The aceE involves in mycolic acid synthesis and biofilm formation in Mycobacterium smegmatis

Background The integrity of cell wall structure is highly significant for the in vivo survival of mycobacteria. We hypothesized that changes in morphology may indicate changes in cell wall metabolism and identified an aceE gene mutant (aceE-mut) which presented a deficient colony morphology on 7H10 agar by screening transposon mutagenesis in Mycolicibacterium smegmatis, basonym Mycobacterium smegmatis (M. smegmatis). This study aimed to identify the functional role of aceE gene in cell wall biosynthesis in M. smegmatis. Results We observed that the colony morphology of aceE-mut was quite different, smaller and smoother on the solid culture medium than the wild-type (WT) strain during the transposon library screening of M. smegmatis. Notably, in contrast with the WT, which aggregates and forms biofilm, the aceE-mut lost its ability of growing aggregately and biofilm formation, which are two very important features of mycobacteria. The morphological changes in the aceE-mut strain were further confirmed by electron microscopy which indicated smoother and thinner cell envelope images in contrast with the rough morphology of WT strains. Additionally, the aceE-mut was more fragile to acidic stress and exhibited a pronounced defects in entering the macrophages as compared to the WT. The analysis of mycolic acid (MA) using LC-MS indicated deficiency of alpha-MA and epoxy-MA in aceE-mut strain whereas complementation of the aceE-mut with a wild-type aceE gene restored the composition of MA. Conclusions Over all, this study indicates that aceE gene plays a significant role in the mycolic acid synthesis and affects the colony morphology, biofilm formation of M. smegmatis and bacteria invasion of macrophage.


Background
The mycobacterial cell wall is mainly composed of three types of macromolecules i.e. peptidoglycan, arabinogalactan and mycolic acids (specific components shared by the members of the order Corynebacterineae (e.g., mycobacteria, nocardia, and rhodococci)) [1]. Glycolipids, porins and lipoarabinomannan as well as its variants, which are anchored to the cell membrane by diacylglycerol, are also essential components [2]. The carboxyl group of mycolic acids is vertically covalently linked to the hydroxyl group of arabino-galactan by ester bond, arabino-galactan is linked to the peptidoglycan layer by phospholipid bond, whereas other glycolipids and free lipids are regularly distributed in the thicker layer of mycolic acids [3][4][5]. These collectively form a thick, dense, poorly permeable cell wall, which not only allows mycobacteria to resist the dry environment and harmful chemicals but also allows it to reproduce in the macrophages [4,6]. Therefore, the molecules involved in the cell wall biosynthesis of TB bacilli have been considered as important anti-TB drug targets. Among the existing anti-TB drugs, isoniazid, ethionamide and ethambutol target the cell wall synthesis pathways, among which isoniazid and ethambutol are the first-line anti-TB drugs.
Pyruvate dehydrogenase (PDH) is an enzyme complex that catalyzes the conversion of pyruvate into acetyl-coA in vivo. The complex mainly consists of three enzymes, which are respectively called E1, E2 and E3 components of PDH according to the order in which they participate in the reactions. Through a series of chemical reactions of pyruvate decarboxylation, the glycolytic pathway (the final product is pyruvate) and the tricarboxylic acid cycle (the initial reactant is acetyl-coA) can be effectively connected [7]. As an important intermediate metabolite, acetyl-coA not only participates in tricarboxylic acid cycle as well as the glyoxylate cycle but also provides carbon source for the synthesis of mycolic acid and lipids [8,9]. Studies have also found that the genes involved in gluconeogenic pathway and glyoxylate cycle are up-regulated in Mtb isolated from macrophages, mouse lung tissues and tissue samples of patients, suggesting that the compensatory metabolism of acetyl-coA is necessary for intracellular growth and persistence in vivo [10,11].
In the present study, we found a mutant strain in which the aceE gene, encoding the E1 component of PDH, was inactivated by Himar1 transposon insertion (aceE-mut). This mutant had obvious differences in colony morphology (smaller plaque, edge smooth and round) and defects in biofilm formation in contrast with the wild-type (WT) strains. Further analyses indicated that the aceE gene deficiency affected the cell mycolic acid profile of M. smegmatis.

Results
The aceE-Mut exhibited unusual colony morphology By screening the M. smegmatis transposon library, a transposon mutant showed obvious differences in colony morphology (small, smooth without jagged edges and yellow color) on agar plate when compared to the parental WT M. smegmatis strain (Fig. 1a). DNA sequencing and analysis of the MycoMar/M. smegmatis chromosomal junction revealed that the transposon mutant had an insertion at a TA dinucleotide within the aceE gene and the distance of the Tn insertion from the start codon of the aceE gene is 2243 bp (Fig. 1b and c). No apparent difference in growth rate was found between the WT and aceE-mut strain with the Middlebrook 7H9 neutral medium culturing ( Fig. 2a and b). However, we observed that the aceE-mut strain dispersed uniformly in the broth without Tween-80, while the WT M. smegmatis formed more clumps in the same medium. Notably, the growth rate of aceE-mut was significantly Fig. 1 The identification of M. smegmatis aceE-mut. a. The aceE-mut showed smoother colony morphology in contrast to WT. The culture of M. smegmatis mc 2 155 and aceE-mut were 10× diluted and 10 μL culture aliquots were spotted on the Middlebrook 7H10 medium supplemented with 0.2% glycerol. The images were taken after incubation at 37°C for 3 days on 7H10 plates. b. HimarI transposon insertion site in aceE gene; c. PCR verification of the aceE transposon mutant. WT: M. smegmatis mc 2 155; aceE-mut: aceE gene deficiency mutant selected from M. smegmatis mc 2 155 transposon library lower than the WT strain in acidified 7H9 media during logarithmic and stationary phases ( Fig. 2c and d). Comp strain demonstrated similar growth dynamics as WT in both culture conditions. These results demonstrated that aceE gene is dispensable in M. smegmatis and the disruption of aceE gene in M. smegmatis renders the bacteria more sensitive to acid stress, and this feature could be complemented with the wild-type copy of aceE gene.

Pellicle and biofilm formation defected in aceE-Mut
In contrast to the significant pellicle growth that appeared on the air-liquid interface in the WT strain culture, pellicle was absent from the aceE-mut strain culture when grown in 7H9 medium without Tween-80 supplement and shaking (Fig. 3a). In order to better quantify the biofilm formation defects in aceE-mut, strains were cultivated in M63-based liquid medium and biofilm formation was analyzed visibly as well as using the crystal violet assay. Consistently, aceE-mut did not form biofilm in the M63-based liquid medium either. Notably, the complementary expression of aceE gene in aceE-mut strain restored the pellicle and biofilm formation to the WT levels (Fig. 3b). Taken together, these results indicated that aceE gene is involved in pellicle and biofilm formations. The surface morphology of WT and aceE-mut was observed by scanning electron microscopy. The aceE-mut cells were slightly slender than WT cells, and their surfaces were smoother than WT cells (Fig. 4). These microscopic observations were consistent with the smooth phenotype on agar plate. The cell morphology and cell wall architecture were further examined by transmission electron microscopy, which showed that the cell wall of the mutant was thinner (P < 0.01) than WT and Comp strains (Fig. 5).

Cell wall permeability analysis
In order to analyze the effect of disruption of the aceE gene on cellular response against stress conditions in vitro, the aceE-mut and WT strains were treated with commonly used anti-TB drugs and several antimicrobial agents in vitro. No significant differences in antimicrobial sensitivity tests between WT and mutant strains were observed (Supplementary Table S1 and Supplementary Fig. S1).

aceE-Mut has a defect in invasion of macrophages
The role of aceE in cell wall integrity and the M. smegmatis resistance against acidic environment suggests that it may also protect the bacilli from microbicidal activity Fig. 2 The aceE-mut strain was sensitive to acid stress. a and b. Bacterial strains were grown in the neutral Middlebrook 7H9 medium supplemented with 0.05% Tween-80 and 0.2% glycerol; c and d. Bacterial strains were grown in the acified Middlebrook 7H9 medium (pH 5.0) supplemented with 0.05% Tween-80 and 0.2% glycerol. The OD 600 and CFU were determined at an interval of 24 h. The graph is a representation of one of three independent experiments. The mean ± SD of triplicate experimental samples is shown from one experiment, Error bars represent SD of macrophages. Therefore, we investigated and compared the impact of aceE inactivation on invasion and intracellular survival of mutants with its corresponding complemented strain (Comp) and WT strain. As shown in Fig. 6, aceE-mut strains and its parent strain showed different abilities to invade THP-1 macrophages after 2 h of incubation. aceE-mut exhibited pronounced defects in entering the macrophages as compared to the WT and Comp strains. A significant difference between the mutant and WT strains of approximately 0.5 log unit was observed at t = 0 h post-infection (P < 0.05 at MOI = 1:1; P < 0.01 at MOI = 1:10). However, CFU of the mutant infection group was not lower than the WT infection group at t = 24 h.

Disruption of aceE gene affected the mycolic acid composition
The mycolic acid (C 60~C90 ) composition was analyzed by HPLC using Mycobacteria Identification System. In contrast to WT strain, the mutant strain possessed higher proportion of short-chain mycolic acids but lower proportion of long-chain mycolic acids ( Supplementary  Fig. S2). The percentage of mycolic acid with equivalent carbon length of 5 to 8 in the mutant strain was significantly lower than the wild strain (P < 0.05), suggesting a potential role of aceE gene in mycolic acid metabolism. A further LC-MS-based systematic analysis of mycolic acid and lipids did not identify any obvious differences in the phospholipid and glycolipid compositions (P > 0.05) between the WT and aceE-mut strains (Fig. 7a), whereas certain kinds of α-mycolic acids were deficient in aceE-mut strains (P < 0.05) and these changes were fully restored upon complementation ( Fig. 7b and c).

Discussion
Transposon mutagenesis has been used extensively as a useful tool for studying gene function of mycobacteria. In this study, we identified an aceE gene mutant presenting deficient colony morphology on 7H10 agar using transposon mutagenesis method. Hence, it is reasonable to speculate defective cell wall biosynthesis in the mutant. Changes in the structure of cell envelope components may affect the normal physiological metabolic processes in mycobacteria including, transportation of antimicrobial agents across the plasma membrane and mycobacterial survival in stress conditions, etc.
Previous studies have shown that mycobacterial PDH genes (aceE, dlaT and lpd) are not only involved in pyruvate metabolism, but may also have more complex biological functions [12][13][14][15][16]. It has been observed that dlaT gene knockout significantly affects the in vitro growth of TB bacilli in the standard medium. The mutant is not only sensitive to reactive nitrogen intermediates but its virulence is also reduced in the infected mice. In addition, the study found that dlaT inhibitors can selectively kill non-self-replicating TB bacilli, suggesting that dlaT gene may be associated with latent TB infection in host cells [12,14,15]. Study also showed that the disruption of lpd gene in Mtb leads to decreased virulence with lower bacterial load in lung and other organs in infected mice. Other study revealed that the LPD is a component of branched chain ketones acid dehydrogenase, which takes part in succinyl CoA metabolism of amino acids such as valine (Val), isoleucine (Ile) [16].
However, the biological function of the E1 component of mycobacterial PDH remains poorly understood. In 2005, Tian et al., confirmed for the first time that aceE is Fig. 3 The effect of aceE deficiency on pellicle and biofilm formation of mycobacteria. a. aceE-mut is defective in pellicle formation, while the complementary expression of aceE gene in aceE-mut stain can recover the formation of pellicle in 7H9 medium without Tween-80 in standing culture. Each experiment performed in triplicate. b. Quantification of the biofilm formation after crystal violet staining. Mean optical density for five biological replicates per strain ± SD for a representative experiment from 3 experiments is shown. Significant differences were determined by One Way ANOVA and are indicated by *** (P < 0.001). Error bars represent SD the gene encoding the E1 component of PDH complex in Mtb [17]. In 2008, Li et al., found that the expression level of aceE gene in H37Rv was significantly higher than that of H37Ra during the course of macrophages infection [18]. This result is consistent with the results of gene chip analysis conducted by Manganelli et al., in 2001 [19]. Earlier studies had shown that H37Rv respiration is stronger than H37Ra. Although both, virulent H37Rv and the non-virulent H37Ra strains, rely on glycolysis and aerobic respiration for glucose metabolism, it is believed that glucose metabolism in H37Rv may be more dependent on the glycolytic pathway [20,21]. Therefore, it is speculated that the up-regulated expression of aceE gene enables it to not only participate in the aerobic respiration as a component of PDH, but also guarantees successful glycolysis under the hypoxic conditions and thus provides excellent energy supply for growth of the virulent H37Rv strain. Another study showed that AceE component of PDH forms a fourcomponent peroxidase system with DlaT/AhpD/AhpC, which assists the reductase action using pyruvate as a source of electrons [22], and thus protect the pathogen by participating in antioxidant of antinitraxidative defense. All of the above studies suggest that aceE gene plays an important role in Mtb metabolism in vitro and in vivo.
In the present study, we screened the random transposon mutants and found that the inactivation of aceE gene affects the colony morphology and biofilm formation in M. smegmatis which suggests its potential to affect the lipid metabolism and cell wall biosynthesis. The findings obtained in our study are in agreement with Viswanathan et al. [23]. Integrity of the cell wall has important biological significance for the in vitro and in vivo survival of bacteria. To further investigate the characteristics of the aceE-mut, series of assays were performed to compare the phenotypes of aceE-mut with the WT strains. We further analyzed the key composition of mycolic acids in each strain and identified altered composition in the aceE-mut which suggests that this gene may be involved in the metabolism of mycolic acid in mycobacteria. Therefore, it was necessary to thoroughly analyze the cell wall composition of both WT and aceE-mut strains and explore the role of aceE gene in the process of cell wall synthesis and metabolism. Comparative analysis of the lipid and mycolic acid profiles of the M. smegmatis aceE-mut, M. smegmatis WT and the Comp using LC-MS indicated that the metabolism of certain alpha-MA and epoxy-MA was deficient in aceE-mut, which demonstrated that the role of aceE was associated with mycolic acids, but not with other fatty acids found in M. smegmatis. In addition, we observed that the aceE-mut possess a distinct cell morphology and ultrastructural appearance compared with the WT strain when grown in broth culture, that may result from the inability of the mutants to synthesize certain kinds of alpha-MA. Thus, the inactivation of aceE also impacts bacterial physiology that ranges from reduced biofilm formation to changes in the cell morphology and cell wall thickness. Consistent with other studies, our study also found that the loss of long chain mycolic acids or oxygenated mycolic acids effects biofilm formation [24][25][26]. Previous study by Trivedi et al., showed that H37Rv are of marginally smaller size during biofilm formation, however, it was difficult to compare the differences in cell size between the aceE-mut and WT strains in our study because of the morphological differences between these two strains (aceE-mut strains is more slender than WT strains). Most importantly, aceEmut was more susceptible to acidic environments than the parental M. smegmatis, suggesting plausible role of aceE gene in stress tolerance inside the host.
aceE gene is more readily expressed in the virulent Mtb H37Rv (than H37Ra) throughout the course of infection, that not only suggests its important role in the virulence, survival and persistence of Mtb, but also makes this gene a potential target for the development of newer vaccines and anti-TB drugs. In M. smegmatis, the full length of the MSMEG4323 (aceE) gene is 2790 bp. Using blast tools, aceE gene sequence of M. smegmatis was analyzed which highlighted its conservation in M. smegmatis and a similarity of 82% with the aceE gene in Mtb. As a model bacterium widely used in the study of functional genes of mycobacterium, M. smegmatis can also be used to investigate the potential role of aceE gene in mycobacterial cell wall biosynthesis. In the present study, the stress assay demonstrated that aceE gene helped mycobacteria to withstand acidic stress environment that also suggests its plausible role in stress tolerance inside the host. However, macrophage infection study showed that the inactivation of aceE gene in M. smegmatis does not affect bacterial proliferation in macrophages, but rather affects the ability of M. smegmatis to invade the macrophages. Therefore, we suggest that the aceE gene is a virulence factor of M. smegmatis that may be important in the initiation of infection in vivo. Since M. smegmatis does not have pathogenicity, the study using this bacterium model cannot reveal the possible function of aceE gene in Mtb pathogenesis. Therefore, the role of aceE gene in pathogenesis requires to be further explored using the virulent Mtb H37Rv strain.

Conclusions
An aceE mutant M. smegmatis mc 2 155 strain selected from transposon library presented small, smooth morphology without jagged edges. Compared with its parental WT strain, aceE-mut lost the ability of growing aggregately as well as biofilm formation, and became more fragile to acidic stress. Additionally, alteration of the mycolic acid profile in aceE-mut may directly impact the overall cell wall morphology and acid sensitivity. All these changes of the mutant strain demonstrate that aceE gene inactivation reduces biosynthesis of α-MA, affects the integrity of mycobacterial cell wall, and decreases invasion of macrophage. Since acetyl-coA is an important precursor for the biosynthesis of mycolic acid, we hypothesized that the inactivation of aceE gene may lead to the restriction of acetyl-coA synthesis, thus affecting the biosynthesis of certain mycolic acids. Further study should be conducted to address this issue.

Strains, medium, condition
A transposon library was generated using M. smegmatis mc 2 155 as previously described [27,28] and plated on 7H10 agar containing 20 mg/L kanamycin. Approximately, 1000 single colonies of variable sizes were randomly placed into 96 deep well plates containing 0.5 mL of Middlebrook 7H9 medium (BD Difco) containing kanamycin and grown at 37°C. After 5 days of growth, cultures from each well were spotted on the Middlebrook 7H10 agar (BD Difco) plates, and more than 6 mutants were found for colony morphology defects. E. coli strain DH5α pir 116 (kindly provided by Dr. Kaixia Mi) was used to identify the insertion site of the transposon mutant. E. coli strain Top10 (TransGen Biotech, China) was used to clone specific DNA fragments into pSMT3 plasmid (Table 1). When required, kanamycin (50 mg/L for E. coli and 20 mg/L for mycobacteria) and hygromycin (150 mg/L for E. coli and 75 mg/L for mycobacteria) were added to the growth medium.

Transposon identification
To identify mutants with growth defects, genomic DNA was prepared from the selected transposon mutant. The genomic DNA was randomly digested with BamHI (Fermentas International Inc.) and then purified with a DNA extraction kit (Fermentas International Inc.). The purified DNA was ligated and transformed into DH5α pir116 competent cells. The plasmids from the kanamycinselected positive colony were isolated and sequenced with following primers: TLP1 5′-GCTGACCGCTTCCTCGTG CTTTA-3′; TLP2 5′-GCAGCGCATCGCCTTCTATC-3′.

Construction of complemented strain of aceE-Mut
For complementation of aceE-mut strain, 2.79 kb fulllength aceE gene (MSMEG_4323) from M. smegmatis was cloned into the mycobacterial shuttle vector pSMT3 [30] using NEBuilding pfu kit (New England Biolabs, Ipswich, MA), and pSMT3-aceE was generated ( Table 1). The plasmid pSMT3-aceE was subsequently transformed into  Table 1). The transformants were selected on 7H10 agar plates, supplemented with 20 mg/L kanamycin and 75 mg/L hygromycin, followed by incubation at 37°C for 3-4 days. The positively grown colonies were picked and identified by PCR-sequencing methods using following primers:

Colony morphology observation
To compare the colony sizes for different mycobacterial strains, log phase cultures were 10× serially diluted (1:10), grown on 7H10 medium at 37°C and examined visually for any change. Photographs were taken after 3-4 days of incubation using stereo microscope (Leica MZ APO).

Morphological observation by electron microscopy
Mycobacteria from log phase were harvested and washed with 0.1 M phosphate buffer (PBS). Cells were subsequently fixed using 2.5% glutaraldehyde. Post fixation was carried out in 1% osmic acid. Following several rinses with ddH 2 O, samples were dehydrated in a series of different concentrations of ethanol and 100% acetone. For transition solvent, resin: acetone (2:1) were used overnight. Epoxy resin-812 was used for 1 h for embedding. 90 nm sections were cut and stained with uranyl acetate and Reynold's lead citrate (Ted Pella, Inc). After drying, transmission electron microscopic (TEM) images of the sections were taken using TEM-1400plus. The cell wall thickness was measured for each strain as follows: 100 mycobacteria in the visual field were randomly selected, the largest distance between the outer membrane and the inner membrane of each cell was measured and the data were statistically analyzed using One Way ANOVA with Bonferroni correction. For scanning electron microscopy (SEM), ethanol dehydrated samples were dried in freeze-drier and coated with 10 nm gold film using ion sputter. Scanning electron microscopic images were taken using HITACHI SU8010.

Estimation of pellicle and biofilm formation
For pellicle formation assay, mycobacteria were inoculated in 4~5 mL of Middlebrook 7H9 medium without Tween-80 and grown at 37°C without shaking. Biofilm formation was measured in M63-based liquid medium as previously described [31][32][33][34]. Biofilms of all three strains were grown in 96-well polystyrene plates or glass tubes containing M63-based liquid medium complemented with casein hydrolysate and glucose (without Tween-80), inoculated with 0.1% log phase culture, and incubated at 30°C for 5-7 days under static conditions [33,34]. The biofilm formation in each of the liquid cultures was qualitatively analyzed by photography and the images were processed using Adobe Photoshop CS5 software, and quantified with crystal violet staining, as previously described [31,32].

Growth profile of strains
The growth characteristics of M. smegmatis mc 2 155, aceE-mut, Vector (aceE-mut:Vector) and Comp (aceEmut:aceE) strains were studied in neutral (pH 6.8) or acidified (pH 5.0, the pH was adjusted with hydrochloric acid) 7H9 medium. The cultures were inoculated with an initial optical density at 600 nm (OD 600 ) of 0.01 and incubated at 37°C with constant shaking at 200 rpm. OD 600 was measured at specified time intervals and 10fold serial dilutions were plated on 7H10 agar plates for colony forming unit (CFU) counts.

Stress assays
To

Antimicrobial susceptibility testing
Minimal Inhibitory Concentration (MIC) determination was performed by using the alamar blue microtiter assay as recommended in CLSI guidelines [6]. Alamar blue dye was added in each well and the plates were re-incubated at 37°C for 24 h. The color change from blue to pink indicated bacterial growth. The MIC was defined as the minimal concentration of the drug showed no color changes, which was the lowest concentration of drug that can inhibit the visible growth of the bacterium.
Mycolic acid analysis using HPLC-Sherlock mycobacterium identification system Bacterial cultures were collected for isolation of mycolic acids in the cell walls by extraction, saponification, and derivation according to instructions for the Sherlock Mycobacteria Identification System (SMIS; MIDI, Inc.). Mycolic acid composition of each sample was analyzed by SMIS using HPLC.

Analysis of the composition of mycolic acids and lipids using LC-MS
Total lipids were extracted from the samples using an improved Bligh/Dyer extraction method (double extraction) and appropriate internal standards were added as previously described [16]. Analysis of mycolic acids and lipids was carried out using normal-phase LC-MS as previously described, with minor modification [35,36].

Macrophage infections
Bacteria at log phase was collected and washed with RPMI1640 before infection. Infection of THP-1 cell (ATCC TIB-202) was performed at a multiplicity of infection (MOI) of 10:1 and 1:1 (bacteria: macrophage), using the following conversion: an OD of 1 = 3 × 10 8 CFU/mL. After 2 h of incubation at 37°C, the extracellular mycobacteria were removed by three washings with 1 × PBS and RPMI1640 complete medium containing 100 μg/ml gentamycin was added to inhibit growth of exogenous mycobacteria in infected wells. At 0 h, 2 h and 24 h, infected macrophages were harvested and lysed with 0.1%Tween-80. Then the lysates were serially diluted with 0.05% Tween-80, and plated on 7H10 agar plates with or without the antibiotic. The plates were incubated at 37°C until colonies could be counted.

Statistical analysis
All statistical analyses were performed using SPSS statistics 21. Statistical differences were determined by One Way ANOVA with Bonferroni correction when comparing more than two groups. For mycolic acid and lipid quantification experiments, One Way ANOVA with Tukey's HSD correction was used. Student t test was used to compare the statistical differences between two groups. Only P values < 0.05 were considered as statistically significant.
Additional file 1: Figure S1. Different growth of M. smegmatis mc 2 155 (WT) and aceE-mut after treatment with different chemical agents.
Additional file 2: Figure S2. The effect of aceE deficiency on mycolic acid composition in mycobacterium.