Abstract The unique cell envelope structure of Mycobacterium tuberculosis is fundamental to its pathogenesis. Phosphatidylinositol (PI)-anchored glycolipids, such as phosphatidylinositol mannosides (PIMs), lipomannan and lipoarabinomannan, are essential components of the cell envelope widely conserved among mycobacteria, but their roles in the cell envelope integrity are not fully understood. We previously identified PimE in Mycobacterium smegmatis, a nonpathogenic model organism, as a mannosyltransferase that catalyzes the fifth mannose transfer for the biosynthesis of hexamannosyl PIMs. Our analyses, reported here, further demonstrate that the growth of the pimE deletion mutant (ΔpimE) is defective in the presence of copper. We first found that the small colony phenotype of ΔpimE on a solid Middlebrook 7H10 agar surface was alleviated when grown on M63 agar. Comparative analysis of the two media led us to identify copper in Middlebrook 7H10 as the cause of growth retardation seen in ΔpimE. We further demonstrated that ΔpimE is sensitized to several antibiotics, but the increased sensitivities were independent of the presence of copper. We conclude that the deletion of the pimE gene does not cause growth defects under optimal growth conditions, but makes the cell envelope vulnerable to toxic compounds such as copper and antibiotics. biofilm, cell envelope, glycolipids, mannosyltransferase, mycobacteria, copper INTRODUCTION An important factor that contributes to the intrinsic drug resistance of the tuberculosis pathogen, Mycobacterium tuberculosis, is its unique cell envelope (Jarlier and Nikaido 1990). The mycobacterial cell envelope is a multilayered structure containing a thick, waxy outer membrane that forms an impermeable barrier to many bactericidal compounds (Kaur et al.2009; Jankute et al.2015). The cell envelope is rich in a family of phosphatidylinositol (PI)-anchored glycolipids known as phosphatidylinositol mannosides (PIMs), lipomannan (LM) and lipoarabinomannan (LAM), which are involved in various immunomodulatory activities in the pathogenic species (Mishra et al.2011; Ishikawa, Mori and Yamasaki 2017). However, these glycolipids are ubiquitously found in both pathogenic and nonpathogenic species, suggesting that they play more fundamental structural roles within the cell envelope, perhaps similar to teichoic and lipoteichoic acids in Gram-positive bacteria (Weidenmaier and Peschel 2008). Indeed, PIMs/LM/LAM are abundant, with both LM and LAM being present at approximately 104 molecules per cell, and PIMs being roughly 10–100 times more abundant than LM/LAM (Morita et al.2011). We have previously created mutants defective in LM/LAM structures in both M. tuberculosis and nonpathogenic M. smegmatis (Sena et al.2010), and showed that these mutations result in the loss of cell envelope integrity (Fukuda et al.2013), being consistent with the idea that these glycolipids are structurally important for the cell envelope. The physiological roles of PIMs in the cell envelope remain poorly characterized. The major PIM species are PI dimannosides (AcPIM2 and Ac2PIM2) and PI hexamannosides (AcPIM6 and Ac2PIM6). AcPIM2 and AcPIM6 carry a third fatty acid attached to a mannose residue in addition to diacyl PI, while Ac2PIM2 and Ac2PIM6 carry a fourth fatty acid attached to the inositol. PIM species can represent ∼5% of total membrane phospholipids (Jackson, Crick and Brennan 2000), or substantially more: a recent study using a new lipid extraction method revealed that more than half of the plasma membrane hydrocarbon chains comes from PIM2 species (Bansal-Mutalik and Nikaido 2014). PIM biosynthesis is achieved by sequential additions of mannoses onto PI (Fig. 1), and the first two mannosyltransferases and the acyltransferase have been identified as PimA, PimB’ and PatA, respectively (Korduláková et al.2002, 2003; Lea-Smith et al.2008; Guerin et al.2009; Albesa-Jové et al.2016). Figure 1. View largeDownload slide Biosynthetic pathway of PIMs. PimE mediates the fifth mannose transfer. AcPIM4 is the branch point for AcPIM6 and LM/LAM biosynthesis. PimA, PimB’ and PatA are enzymes that mediate the initial three reactions as described in the text. Only key intermediates and products are shown. Figure 1. View largeDownload slide Biosynthetic pathway of PIMs. PimE mediates the fifth mannose transfer. AcPIM4 is the branch point for AcPIM6 and LM/LAM biosynthesis. PimA, PimB’ and PatA are enzymes that mediate the initial three reactions as described in the text. Only key intermediates and products are shown. We have previously identified PimE as the fifth mannosyltransferase of AcPIM6/Ac2PIM6 biosynthesis in M. smegmatis, and demonstrated that the ΔpimE mutant lacks AcPIM6/Ac2PIM6 and accumulates the precursors AcPIM4/Ac2PIM4 (Morita et al.2006). AcPIM4 is an intermediate positioned at the branch point of AcPIM6 and LM/LAM pathways, and a lipoprotein LpqW has been proposed to function as a regulator controlling the metabolic flux into either AcPIM6 or LM/LAM (Kovacevic et al.2006). Indeed, M. smegmatis lpqW deletion mutant is severely compromised in LM/LAM biosynthesis, and shows a slow growth phenotype on the PPLO agar medium. Interestingly, this mutant is outgrown by a suppressor mutant that restores LM/LAM biosynthesis and accumulates AcPIM4 instead of AcPIM6. It was later shown that the suppressor phenotype was due to loss-of-function mutations in pimE (Crellin et al.2008), highlighting the importance of controlling the metabolism at this branch point for the optimal growth of M. smegmatis. In the current study, we investigated the phenotypes of ΔpimE mutant further to delineate the factors affecting the observed growth defects of ΔpimE mutant, and identified copper in the medium as a factor affecting the growth of this mutant. MATERIALS AND METHODS Bacterial strains and growth conditions Wild-type (WT) M. smegmatis mc2155 (Snapper et al.1990), ΔpimE mutant strains and their derivatives (ΔpimE::pimE-FLAG, ΔpimE::pimE[D58A]-FLAG, ΔpimE::pimE and ΔpimE::empty vector) (Morita et al.2006) were grown in various media as described in Table S1, Supporting Information. Middlebrook 7H9 and 7H10 medium bases were from Becton Dickinson. Starter cultures were grown with appropriate antibiotics, 20 μg/mL kanamycin for ΔpimE or 20 μg/mL kanamycin plus 50 μg/mL streptomycin for complemented ΔpimE strains, and were diluted 100-fold to initiate the experimental cultures. For planktonic growth, cells were grown at 30°C with shaking. For pellicle formation, cells were grown at 37°C in 3 mL of the indicated medium (without Tween-80) in a 12-well microtiter plate (Olympus, flat bottom, tissue culture-treated). Colony morphology and size measurement Cells were grown on agar plates for 4 days at 37°C in triplicate. For size measurement, 25–50 colonies, which were at least 3 mm away from surrounding colonies or the edge of the Petri dish, were randomly selected from each plate, and the diameter was measured using the measurement function of FIJI (Schindelin et al.2012). Disk diffusion assay Sensitivity to hydrogen peroxide was tested using a previously established method (Wonderling, Wilkinson and Bayles 2004) with modifications. Briefly, 10 μl of 1.5% (w/v) H2O2 was spotted onto 7 mm Whatman paper discs, and the discs were placed on an M63 agar spread with mycobacterial cells. Zones of inhibition were measured in triplicate for each strain after 4-day incubation at 37°C. Ethidium bromide uptake assay Previously published protocol was followed with modifications (Danilchanka, Mailaender and Niederweis 2008). Briefly, log phase (OD600 = 0.5–1.0) cells grown in M63 broth were centrifuged and pellets were resuspended at an equal OD600 reading in 50 mM KH2PO4 (pH 7.0) and 5 mM MgSO4. Cells were then incubated for 5 min with 25 mM glucose, transferred to a 96-well microtiter plate (Brand Tech Scientific), and mixed with 20 μM of ethidium bromide. Fluorescence was measured with an excitation wavelength of 530 nm and an emission wavelength of 590 nm. Antibiotic resistance assay Frozen stocks with known colony forming units (cfu) were prepared for all tested strains by growing cells to an OD600 reading between 0.5 and 1.0 in Middlebrook 7H9 unless otherwise indicated, and freezing in aliquots with a final concentration of 15% (w/v) glycerol at –80°C. In 96-well microtiter plates, antibiotics were serially diluted in 100 μl of media and mixed with cells from the frozen stocks to achieve the final density of 5.0 × 103 cfu/mL. The plates were incubated in a humidity chamber at 37°C. After 24-h incubation, 20 μL of filter-sterilized 0.015% (w/v) resazurin solution was added to each well to initiate colorization. After additional 8 h incubation, the plates were read on a spectrophotometer at 570 and 600 nm. Percent difference in cell viability between antibiotic-treated and control cells was calculated using the formula: (O2 × A1 – O1 × A2)/(O2 × P1 – O1 × P2) × 100, where O1 and O2 are molar extinction coefficient of resazurin (oxidized form) at 570 and 600 nm, respectively; A1 and A2 are absorbance of test wells at 570 and 600 nm, respectively; and P1 and P2 are absorbance of positive control well at 570 and 600 nm, respectively. The IC90 values were calculated using OriginPro 9.1 data analysis software. RESULTS Impact of ΔpimE mutation on planktonic growth We have previously reported that the ΔpimE mutant shows little defect in planktonic growth (Morita et al.2006). However, a closer examination of the logarithmic phase revealed a small but reproducible delay in the doubling time of ΔpimE compared with that of the WT cells (Table S2, Supporting Information, see Middlebrook 7H9). The ΔpimE mutant complemented with the PimE or PimE-FLAG expression vector (ΔpimE::pimE or ΔpimE::pimE-FLAG, respectively), but not with the empty vector (ΔpimE::empty vector), restored the doubling time to the values closer to that of the WT cells. Furthermore, when ΔpimE was complemented with the gene encoding enzymatically inactive PimE[D58A]-FLAG (ΔpimE::pimE[D58A]-FLAG) (Morita et al.2006), the delay in the doubling time was not restored, indicating that the catalytic activity of PimE is required for the WT level of planktonic growth. Small size and aberrant morphology of the ΔpimE mutant colony As we reported briefly in our recent study (Rahlwes et al.2017), a more prominent defect was observed when ΔpimE mutant was grown on a Middlebrook 7H10 agar plate (Fig. 2a). WT, ΔpimE::pimE, and ΔpimE::pimE-FLAG formed normal large colonies, whereas ΔpimE, ΔpimE::empty vector and ΔpimE::pimE[D58A]-FLAG formed significantly smaller colonies. After a 4-day incubation at 37°C on Middlebrook 7H10 agar plates, colony sizes of the WT were 4.17 mm in diameter on average while those of ΔpimE were 1.31 mm (Table 1). The WT colonies showed typical wrinkled surface with irregular edges, whereas the mutant colonies showed lumpy cell aggregates without wrinkled spreading or irregular edges (Fig. 2b). The undulating pattern of WT biofilm formation was also evident along the heavily streaked region of the plate, while the mutant strains lacked such patterns completely (Fig. 2c). The morphological features of ΔpimE::pimE[D58A]-FLAG on the agar surface were comparable to those of ΔpimE or ΔpimE::empty vector, suggesting the requirement of PimE catalytic activity. Figure 2. View largeDownload slide Colony growth phenotype of ΔpimE mutants on Middlebrook 7H10 agar after 4 days at 37°C. (a) Whole plate views. (b) Magnified views of the WT and ΔpimE mutant colonies. Scale bars = 1 mm. (c) Enlargement of the plates shown in the panel a, focusing on the biofilm formation of the heavily streaked regions. These experiments were repeated more than twice and representative images are shown. Figure 2. View largeDownload slide Colony growth phenotype of ΔpimE mutants on Middlebrook 7H10 agar after 4 days at 37°C. (a) Whole plate views. (b) Magnified views of the WT and ΔpimE mutant colonies. Scale bars = 1 mm. (c) Enlargement of the plates shown in the panel a, focusing on the biofilm formation of the heavily streaked regions. These experiments were repeated more than twice and representative images are shown. Table 1. Colony size of M. smegmatis ΔpimE. Strain Middlebrook 7H10 M63 DIY 7H10 (+CuSO4) DIY 7H10 (–CuSO4) WT 4.17 ± 0.91 3.33 ± 1.04 4.39 ± 0.76 3.62 ± 0.67 ΔpimE 1.31 ± 0.45 3.16 ± 1.04 0.75 ± 0.19 2.61 ± 0.51 Strain Middlebrook 7H10 M63 DIY 7H10 (+CuSO4) DIY 7H10 (–CuSO4) WT 4.17 ± 0.91 3.33 ± 1.04 4.39 ± 0.76 3.62 ± 0.67 ΔpimE 1.31 ± 0.45 3.16 ± 1.04 0.75 ± 0.19 2.61 ± 0.51 Colony size (mm) after 4-day incubation on Middlebrook 7H10 and DIY 7H10 agar plates, or 3-day incubation on M63 agar plates, at 37°C. N = 50 for Middlebrook 7H10 and M63, or N = 25 for DIY 7H10. View Large Defective pellicle formation of ΔpimE is medium-dependent We next examined the formation of pellicles, a bacterial community at the liquid–air interface (Islam, Richards and Ojha 2012). Using a previously described protocol (Ojha and Hatfull 2007), we observed a delay in ΔpimE pellicle formation on M63 broth at day 3. Nevertheless, both WT and ΔpimE reached a dense and highly wrinkled pellicle by day 5 (Fig. 3a), suggesting that ΔpimE is not significantly affected in pellicle formation. However, an alternative interpretation is that the differential outcomes of ΔpimE community formations (i.e. colony vs. pellicle) might originate from the use of the different media (i.e. Middlebrook 7H10 agar vs. M63 broth). To distinguish these possibilities, we tested pellicle formation on Middlebrook 7H9 broth, which is similar to Middlebrook 7H10 agar in composition. Strikingly, we found defective pellicle formation of ΔpimE and ΔpimE::pimE[D58A]-FLAG on 7H9 broth in comparison to WT and ΔpimE::pimE-FLAG (Fig. S1, Supporting Information). The observed defect remained evident even after a prolonged 7-day culture, suggesting that it is not a delay but a defect in pellicle formation (Fig. S1, Supporting Information). These data indicated that the normal pellicle formation of ΔpimE is dependent on the use of M63 broth. Figure 3. View largeDownload slide Growth of ΔpimE and WT on M63. (a) Pellicles formed on the surface of M63 broth. Tween-80 was omitted from the medium. (b) Whole plate view of colonies grown on solid M63 agar after 4 days at 37°C. (c) Magnified views of the WT and ΔpimE mutant colonies. Scale bars = 1 mm. (d) Enlargement of the plates shown in the panel a, focusing on the biofilm formation of the heavily streaked regions. These experiments were repeated more than twice and representative images are shown. Figure 3. View largeDownload slide Growth of ΔpimE and WT on M63. (a) Pellicles formed on the surface of M63 broth. Tween-80 was omitted from the medium. (b) Whole plate view of colonies grown on solid M63 agar after 4 days at 37°C. (c) Magnified views of the WT and ΔpimE mutant colonies. Scale bars = 1 mm. (d) Enlargement of the plates shown in the panel a, focusing on the biofilm formation of the heavily streaked regions. These experiments were repeated more than twice and representative images are shown. Other growth defects of ΔpimE are alleviated in M63 medium These results prompted us to examine if the other growth defects described so far are also dependent on the medium. First, we compared the doubling time of planktonic growth in M63, and found that the doubling time of ΔpimE was not substantially different from that of the WT (Table S2, Supporting Information). Next, we grew the cells on M63 agar for 4 days, and found that ΔpimE formed colonies that are similar in size to WT (Fig. 3b and c) (Table 1). In addition, the undulating pattern of biofilm along the heavily streaked region was also restored in ΔpimE grown on the M63 plate (Fig. 3d). Taken together, all types of growth defects observed in ΔpimE were dependent on Middlebrook media. Copper in the Middlebrook media is a cause of the ΔpimE growth defects To systematically analyze and identify the factor(s) that contribute to the observed growth defect of ΔpimE, we first reproduced the growth defect of ΔpimE in our custom-made Middlebrook 7H10 medium herein referred as DIY 7H10 (Table 1, Fig. 4a). One difference between Middlebrook 7H10 and M63 is the main carbon source: 0.2% glucose and 0.5% glycerol for Middlebrook 7H10 and 2% glucose for M63. The presence of glycerol or the relatively reduced level of carbon source in Middlebrook 7H10 could explain the small colony of ΔpimE. We therefore replaced the carbon sources of Middlebrook 7H10 with 2% glucose used in M63. The increased concentration of glucose and the lack of glycerol in this modified Middlebrook 7H10 medium had no effect on the colony size of WT or ΔpimE (Table S3, Supporting Information), indicating that the small colony phenotype of ΔpimE is not due to the carbon sources. Middlebrook 7H10 and M63 are also different in concentrations of metal ions. Therefore, in another series of the experiment, we considered the toxicity of metal ions and reduced the concentration of each metal ion in DIY 7H10 to 1% or 0% of the standard concentration. When copper was reduced to 0% in DIY 7H10, we found that the colony size of ΔpimE was comparable to that of WT (Table 1; Fig. 4a). This improvement was specific to copper and reduction in other metal ions did not restore the colony size of ΔpimE significantly (data not shown). Next, we tested if removing copper could also restore the defect in planktonic growth. The removal of copper from DIY 7H9 broth allowed both WT and ΔpimE to grow faster (Table S4, Supporting Information). Specifically, the doubling time of ΔpimE was reduced by 20% (from 5.15 to 4.14 h), while that of WT was reduced by 11% (from 4.02 to 3.56 h). These results indicate that copper inhibits the growth of M. smegmatis at the concentration used in Middlebrook media, and the planktonic growth of ΔpimE improves to a greater extent than that of WT by removing this metal ion. We further tested the effect of removing copper from DIY 7H9 on the pellicle formation. Interestingly, pellicle formation was not restored by the removal of copper, suggesting that other factors are involved in the restoration of pellicle formation on M63 (Fig. S2, Supporting Information) (see Discussion). Figure 4. View largeDownload slide Effect of copper on the colony size of ΔpimE. (a) Colony growth of ΔpimE and WT on DIY 7H10 agar with or without copper. (b) Impact of copper on the colony size of ΔpimE grown on M63. Colonies were grown for 4 days at 37°C. Box plot from N = 25. ***P < 0.0001 by one-way ANOVA against 0 μM datasets. (c) Ethidium bromide uptake assay. Each data point represents the average of quadruplicate ± standard deviation. All experiments were repeated more than twice and representative data are shown. Figure 4. View largeDownload slide Effect of copper on the colony size of ΔpimE. (a) Colony growth of ΔpimE and WT on DIY 7H10 agar with or without copper. (b) Impact of copper on the colony size of ΔpimE grown on M63. Colonies were grown for 4 days at 37°C. Box plot from N = 25. ***P < 0.0001 by one-way ANOVA against 0 μM datasets. (c) Ethidium bromide uptake assay. Each data point represents the average of quadruplicate ± standard deviation. All experiments were repeated more than twice and representative data are shown. To confirm that copper is the main contributor of the growth defect on Middlebrook 7H10, we compared the sizes of WT and ΔpimE colonies on M63 agar in the presence of various metal ions at concentrations used in Middlebrook 7H10. As shown in Table S5, Supporting Information, and Fig. 4b, increasing concentrations of copper made the colony size of ΔpimE progressively smaller, while changing the concentrations of other metal ions to those found in Middlebrook 7H10 showed no inhibitory effect. Because copper is a redox-active metal ion, we also tested the sensitivity of ΔpimE to reactive oxygen species. We grew WT and ΔpimE on M63 medium with a filter disk containing 1.5% hydrogen peroxide. The zone of inhibition was 25.3 ± 2.1 and 25.7 ± 2.3 mm (average of triplicate ± standard deviation) for WT and ΔpimE, respectively, suggesting that the increased sensitivity of ΔpimE to copper may not be due to the hypersensitivity to oxidative stress. We next considered if the increased copper sensitivity of ΔpimE is due to the increased permeability of the ΔpimE cell envelope. Ethidium bromide permeability assay is a frequently used method to test the permeability of mycobacterial cell envelope (Danilchanka, Mailaender and Niederweis 2008). When we compared the uptake of ethidium bromide by WT and ΔpimE, we found that ΔpimE was approximately three times faster in accumulating the dye (Fig. 4c). This result suggests that the defective permeability of the ΔpimE cell envelope contributes to the increased sensitivity of ΔpimE to copper. Sensitivity to antibiotics We have previously demonstrated that mutants with structural alterations in LM/LAM show increased sensitivities to various antibiotics (Fukuda et al.2013). We have also shown that ΔpimE is hypersensitive to toxic chemicals such as sodium dodecyl sulfate, malachite green and crystal violet (Fukuda et al.2013), suggesting that the cell envelope integrity is compromised. In the previous section, we further demonstrated that the permeability barrier of ΔpimE appears to be compromised. Extending these previous observations, we examined the antibiotic sensitivities of ΔpimE in Middlebrook 7H9 broth and found that the mutant is more sensitive to specific antibiotics such as vancomycin, cefotaxime and erythromycin (Table S6, Supporting Information). Furthermore, in the cases of vancomycin and cefotaxime, the resistance to antibiotics was at least partially restored in ΔpimE::pimE-FLAG but not in ΔpimE::pimE[D58A]-FLAG. Notably, the increased sensitivity of ΔpimE to erythromycin was not restored in ΔpimE::pimE-FLAG, suggesting either that the increased sensitivity is not dependent on pimE deletion or that the level of expression from the artificial HSP60 promoter is not optimal for restoration even though hexamannosyl PIM formation is completely restored (Morita et al.2006). Isoniazid and clarithromycin were effective against WT and sensitivities to these compounds were not significantly affected by pimE deletion. Taken together, we conclude that ΔpimE is moderately sensitive to certain antibiotics, likely arising from defective cell envelope lacking hexamannosyl PIM species. Because the planktonic growth of ΔpimE was restored in M63 or DIY 7H9 broth, we tested if the antibiotic resistance could be restored when the mutant is grown in these media. As shown in Table 2, both WT and ΔpimE became more resistant to vancomycin, cefotaxime, clarithromycin and erythromycin when tested in M63 medium (see columns 1 and 2, compare with Table S6, Supporting Information). While both WT and ΔpimE became completely resistant to cefotaxime, clarithromycin and erythromycin (i.e. IC90 > 100 μg/ml), their vancomycin sensitivity in M63 medium remained within our assay range. Importantly, IC90 of ΔpimE for vancomycin in M63 medium was nearly five times lower than that of WT, suggesting that antibiotic resistance of ΔpimE was not restored to the WT level, while cells are generally more resistant to antibiotics in this medium. Table 2. IC90 of M. smegmatis ΔpimE in various media. Column no. 1 2 3 4 5 6 7 8 Growth medium M63 M63 M63 M63 DIY 7H9 (–CuSO4) DIY 7H9 (–CuSO4) 7H9 7H9 Strains WT ΔpimE WT ΔpimE WT ΔpimE WT ΔpimE Stock DIY 7H9 DIY 7H9 DIY 7H9 DIY 7H9 DIY 7H9 DIY 7H9 prepared in 7H9 7H9 (–CuSO4) (–CuSO4) (–CuSO4) (–CuSO4) (–CuSO4) (–CuSO4) Vancomycin 15.03 ± 3.99 3.04 ± 0.66 5.93 ± 0.47 1.87 ± 0.18 2.54 ± 0.39 1.97 ± 0.36 0.46 ± 0.01 0.26 ± 0.01 Cefotaxime >100 >100 >100 >100 >100 >100 56.9 ± 1.7 30.2 ± 3.3 Clarithromycin >100 >100 0.30 ± 0.03 0.09 ± 0.01 0.15 ± 0.01 0.09 ± 0.01 0.20 ± 0.01 0.09 ± 0.01 Erythromycin >100 >100 6.63 ± 0.94 0.55 ± 0.08 1.18 ± 0.01 0.34 ± 0.03 0.50 ± 0.04 0.19 ± 0.02 Column no. 1 2 3 4 5 6 7 8 Growth medium M63 M63 M63 M63 DIY 7H9 (–CuSO4) DIY 7H9 (–CuSO4) 7H9 7H9 Strains WT ΔpimE WT ΔpimE WT ΔpimE WT ΔpimE Stock DIY 7H9 DIY 7H9 DIY 7H9 DIY 7H9 DIY 7H9 DIY 7H9 prepared in 7H9 7H9 (–CuSO4) (–CuSO4) (–CuSO4) (–CuSO4) (–CuSO4) (–CuSO4) Vancomycin 15.03 ± 3.99 3.04 ± 0.66 5.93 ± 0.47 1.87 ± 0.18 2.54 ± 0.39 1.97 ± 0.36 0.46 ± 0.01 0.26 ± 0.01 Cefotaxime >100 >100 >100 >100 >100 >100 56.9 ± 1.7 30.2 ± 3.3 Clarithromycin >100 >100 0.30 ± 0.03 0.09 ± 0.01 0.15 ± 0.01 0.09 ± 0.01 0.20 ± 0.01 0.09 ± 0.01 Erythromycin >100 >100 6.63 ± 0.94 0.55 ± 0.08 1.18 ± 0.01 0.34 ± 0.03 0.50 ± 0.04 0.19 ± 0.02 All data are given in μg/mL as average ± standard deviation from triplicate data. View Large Copper is known to induce global transcriptional activation, and M. smegmatis becomes resistant to antibiotics when exposed to copper (Rao et al.2012). In the antibiotic sensitivity assays described above, we prepared the frozen stocks by growing M. smegmatis in Middlebrook 7H9 that is rich in copper. Therefore, we wondered if the increased resistance of both WT and ΔpimE in M63 medium could be partially due to the prior exposure of our frozen stock to copper, followed by the growth assay in copper-free M63. To test this possibility, we prepared another set of frozen stocks in DIY 7H9 (–CuSO4) and tested antibiotic sensitivity in M63. Consistent with our prediction, we found a modest increase in sensitivity (i.e. decrease in IC90) to vancomycin and more striking increases in sensitivity to clarithromycin and erythromycin (Table 2, columns 3 and 4). These observations are consistent with the published report, proposing that the exposure to copper makes the cells resistant to antibiotics (Rao et al.2012). Similar levels of inhibitions were observed when cells prepared in DIY 7H9 (–CuSO4) were assayed in DIY 7H9 (–CuSO4) (Table 2, columns 5 and 6). Furthermore, cells prepared in DIY 7H9 (–CuSO4) were more sensitive to the antibiotics when assayed in regular Middlebrook 7H9 (Table 2, columns 7 and 8), presumably because of the additive toxicity of the high copper concentration in Middlebrook 7H9. Most importantly, in all cases, ΔpimE remained more sensitive to antibiotics than WT, even though its growth phenotypes are significantly improved in the copper-free media. DISCUSSION In this report, we revealed that the high concentration of copper in Middlebrook media caused the observed defects of ΔpimE in planktonic growth and colony formation. PimE is a dual-functional enzyme that also mediates terminal α1,2 mannose additions of protein O-mannosylation (Liu et al.2013). However, the lack of protein O-mannosylation has no effects on M. smegmatis growth. Specifically, the deletion of the gene encoding the protein O-mannosyltransferase, MSMEG_5447, resulted in complete abrogation of protein O-mannosylation in M. smegmatis (Liu et al.2013). This mutant shows no defects in PIM biosynthesis, and the in vitro growth in Middlebrook 7H9 was unaffected, suggesting that the growth defects of ΔpimE observed in our current study are due to the defects in hexamannosyl PIM biosynthesis. AcPIM6 and Ac2PIM6 are thought to be present predominantly in the plasma membrane, and only a minor fraction of PIM species is present on the outer membrane (Bansal-Mutalik and Nikaido 2014). Nevertheless, antibiotics such as cefotaxime and vancomycin that have periplasmic targets showed increased efficacy against ΔpimE compared with WT. These findings collectively suggest that either the lack of AcPIM6 or the accumulation of AcPIM4 in the plasma membrane leads to a more global defect of the cell envelope as a permeability barrier. Our current study revealed a link between glycolipid-dependent cell envelope integrity and copper toxicity, and our ethidium bromide permeability assay supported the possibility that the increased sensitivity of ΔpimE to copper is due to the compromised cell envelope permeability barrier. However, resistance of mycobacteria to copper is a highly regulated process mediated by a complex molecular network involving many proteins (Rowland and Niederweis 2012; Darwin 2015). For example, MctB confers mycobacteria resistance to copper (Wolschendorf et al.2011), and is proposed to be a plasma membrane-associated periplasmic protein (Rowland and Niederweis 2012). Thus, the plasma membrane defect caused by the pimE deletion could have an indirect impact on the function of MctB or other proteins involved in copper resistance. Finally, copper is not the only factor associated with the growth defects observed in the ΔpimE mutant. Specifically, the removal of copper from DIY 7H9 did not rescue the pellicle formation of ΔpimE. This observation indicates that another medium component in either DIY 7H9 or M63 is inhibitory or beneficial to the pellicle formation of the ΔpimE mutant. We speculate that the colony formation on a solid agar surface and the pellicle formation on a liquid surface are two different phenomena requiring distinct environmental conditions. Precise factor(s) promoting or inhibiting the pellicle formation of ΔpimE remain to be determined. The ΔpimE mutant also showed increased sensitivities to various antibiotics, and this effect was specifically due to the lack of the enzyme's catalytic activity because functional PimE, but not inactive PimE, could restore the resistance to these antibiotics. The only exception was that erythromycin sensitivity of ΔpimE could not be restored by the complemented strain. Since ΔpimE::pimE[D58A]-FLAG showed more sensitivity to erythromycin than ΔpimE, we suspect that the overexpression of the PimE protein has a subtle impact on the cell envelope integrity, affecting the resistance of the complement to erythromycin. Selective sensitivities of certain cephalosporins have been observed in LM/LAM mutants of M. smegmatis, where LM/LAM mutants are sensitive to cefotaxime more than 16 times, while the mutant's resistance to other cephalosporins was unaffected (Fukuda et al.2013). In the current study, ΔpimE mutant showed a similar result, in that cefotaxime sensitivity was increased up to approximately two times in ΔpimE while other cephalosporins such as cefalothin and cefamandole had no effect. A possible explanation could be that cefotaxime is a poor substrate of β-lactamase making it more effective than other cepholosporins (Quinting et al.1997). Taken together, our current report suggests that PimE-mediated mannosylation plays crucial roles in maintaining cell envelope integrity in M. smegmatis, and is physiologically linked to the formation of biofilms such as colonies and pellicles. Biofilms are implicated in the pathogenesis of tuberculosis (Islam, Richards and Ojha 2012), and PimE is predicted to be an essential gene in M. tuberculosis (Griffin et al.2011). Therefore, it is important to clarify the precise roles of this enzyme in M. tuberculosis during the biofilm formation and the establishment of the infection. SUPPLEMENTARY DATA Supplementary data are available at FEMSLE online. Acknowledgements We thank Manju Sharma for support, and Corelle Rokicki for help with the graphical abstract and critical reading of the manuscript. FUNDING This work was supported by grants from the Pittsfield Anti-Tuberculosis Association and the American Lung Association [Biomedical Research Grant RG-414805] to YSM. Conflict of interest. None declared. REFERENCES Albesa-Jové D, Svetlikova Z, Tersa M et al. Structural basis for selective recognition of acyl chains by the membrane-associated acyltransferase PatA. Nat Commun 2016; 7: 10906. Google Scholar CrossRef Search ADS PubMed Bansal-Mutalik R, Nikaido H. Mycobacterial outer membrane is a lipid bilayer and the inner membrane is unusually rich in diacyl phosphatidylinositol dimannosides. P Natl Acad Sci USA 2014; 111: 4958– 63. Google Scholar CrossRef Search ADS Crellin PK, Kovacevic S, Martin KL et al. Mutations in pimE restore lipoarabinomannan synthesis and growth in a Mycobacterium smegmatis lpqW mutant. J Bacteriol 2008; 190: 3690– 9. Google Scholar CrossRef Search ADS PubMed Danilchanka O, Mailaender C, Niederweis M. Identification of a novel multidrug efflux pump of Mycobacterium tuberculosis. Antimicrob Agents Ch 2008; 52: 2503– 11. Google Scholar CrossRef Search ADS Darwin KH. Mycobacterium tuberculosis and copper: a newly appreciated defense against an old foe? J Biol Chem 2015; 290: 18962– 6. Google Scholar CrossRef Search ADS PubMed Fukuda T, Matsumura T, Ato M et al. Critical roles for lipomannan and lipoarabinomannan in cell wall integrity of mycobacteria and pathogenesis of tuberculosis. mBio 2013; 4: e00472-12. Google Scholar CrossRef Search ADS PubMed Griffin JE, Gawronski JD, DeJesus MA et al. High-resolution phenotypic profiling defines genes essential for mycobacterial growth and cholesterol catabolism. PLoS Pathog 2011; 7: e1002251. Google Scholar CrossRef Search ADS PubMed Guerin ME, Kaur D, Somashekar BS et al. New insights into the early steps of phosphatidylinositol mannoside biosynthesis in mycobacteria: PimB' is an essential enzyme of Mycobacterium smegmatis. J Biol Chem 2009; 284: 25687– 96. Google Scholar CrossRef Search ADS PubMed Ishikawa E, Mori D, Yamasaki S. Recognition of mycobacterial lipids by immune receptors. Trends Immunol 2017; 38: 66– 76. Google Scholar CrossRef Search ADS PubMed Islam MS, Richards JP, Ojha AK. Targeting drug tolerance in mycobacteria: a perspective from mycobacterial biofilms. Expert Rev Anti Infect Ther 2012; 10: 1055– 66. Google Scholar CrossRef Search ADS PubMed Jackson M, Crick DC, Brennan PJ. Phosphatidylinositol is an essential phospholipid of mycobacteria. J Biol Chem 2000; 275: 30092– 9. Google Scholar CrossRef Search ADS PubMed Jankute M, Cox JAG, Harrison J et al. Assembly of the mycobacterial cell wall. Annu Rev Microbiol 2015; 69: 405– 23. Google Scholar CrossRef Search ADS PubMed Jarlier V, Nikaido H. Permeability barrier to hydrophilic solutes in Mycobacterium chelonei. J Bacteriol 1990; 172: 1418– 23. Google Scholar CrossRef Search ADS PubMed Kaur D, Guerin ME, Škovierová H et al. Chapter 2: Biogenesis of the cell wall and other glycoconjugates of Mycobacterium tuberculosis. Adv Appl Microbiol 2009; 69: 23– 78. Google Scholar CrossRef Search ADS PubMed Korduláková J, Gilleron M, Mikušová K et al. Definition of the first mannosylation step in phosphatidylinositol mannoside synthesis. PimA is essential for growth of mycobacteria. J Biol Chem 2002; 277: 31335– 44. Google Scholar CrossRef Search ADS PubMed Korduláková J, Gilleron M, Puzo G et al. Identification of the required acyltransferase step in the biosynthesis of the phosphatidylinositol mannosides of Mycobacterium species. J Biol Chem 2003; 278: 36285– 95. Google Scholar CrossRef Search ADS PubMed Kovacevic S, Anderson D, Morita YS et al. Identification of a novel protein with a role in lipoarabinomannan biosynthesis in mycobacteria. J Biol Chem 2006; 281: 9011– 7. Google Scholar CrossRef Search ADS PubMed Lea-Smith DJ, Martin KL, Pyke JS et al. Analysis of a new mannosyltransferase required for the synthesis of phosphatidylinositol mannosides and lipoarbinomannan reveals two lipomannan pools in Corynebacterineae. J Biol Chem 2008; 283: 6773– 82. Google Scholar CrossRef Search ADS PubMed Liu C-F, Tonini L, Malaga W et al. Bacterial protein-O-mannosylating enzyme is crucial for virulence of Mycobacterium tuberculosis. P Natl Acad Sci USA 2013; 110: 6560– 5. Google Scholar CrossRef Search ADS Mishra AK, Driessen NN, Appelmelk BJ et al. Lipoarabinomannan and related glycoconjugates: structure, biogenesis and role in Mycobacterium tuberculosis physiology and host-pathogen interaction. FEMS Microbiol Rev 2011; 35: 1126– 57. Google Scholar CrossRef Search ADS PubMed Morita YS, Fukuda T, Sena CBC et al. Inositol lipid metabolism in mycobacteria: biosynthesis and regulatory mechanisms. BBA-Gen Subjects 2011; 1810: 630– 41. Google Scholar CrossRef Search ADS Morita YS, Sena CBC, Waller RF et al. PimE is a polyprenol-phosphate-mannose-dependent mannosyltransferase that transfers the fifth mannose of phosphatidylinositol mannoside in mycobacteria. J Biol Chem 2006; 281: 25143– 55. Google Scholar CrossRef Search ADS PubMed Ojha A, Hatfull GF. The role of iron in Mycobacterium smegmatis biofilm formation: the exochelin siderophore is essential in limiting iron conditions for biofilm formation but not for planktonic growth. Mol Microbiol 2007; 66: 468– 83. Google Scholar CrossRef Search ADS PubMed Quinting B, Galleni M, Timm J et al. Purification and properties of the Mycobacterium smegmatis mc2155 beta-lactamase. FEMS Microbiol Lett 1997; 149: 11– 5. Google Scholar CrossRef Search ADS PubMed Rahlwes KC, Ha SA, Motooka D et al. The cell envelope-associated phospholipid-binding protein LmeA is required for mannan polymerization in mycobacteria. J Biol Chem 2017; 292: 17407– 17. Google Scholar CrossRef Search ADS PubMed Rao M, Liu H, Yang M et al. A copper-responsive global repressor regulates expression of diverse membrane-associated transporters and bacterial drug resistance in mycobacteria. J Biol Chem 2012; 287: 39721– 31. Google Scholar CrossRef Search ADS PubMed Rowland JL, Niederweis M. Resistance mechanisms of Mycobacterium tuberculosis against phagosomal copper overload. Tuberculosis 2012; 92: 202– 10. Google Scholar CrossRef Search ADS PubMed Schindelin J, Arganda-Carreras I, Frise E et al. Fiji: an open-source platform for biological-image analysis. Nat Methods 2012; 9: 676– 82. Google Scholar CrossRef Search ADS PubMed Sena CBC, Fukuda T, Miyanagi K et al. Controlled expression of branch-forming mannosyltransferase is critical for mycobacterial lipoarabinomannan biosynthesis. J Biol Chem 2010; 285: 13326– 36. Google Scholar CrossRef Search ADS PubMed Snapper SB, Melton RE, Mustafa S et al. Isolation and characterization of efficient plasmid transformation mutants of Mycobacterium smegmatis. Mol Microbiol 1990; 4: 1911– 9. Google Scholar CrossRef Search ADS PubMed Weidenmaier C, Peschel A. Teichoic acids and related cell-wall glycopolymers in Gram-positive physiology and host interactions. Nat Rev Micro 2008; 6: 276– 87. Google Scholar CrossRef Search ADS Wolschendorf F, Ackart D, Shrestha TB et al. Copper resistance is essential for virulence of Mycobacterium tuberculosis. P Natl Acad Sci USA 2011; 108: 1621– 6. Google Scholar CrossRef Search ADS Wonderling LD, Wilkinson BJ, Bayles DO. The htrA (degP) gene of Listeria monocytogenes 10403S is essential for optimal growth under stress conditions. Appl Environ Microb 2004; 70: 1935– 43. Google Scholar CrossRef Search ADS © FEMS 2018. All rights reserved. For permissions, please e-mail: firstname.lastname@example.org
FEMS Microbiology Letters – Oxford University Press
Published: Mar 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
All the latest content is available, no embargo periods.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud