocalization with Rab5 did not alter (Fig. 7A). To confirm the confocal microscopy data, we next isolated phagosomes from Rab7l1-KD and control THP-1 macrophages, both infected with Msmeg-pVV16, and recruitment of various phagosomal and lysosomal markers was examined by immunoblotting. We observed reduced recruitment of EEA1, Rab7, and LAMP2 in the phagosome fraction isolated from Msmeg-pVV16–infected Rab7l1-KD THP-1 macrophages as compared with those isolated from Msmeg-pVV16–infected control macrophages, and no change was observed in Rab5 recruitment (Fig. 7B). These results suggest that recruitment of Rab7l1 on phagosomes is important in the subsequent recruitment of P–L markers and induction of P–L fusion during infection. Because PknG interferes with Rab7l1 recruitment to phagosomes and is known to inhibit P–L fusion, we next investigated whether acquisition of P–L markers by phagosomes of M. bovis BCG–infected macrophages was similar to that of phagosomes of Msmeg-pVV16–infected Rab7l1-KD THP-1 macrophages in which P–L fusion was inefficient because of reduced recruitment of Rab7l1. Accordingly, phagosomes were isolated from M. bovis BCG– and M. bovis BCG∆PknG–infected THP-1 macrophages, and recruitment of Rab5, EEA1, Rab7, and LAMP2 was compared between these two groups by immunoblotting. It was found that recruitment of EEA1, Rab7, and LAMP2 to phagosomes isolated from M. bovis BCG–infected macrophages was lesser as compared with those isolated from M. bovis BCG∆PknG, whereas recruitment of Rab5 to phagosomes remained unaffected in both groups (Fig. 7C). These data suggest that Rab7l1 plays an important role in the recruitment of markers important for the phagosome maturation process, and PknG targets the Rab7l1 signaling cascades to arrest recruitment of key phago-lysosomal markers, thus resulting in inhibition of P–L fusion.

Intracellular survival of the pathogens, at least in part, is associated with their ability to inhibit phagosomal maturation and its fusion with lysosomes (22). Earlier studies indicate that PknG, a mycobacterial serine/threonine kinase, is released in the cytoplasm of infected macrophages and blocks P–L fusion to favor intracellular survival of mycobacteria (12, 23). Kinase activity was found to be crucial for the ability of PknG to promote mycobacterial survival (12, 13, 15, 16). Using Y2H library screening, we were able to identify that PknG interacts with Rab7l1, which is involved in the regulation of P–L fusion. Interestingly, we found that functional kinase activity was not essential for interaction with Rab7l1, but the kinase function of PknG was critical for regulating Rab7l1 function as PknG-WT inhibited GTPase activity of Rab7l1 by blocking the conversion of inactive Rab7l1-GDP to active Rab7l1-GTP. In contrast, PknG-K181M could interact with Rab7l1-GDP but failed to inhibit its GTPase activity.

Rab7l1, a small GTP-binding protein (24) of the Ras superfamily, has been implicated in maintaining the integrity of the endosome-TGN structure (20, 21). In the current study, to the best of our knowledge, we for the first time found a novel role of Rab7l1 in the phagosomal maturation. M. smegmatis that does not express PknG as well as M. bovis BCG∆PknG could survive inside the Rab7l1-KD THP-1 macrophages, which were otherwise eliminated efficiently by the control macrophages. We observed an increased recruitment of Rab7l1-GTP in the PknG-deficient mycobacterial phagosomes, and this may constitute an important step for the normal phagosomal maturation process. However, Rab7l1-GTP recruitment was significantly reduced in PknG-sufficient mycobacterial phagosomes with a concurrent decrease in recruitment of other endo-lysosomal markers, indicating that Rab7l1 may influence recruitment of phago-lysosome markers important for the phagosome maturation process. Mycobacterial phagosomes of Rab7l1-KD macrophages had decreased recruitment of EEA1, Rab7, and LAMP2; however, the levels of Rab5 remained unchanged, indicating that Rab7l1 acts upstream of EEA1 recruitment in the maturation sequence. Vesicular trafficking from TGN is shown to be crucial for the formation of phago-lysosomal organelles (25, 26), and small GTPases belonging to the Rab family and their effectors such as the membrane-tethering molecule EEA1 play crucial roles in vesicle tethering and fusion machinery (27). Recruitment of EEA1 to the early endosome is critical to activate the machinery to form late endosomes in a unidirectional manner (28, 29). Rab7l1 was also found to be important in assisting recruitment of markers like the mannose-6 phosphate receptor from TGN to endosome, and deficiency of Rab7l1 resulted in endo-lysosomal and Golgi apparatus sorting defects in Parkinson disease (20, 21). Studies indicate that M. tuberculosis and M. bovis BCG inhibit EEA1 recruitment to phagosomes in infected macrophages and block the formation of phago-lysosomes (28, 29) by inhibiting the steps associated with Rab5 to Rab7 conversion (30). Rab5 influences early endosomes to undergo homotypic fusion (31) and initiates subsequent binding of other effector molecules like EEA1 and vps34 (PI3K3) (32, 33). Our data hint that PknG inhibits recruitment of Rab7l1-GTP to the early phagosomes, leading to inhibition of EEA1 recruitment and subsequent collapsing of phagosome maturation. Although a direct role of Rab7l1 in P–L fusion is not reported till now, one study demonstrates that Salmonella typhimurium factor Gtg-E can proteolytically cleave Rab7l1 and inhibit acidification of lysosomes to favor intracellular survival of the bacteria (34). Rabs are known to be associated with GDP dissociation inhibitor, guanosine exchange factor, and GTPase activating protein to control the Rab-GDP to Rab-GTP transition (35). It is possible that PknG targets any of these factors to inhibit Rab7l1-GDP/GTP transition. As classical guanosine exchange factor, GTPase activating protein, and GDP dissociation inhibitor are yet to be reported for Rab7l1, further studies need to be carried out to understand exactly how PknG manipulates Rab7l1 signaling to inhibit Rab7l1-GDP/GTP transition.

We also observed localization of PknG to the TGN (20, 21, 36), which possibly allows its interaction with Rab7l1-GDP. PknG is secreted in the cytoplasm of infected macrophages (12), and our observations suggest that cytosolic PknG can be transported to TGN; however, the exact mechanism by which it is translocated to the TGN is not well understood. Although our data point to the probable mechanisms by which PknG averts P–L fusion, it was also found to enhance mycobacterial survival by regulating glutamine metabolism as a nutritional stress sensor (14). In addition, the trx motif of PknG was also implicated to be involved in redox sensing (15). Therefore, the role of PknG in the mycobacterial pathogenesis appears to be pleiotropic in nature, which regulates multiple mechanisms to favor better survival of the bacilli inside macrophages.

In summary, our data hint at the crucial role of Rab7l1 in the phagosome maturation process, which is targeted by PknG to block conversion of Rab7l1-GDP to Rab7l1-GTP transition, leading to reduced recruitment of Rab7l1-GTP to the mycobacterial phagosomes. This resulted in decreased recruitment of key markers involved in phagosomal maturation and subsequent degradation of intracellular mycobacteria. To the best of our knowledge, for the first time, we describe a novel mechanism of PknG and survival strategy of the bacilli inside the host, which may be helpful in understanding the host–pathogen interaction during mycobacterial infection.

We thank Dr. G. Srijeet for technical help in Y2H assay; Prof. J. Tyagi (All India Institute of Medical Sciences, New Delhi) for providing GFP–M. smegmatis strain; Dr. V. Nandicoori (National Institute of Immunology, New Delhi) for providing anti-PknG Ab, pQE-2-PknG-WT, and pQE-2-PknG-K181M plasmids; and Dr. J. Pieters (University of Basel, Switzerland) for providing M. bovis BCG and M. bovis BCG∆PknG strains.

The authors have no financial conflicts of interest.