CpsA, a LytR-CpsA-Psr Family Protein in Mycobacterium marinum, Is Required for Cell Wall Integrity and Virulence

Abstract

LytR-CpsA-Psr family proteins play an important role in bacterial cell wall integrity. Although the pathogenic relevance of LytR-CpsA-Psr family proteins has been studied in a few bacterial pathogens, their function in mycobacteria remains uncharacterized. In this work, a transposon insertion mutant (cpsA::Tn) of Mycobacterium marinum was studied. We found that inactivation of CpsA altered bacterial colony morphology, sliding motility, cell surface hydrophobicity, and cell wall permeability. Besides, the cpsA mutant exhibited a decreased arabinogalactan content, indicating that CpsA plays a role in cell wall assembly. Moreover, the mutant shows impaired growth within macrophage cell lines and is severely attenuated in zebrafish larvae and adult zebrafish. Taken together, our results indicated that CpsA, a previously uncharacterized protein, is important for mycobacterial cell wall integrity and is required for mycobacterial virulence.

INTRODUCTION

Mycobacterium tuberculosis, the causative agent of tuberculosis (TB), is a highly successful human pathogen that has infected one-third of the world's population, causing approximately 8 million new infections and 1.4 million deaths annually (1). The success of M. tuberculosis as an infectious pathogen relies partly on its strategies to evade the host immune defense and proliferate within the hostile host environment. Various genes are involved in these strategies (2), and a deeper understanding of the roles of mycobacterial virulence genes is required for developing new strategies to cure TB.

The LytR-CpsA-Psr (LCP) family proteins are putative transmembrane proteins, distributed mainly in Gram-positive bacteria, that typically consist of a very short intracellular N-terminal region, a transmembrane helix, and a large extracellular portion carrying the so-called LytR-CpsA-Psr domain (3). Different bacterial species contain 1 to 11 LCP proteins (3). LytR was first described in Bacillus subtilis as a transcriptional attenuator of both itself and the divergently transcribed lytABC operon, which encodes three proteins associated with bacterial autolysis (4). However, Kawai et al. provided genetic and enzymatic evidence suggesting that LytR catalyzes the final attachment of wall teichoic acid (WTA) to peptidoglycan (PG) (5). CpsA is considered a transcription activator of the capsular polysaccharide synthesis operon of Streptococcus pneumoniae (6). Psr was initially proposed to be a repressor of the synthesis of penicillin-binding protein 5 (PBP5) in Enterococcus hirae, because an 87-bp deletion in the psr gene was found in E. hirae strain R40, which overproduces PBP5 (7), but the role of Psr in PBP5 synthesis has not been confirmed (8). A conserved amino acid sequence region, namely, the LytR-CpsA-Psr region, is shared by these three proteins (3).

LCP proteins generally appear to be involved in cell envelope maintenance and bacterial virulence, although their mechanistic functions remain unclear. BrpA, an LCP protein of Streptococcus mutans, plays critical roles in biofilm formation, autolysis, and cell division, and the loss of BrpA can significantly affect acid and oxidative stress tolerance (9). Moreover, the brpA mutant strain of S. mutans caused a stronger immune response than the wild type and displayed attenuated virulence in the rat (10). CpsA in streptococcal pathogens is involved in the synthesis of capsule polysaccharide, shielding the pathogens from host immune system mechanisms including phagocytosis, complement deposition, and clearance by mucus (11–14). In Streptococcus agalactiae, a strain that lacks the cpsA gene is significantly attenuated in zebrafish (15). LCP proteins in Staphylococcus aureus are essential for cell division, autolysis, and β-lactam resistance (16), and the loss of MsrR (an LCP protein) results in decreased virulence in a nematode-killing assay as well as in the rat experimental endocarditis model (17). An LCP protein in an Anabenea sp., ConR, is also involved in vegetative cell septum formation under specific growth conditions (18).

The annotated M. tuberculosis H37Rv genome contains four genes encoding proteins with LCP domains: Rv0822c, Rv3267, Rv3484 (cpsA), and Rv3840. The function of these genes is not known at present. Mycobacterium marinum contains orthologs of all four M. tuberculosis LCP proteins: MMAR_4858, MMAR_1274, MMAR_4966 (cpsA), and MMAR_5392, respectively. In this study, we characterized the role of CpsA in M. marinum under both in vitro and in vivo conditions. The results indicate that this LCP family protein of M. marinum is involved in cell wall assembly and plays a critical role in mycobacterial virulence.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

M. marinum strain M (ATCC BAA-535) was used as the wild-type (WT) strain in this study. The cpsA (MMAR_4966) mutant strain (cpsA::Tn) was identified from a transposon insertion library from our laboratory (19). M. marinum strains were routinely cultivated either in Middlebrook 7H9 medium with 10% OADC (oleic acid–albumin–dextrose–catalase), 0.5% glycerol, and 0.05% Tween 80 or on Middlebrook 7H10 agar supplemented with 10% OADC and 0.5% glycerol. When necessary, 20 μg/ml of kanamycin or 50 μg/ml of hygromycin was included in the medium. For growth measurement in different carbon sources, strains were grown either in a rich medium comprising 7H9 broth plus 0.5% glycerol and 10% OADC or in a 7H9 salt solution containing either 0.4% glycerol or 0.4% acetate as the sole carbon source. In addition, 0.05% tyloxapol was added to these media to reduce bacterial clumping. Aliquots (500 μl) of bacterial cultures were collected at various time points, immediately boiled at 100°C for 10 min, and stored at −80°C before usage. Cell suspensions (50 μl) were mixed with an equal volume of the BacTiter-Glo reagent and were incubated for 5 min in the dark according to the manufacturer's protocol. Luminescence was detected by a PerkinElmer X5 plate reader and is expressed in relative luminescence units (RLU).

DNA manipulation.

The cpsA (MMAR_4966) gene of M. marinum was PCR amplified using forward primer 5′-AGATCTAGGAGGAGGCTCGTAATGGGACGTACCAAAGC-3′ and reverse primer 5′-ACTAGTCGGAACTCACGCCTGACTCGGGCTAG-3′, which contain a BglII and an SpeI site, respectively (underlined). Then, the BglII-SpeI fragment was cloned into the BamHI- and SpeI-cut pSMT3LxEGFP vector, creating pMMcpsA, in which the open reading frame (ORF) of the cpsA gene was cloned downstream of the hsp60 promoter. Similarly, the cpsA gene (Rv3484) of M. tuberculosis was cloned using PCR primers 5′-AGATCTAGGAGGAGGCATATGATGGCGCGTTCTGAGGG-3′ (forward) and 5′-ACTAGTGTGGAGTGGTCGGATAACTTACC-3′ (reverse) to generate pRvcpsA. These constructs were electroporated into the cpsA mutant of M. marinum, and transformants were selected on 7H10 agar containing 50 μg/ml of hygromycin.

Assays to determine cell surface properties.

The sliding motility assay was carried out according to a published protocol (20). The relative hydrophobicity of the cell surface was assessed by a two-phase aqueous hexadecane system (21). As described previously (22), the permeability of the cell wall was assessed by measuring the accumulation of Nile red and ethidium bromide (EtBr), representatives of hydrophobic and hydrophilic compounds, respectively.

Antibiotic susceptibility assay.

For the antibiotic susceptibility assay, bacteria in 96-well plates were incubated with various antibiotics at serial 2-fold dilutions. The MIC was defined as the minimal concentration at which no visible growth was observed after 7 days of cultivation.

Sugar analysis.

To characterize the monosaccharide composition of the cell wall, 200-ml cultures were harvested at the exponential phase (optical density at 600 nm [OD₆₀₀], ≈0.8) and stationary phase (OD₆₀₀, ≈3.0 for glycerol cultures and ≈1.0 for acetate cultures) for each strain. Pellets were extracted overnight using chloroform-methanol (1:2, vol/vol), followed by centrifugation at 3,000 × g for 3 min. Pellets were extracted again using chloroform-methanol (2:1, vol/vol) overnight. Delipidated cells were further treated overnight, three times, with 50% ethanol (EtOH) at 60°C to remove lipoarabinomannan (LAM). The pellets were dried under a nitrogen stream. Ten milligrams of the resulting pellets for each strain was treated with 2 M trifluoroacetic acid (TFA) (100 μl) for 2 h at 120°C after the addition of the internal standard 3-O-methyl glucose. After the samples were cooled down to room temperature, they were extracted twice with 1 ml chloroform. The aqueous extracts were dried totally and were reduced with NaBD₄ (10 mg/ml in 1 M ammonium hydroxide–95% ethanol [1:1]) overnight. Excess NaBD₄ was removed by adding a few drops of glacial acetic acid; the tubes were dried under N₂; 10% acetic acid in methanol was added; and the samples were dried three times, followed by three methanol-drying steps. The samples were peracetylated using 100 μl of acetic anhydride and were heated for 1 h at 100°C. The alditol acetates were extracted with chloroform-water (2:1), and the chloroform layer was dried and was analyzed by gas chromatography (GC)-mass spectrometry (MS) as described previously (23).

Macrophage infection experiments.

Cells of the murine macrophage cell line RAW 264.7 (ATCC TIB71) were infected with M. marinum as described previously (24). Briefly, RAW 264.7 cells were cultured at 37°C under 5% CO₂ in Dulbecco's modified Eagle medium (DMEM) (Gibco) supplied with 10% fetal bovine serum (FBS) (Gibco). Cells were seeded into 24-well tissue culture plates (Corning) at a density of 10⁵ per well for 24 h before infection. The single-bacterial-cell suspension was prepared by subjecting bacterial cultures to a vortex with sterile glass beads and passing the product through a 5-μm syringe filter to remove bacterial aggregates; the concentration of bacteria in each suspension was then evaluated by CFU counts. RAW 264.7 cells were infected with M. marinum at a multiplicity of infection (MOI) of 1. The infection was allowed to proceed at 32°C under 5% CO₂ for 4 h, after which the monolayers infected with M. marinum were washed twice with sterile phosphate-buffered saline (PBS) to remove extracellular bacteria. Extracellular M. marinum bacteria were further killed by incubation with DMEM supplemented with 200 μg/ml gentamicin (Amresco) for 2 h. Subsequently, cells were incubated at 32°C under 5% CO₂ in fresh medium with 20 μg/ml gentamicin to limit the growth of extracellular bacteria. On days 0 (6 h), 1, 2, 3, and 4, the infected macrophage monolayers (two wells per strain) were washed twice with fresh medium and were lysed with 0.1 ml of 1% Triton X-100 (Sigma) to release intracellular bacteria. The intracellular mycobacteria were enumerated by spreading appropriate dilutions on Middlebrook 7H10 agar plates with appropriate antibiotics.

To measure the ratio of invasion of macrophages by M. marinum, infection experiments were performed with an MOI of 10. Immediately following infection (4 h at 32°C), macrophages were washed three times with PBS and were cultured in DMEM supplemented with 200 μg/ml gentamicin (Amresco) for 2 h to remove extracellular bacteria. The CFU recovered from macrophages were enumerated by plating on 7H10 agar.

To determine cytokine levels, RAW 264.7 cells were infected in triplicate at an MOI of 10 for 4 h. Extracellular bacteria were removed by washing. Cells were then maintained in DMEM supplemented with 10% FBS and 20 μg/ml gentamicin for the times indicated in Fig. 5. At 4, 24, and 48 h postinfection, cell-free supernatants were collected, and cytokine levels were measured by sandwich enzyme-linked immunosorbent assays (ELISA) according to the manufacturer's instructions (BD Biosciences).

FIG 5.

CpsA affects the cytokine profiles of infected macrophages. Shown is the secretion of TNF-α (A) and IL-6 (B) by macrophages infected with M. marinum. RAW 264.7 macrophages were infected with the WT, cpsA mutant, or complemented strain at an MOI of 10. Supernatants of infected cells were collected 4, 24, and 48 h postinfection, and levels of secreted TNF-α and IL-6 were measured by ELISA. Data were generated from three independent experiments. Asterisks indicate significant differences (**, P < 0.01) as determined by an unpaired Student t test. ns, not significant.

Colocalization of M. marinum with LAMP-1 protein.

RAW 264.7 macrophages (1 × 10⁵) attached to glass coverslips were infected with green fluorescent protein (GFP)-labeled M. marinum. The macrophages were fixed, permeabilized, blocked, and incubated with a rabbit polyclonal antibody to LAMP-1 (1:1,000 dilution; Abcam), followed by Alexa Fluor 594-conjugated goat anti-rabbit IgG. Cell nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) (Invitrogen). Florescence signals were detected by confocal microscopy.

Zebrafish infection experiments.

Zebrafish larvae were obtained from pairs of zebrafish AB line adults by natural spawning and were raised at 28°C in tank water. At 30 h postfertilization (hpf), larvae were injected via the caudal vein with about 100 CFU M. marinum carrying the pTEC15 plasmid (a gift from L. Ramakrishnan, University of Washington, USA [25]) by using phenol red as a visual marker, as described previously (26). The infected larvae were observed and photographed using an EVOS FL color imaging system (Invitrogen). Bacterial burdens in larvae were determined by fluorescence pixel counts through ImageJ software. Larvae began to die after 5 days postinfection (dpi). Dead larvae were removed and recorded, and survival curves were calculated.

Adult zebrafish were infected with M. marinum as described previously (27). Briefly, after being anesthetized with 0.1% 3-aminobenzoic acid ethyl ester (tricaine), adult zebrafish were infected by intraperitoneal injection with 10 μl of single-cell bacterial suspensions at a dosage of 10⁴ CFU per fish. For survival studies, 20 fish per group were infected with individual strains for each experiment. For the histopathological study, three fish were sacrificed at 10 days postinfection. Fish were first fixed for at least 96 h in 10% formalin, then dehydrated with ethanol, and finally embedded in paraffin. Serial paraffin sections (thickness, 5 μm) were prepared and were stained with a hematoxylin-eosin solution or with modified Ziehl-Neelsen staining (BA-409 TB stain kit; Baso, China) according to the manufacturer's instructions. Sections were examined under an Olympus BH2 microscope (Tokyo, Japan), and images were collected with a digital camera (TK-C1481BEC; JVC, Tokyo, Japan).

RESULTS

Interruption of cpsA influences colony morphology and growth in vitro.

To analyze the role of the cpsA gene, we utilized a transposon insertion mutant that had been generated in our laboratory previously (19). This mutant strain contains the Himar1 transposon with a kanamycin resistance marker inserted at nucleotide position 1204 in the cpsA gene (genome location, positions 6028733 to 6030268), just between the predicted LytR-CpsA-Psr domain and the LytRc domain (function unknown; this family domain appears at the C terminus of some LCP family proteins) of the protein encoded by cpsA (Fig. 1A). Compared to the WT parental strain of M. marinum, the cpsA mutant exhibited a rough, dry colony morphology (Fig. 1B). To confirm that the phenotype observed was a result of cpsA inactivation, the intact cpsA gene from M. marinum was cloned into the pSMT3LxEGFP shuttle vector to generate plasmid pMMcpsA. The complemented strain containing pMMcpsA displayed WT colony morphology.

FIG 1.

The M. marinum cpsA mutant has altered colony morphology and a growth defect in vitro. (A) Genetic organization of the cpsA locus in the genomes of M. marinum and M. tuberculosis, and the transposon insertion site (inverted triangle) in the cpsA mutant. Dark shaded arrows represent orthologs between M. marinum and M. tuberculosis. (B) Colony morphology of M. marinum strains cultured on Middlebrook 7H10 plus 10% OADC plates. The cpsA mutant colony is rough and dry, with a translucent border. (C to E) Bacterial growth was evaluated using the BacTiter-Glo assay. M. marinum strains were grown in a rich medium (7H9 broth plus 0.5% glycerol and 10% OADC) (C) or in a 7H9 salt solution containing 0.4% glycerol (D) or 0.4% acetate (E) as the sole carbon source. Samples were taken at various time points, and the BacTiter-Glo assay was performed according to the manufacturer's protocol. Luminescence readings (expressed as RLU) were recorded using a PerkinElmer X5 plate reader. Growth curves from two independent experiments are shown with standard deviations. Asterisks indicate significant differences (*, P < 0.05; **, P < 0.01; ***, P < 0.001), as determined by an unpaired Student t test, between wild-type M. marinum and the cpsA mutant strain.

To determine whether the inactivation of cpsA affected bacterial growth, we examined bacterial growth in vitro using the BacTiter-Glo assay. The cpsA mutant displayed a lower growth rate than the wild type in a rich medium (Middlebrook 7H9 plus 10% OADC) or in a defined synthetic medium containing glycerol or acetate as the sole carbon source (Fig. 1C to E). It is worth noting that when acetate was used as the carbon source, the growth difference between the WT and the cpsA mutant strain was much reduced (Fig. 1E).

The cpsA mutant exhibits altered cell surface properties.

The altered colony morphology of the cpsA mutant suggested that the inactivation of cpsA might be associated with an alteration of cell surface properties. First, the mutant strain exhibited a lower level of sliding motility than the WT strain (Fig. 2A and B). Second, the cpsA mutant aggregated more than the WT strain, and clumped, in the 7H9 culture (Fig. 2C). These phenomena could be linked to a change in cell surface hydrophobicity, which is known to play a major role in the intercellular adhesion of bacteria. Thus, we tested the relative hydrophobicity of the cell surface of each strain by using a two-phase aqueous hexadecane system. As shown in Fig. 2D, the cpsA mutant strain showed a higher affinity for hexadecane than the WT strain. These data indicated that the interruption of the cpsA gene affected the cell surface properties of M. marinum.

FIG 2.

The cpsA mutant exhibits altered cell surface properties. (A and B) The cpsA mutant is defective in sliding motility. (A) Illustration of sliding motility among M. marinum strains on 0.3% agarose plates. (B) Quantification of sliding motility among M. marinum strains. The diameter of each halo was measured. Data were collected from four plates. Asterisks indicate significant differences (**, P < 0.01; ***, P < 0.001) as determined by an unpaired Student t test. (C) Aggregation of cpsA mutant cells in 7H9 broth. (D) The hydrophobicity of the cell surface of the cpsA mutant is stronger than that of the wild-type strain. Data are from three independent experiments. Asterisks indicate significant differences (*, P < 0.05) as determined by an unpaired Student t test. (E) Whole-cell accumulation of Nile red by M. marinum was measured. Results are expressed as relative fluorescence units. Data are from three biological repeats. Asterisks indicate significant differences (**, P < 0.01; ***, P < 0.001) as determined by an unpaired Student t test. The cpsA mutant strain exhibited a more rapid and higher level of Nile red accumulation than the wild-type strain.

CpsA plays a role in cell wall permeability.

Since CpsA is associated with cell surface properties, the loss of CpsA could potentially compromise the integrity of the cell surface. Thus, we next investigated whether the cpsA mutant has altered susceptibility to various antibiotics. The WT, cpsA mutant, complemented strain, and vector control were used for this experiment. Interestingly, it was found that the cpsA mutant is more susceptible than the wild type to a hydrophobic drug, erythromycin, as well as to antibiotics targeting peptidoglycan synthesis (vancomycin and penicillin), which are hydrophilic (Table 1). However, the susceptibilities of the mutant strain to other antimicrobials, including isoniazid, streptomycin, ethambutol, and rifampin, remained unchanged (Table 1).

TABLE 1.

Susceptibilities of WT and mutant strains to antibiotics

Strain/plasmid	MIC (μg/ml)
	Erythromycin	Vancomycin	Penicillin	Streptomycin	Isoniazid	Ethambutol	Rifampin
WT	12.5	5	250	1.25	25	0.625	0.16
CpsA::Tn mutant	0.4	0.625	62.5	1.25	25	0.625	0.16
CpsA::Tn mutant/pMMCpsA	12.5	5	250	1.25	25	0.625	0.16
CpsA::Tn mutant/pSMT3	0.4	0.625	62.5	1.25	25	0.625	0.16

To determine whether the increased drug susceptibility of the cpsA mutant strain is caused by a general increase in cell wall permeability, we used fluorescence spectroscopy to measure whole-cell accumulations of ethidium bromide (EtBr) and Nile red, representatives of hydrophilic and hydrophobic compounds, respectively. It was shown that Nile red accumulated to a higher level in the cpsA mutant than in the WT strain (Fig. 2E), whereas there was no difference between the two in the accumulation of EtBr (data not shown). Together with the changes observed in sliding motility, colony morphology, and cell surface hydrophobicity, these findings suggest that the cspA mutant displays an altered cell envelope structure or composition that affects its permeability to hydrophobic antibiotics, including erythromycin.

The cpsA mutant has decreased AG content.

The phenotypes of the cpsA mutant presented above prompted us to investigate which cell wall component(s) had changed in the cpsA mutant. On the basis of genetic and structural evidence, Kawai and colleagues proposed a model in which LCP proteins in Bacillus subtilis catalyze the transfer of bactoprenol phosphate-linked teichoic acid precursors to peptidoglycan (5), and several reports on LCP proteins in various bacterial species supported this model (17, 28–30). However, the cell wall of M. marinum (or M. tuberculosis) lacks teichoic acids; instead, it has arabinogalactan (AG), which is synthesized on a bactoprenol phosphate carrier lipid prior to being attached to peptidoglycan as teichoic acids are. Therefore, we hypothesized that mycobacterial CpsA might be involved in the ligation of arabinogalactan to peptidoglycan. This hypothesis was tested by analyzing the glycosyl composition of WT and cpsA mutant cells collected at logarithmic or stationary phase upon growth on different carbon sources. The glycosyl composition was standardized to the N-acetylglucosamine content of the cells, since N-acetylglucosamine is the major glycosyl component of peptidoglycan. As shown in Fig. 3A to D, the cpsA mutant showed reproducible reductions in the levels of arabinosyl, galactosyl, and rhamnosyl residues relative to the N-acetylglucosamine level. Moreover, the Ara/Gal ratios in the WT and cpsA mutant strains were equivalent, suggesting that the AG produced by the mutant was similar in structure overall to that produced by the WT strain. These results point to a possible role of CpsA in the transfer of arabinogalactan to peptidoglycan.

FIG 3.

Glycosyl composition of the mycobacterial cell wall. (A to D) The contents of arabinosyl, galactosyl, and rhamnosyl residues relative to that of N-acetylglucosamine in the mycobacterial delipidated cell wall were analyzed by GC-MS. (A and B) Bacterial cells collected from logarithmic-phase (A) or stationary-phase (B) cultures with 0.4% glycerol as the sole carbon source. (C and D) Bacterial cells collected from logarithmic-phase (C) or stationary-phase (D) cultures with 0.4% acetate as the sole carbon source. (Insets) Changes in rhamnose contents are shown on a smaller scale for legibility. Three separate samples from M. marinum strains were analyzed. Data were generated from three analyses. Asterisks indicate significant differences (*, P < 0.05; **, P < 0.01; ***, P < 0.001) as determined by an unpaired Student t test.

The cpsA mutant exhibits impaired growth within macrophages.

Macrophages are the primary target of pathogenic mycobacteria, and the interaction between bacteria and macrophages in the early phase of infection is crucial for the pathogenesis of the disease (31). First, macrophage infection experiments were performed, and the intracellular growth of the WT was compared with that of the mutant. Intracellular growth was assayed by enumerating the CFU at different time points postinfection. The results showed that the level of replication of the cpsA mutant in macrophages was significantly lower than that of the WT over the course of a 4-day infection (Fig. 4A). In contrast, the mutant strain complemented with an intact cpsA gene from either M. marinum or M. tuberculosis proliferated similarly to the wild-type strain, showing that growth inside macrophages was attenuated by the inactivation of cpsA (Fig. 4A).

FIG 4.

The cpsA mutant shows attenuated intracellular growth and invades macrophages more efficiently. (A) Growth of the cpsA::Tn strain in RAW 264.7 macrophages. Macrophages were infected with the M. marinum WT, cpsA::Tn mutant, or complemented strain at an MOI of 1. Infected cells were lysed at the indicated time points for CFU counts. Data were generated from four independent experiments. A significant difference between the M. marinum WT and cpsA::Tn strains (P < 0.0001) was determined by an unpaired Student t test. (B) Invasion ratios of the M. marinum WT, cpsA::Tn mutant, and complemented strains at day 0 (6 h postinfection). An MOI of 10 was used for infection. Data were generated from three independent experiments. Asterisks indicate significant differences (*, P < 0.05; **, P < 0.01) as determined by an unpaired Student t test.

Then the invasion ratio was measured in order to determine whether CpsA plays a role in phagocytosis. The results showed that inactivation of cpsA significantly increased the phagocytosis of M. marinum by macrophages, since the CFU count recovered from macrophages infected with the mutant strain was significantly higher (P < 0.05) than that for the WT strain (Fig. 4B). This phenotype could be complemented by expressing the cpsA gene of M. marinum or M. tuberculosis in the mutant strain.

To explore whether CpsA affects the cytokine profiles of infected macrophages, macrophages were infected with various strains, and cell culture supernatants were collected and were analyzed by ELISA. Macrophages infected with the cpsA mutant strain secreted significantly lower levels of the proinflammatory cytokines tumor necrosis factor alpha (TNF-α) and interleukin 6 (IL-6) than cells infected with the WT strain (Fig. 5A and B), whereas gamma interferon (IFN-γ), IL-17A, and IL-10 were not at detectable levels (data not shown).

CpsA is functional in arresting macrophage phagosome maturation.

Phagosome-lysosome fusion usually leads to mycobacterial killing. However, pathogenic mycobacteria have the ability to resist the delivery of cellular phagosomes to lysosomes, thereby creating a protective environment that facilitates bacterial survival and replication (32). Thus, we sought to determine if the attenuation of the cpsA mutant was due partly to increased fusion of bacillus-containing phagosomes with lysosomes. Macrophages were infected with M. marinum strains carrying GFP and were examined by immunostaining for colocalization of bacterium-containing phagosomes with LAMP-1, a marker for late endosomes and lysosomes. As shown in Fig. 6A and B, more fusion with lysosomes was observed for phagosomes containing the cspA mutant than for phagosomes containing the WT. This result indicated that the cpsA mutant is not as proficient as the WT strain at preventing the fusion of phagosomes with lysosomes, a defect that could partially contribute to the proliferation defect of the cpsA mutant in macrophages.

FIG 6.

CpsA is involved in arresting phagosome maturation. (A and B) Confocal microscopy analysis of colocalization of LAMP-1 with GFP-labeled bacteria. (A) Representative micrographs of RAW 264.7 macrophages infected with either WT M. marinum, the cpsA mutant, or a complemented strain. LAMP-1 was visualized by staining with a primary antibody followed by an Alexa Fluor 594-conjugated secondary antibody. (B) The colocalization of GFP-labeled bacteria and LAMP-1 was quantified at 4 h postinfection for three independent experiments. At least 200 macrophages were scored for each experiment, Asterisks indicate significant differences (**, P < 0.01) as determined by an unpaired Student t test.

CpsA is required for the virulence of M. marinum in zebrafish.

To evaluate the virulence of the cpsA mutant in vivo, studies were performed first in the zebrafish larva infection model. The zebrafish larva, which has only innate immunity in the experimental period, is transparent, so examination of the real-time progression of microbial infections is feasible. Approximately 100 CFU of GFP-labeled mycobacteria was microinjected into the caudal veins of larvae at approximately 30 h postfertilization. The infected larvae were monitored for survival and bacterial loads. Over a 2-week time course, it was found that the cpsA mutant is significantly attenuated relative to the WT strain, as evidenced by the fact that the mean time to death for larvae infected with the mutant strain was significantly longer than that for WT-infected larvae (Fig. 7C). In agreement with this finding, the replication of the cpsA mutant in larvae was also severely impaired (Fig. 7A and B).

FIG 7.

The cpsA mutant is attenuated in zebrafish. (A to C) Infection of zebrafish larvae with GFP-labeled M. marinum WT and cpsA mutant strains. (A) Representative fluorescent images of individual infected larvae at 5 days postinfection (right) and their overlays with transmission images (left). (B) Bacterial burdens of infected larvae were determined by green pixel intensity counts. Asterisks indicate significant differences (**, P < 0.01; ***, P < 0.001) as determined by an unpaired Student t test. (C) Survival of larvae (n, 25 for each strain) infected with 100 CFU of the wild-type or cpsA mutant strain. The hazard ratio was 3.89, and the P value was <0.0001, by the Kaplan-Meier method with a log rank test. These results are representative of two independent experiments. (D to F) The cpsA mutant is attenuated in adult zebrafish. (D) Images of adult zebrafish at 10 days after infection with WT M. marinum or the cpsA mutant. (E) Survival curves of adult zebrafish during a 30-day infection. Fish were intraperitoneally infected with 10⁴ CFU of WT M. marinum, the cpsA mutant, or the cpsA::Tn mutant complemented with the cpsA gene from M. tuberculosis or M. marinum. The results are from one experiment representative of three independent experiments. The difference between the WT- and cpsA mutant-infected groups was significant (P < 0.0001) as determined by the Kaplan-Meier method with a log rank test. (F) Histopathology of zebrafish at 10 days after infection with the WT, cpsA mutant, or complemented strain. Three fish from each group were sacrificed and were subjected to histological analysis with hematoxylin-and-eosin staining and Ziehl-Neelsen staining. Arrows indicate bacteria inside granulomas. Magnification, ×40 for all sections.

To further assess the role of CpsA in virulence, adult zebrafish, which have both innate and adaptive immune systems, were infected with either wild-type M. marinum, the cpsA mutant, or the mutant complemented with the intact cpsA gene or a cloning vector (20 fish infected per strain), and the survival curves of each group were calculated (Fig. 7E). Zebrafish infected with the wild-type strain or the complemented strain with the intact cpsA gene died within 16 days postinfection, whereas zebrafish infected with the cpsA mutant remained alive at the time when the experiment was terminated (1 month postinfection) (Fig. 7E). In the group infected with the mutant complemented with the cloning vector, only one fish died, on the 15th day postinfection; this may have been an accidental death (Fig. 7E). Moreover, the gross pathology and histopathology of infected fish showed a distinct difference between the wild-type- and cpsA mutant-infected groups (Fig. 7D and F). Fish infected with the wild type or the complemented strain exhibited extensive inflammation and granuloma formation associated with a larger number of bacteria (Fig. 7F, left and right). In contrast, no pathological change was observed in zebrafish infected with the cpsA mutant; no granulomas formed; and fewer bacteria were detected (Fig. 7F, center). Taken together, these differences indicate that CpsA is required for the virulence of M. marinum in zebrafish.

DISCUSSION

The cpsA gene of M. marinum (or its ortholog in M. tuberculosis) is annotated to encode a hypothetical protein containing a LytR-CpsA-Psr domain, but its function remains a mystery. In this study, we explored the biological functions of the cpsA gene in M. marinum. Disruption of cpsA resulted in decreased arabinogalactan levels in the bacterial cell wall and changes in colony morphology, sliding motility, cell wall hydrophobicity, and susceptibility to antibiotics. The cpsA mutant was not as proficient as the WT strain at preventing the fusion of phagosomes with lysosomes and induced significantly lower levels of the proinflammatory cytokines TNF-α and IL-6. Furthermore, it was shown that cpsA is essential for mycobacterial proliferation in macrophages and for virulence in zebrafish. Since CpsA is conserved between M. marinum and M. tuberculosis, and the virulence of the cpsA mutant can be complemented by the M. tuberculosis ortholog Rv3484, we speculate that CpsA also plays an important role in pathogenesis caused by M. tuberculosis.

In our study, a decrease in the arabinogalactan content was observed in the M. marinum cpsA mutant. The arabinogalactan in mycobacteria is synthesized on a decaprenyl phosphate carrier lipid prior to being attached to peptidoglycan (PG). Intriguingly, previous studies had reported that LytR-CpsA-Psr family proteins in Bacillus subtilis catalyze the attachment of teichoic acid to cell wall peptidoglycan (5). In agreement with this model, the crystal structure of recombinant Cps2A in Streptococcus pneumoniae showed that the LCP domain interacts strongly with undecaprenyl phosphate (5). These data suggest that CpsA is likely to participate in the ligation of AG to PG in mycobacteria (Fig. 8). Besides, it should be noted that the cpsA mutant loses AG only partly, suggesting that compensatory activities must exist in this mutant. Previous studies with other bacteria also showed some functional overlap between LCP proteins (5, 16, 28, 33). However, the cpsA mutant exhibited many defects, including the attenuation of virulence, indicating that other mycobacterial LCP proteins only partially compensate for the function of CpsA in M. marinum. Further studies are needed to figure out the detailed functional differences between CpsA and other LCP proteins in mycobacteria.

FIG 8.

Schematic representation of the function of CpsA and decaprenyl phosphate recycling. R, unfinished arabinan chain; PIMs, phosphatidylinositol mannosides; LOS, lipooligosaccharides; IPP, isopentenyl pyrophosphate; DMAPP, dimethylallyl pyrophosphate; MEP, methylerythritol phosphate.

Members of the LCP family proteins are generally associated with cell surface properties. For example, the loss of BrpA, an LCP protein, in S. mutans results in increased cell surface hydrophobicity (34), and an msrR mutant of S. aureus exhibits enhanced aggregation in brain heart infusion (BHI) broth and reduced surface spreading on BHI agar plates (17). In our study, we also found that inactivation of cpsA in M. marinum altered bacterial cell surface properties, including sliding motility, aggregation, and cell surface hydrophobicity. Interestingly, cpsA disruption resulted in increased susceptibility to the hydrophobic antibiotic erythromycin and increased uptake of the hydrophobic stain Nile red, which could be due to the partial loss of hydrophilic arabinogalactan in the cell wall of the cpsA mutant. In agreement with this view, a previous study showed that mycobacterial cells were more susceptible to erythromycin and more permeable to a hydrophobic compound (β-sitosterol) after treatment with ethambutol, an inhibitor of arabinogalactan biosynthesis (35). Moreover, in agreement with the reduced β-lactam resistance shown in S. aureus LCP mutants (17, 36), the cpsA mutant of M. marinum exhibited higher susceptibilities to antibiotics targeting peptidoglycan synthesis (vancomycin and penicillin) than to other hydrophilic antibiotics, including streptomycin, isoniazid, and ethambutol, which could be caused by increased exposure of the targets of vancomycin and penicillin after the impairment of the attachment of arabinogalactan to peptidoglycan.

Polyprenyl phosphate is a low-abundance lipid, and its availability may be rate limiting for bacterial cell wall peptidoglycan synthesis (37). The failure of recycling of mycobacterial decaprenyl phosphate also causes a severe growth defect due to interference with peptidoglycan precursor (lipid II) biosynthesis (38). We found that the cpsA mutant exhibited a growth defect in vitro in this study and that MMAR_1274 (another LCP gene) is an essential gene for M. marinum growth in vitro (unpublished data). It is likely that the impairment of the ligation between AG and PG by the inactivation of LCP proteins causes an accumulation of decaprenyl-linked intermediates, thus disrupting the recycling of decaprenyl phosphate (Fig. 8). This is consistent with a previous study which showed that an LCP mutant of S. aureus is hypersensitive to bacitracin, a polypeptide that inhibits the recycling of undecaprenyl phosphate by preventing the dephosphorylation of undecaprenyl pyrophosphate, and that the inhibition of TagO (the first teichoic acid synthesis enzyme) can partly rescue the growth defect caused by the loss of LCP proteins (enzymes that transfer teichoic acids from undecaprenyl phosphate to peptidoglycan) in S. aureus by eliminating the accumulation of undecaprenyl-teichoic acid intermediates (39).

The cell envelopes of mycobacteria can contribute significantly to their virulence, and many mycobacterial mutants that possess an impaired cell envelope show attenuated virulence in vitro or in vivo (22, 40–46). A previous study comprehensively identified the genes required by M. tuberculosis for optimal growth by using transposon sequencing (Tnseq), and those investigators speculated that Rv3484 (cpsA) is an essential gene for M. tuberculosis growth in vivo (47). Our study also strongly demonstrated that cpsA is crucial for the virulence of M. marinum in zebrafish. Several potential mechanisms contribute to this outcome. First, a growth defect of the cpsA mutant in vitro or in macrophages could be associated with its attenuation in zebrafish. Second, the cpsA mutant is less efficient than the wild type at arresting the fusion of phagosomes and lysosomes. Thus, the cpsA mutant might end up in a compartment that contains higher concentrations of host defense molecules. Third, it is possible that the cpsA mutant is more sensitive than the wild type to the innate antimicrobial mechanisms of macrophages, including acid killing, reactive oxygen species (ROS), reactive nitrogen species (RNS), defensins, and lysozyme. We find that the mutant is more susceptible than the wild type to vancomycin, whose target is peptidoglycan, a target similar to that of lysozyme. It has also been reported that an S. mutans LCP mutant displayed drastic reductions in acid and oxidative stress tolerance responses (9). Finally, excessive proinflammatory responses can lead to deleterious effects on the host, and emerging evidence suggests that a high level of TNF-α could accelerate the disease progression of tuberculosis (48–50). Therefore, the higher levels of TNF-α and IL-6 production induced by the WT strain might contribute to more-extensive tissue damage and bacterial dissemination in zebrafish infected with the WT than in those infected with the mutant strain.

In conclusion, the CpsA protein is indispensable for maintaining the integrity of the mycobacterial cell wall and plays an important role in mycobacterial pathogenesis. Considering both its essential role for mycobacterial growth in vivo and the extracytoplasmic membrane location of the active sites of LCP proteins, CpsA could be an attractive new target for antimycobacterial drugs.