Survival of pathogenic mycobacteria in macrophages is mediated through autophosphorylation of protein kinase G

Abstract

Pathogenic mycobacteria survive within macrophages through the inhibition of phagosome-lysosome fusion. A crucial factor for avoiding lysosomal degradation is the mycobacterial serine/threonine protein kinase G (PknG). PknG is released into the macrophage cytosol upon mycobacterial infection, suggesting that PknG might exert its activity by interfering with host signaling cascades, but the mode of action of PknG remains unknown. Here, we show that PknG undergoes autophosphorylation on threonine residues located at the N terminus. In contrast to all other mycobacterial kinases investigated thus far, autophosphorylation of PknG was not involved in the regulation of its kinase activity. However, autophosphorylation was crucial for the capacity of PknG to promote mycobacterial survival within macrophages. These results will contribute to a better understanding of the virulence mechanisms of pathogenic mycobacteria and may help to design improved inhibitors of PknG to be developed as antimycobacterial compounds.

FIG. 1.

Autophosphorylation of PknG. (A) Wild-type PknG or PknG(K181M) was incubated for 30 min at 37°C in kinase reaction buffer containing 1 μCi of γ-ATP. Samples were electrophoresed on a 12.5% SDS-PAGE gel and analyzed by autoradiography. M_r values are in thousands. (B) Phosphorylated PknG separated by SDS-PAGE as shown in panel A was transferred to a polyvinylidene difluoride membrane, excised, and hydrolyzed. For TLC, samples were applied to TLC cellulose plates, stained by ninhydrin, and autoradiographed. (C) Endoproteinase Lys-C-digested PknG was applied to a C₁₈ nano-HPLC column coupled to the ESI-MS. The LC/MS run was further analyzed for phospho-peptides. The elution profile indicates the K2 peptide and its phosphorylated derivatives (1P to 3P) for the scan numbers between 450 to 600. (D) Original MS spectrum showing the series of different m/z ratios observed for unphosphorylated and singly, doubly, and triply phosphorylated K2 fragment. (E) Deconvoluted MS spectrum indicating the relative ratio of the unphosphorylated and the differentially phosphorylated forms of the K2 fragment. Chrom, chromatogram.

FIG. 2.

Identification of the autophosphorylated PknG fragment. (A) PknG was subjected to proteolytic digestion using endoproteinase Lys-C in solution, and the digested peptides of ³²P-labeled PknG were separated on a C₁₈ solid-phase HPLC column. Fractions were collected and analyzed by scintillation counting to determine the relative amount of ³²P. (B) The radioactive fractions were pooled, loaded on a 16% Tricine gel, and analyzed by autoradiography. M_r values are in thousands. (C) Sequence of the K2 ECL fragment. CPM, counts per minute.

FIG. 3.

Identification of the autophosphorylated residues. (A) The K2 fragment was digested with either trypsin or endoproteinase Asp-N, and the resulting peptides were identified by ESL-LC/MS. The regions covering the corresponding phospho-peptides are underlined and highlighted in green (trypsin) or blue (endoproteinase Asp-N). Threonines are highlighted in red. (B) Sequence of the K2 fragment with the potential autophosphorylation sites (T₂₁, T₂₃, T₂₆, T₃₂, T₆₃, and T₆₄) as identified in the experiments described above. Threonine residues are highlighted in red. (C) Autophosphorylation of PknG wild-type and mutant proteins containing the indicated point mutations. Kinase reaction mixtures were analyzed by 12.5% SDS-PAGE and autoradiography as well as by Coomassie staining. M_r values are in thousands. (D) Sequence of the K2 fragment with the identified autophosphorylation sites (T₂₃, T₃₂, T₆₃, and T₆₄). Threonine residues are highlighted in red.

FIG. 4.

Analysis of PknG autophosphorylation in vitro. (A) Domain structure of wild-type PknG, PknG-ΔN, PknG-Pmut, and PknG-N-term. PknG-ΔN lacks the first 73 amino acids at the N terminus and is devoid of all potential phosphorylation sites. For construction of the PknG-Pmut, all potential autophosphorylation sites identified were mutated to alanine residues. (B) Autophosphorylation activity of PknG wild type (WT), PknG-ΔN (ΔN), and PknG-Pmut (P). (C) Kinase activity of PknG wild type (WT), PknG-ΔN (ΔN), and PknG-Pmut (P). M_r values are in thousands. TPR, tetratricopeptide repeat.

FIG. 5.

Analysis of PknG autophosphorylation upon infection of macrophages. (A) Intracellular trafficking of mycobacteria analyzed by immunofluorescence microscopy. Bone marrow-derived macrophages were infected with M. bovis BCG wild type (wt), M. bovis BCG ΔPknG (ko), and M. bovis BCG ΔPknG overexpressing PknG-ΔN (ΔN) and PknG-Pmut (P-mut) for 1 h, followed by a 3-h chase. The infected cells were stained for lysosomes using rat anti-LAMP and for mycobacteria using polyclonal rabbit anti-BCG antibodies, visualized using Alexa Fluor 488- and Alexa Fluor 568-conjugated anti-rat and anti-rabbit antibodies, respectively, and analyzed by confocal microscopy. (B) Quantitative evaluation of the immunostaining described in panel A. P values of a student's t test (paired two sample for means) are 0.000274 for M. bovis BCG wild type (wt) versus M. bovis BCG ΔPknG (ko), 0.001595 for wild type versus M. bovis BCG ΔPknG overexpressing PknG-ΔN (dN), and 0.001849 for wild type versus M. bovis BCG ΔPknG overexpressing PknG-Pmut (P-mut). (C) Intracellular trafficking of mycobacteria analyzed by cell organelle electrophoresis. M. bovis BCG wild type, M. bovis BCG PknG knockout (ko), M. bovis BCG ΔPknG overexpressing PknG-Pmut, and M. bovis BCG ΔPknG overexpressing PknG-ΔN (ΔN) were phagocytosed for 3 h by bone marrow-derived macrophages. The cells were then washed and homogenized, and the organelles were separated by charge on a Ficoll gradient. The lysosomal fractions are represented in black, and the phagosomal fractions are shown in red. The mycobacterial repartition was determined by acid-fast staining after cytospinning of the fractions. β-hex, β-hexosaminidase. (D) Survival of internalized mycobacteria. Bone marrow-derived macrophages (BMM) were infected for 1 h with M. bovis BCG wild type (wt) and mutant strains (ko, dN, and P-mut as described in panel B). Extracellular bacteria were killed by amikacin treatment, and the macrophages were washed and lysed using 0.15% saponin at the indicated chase times. Mycobacterial survival was determined by bacterial incorporation of tritiated uracil for 24 h, followed by scintillation counting.