All bacterial strains and plasmids used in this study are listed in Table 1. All primers were synthesized by Sangon Biotech (Shanghai, China) and listed in Table S1. For normal growth, S. oneidensis and E. coli were cultivated aerobically in Luria-Bertani (LB) medium (Difco, Detroit, MI) at 30°C and 37°C, respectively. Where needed, the growth medium was supplemented with chemicals at the following concentrations: 2,6-diaminopimelic acid (DAP), 0.3 mM; ampicillin (AMP), 100 μg/mL; kanamycin (Km), 50 μg/mL; and gentamicin (Gm), 15 μg/mL. All chemicals were purchased from Sigma-Aldrich (Shanghai, China) unless otherwise noted.

In-frame deletion strains of S. oneidensis MR-1 were constructed by the att-based fusion PCR method (Jin et al., 2013). In brief, two fragments flanking the targeted gene were amplified by PCR with primers containing attB sequence or gene-specific sequence, and then joined together by an overlapping PCR method. The fusion fragments were introduced into plasmid pHGM01 by site-specific recombination, using BP Clonase (Invitrogen), and then transformed into E. coli WM3064 strain (DAP auxotroph). The resulting recombinant plasmids were transferred from E. coli WM3064 into the appropriate S. oneidensis strains via conjugation. Integration of the recombinant plasmid into the chromosome was selected by resistance to gentamicin and verified by PCR. The correct transconjugants were grown in LB broth in the absence of NaCl and plated onto LB medium supplemented with 10% sucrose. Gentamicin-sensitive and sucrose-resistant colonies were screened by PCR for deletion of the targeted gene. The deletion mutations were then verified by sequencing.

Plasmid pHG101 was used for genetic complementation of mutants (Wu et al., 2011). A fragment containing the gene of interest together with its native promoter was amplified by PCR and cloned into pHG101. The correct recombinant plasmids were transferred into its corresponding mutant strains via conjugation. The presence of recombinant plasmid was confirmed by plasmid purification and restriction enzyme digestion.

For measuring growth of different S. oneidensis strains, Overnight cultures were diluted 1:100 in 3 mL fresh LB medium and incubated at 200 rpm 30°C. The optical density at 600 nm (OD600) was recorded every 1 h. Then the generation times (G) were calculated according to the OD600 at the exponential stage.

Antibiotic susceptibility of S. oneidensis was determined with both liquid and solid cultures. Liquid cultures were utilized to determine the minimum inhibitory concentration (MIC) of the antibiotics (Yin et al., 2014). Antibiotics used in the MICs assay were AMP, cefotaxime (CTX), and imipenem (IPM). All MICs were determined at least in triplicate. Susceptibility assays on plates were also used to compare differences in AMP resistance among S. oneidensis strains. In this case, decimal dilution series were prepared. Three μL of each dilution was placed onto LB plates supplemented with AMP at different concentrations. The plates were incubated for 18 h at 30°C and then photographed.

The activity of the promoter for each LT and blaA genes was determined by a markerless integrative lacZ reporter system, which we used to measure the activity of the blaA promoter (Yin et al., 2014, 2015). In brief, the wild type strains harboring different transcriptional fusion vector were cultivated to the exponential-phase ([OD600] of ≈ 0.4), then harvested by centrifugation at 4°C. After washed with phosphate-buffered saline (PBS) and treated with lysis buffer [0.25 M Tris-HCl (pH = 7.5), 0.5% Triton-X-100]. β-Galactosidase activity was determined by monitoring optical density at 420 nm using a Infinite M200 Pro microplate reader (Tecan) as previously described (Yin et al., 2014, 2015).

The β-lactamase activity was measured directly by nitrocefin hydrolysis method, as described previously (Yin et al., 2014). In brief, bacterial cultures were cultivated to the early-exponential-phase [(OD600) of ≈ 0.2], and then 0 or 200 μg/mL AMP was added for blaA induction. The cultures were incubated for an additional 2 h at 30°C and 200 rpm. After incubation, 1 mL culture was centrifuged at 2,500 × g for 5 min, washed once with 1 mL 50 mM phosphate buffer (pH 7.0), and resuspended in 1 mL of the same buffer. Crude cell extracts were prepared by sonication (pulse on, 3 s; pulse off, 4 s; 10 times). The protein content of the crude extracts was determined using a Bradford assay with BSA as a standard (Bio-Rad). The nitrocefin hydrolysis assays were performed in 200 μL reaction mixtures containing 8 μg total protein and 4 μg nitrocefin (Calbiochem, San Diego, CA). Nitrocefin hydrolysis was measured every minute for 10 min at 25°C by absorbance at 486 nm. The specific BlaA β-lactamase activity was expressed in nanomoles of nitrocefin hydrolyzed per minute per milligram of protein. All experiments were performed in triplicate, and the results presented are averages for the three experiments.

The experimental values were subjected to statistical analyses and are presented as means ± standard deviations. Student's t-test was performed for pairwise comparisons of groups.

According to genome annotation, seven genes [mltB (SO1166), SO1994, sltY (SO2040), SO2564, mltF (SO3288), mltD (SO4017), and SO4660] are predicted to encode LTs in S. oneidensis MR-1. Except for SO4660, all LTs have a homolog in E. coli and P. aeruginosa (Table 2). SltY and MltF share 29 and 41% sequence identity with E. coli Slt70 and MltF, respectively. Interestingly, both MltB and SO1994 are E. coli MltB homologs (36 and 35% identity respectively); both SO2564 and MltD are homologous with E. coli MltD (34 and 33% identity respectively). These results indicate that S. oneidensis may produce two MltB and MltD isozymes. SO4660 is a protein of 239 amino acids (aa), which contains a transglycosylase SLT domain at the C-terminal region (115–217 aa). Blastp results demonstrate that SO4660 shares 37% identity and 51% similarity with P. aeruginosa Slt only at the C-terminal region (Query cover: 52%; E-value: 6e-17). The amino acids of SltY and SO4660 do not include a predicted lipoprotein process site, suggesting that they are soluble LTs. However, the rest LTs contain a predicted lipidation site [L(A/S)(G/A)C], a hallmark for lipoprotein (Tokuda and Matsuyama, 2004), implicating that these enzymes are membrane-bound lipoprotein. E. coli Slt70, MltF, MltB, MltD, and P. aeruginosa Slt belong to family 1A, 1E, 3, 1D, and 1A, respectively (Blackburn and Clarke, 2001). The consensus motifs of each family are also found in S. oneidensis counterparts (Figure S1). Therefore, based on sequence similarities and consensus motifs, we designated SO1994, SO2564, and SO4660 of S. oneidensis as mltB2, mltD2, and slt2, respectively. In total, S. oneidensis produces seven LTs, two from families 1A (SltY, Slt2), 1D (MltD and MltD2), and 3 (MltB and MltB2), one from family 1E (MltF).

To determine whether all these LTs are expressed, a markerless integrative lacZ-reporter system was employed to measure the promoter activity of each LT genes (Fu et al., 2014; Figure 1). Compared to the control, the expression of the β-galactosidase driven by each promoter was detectable in the wild type background. Among these, the promoter of mltD2 (P_mltD2) had the highest activity to induce the expression of lacZ, while the promoter of mltF (P_mltF) had the lowest activity. These results suggest that all LT genes are effectively expressed in our tested conditions.

To understand the roles of LTs in S. oneidensis, strains lacking each gene were constructed individually. Firstly, we measured growth of each mutant in LB broth medium (Figure 2). Deletion of the mltB, mltD, mltD2, and slt2 genes had little impact on cell growth. In contrast, the other three mutants (the ΔsltY, ΔmltF, and ΔmltB2 strains) displayed obviously defects when grown in LB broth. Compared to the wild type, the generation time of the three mutants increased about 20 percent. Genetic complementation was then carried out with the multicopy plasmid pHG101 harboring a copy of sltY, mltF, or mltB2 under the control of their native promoters (Wu et al., 2011). The growth defect of the three mutants was fully recovered by expression of the corresponding genes in trans (Figure 2), confirming that growth defects observed in the three mutants were due to the mutation per se.

To assess whether growth defects are involved in impaired cell wall integrity, we observed cell shape and determined the susceptibility to hyposmolality. S. oneidensis is a typical rod-shaped bacterium and can grow normally in LB minus NaCl. Both cell shape and susceptibility to hyposmolality in all tested mutants were similar to those in the wild type (Figure S2). These results suggest that loss of single LT does not affect the cell wall integrity, growth defects of the ΔsltY, ΔmltF, and ΔmltB2 strains are result from other unknown reasons.

Our previous study showed that major peptidoglycan recycling enzymes are involved in β-lactamase BlaA expression and β-lactam resistance in S. oneidensis (Yin et al., 2014). To study whether LTs are also play important roles in this process, the resistance to β-lactams in these LTs mutants were determined. Deletion of the mltF, mltD, mltD2 and slt2 genes did not affect the susceptibility to AMP. In contrast, the ΔsltY, ΔmltB2, and ΔmltB strains displayed significantly improved resistance to AMP (Figure 3A). Compared to the wild type, the ΔsltY, ΔmltB2, and ΔmltB strains had 8-, 8-, and 4-fold increases in resistance to AMP, respectively. Besides, the ΔsltY and ΔmltB2 strains had 2-fold increases in resistance to another two tested β-lactams, ceftotaxime (CTX) and imipenem (IMP) (Table 3).

More importantly, expression of the sltY, mltB2, and mltB genes under the control of their native promoters recovered all β-lactam susceptibility to the level of the wild type, indicating that the phenotypes observed in these mutants was due to the intended mutations (Figure 3A and Table 3). These data, collectively, indicate that certain LTs (SltY, MltB2 and MltB) are important players mediating the β-lactam resistance in S. oneidensis.

Production of β-lactamase BlaA confers S. oneidensis with natural resistance to certain β-lactams (Yin et al., 2013). In strains lacking AmpG or NagZ, two major peptidoglycan recycling enzymes, BlaA expression is significantly increased, leading to improved β-lactam resistance (Yin et al., 2014). It is reasonable that LTs-mediated β-lactam resistance is also dependent on BlaA.

To test this possibility, the lacZ reporter system was employed to determine the activity of the blaA promoter (P_blaA) (Yin et al., 2014, 2015). As shown in the Figure 3B, the expression levels of the β-galactosidase driven by the P_blaA in the ΔsltY, ΔmltB2 and ΔmltB strains were substantially higher than that in the wild type under all tested conditions. To confirm, the BlaA activity in all LTs mutants was measured by the nitrocefin hydrolysis method (Table 3; Yin et al., 2014). Our previously study found that β-lactamase activity was hardly detected in the ΔblaA strain under all tested conditions, implicating that nitrocefin assay is specific for BlaA in S. oneidensis. As expected, deletion of the mltF, mltD, mltD2, and slt2 genes did not affect the BlaA activities under all tested conditions. However, the ΔsltY, ΔmltB2, and ΔmltB strains had higher levels of BlaA activity. Compared to the wild type, the basal levels of BlaA activity of all these three mutants were increased ~2-fold, while the induced levels were increased 1.2~1.6-fold. Notably, expression of the sltY, mltB2 and mltB in the corresponding mutants significantly decreased the BlaA activities under all tested conditions.

In addition, the blaA gene was deleted from the ΔsltY, ΔmltB2, and ΔmltB strains. Similar to the ΔblaA strain, loss of BlaA completely abolished the β-lactam resistance and BlaA activities of the three LTs mutants (Table 3). Based on these results, we can conclude that the improved β-lactam resistance in the ΔsltY, ΔmltB2, and ΔmltB strains is dependent on BlaA.

Given that different LTs appear to play functionally redundant roles in peptidoglycan degradation, we supposed that LTs may have additive effects on blaA expression. To confirm, the ΔmltBΔmltB2 and ΔsltYΔmltB2 strains were constructed and subjected to β-lactam susceptibility testing. As shown in Figure 4 and Table 3, both two strains had increased β-lactam resistance compared to the single mutants (ΔmltB, ΔmltB2, and ΔsltY). When compared to the ΔmltB and ΔmltB2 strains, the MICs of the ΔmltBΔmltB2 strain for all three β-lactams (AMP, CTX, and IMP) were increased 4- and 2-fold, respectively. The ΔsltYΔmltB2 strain had 2-fold increases in resistance to these β-lactams compared to the single mutants.

Consistently, both the basal and induced levels of BlaA activity were increased in the two double mutants. The ΔmltBΔmltB2 and ΔsltYΔmltB2 strains had basal levels of BlaA activity ~2.2- and ~1.7-fold higher than either of its single-mutation parent strains, respectively. By induction, the BlaA activity of the ΔmltBΔmltB2 strain was 1.3-fold higher than that of the ΔmltB strain. Similarly, the BlaA activity of the ΔsltYΔmltB2 strain was ~1.4-fold higher than that of the ΔsltY strain. These results suggest that SltY, MltB and MltB2 have additive effect on blaA expression and β-lactam resistance.

In S. oneidensis, loss of AmpG increases the basal level of blaA expression, implying that intermediate products transported across AmpG permease are likely to be the repressors rather than inducers for blaA expression (Yin et al., 2014). To evaluate the effects of AmpG on LTs-mediated blaA expression, the ΔmltBΔmltB2ΔampG and ΔsltYΔmltB2ΔampG strains were constructed for antibiotic susceptibility testing (Table 3). The two strains, together with the ΔampG strain, displayed similar levels of resistance to all tested β-lactams, which had 2-fold increases in resistance to AMP when compared to the strains lacking two LTs (ΔmltBΔmltB2 and ΔsltYΔmltB2). Notably, deletion of ampG from the two double mutants increased the basal levels of BlaA activity. Therefore, loss of AmpG further increases the basal levels of blaA expression and β-lactam resistance in strains lacking LTs.

Our previous study demonstrated that deletion of mrcA (encoding PBP1a) drastically enhances blaA expression in S. oneidensis even in the absence of β-lactams, which is independent of AmpG (Yin et al., 2015). Although significant, the increase in β-lactamase activity after inactivating LTs is much lower than that produced by PBP1a inactivation. To explore the involvement of LTs in PBP1a-mediated blaA expression, we deleted the mrcA gene from both ΔmltBΔmltB2 and ΔsltYΔmltB2 strains and performed the antibiotic susceptibility testing (Table 3). The resistance to β-lactams of both the ΔmltBΔmltB2ΔmrcA and ΔsltYΔmltB2ΔmrcA strains was identical to those of the ΔmrcA strain. Compared to the strains lacking two LTs, both the two triple mutants had 2-fold increases in resistance to AMP. Consistently, deletion of mrcA significantly increased the basal levels of BlaA activity of the strains lacking two LTs. Both the basal and induced levels of BlaA activity of the triple mutants were comparable to those of the ΔmrcA strain (Table 3). These results suggest that PBP1a-mediated blaA expression is independent of LTs. In other words, PBP1a functionally overrides LTs in blaA expression.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.