Bacillus anthracis Sterne strain 7702 (pXO1⁺, pXO2^-, American Type Culture Collection, Manassas, VA, USA) was used in these experiments and designated as parent strain. B. anthracis ΔftsX and B. anthracis ftsE(K123A/D481N) were derived from the Sterne strain as previously described (Crawford et al., 2011; Margulieux et al., 2016). Vegetative cells were prepared by inoculating B. anthracis parent strain spore stocks or B. anthracis mutant strain vegetative cell frozen stocks into 10 ml of brain heart infusion (BHI) broth (Difco, Franklin Lakes, NJ, USA) and incubating overnight at 37°C with shaking (250 rpm). Mid-log phase cultures were prepared the following day by diluting the overnight culture 1:20 in fresh BHI broth followed by incubation at 37°C with shaking for ∼2 h until the culture reached an optical density at 600 nm (OD₆₀₀) between 0.6 and 0.65, at which time the cells were used for experiments. All laboratory work involving B. anthracis Sterne or Sterne-derived strains was approved through the University of Virginia Institutional Biosafety Committee. Biosafety level 2 practices were used for all work involving B. anthracis Sterne or Sterne-derived strains.

Recent advances in the field of peptidoglycan labeling have resulted in the development of a variety of fluorescent probes for amino acids that are incorporated into the peptide side chain of peptidoglycan during cell growth and division (Kuru et al., 2012, 2015; Popham, 2013; Liechti et al., 2014, 2016; Boersma et al., 2015). Ethynyl-D-alanine (EDA) is a D-alanine molecule with an additional alkyne group that can be “clicked” together, using the principles of “click” chemistry, with fluorophores containing a reactive azide group, through a copper-mediated reaction, resulting in a stable, coupled conjugate (Kolb et al., 2001).

EDA [(S)-2-Amino-4-pentynoic acid] (Boaopharma, Inc., Natick, MA, USA) was dissolved in H₂O and used at a concentration of 1 mM. EDA was used in conjunction with the Click-iT^® Cell Reaction Buffer Kit (Invitrogen, Carlsbad, CA, USA) and ‘clickable’ Alexa Fluor 488 Azide (Invitrogen, Carlsbad, CA, USA) per manufacturer’s instructions. Overnight bacterial cultures of B. anthracis parent strain, ΔftsX strain, or ftsE(K123A/D481N) strain were diluted 1:20 in fresh BHI broth followed by a 1 h incubation at 37°C with shaking at 250 rpm. After 1 h, 100 mM EDA stock solution was added to the culture to a final concentration of 1 mM, and incubation was continued until an OD₆₀₀ of between 0.6 and 0.65 was reached. The entire culture was centrifuged for 5 min at 3600 rpm to collect a pellet of the log-phase bacteria. The supernatant was discarded, and a total volume of 5 μl of the resulting pellet was placed into 500 μl Dulbecco’s modified Eagle’s medium (DMEM) (Gibco-Invitrogen, Carlsbad, CA, USA) + 10% fetal bovine serum (FBS) + 1 mM EDA and incubated for 1 h at 37°C in 5% CO₂. After 1 h incubation with EDA, the bacterial culture was centrifuged for 8 min at 15,500 rpm, and the supernatant was discarded. A 1 ml volume of ice-cold 70% ethanol was added to the cell pellet and incubated for 1 h on ice. The ethanol solution was removed and the bacterial cells were resuspended and incubated for 1 h at room temperature in 1 ml PBS containing 3% BSA. Cells were then incubated for 1 h at room temperature in 0.5 ml of Click-iT^® Cell Reaction Buffer according to the manufacturer’s directions with Alexa Fluor 488 Azide (ThermoFisher Scientific, Waltham, MA, USA) to fluorescently label the EDA incorporated into the cell wall peptidoglycan. The bacterial cells were re-suspended in PBS and placed on glass slides with a coverslip and imaged using a Zeiss LSM 880 confocal microscope (Zeiss, Oberkochem, Germany) using 488 nm (Alexa Fluor 488) wavelength filter at 63X magnification.

TAMRA-D-lysine (TDL) was synthesized according to the protocol described previously (Kuru et al., 2012, 2015). Briefly, 5-carboxytetramethylrhodamine succinimidyl ester (Anaspec, Fremont, CA, USA) and N-α-BocD-Lys-OH (Chem-Impex, Bensenville, IL, USA) were mixed to form TDL. The product was purified using reverse phase HPLC on a Waters ZQ 2000 LCMS system (Waters, Milford, MA, USA) with a gradient of acetonitrile in water with 0.1% (vol/vol) formic acid 5–95% (vol/vol) acetonitrile for 20 min, eluting at 20 ml min^-1. The two isomers were eluted on a Waters Sunfire Prep C18 OBD column with a 5 μm particle size, 19 mm internal diameter and length of 100 mm (Waters, Milford, MA, USA, product number 186008153). The solution containing the purified product was then lyophilized overnight to form a stable TDL powder, which was stored at -20°C and protected from light.

To perform dual labeling experiments, overnight cultures were diluted 1:20 in fresh BHI broth followed by a 1 h incubation at 37°C with shaking at 250 rpm. After 1 h, 100 mM EDA stock solution was added to the culture to a final concentration of 1 mM, and incubation was continued until an OD₆₀₀ of between 0.6 and 0.65 was reached. The entire culture was centrifuged for 5 min at 3600 rpm to collect a pellet of the log-phase bacteria. The supernatant was discarded, and a total volume of 5 μl of the resulting pellet was placed into 500 μl DMEM (Gibco-Invitrogen, Carlsbad, CA, USA) + 10% FBS + 1 mM EDA and incubated for 1 h at 37°C in 5% CO₂. After this 1 h incubation with EDA, the bacterial culture was centrifuged for 8 min at 15,500 rpm, and the supernatant discarded. A total volume of 5 μl of the resulting EDA-labeled pellet was placed into 500 μl DMEM + 10% FBS + 0.5 mM TDL, -/+ penicillin (35 μg/ml), tetracycline (40 μg/ml), CXCL10 (1.4 μM), or CTTC (2.8 μM) and incubated 1 h at 37°C in 5% CO₂. An aliquot of each sample was transferred to individual 1.5 ml Eppendorf tubes and centrifuged for 8 min at 15,500 rpm after which the supernatant was discarded and the samples were processed and fixed for imaging as described above.

The bacterial cells were re-suspended in PBS and placed on glass slides with a coverslip and imaged using a Zeiss LSM 880 confocal microscope (Zeiss, Oberkochem, Germany) using 488 nm (Alexa Fluor 488) and 514 nm (TDL) wavelength filters at 63X magnification. Phase contrast light microscopy images and fluorescence images were collected using settings, including brightness and contrast, which were optimized and fixed based on the corresponding untreated control group at the beginning of the acquisition session. Further analyses were completed using Image J (Schneider et al., 2012). Relative fluorescent intensity was calculated as a percent control of the untreated sample. Ten individual fields per experimental run were analyzed to determine mean fluorescence intensity +/- standard deviation, as compared to results for untreated control sample of the same strain. Control and test samples were not compared between runs to account for experimental variation in the percent control determination.

Overnight cultures were diluted 1:20 in fresh BHI broth followed by a 2 h incubation at 37°C with shaking at 250 rpm. Vegetative cells were diluted to approximately 5.0 × 10⁶ CFU ml^-1 in fresh DMEM with high glucose, no glutamine, no phenol red (Gibco-Invitrogen, Carlsbad, CA, USA) that also contained 4 mM GlutaMAX^TM supplement, 10% (v/v) FBS (HyClone, Logan, UT, USA), and the concentration of penicillin, CXCL10, or CTTC to be tested. Aliquots of 400 μl were placed into 48 well plates and incubated at 37°C with 5% CO₂ for 0 or 2 h. Penicillin (0.5 μg ml^-1) was used as a positive control. CXCL10 was tested at a concentration of 1.4 μM, and CTTC was tested at 2.8 μM. At time 0 h or after a 2 h incubation, the supernatants for each sample group were collected and transferred into 1.5 ml Eppendorf tubes and centrifuged for 5 min at 15,000 rpm. The resulting supernatant was sterilized by filtration through a 13 mm 0.22 μM filter (ThermoFisher Scientific, Waltham, MA, USA) and then heat-inactivated for 5 min at 95°C. Concurrent CFU ml^-1 measurements were determined for each sample. The HEK-Blue^TM hNOD2 mammalian cells (Invivogen, San Diego, CA, USA) were grown and maintained in DMEM medium containing 10% FBS, 50 U/ml penicillin, 50 μg/ml streptomycin, 100 μg/ml Normocin, 30 μg/ml blasticidin, and 100 μg/ml Zeocin at 37°C with 5% CO₂ for up to no more than 10 passages and used as per manufacturer’s instructions. Briefly, the cells were sub-cultured into 96-well plates at a concentration of 50,000 cells well^-1 in 180 μl total volume and were incubated with 20 μl of the filtered supernatant in triplicate for 20 h at 37°C with 5% CO₂. Finally, 20 μl of the resulting HEK-Blue^TM hNOD2 supernatant was added to 180 μl HEK-Blue^TM detection medium (Invivogen, San Diego, CA, USA) and incubated for 24 h at 37°C with 5% CO₂. Muramyl dipeptide (MDP) (Invivogen, San Diego, CA, USA) at a final concentration of 15 ng/ml reconstituted as a stock solution of 0.1 mg/ml in pure H₂O was used as a positive control. Filtered DMEM was used as a medium control and pure H₂O was used as a second negative control. The colorimetric change of the detection medium [due to secreted embryonic alkaline phosphatase (SEAP) enzyme released upon NF-κB reporter system activation] was measured at OD₆₅₀ using an ELx800 Absorbance Reader (BioTek, Winooski, VT, USA).

Statistical analyses were performed using GraphPad Prism 6.0 software. Experimental groups were analyzed using an unpaired, two-tailed Student’s t-test or Mann–Whitney test as noted in the text or figure legends. Significant differences were determined to have a p-value of ≤ 0.05.

Previous studies with B. anthracis ΔftsX and ftsE(K123A/D481N) showed that they have about a 30 min lag phase prior to initiation of log phase growth in BHI broth media compared to the parent strain (Crawford et al., 2011; Margulieux et al., 2016). However, their growth rates were similar once log phase was achieved, so all of the following experiments were conducted with mid-log phase cultures. All B. anthracis strains used in this study exhibited marked incorporation of EDA into the septa after 2 h of incubation, indicative of active cellular division. EDA incorporation was also noted along the cell wall of each vegetative cell, suggesting that the modified D-alanine was incorporated into the cell wall of the growing/dividing bacteria (Figure 1A). Notably, significant morphological differences in cell length, cell width, number of septa in a given distance, and bacterial chain angle were observed between the B. anthracis parent strain and the ΔftsX and ftsE(K123A/D481N) strains (Figures 1A–E). In our present study, B. anthracis parent strain appeared as regular, even, and smooth chains of vegetative cells while the mutant strains showed a “kinked” phenotype (Figure 1A). Cells of B. anthracis ΔftsX and ftsE(K123A/D481N) were significantly shorter in length (measured as the septum-to-septum distance of bacterial cells in a chain) and wider (measured as the width across the midpoint of the cell) compared to the parent strain (Figures 1B,C). These morphological differences were consistent with results reported for Bacillus subtilis mutants of the FtsE/X complex (Garti-Levi et al., 2008; Domínguez-Cuevas et al., 2013; Meisner et al., 2013). B. anthracis ΔftsX and ftsE(K123A/D481N) also exhibited an increased number of septal division points over the arbitrarily chosen distance of 10 μm, which suggested an altered or dysregulated process of cellular division and/or elongation in these mutants (Figures 1A,D). The “kinked” morphology, noted above, was quantified through measurement of vegetative cell chain angles from end-middle-end of the entire chain. These measurements indicated that, compared to the B. anthracis parent strain, the B. anthracis ΔftsX and ftsE(K123A/D481N) mutant strains exhibited a wide array of more acute angles within chains of bacteria (Figure 1E).

In order to examine if the antimicrobial effect of CXCL10 and CTTC against vegetative cells of B. anthracis was due to an alteration of peptidoglycan integrity and/or inhibition of new synthesis, we performed dual labeling experiments using EDA and TDL as described in Section “Materials and Methods” to determine if there was a difference in peptidoglycan remodeling. Dual labeling allowed us to determine effects on existing peptidoglycan (EDA-labeled) and new peptidoglycan subunit incorporation (TDL-labeled) in the absence or presence of CXCL10, CTTC, or other antimicrobial agents. We used an untreated sample to standardize the baseline microscopy settings for the treated samples. Images of the untreated and treated sample groups for one bacterial strain were obtained in one session to ensure comparative fluorescent levels. The sequential incorporation of EDA and TDL into the septa of vegetative cells was observed with the untreated parent strain (Figure 2A). Penicillin was used as a positive control as it is a beta-lactam antibiotic that targets the bacterial cell wall and blocks the final cross-linking step, resulting in a loss of peptidoglycan integrity and cell lysis (Figure 2B) (Tipper and Strominger, 1965; Wise and Park, 1965; Novak et al., 2000). Tetracycline was used as a negative control as it is a bacteriostatic agent that inhibits protein synthesis by targeting the bacterial aminoacyl-tRNAs and 30S ribosome complex (Chopra and Roberts, 2001). In the presence of penicillin (35 μg/ml), the parent strain exhibited reduced fluorescence signal from previously incorporated EDA (54.8 + 4.5% of untreated control) and an absence (4.5 + 0.5% of untreated control) of newly incorporated TDL-labeled peptidoglycan subunits (Figure 2B). In contrast, incubation of B. anthracis parent strain with tetracycline (40 μg/ml) had a lesser impact on the incorporation of both EDA (71.0 + 3.4% of untreated control) and TDL (54.7 + 3.6% of untreated control) into the peptidoglycan (data not shown), which likely reflected that this bacteriostatic drug inhibited bacterial growth during the incubation period. In the presence of CXCL10 (1.4 μM), the B. anthracis parent strain exhibited decreased levels of EDA-labeled peptidoglycan (55.4 + 4.9% of untreated control) and reduced levels of TDL-labeled peptidoglycan (25.1 + 3.1% of untreated control) (Figure 2C). Incubation of B. anthracis parent strain with CTTC (2.8 μM) resulted in a similar loss of EDA and TDL fluorescence, 52.4 + 5.2% and 30.4 + 3.7% of untreated control, respectively (Figure 2D).

Bacillus anthracis ΔftsX is less susceptible to CXCL10 than the parent strain and is completely resistant to CTTC (Margulieux et al., 2016). Killing of the B. anthracis ΔftsX strain by higher concentrations of CXCL10 appears to occur by an FtsE/X-independent pathway through general membrane depolarization and disruption via the C-terminal amphipathic α-helix of CXCL10 (Margulieux et al., 2016). The data obtained with the B. anthracis parent strain indicates that CXCL10 impacts peptidoglycan processing via the FtsE/X-dependent pathway; however, the B. anthracis ΔftsX mutant lacks the FtsX target necessary for interaction with CXCL10. Thus, the only pathway by which CXCL10 can exert its effect is the previously described FtsE/X-independent mechanism that results in cellular lysis through generalized membrane depolarization. The B. anthracis ΔftsX untreated control strain incorporated both EDA and TDL into the peptidoglycan (Figure 3A). Penicillin treatment resulted in an absence of TDL incorporation (6.3 + 1.0% of untreated control), i.e., newly synthesized peptidoglycan (Figure 3B). Notably, the EDA fluorescence (93.4 + 10.3% of untreated control) was unaffected by penicillin treatment compared to the penicillin-treated parent strain (Figure 2B), which suggested a possible difference between these strains to the effects of penicillin. B. anthracis ΔftsX incubation with tetracycline gave similar results to those observed with the parent strain (data not shown). Incubation of B. anthracis ΔftsX with CXCL10 resulted in CXCL10 having only a minor effect on the integrity of the existing (pre-treatment) peptidoglycan based on the fluorescence intensity due to EDA (89.5 + 6.6% of untreated control) and an intermediate effect on newly synthesized peptidoglycan based on reduction in TDL signal intensity (44.0 + 3.8% of untreated control) (Figure 3C). Under light microscopy, vegetative cells of ΔftsX exhibited some cell lysis, perhaps due to membrane depolarization and cell death via the FtsX-independent pathway, which may have had some effect on fluorescence intensity compared to the untreated ΔftsX control. CTTC treatment had no impact on EDA (116.8 + 7.2% of untreated control) or TDL (101.4 + 6.3% of untreated control) incorporation/uptake into the peptidoglycan of B. anthracis ΔftsX (Figure 3D). These data suggested that in the absence of FtsX, the N-terminal region of CXCL10 was not able to initiate a bactericidal effect through disruption of peptidoglycan integrity or synthesis.

An FtsE mutant ftsE(K123A/D481N) was previously constructed in which select point mutations in the Walker sites eliminated ATP hydrolysis rendering the FtsE/X complex functionally inactive (Margulieux et al., 2016). Like ΔftsX, the B. anthracis ftsE(K123A/D481N) mutant strain was also found to be less susceptible to CXCL10 and completely resistant to CTTC (Margulieux et al., 2016). Dual labeling of untreated B. anthracis ftsE(K123A/D481N) yielded similar results for EDA and TDL incorporation (Figure 4A) as observed with B. anthracis ΔftsX (Figure 3A). Incubation of ftsE(K123A/D481N) with penicillin resulted in no loss of EDA relative fluorescence intensity (102.8 + 7.1% of untreated control), but a significant reduction in newly synthesized peptidoglycan as evidenced by reduced incorporation of TDL (7.9 + 1.2% of untreated control) (Figure 4B). Tetracycline treatment of the B. anthracis ftsE(K123A/D481N) strain gave results that were similar to those obtained with ΔftsX (data not shown). CXCL10 had little effect on the incorporation of EDA in ftsE(K123A/D481N), while TDL incorporation (48.4 + 5.3% of control) in this mutant strain (Figure 4C) was similar to the results for ΔftsX (Figure 3C). Similar to the findings for the ΔftsX strain in Figure 3, CTTC treatment of the ftsE(K123A/D481N) mutant strain had no apparent impact on existing peptidoglycan integrity as measured by EDA incorporation (97.5 + 3.9% of untreated control) or on newly synthesized peptidoglycan as measured by TDL incorporation (106.3 + 8.0% of untreated control) (Figure 4D).

A mammalian reporter cell line that overexpressed the immune cell receptor NOD2 was used to quantitate the amount of peptidoglycan released (Packiam et al., 2015; Liechti et al., 2016). This assay provided a sensitive method for quantitating the amount of peptidoglycan released by bacteria exposed to antimicrobial agents (Girardin et al., 2003a,b; Loving et al., 2009; Philpott et al., 2014; Packiam et al., 2015). The NOD2 receptors expressed by the reporter cell line are activated by binding fragments containing N-acetylmuramic acid released from bacterial cell walls (Girardin et al., 2003a,b). The activated NOD2 receptor subsequently activates an NF-κB construct in the mammalian cell that results in the production of a SEAP reporter (Packiam et al., 2015; Liechti et al., 2016). SEAP activity causes a colorimetric change to the detection medium such that the enzymatic activity of SEAP is proportional to the amount of MDP in the original sample. An antibiotic that targets the cell wall (penicillin) was used as a positive control in this study. Tetracycline was used as a negative control. The parent strain exhibited a significant increase in peptidoglycan release, as compared to the untreated parent strain control, when exposed to penicillin (0.5 μg/ml), CXCL10 (1.4 μM), or CTTC (2.8 μM) (Figure 5A). In contrast, no increase in the release of peptidoglycan was observed after incubation with tetracycline (10 μg/ml) (data not shown). With the B. anthracis ΔftsX strain, a statistically significant increase in peptidoglycan release was observed in the presence of penicillin or CXCL10 (Figure 5B). However, no increase in peptidoglycan release by ΔftsX was observed in the presence of CTTC, suggesting that CTTC did not disrupt peptidoglycan processing. Corresponding viable bacterial cell counts (colony forming units [CFU] per ml) were also obtained for both the parent and ΔftsX strains to quantitate the degree of bacterial killing under these experimental conditions (Figures 5C,D). Upon exposure to CXCL10, only 0.02% of the parent strain cells remained viable after 1 h while 70% of the cells in the presence of CTTC remained viable compared to the initial inoculum concentration of bacteria. Comparatively, CXCL10- or CTTC-treated samples of ΔftsX were able to grow such that the CFU/ml results matched or exceeded those of the initial bacterial inoculum for the strain.