Deficiency of D‐alanyl‐D‐alanine ligase A attenuated cell division and greatly altered the proteome of Mycobacterium smegmatis

Abstract D‐Alanyl‐D‐alanine ligase A (DdlA) catalyses the dimerization of two D‐alanines yielding D‐alanyl‐D‐alanine required for mycobacterial peptidoglycan biosynthesis, and is a promising antimycobacterial drug target. To better understand the roles of DdlA in mycobacteria in vivo, we established a cell model in which DdlA expression was specifically downregulated by ddlA antisense RNA by introducing a 380 bp ddlA fragment into pMind followed by transforming the construct into nonpathogenic Mycobacterium smegmatis. The M. smegmatis cell model was verified by plotting the growth inhibition curves and quantifying endogenous DdlA expression using a polyclonal anti‐DdlA antibody produced from the expressed DdlA. Scanning electron microscopy and transmission electron microscopy were used to investigate mycobacterial morphology. Bidimensional gel electrophoresis and mass spectrometry were used to analyze differentially expressed proteins. Consequently, the successful construction of the M. smegmatis cell model was verified. The morphological investigation of the model indicated that DdlA deficiency led to an increased number of Z rings and a rearrangement of intracellular content, including a clear nucleoid and visible filamentous DNA. Proteomic techniques identified six upregulated and 14 downregulated proteins that interacted with each other to permit cell survival by forming a regulatory network under DdlA deficiency. Finally, our data revealed that DdlA deficiency inhibited cell division in mycobacteria and attenuated the process of carbohydrate catabolism and the pathway of fatty acid anabolism, while maintaining active protein degradation and synthesis. N‐Nitrosodimethylamine (NDMA)‐dependent methanol dehydrogenase (MSMEG_6242) and fumonisin (MSMEG_1419) were identified as potential antimycobacterial drug targets.


| INTRODUC TI ON
, an infectious disease caused by Mycobacterium tuberculosis, remains a major threat to human health in terms of worldwide morbidity and mortality (World Health Organization, 2018). D-Alanyl-D-alanine ligase A (DdlA) is an attractive potential anti-TB drug target (Yang et al., 2018). DdlA catalyses the ATP-dependent dimerization of two D-alanine molecules to yield D-alanyl-D-alanine, which is required for the biosynthesis of peptidoglycan (PG); PG is essential for preserving cell integrity by preserving internal osmotic pressure and maintaining the defined cell morphology and cell division (Bruning, Murillo, Chacon, Barletta, & Sacchettini, 2011;Kieser et al., 2015;Yang et al., 2018). Theoretically, the biosynthesis of PG is an ideal target for anti-TB drug design because the complete pathways of PG biosynthesis are not present in mammalian cells. PG is composed of linear glycan chains consisting of repeating disaccharide units consisting of β-1,4-linked N-acetylglucosamine-N-acetylmuramic acid moieties and crosslinked tetrapeptide chains consisting of an L-alanyl-D-isoglutaminyl-meso-diaminopimelyl-D-alanine moiety that is bound to N-acetylated muramic acid. Crosslinks between two adjacent oligopeptide chains are formed by the binding of meso-diaminopimelate tetrapeptide side chains to a D-alanine residue on the adjacent tetrapeptide, in turn releasing the terminal D-alanine residue (Barreteau et al., 2008;Macheboeuf, Contreras-Martel, Job, Dideberg, & Dessen, 2006;Yang et al., 2018). The crosslinked PG confers tensile strength to the bacterial cell wall. Thus, the inhibition of PG crosslinking causes extensive cell wall weakening or cell death; DdlA is involved in PG crosslinking by supplying the indispensable D-alanyl-D-alanine. In addition, compounds with structural similarity to D-alanine can competitively inhibit DdlA activity with fewer side effects, because D-alanine is absent in humans. Therefore, in combination with the essentiality of the ddlA gene in M. tuberculosis, the characteristics of DdlA make it an ideal target for anti-TB drugs.
In fact, the particularly interesting nature of DdlA as a novel anti-TB drug target that has been indicated by the clinical application of D-cycloserine as an effective second-line antibiotic against M. tuberculosis (Halouska et al., 2014;McCoy & Maurelli, 2005;Prosser & de Carvalho, 2013). However, the use of D-cycloserine, which targets DdlA and alanine racemase as structural analogues of D-alanine, is restricted in TB treatment because of its serious neurological side effects (Yang et al., 2018). D-Cycloserine overdose has been reported to cause central nervous system side effects such as drowsiness, headaches, vertigo, depression, paraesthesias, dizziness, dysarthria, confusion, hyperirritability, convulsions, psychosis, tremors, paresis, seizures, and coma by activating the N-methyl-D-aspartate-type glutamate receptor (Chen, Uplekar, Gordon, & Cole, 2012;Halouska et al., 2014;Schade & Paulus, 2016). To design a nontoxic compound that targets DdlA and to better understand the roles of DdlA in mycobacteria in vivo, an extensive study on DdlA was performed using a Mycobacterium smegmatis model of DdlA downregulation induced by Sm-ddlA antisense RNA.

| Strains and plasmids
The properties of the plasmids and bacterial strains used in this study are shown in Table 1

| Gene manipulation and protein expression
Mycobacterium smegmatis genomic DNA was used as a PCR template to amplify Sm-ddlA (MSMEG_2395), yielding a 1,122 bp DNA fragment. Then, the purified PCR product was cloned into pMD18-T, generating the pMD18-Sm-ddlA plasmid. Following sequence confirmation, Sm-ddlA was digested for subsequent ligation into pCold II (Takara, China) to generate pCold II-Sm-ddlA, which was transformed into E. coli BL21(DE3) for DdlA expression. The preparation of cell lysates and the purification and detection of DdlA were conducted as described previously (Yang et al., 2018).

| Production of the antibody against M. smegmatis DdlA
Purified DdlA was used to immunize mice to produce the anti-DdlA polyclonal antibody. A 1 mg/ml sample of DdlA (in physiological saline) was emulsified with an equal volume of Freund's incomplete adjuvant (Thermo Scientific) by vortexing vigorously to yield a homogenate of antigen suspension. The prepared antigen suspension was injected subcutaneously into three sites on each female BALB/c mouse (approximately 8 weeks old) on the 1st, 10th, and 17th day. Blood was collected from the orbital venous plexus of the mice on the 24th day. Antiserum was obtained by centrifugation. Then, the specificity of the polyclonal anti-DdlA antibody was evaluated by Western blot (WB) analysis, and antibody titers were determined by enzyme-linked immunosorbent assay (ELISA). The colorimetric visualization of the WB was performed using BCIP/NBT solution, and the ELISA was conducted using p-nitrophenyl phosphate disodium solution.

| Establishment of the cell models of M. smegmatis DdlA downregulation by Sm-ddlA antisense RNA
The DNA fragment (380 bp) of Sm-ddlA-AS, including 100 bp upstream of the Sm-ddlA transcription initiation site and 280 bp of the Sm-ddlA sequence at the 5' end, was amplified from M. smegmatis genomic DNA by PCR using the primers ddlA-F, 5' TAACTAGTGTGACTGCCCCGAACCATC 3' (the SpeI site is underlined), and ddlA-R, 5' ATCATATGCTGGATGGTGCCGTCTTCG 3' (the NdeI site is underlined). The purified PCR fragment was ligated to pMD18-T to yield pMD-Sm-ddlA-AS. Following sequence confirmation, the Sm-ddlA-AS fragment from pMD-Sm-ddlA-AS was cloned into pMind to generate pMind-Sm-ddlA-AS (Table 1), whose expression can be induced by tetracycline. The pMind-Sm-ddlA-AS was electroporated into competent M. smegmatis cells as described previously (Pelicic et al., 1997).
Tetracycline was added as an inducer at gradient concentrations of 0 ng/ml, 10 ng/ml, 20 ng/ml, 30 ng/ml, and 50 ng/ml. The growth of M. in LB broth containing 20 ng/ml tetracycline were used as the control.

| Detection of endogenous DdlA
Mycobacterium smegmatis cells carrying pMind-Sm-ddlA-AS were grown in LB broth containing tetracycline at concentrations of 20 ng/ml and 0 ng/ml. The cells were harvested after shaking for 36 hr, and cell lysates were extracted as described previously (Yang et al., 2018). Samples of total soluble protein (50 μg) were loaded membrane for WB analysis. The transferred membrane was probed by incubating with a polyclonal antibody against DdlA at a 1:2,000 dilution, followed by incubation with an anti-mouse IgG HRP-conjugated secondary antibody. The target DdlA bands were visualized using enhanced chemiluminescence (ECL) reagents. M. smegmatis cells carrying pMind and grown in LB broth containing 20 ng/ml or 0 ng/ml tetracycline were used as the controls.

| Morphological observation
Mycobacterium smegmatis cells carrying pMind-Sm-ddlA-AS and grown in LB broth containing 20 ng/ml tetracycline were harvested.

| Protein processing, bidimensional gel electrophoresis and image analysis
Soluble protein was extracted from the harvested M. smegmatis cells carrying pMind-Sm-ddlA-AS and processed using a clean-up kit according to the manufacturer's instructions (Bio-Rad). The protein pellets were dissolved in 50 μl of isoelectric focusing (IEF) buffer (7 mol/L urea, 2 mol/L thiourea, 2% (w/v) CHAPS, and 30 mmol/L Tris, pH 8.0), and the total soluble protein was quantified using a BCA kit. Fifty micrograms of total soluble protein was subjected to bidimensional difference in gel electrophoresis (2D-DIGE). Proteins were separated in the first dimension according to their pI using Immobiline DryStrips (11 cm, pH 4-7, Bio-Rad) and in the second dimension according to their molecular weight using 12% SDS-PAGE gels. The gels were stained using a ProteoSilver Plus SilverStain Kit (Sigma). Image analysis and statistical quantification of relative protein abundances were performed using PDQuest software. Student's t test was performed to assess the statistical significance of differentially expressed proteins based on the average spot volume ratio. The protein spots that met the criteria of a fold change (FC) ≥2 or ≤0.5 at a 95% confidence level (Student's t test; p < 0.05) were selected for further identification by mass spectrometry (MS). Spots located near the gel borders and small or faint spots were excluded from protein identification. M. smegmatis cells carrying pMind were used as the control in this study. The experiment was performed in triplicate.

| Trypsin digestion, MS/MS analysis and data acquisition
The selected protein spots were excised manually from the gels for MS analysis. The subsequent trypsin digestion and MS/MS analysis were performed as described previously (Yang et al., 2018). Proteomic data were acquired using Mascot 2.2 software.
The obtained MS/MS spectra were searched against the NCBI database. Confidence in the peptide identifications was assessed based on the Mascot sequence assignment score and visual inspection of the molecular mass and pI values of the selected spots from the gels. org/) for protein network construction.

| Anti-DdlA antibody specificity testing and titer determination
The Female BALB/c mice were immunized with the pure recombinant DdlA antigen and sacrificed to obtain serum for testing the specificity of the anti-DdlA antibody. Figure 1c shows that the produced anti-DdlA antibody specifically interacted with the recombinant DdlA protein antigen. In addition, the optimum titer of 1:2,000 for the produced antibody was determined using ELISA. Therefore, the anti-DdlA antibodies produced in this study can respond specifically against the DdlA antigen and are applicable for DdlA detection. Therefore, the TEM images indicate that a rearrangement of intracellular content and an increase in the number of Z rings numbers occur in mycobacteria due to DdlA deficiency.

| Identification of differentially expressed proteins by 2D-DIGE coupled with MS/MS analysis
Proteomic changes were automatically analyzed by PDQuest software. Figure 5 shows the paired 2D-DIGE images, which were labeled manually with the numbers obtained from the PDQuest analysis. As a consequence, 53 filtered protein spots with fold changes in expression of >2 or <0.05 displayed significant changes between the paired groups (p < 0.05). Of the 53 protein spots, 12 were differentially up- Twenty-two spots with the remarkable intensity changes were further extracted from the gel and identified by MS. Table 2 shows F I G U R E 1 Identification of the purified DdlA by SDS-PAGE (a) and WB (b) and specificity testing of the anti-DdlA polyclonal antibody by WB (c). M, PageRuler prestained protein ladder (Fermentas). Colorimetric visualization was performed using BCIP/NBT solution in the WB analysis (a and c) and Coomassie in the SDS-PAGE analysis (b). The monoclonal anti-polyhistidine antibody at a 1:5,000 dilution was used to probe the expressed DdlA (a); the produced polyclonal anti-DdlA antibody at a 1:2,000 dilution was used to probe the purified DdlA (c). a and b, 1-2, the fractions of the purified DdlA. c, 1, purified Mycobacterium smegmatis DdlA Interestingly, NDMA-dependent methanol dehydrogenase (MSMEG_6242) was identified as a differentially expressed protein under DdlA deficiency; it was also identified as a differentially expressed protein when M. smegmatis growth was inhibited by the taurine-5-bromosalicylaldehyde Schiff base compound (data not published). NDMA-dependent methanol dehydrogenase allows gram-negative microorganisms to generate energy from methanol oxidation and to synthesize compounds with carbon-carbon bonds from methanol assimilation (Harm, Harm, & Dijkhuizen, 2000). In this study, MSMEG_6242 was found to directly interact with MSMEG_1735, MSMEG_6008, MSMEG_1665, and In this study, the most highly differentially expressed protein was MSMEG_1419, which was highly upregulated when M.

| CON CLUS IONS
In and fumonisin (MSMEG_1419) were the identified proteins with the greatest differential expression; these proteins have substantial potential to be developed into anti-TB drug targets. Therefore, our dataset provides intriguing insight into antimycobacterial drug design via developing novel anti-TB drug targets which were revealed in this study.

ACK N OWLED G EM ENTS
This work was supported by National Natural Science Foundation of China (81801981, 81272429) and Liaoning Provincial Program for Top Discipline of Basic Medical Sciences.

CO N FLI C T O F I NTE R E S T S
The authors declared no conflict of interest.

AUTH O R S CO NTR I B UTI O N
W.Z. conducted experiments and modified the manuscript. Y.C.
composed and modified the manuscript. Y.X., W.D., and S.Y. contributed to funding support. S.L. processed and analyzed the data. All authors shared in responsibility for the final decision to submit it for publication.

E TH I C S S TATEM ENT
None required.

DATA ACCE SS I B I LIT Y
All data are provided in full in the results section of this paper.