Imbalance of peptidoglycan biosynthesis alters the cell surface charge of Listeria monocytogenes

Graphical abstract

Abstract

The bacterial cell wall is composed of a thick layer of peptidoglycan and cell wall polymers, which are either embedded in the membrane or linked to the peptidoglycan backbone and referred to as lipoteichoic acid (LTA) and wall teichoic acid (WTA), respectively. Modifications of the peptidoglycan or WTA backbone can alter the susceptibility of the bacterial cell towards cationic antimicrobials and lysozyme. The human pathogen Listeria monocytogenes is intrinsically resistant towards lysozyme, mainly due to deacetylation and O-acetylation of the peptidoglycan backbone via PgdA and OatA. Recent studies identified additional factors, which contribute to the lysozyme resistance of this pathogen. One of these is the predicted ABC transporter, EslABC. An eslB mutant is hyper-sensitive towards lysozyme, likely due to the production of thinner and less O-acetylated peptidoglycan. Using a suppressor screen, we show here that suppression of eslB phenotypes could be achieved by enhancing peptidoglycan biosynthesis, reducing peptidoglycan hydrolysis or alterations in WTA biosynthesis and modification. The lack of EslB also leads to a higher negative surface charge, which likely stimulates the activity of peptidoglycan hydrolases and lysozyme. Based on our results, we hypothesize that the portion of cell surface exposed WTA is increased in the eslB mutant due to the thinner peptidoglycan layer and that latter one could be caused by an impairment in UDP-N-acetylglucosamine (UDP-GlcNAc) production or distribution.

Introduction

In recent years, antibiotic resistance became a serious threat for public health. One of the main targets of currently available antibiotics is the bacterial cell wall (Fig. 1). In Gram-positive bacteria, the cell wall consists of a thick layer of peptidoglycan and cell wall polymers, which are either tethered to the cell membrane or covalently linked to peptidoglycan and referred to as lipoteichoic (LTA) and wall teichoic acid (WTA), respectively. Peptidoglycan biosynthesis starts in the cytoplasm by the conversion of UDP-N-acetylglucosamine (UDP-GlcNAc) to UDP-N-acetylmuramic acid (UDP-MurNAc) via MurA and MurB. The activity of the UDP-GlcNAc 1-carboxyvinyltransferase MurA can be specifically inhibited by the antibiotic fosfomycin (Fig. 1) (Kahan et al., 1974). In the subsequent steps, which are depicted in Fig. 1, the peptidoglycan precursor lipid II is synthesized in the cytoplasm and transported across the membrane by the flippase MurJ (Meeske et al., 2015, Ruiz, 2008, Sham et al., 2014). Lipid II is then incorporated into the growing glycan strand by members of the SEDS (shape, elongation, division, sporulation) protein family, namely RodA and FtsW, which act in concert with class B penicillin binding proteins (PBPs) that have a transpeptidase activity (Cho et al., 2016, Leclercq et al., 2017, Taguchi et al., 2019). Recent findings led to the conclusion that class A PBPs, which possess a glycosyltransferase and a transpeptidase domain, are mainly involved in filling gaps and/or repair defects in the peptidoglycan mesh, rather than being the main enzymes involved in peptidoglycan polymerization and crosslinking (Cho et al., 2016, Dion et al., 2019, Vigouroux et al., 2020). The ability of bacterial cells to elongate and divide depends on the activity of peptidoglycan hydrolases. In the Gram-positive model organism Bacillus subtilis, which is closely related to L. monocytogenes, two dl-endopeptidases, CwlO and LytE, are essential for cell elongation. In addition to its role in cell elongation, LytE is also involved in cell separation (Carballido-López et al., 2006, Ohnishi et al., 1999, Vollmer et al., 2008). Both enzymes cleave the peptide bond between d-glutamic acid and meso-diamino pimelic acid of the peptidoglycan peptide, thereby allowing the insertion of new peptidoglycan material (Yamaguchi et al., 2004). Lack of CwlO results in cell shortening, while absence of CwlO and LytE is lethal (Domínguez-Cuevas et al., 2013, Hashimoto et al., 2012). Peptidoglycan biosynthesis and hydrolysis need to be tightly regulated to prevent cell lysis. In B. subtilis, this regulation is partly achieved by controlling expression of cwlO and lytE via the essential two-component system WalRK, whose expression is for instance induced during heat stress (Bisicchia et al., 2007, Dubrac et al., 2008, Takada et al., 2018). On the other hand, activity of CwlO depends on a direct protein–protein interaction with the ABC transporter FtsEX (Meisner et al., 2013). The FtsEX-dependent regulation of B. subtilis CwlO further depends on the presence of two cofactors, SweC and SweD (Brunet et al., 2019, Rismondo and Schulz, 2021).

Fig. 1.

Schematic of the UDP-GlcNAc, peptidoglycan and wall teichoic acid biosynthesis pathway. UDP-GlcNAc, which is required for peptidoglycan and wall teichoic acid (WTA) biosynthesis, is synthesized from Fructose-6-phosphate in a four-step reaction catalyzed by GlmS, GlmM, GlmU and GlmR. MurA, an UDP-GlcNAc-1-carboxyvinyltransferase catalyzing the first step of peptidoglycan biosynthesis, and MurB convert UDP-GlcNAc to UDP-MurNAc. Lipid II is produced in subsequent steps, which are performed by MurCDEF, Alr, Ddl, MraY and MurG, flipped across the membrane by MurJ and inserted into the growing glycan strand by glycosyltransferases (GTases). Finally, peptidoglycan is crosslinked by the action of transpeptidases (TPases) (Pazos and Peters, 2019). UDP-GlcNAc also serves as a substrate of TarO, the first enzyme of the WTA biosynthesis pathway. After the WTA polymer is synthesized by a subset of enzymes, it is transported across the membrane via TarGH (Brown et al., 2013) and, in case of L. monocytogenes 10403S, decorated with GlcNAc residues (Shen et al., 2017). UDP-GlcNAc-consuming enzymes are labelled in red. Antibiotics targeting different steps of the UDP-GlcNAc, peptidoglycan and WTA biosynthesis pathways or degrade peptidoglycan are depicted in bold (Campbell et al., 2011, Pensinger et al., 2021, Sarkar et al., 2017). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

UDP-GlcNAc is synthesized from fructose-6-phosphate via a four-step reaction catalyzed by GlmS, GlmM, GlmU and GlmR and serves as a substrate of MurA. Recent findings suggest that the UDP-GlcNAc biosynthetic pathway can be inhibited by trans-cinnamaldehyde (t-Cin) (Fig. 1)(Pensinger et al., 2021, Sun et al., 2021). UDP-GlcNAc is also consumed by TarO/TagO, which catalyzes the first committed step for the synthesis of WTA (Soldo et al., 2002). The activity of TarO/TagO can be blocked by tunicamycin, which also affects the activity of MraY at high concentrations (Campbell et al., 2011, Hakulinen et al., 2017, Price and Tsvetanova, 2007, Watkinson et al., 1971). In L. monocytogenes 10403S, WTA is composed of a glucose-glucose(Glc-Glc)-glycerol phosphate-GlcNAc-N-acetylmannosamine (ManNAc) linker unit and an anionic ribitol phosphate backbone, which can be modified with rhamnose, GlcNAc and d-alanine residues (Shen et al., 2017). The modification of WTA with positively charged d-alanine residues helps to mask the negative charge of the ribitol phosphate backbone, thereby conferring resistance towards cationic antimicrobial peptides and lysozyme (Brown et al., 2013, Vadyvaloo et al., 2004).

Lysozyme is an enzyme that cleaves the β-1,4-glycosidic bond between MurNAc and GlcNAc of the bacterial peptidoglycan backbone and is found in human body fluids such as tears, saliva and mucus. L. monocytogenes is intrinsically resistant towards lysozyme, which is mainly achieved by modifications of the peptidoglycan. PgdA, an N-deacetylase, and OatA, an O-acetyltransferase, deacetylate and acetylate the GlcNAc and MurNAc residues, respectively (Aubry et al., 2011, Boneca et al., 2007). In addition to peptidoglycan modifying enzymes, lysozyme resistance of L. monocytogenes is affected by the activity of the predicted carboxypeptidase PbpX, the non-coding RNA Rli31 and the transcription factor DegU (Burke et al., 2014). Interestingly, it was also been observed that the lack of components of the putative ABC transporter EslABC leads to a strong reduction in lysozyme resistance (Burke et al., 2014, Durack et al., 2015, Rismondo et al., 2021b). Recently, we could show that the absence of EslB, one of the transmembrane proteins of the EslABC transporter, resulted in the production of a thinner peptidoglycan layer and a reduction in O-acetylation of the peptidoglycan, which likely contributes to the reduced lysozyme resistance. Additionally, we observed that the eslB mutant is unable to grow in media containing high sugar concentrations and that the strain has a cell division defect (Rismondo et al., 2021b, Rismondo and Schulz, 2021).

In the current study, we used a suppressor screen to gain further insights into the role of EslABC on the physiology of L. monocytogenes. This screen revealed that phenotypes of the eslB mutant can either be suppressed by enhancing peptidoglycan biosynthesis, reducing peptidoglycan hydrolysis or by altering WTA production or modification. Using a cytochrome C assay, we further demonstrate that the lack of EslB manifests in a higher negative surface charge, which likely affects the activity of peptidoglycan hydrolases and provides an additional explanation for the increased lysozyme sensitivity of the eslB mutant.

Materials and methods

Bacterial strains and growth conditions

All strains and plasmids used in this study are listed in Table S1. Escherichia coli strains were grown in Luria-Bertani (LB) medium and Listeria monocytogenes strains in brain heart infusion (BHI) medium at 37 °C unless otherwise stated. Where required, antibiotics and supplements were added to the medium at the following concentrations: for E. coli cultures, ampicillin (Amp) at 100 µg ml⁻¹, kanamycin (Kan) at 30 µg ml⁻¹, and for L. monocytogenes cultures, chloramphenicol (Cam) at 10 µg ml⁻¹, erythromycin (Erm) at 5 µg ml⁻¹, Kan at 30 µg ml⁻¹, nalidixic acid (Nal) at 30 µg ml⁻¹, streptomycin (Strep) at 200 µg ml⁻¹ and IPTG at 1 mM.

Strain and plasmid construction

All primers used in this study are listed in Table S2. For the construction of pIMK3-murA, pIMK3-glmR, pIMK3-glmU and pIMK3-glmM, the murA, glmR, glmU and glmM genes were amplified using primer pairs JR90/JR91, JR134/JR135, JR136/JR137 and JR138/JR139, respectively, cut with NcoI and SalI and ligated into plasmid pIMK3 that had been cut with the same enzymes. The resulting plasmids pIMK3-murA, pIMK3-glmR, pIMK3-glmU and pIMK3-glmM were recovered in E. coli XL1-Blue yielding strains XL1 Blue pIMK3-murA (EJR52), XL1 Blue pIMK3-glmR (EJR106), XL1 Blue pIMK3-glmU (EJR107) and XL1 Blue pIMK3-glmM (EJR108). Next, plasmids pIMK3-murA, pIMK3-glmR, pIMK3-glmU and pIMK3-glmM were transformed into E. coli S17-1 yielding strains S17-1 pIMK3-murA (EJR59), S17-1 pIMK3-glmR (EJR132), S17-1 pIMK3-glmU (EJR133) and S17-1 pIMK3-glmM (EJR134). E. coli strain S17-1 pIMK3-murA (EJR59) was used as a donor strain to transfer plasmid pIMK3-murA by conjugation into L. monocytogenes strains 10403S (ANG1263) and 10403SΔeslB₍₂₎ (ANG5662) using a previously described method (Lauer et al., 2002). This resulted in the construction of strains 10403S pIMK3-murA (LJR26) and 10403SΔeslB₍₂₎ pIMK3-murA (LJR27), in which the expression of murA is under the control of an IPTG-inducible promoter. Strains carrying the empty pIMK3 vector were used as controls. For this purpose, pIMK3 was transformed into E. coli S17-1 yielding strain EJR58. S17-1 pIMK3 was subsequently used as a donor strain to transfer plasmid pIMK3 by conjugation into L. monocytogenes strains 10403S (ANG1263) and 10403SΔeslB₍₂₎ (ANG5662), which resulted in the construction of strains 10403S pIMK3 (LJR24) and 10403SΔeslB₍₂₎ pIMK3 (LJR25). For the construction of pIMK3-glmS, the glmS gene was amplified using primer pair JR140/JR141. The resulting PCR product was cut with BamHI and XmaI and ligated into plasmid pIMK3 that had been cut with the same enzymes. Plasmid pIMK3-glmS was recovered in E. coli XL1-Blue and subsequently transformed into E. coli S17-1 yielding strains XL1 Blue pIMK3-glmS (EJR112) and S17-1 pIMK3-glmS (EJR135), respectively. E. coli strains S17-1 pIMK3-glmR (EJR132), S17-1 pIMK3-glmU (EJR133), S17-1 pIMK3-glmM (EJR134) and S17-1 pIMK3-glmS (EJR135) were used to transfer plasmids pIMK3-glmR, pIMK3-glmU, pIMK3-glmM and pIMK3-glmS by conjugation into L. monocytogenes strain 10403SΔeslB₍₂₎ (ANG5662) yielding strains 10403SΔeslB₍₂₎ pIMK3-glmR (LJR63), 10403SΔeslB₍₂₎ pIMK3-glmU (LJR64), 10403SΔeslB₍₂₎ pIMK3-glmM (LJR65) and 10403SΔeslB₍₂₎ pIMK3-glmS (LJR66).

For the construction of a markerless deletion of cwlO (lmrg_01743), 1-kb DNA fragments up- and downstream of cwlO were amplified by PCR with primers LMS106/107 and LMS104/105. The resulting PCR products were fused in a second PCR using primers LMS105/106, the product cut with BamHI and KpnI and ligated into pKSV7 that had been cut with the same enzymes. The resulting plasmid pKSV7-ΔcwlO was recovered in E. coli XL1-Blue yielding strain XL1 Blue pKSV7-ΔcwlO (EJR63). Plasmid pKSV7-ΔcwlO was subsequently transformed into L. monocytogenes 10403S and cwlO deleted by allelic exchange according to a previously published method (Camilli et al., 1993) yielding strain 10403SΔcwlO (LJR37). Since attempts of producing electrocompetent cwlO cells were unsuccessful, plasmid pIMK3-cwlO was first conjugated into strain 10403SΔcwlO (LJR37), which allows for IPTG-inducible expression of cwlO, and resulting in the construction of strain 10403SΔcwlO pIMK3-cwlO (LJR103). For the construction of pIMK3-cwlO, cwlO was amplified using the primer pair LMS226 and LMS227. The fragment was cut with NcoI and BamHI, ligated into pIMK3 that had been cut with the same enzymes and recovered in E. coli XL1-blue and S17-1, yielding strains EJR114 and EJR115, respectively. To generate a cwlO eslB double mutant, plasmid pKSV7-ΔeslB was first transformed into L. monocytogenes strain 10403SΔcwlO pIMK3-cwlO (LJR103) resulting in strain 10403SΔcwlO pIMK3-cwlO pKSV7-ΔeslB (LJR114). The eslB gene was then deleted by allelic exchange. This resulted in the construction of strain 10403SΔcwlO ΔeslB pIMK3-cwlO (LJR119).

For the construction of promoter-lacZ fusions, the plasmid pAC7 was used. The promoter region of the dlt operon was amplified from genomic DNA of the L. monocytogenes wildtype 10403S and the eslB suppressor strain 10403S ΔeslB₍₃₎ P_dlt* (ANG5746) containing a mutation 31 bp upstream of the ATG start codon, using the primer pair LMS93 and LMS94. The PCR products were digested with BamHI and EcoRI and ligated into pAC7. The resulting plasmids pAC7-P_dlt and pAC7-P_dlt* were recovered in XL1 Blue resulting in strains XL1 Blue pAC7-P_dlt (EJR50) and XL1 Blue pAC7-P_dlt* (EJR51). Subsequently, both plasmids were integrated into the amyE site of the B. subtilis wildtype strain 168, resulting in B. subtilis strains 168 pAC7-P_dlt (BLMS4) and 168 pAC7-P_dlt* (BLMS5), respectively.

Generation of eslB suppressors and whole genome sequencing

For the generation of eslB suppressors, overnight cultures of three independently generated L. monocytogenes eslB mutants, strains 10403SΔeslB₍₁₎ (ANG4275), 10403SΔeslB₍₂₎ (ANG5662) and 10403SΔeslB₍₃₎ (ANG5685), were adjusted to an OD₆₀₀ of 1. 100 µl of 10^-1 and 10^-2 dilutions of these cultures were plated on BHI plates containing (A) 0.5 M sucrose and 0.025 µg ml⁻¹ penicillin, (B) 0.5 M sucrose and 0.05 µg ml⁻¹ penicillin or (C) 100 µg ml⁻¹ lysozyme, conditions under which the eslB mutant is unable to grow while the L. monocytogenes wildtype strain 10403S can grow. The plates were incubated at 37 °C overnight and single colonies were re-streaked on BHI plates. This procedure was repeated at least three independent times per condition. The genome sequence of a selection of eslB suppressors was determined by whole genome sequencing (WGS) using an Illumina MiSeq machine and a 150 paired end Illumina kit as described previously (Rismondo et al., 2021b). The reads were trimmed, mapped to the L. monocytogenes 10403S reference genome (NC_017544) and single nucleotide polymorphisms (SNPs) with a frequency of at least 90 % identified using CLC workbench genomics (Qiagen) and Geneious Prime® v.2021.0.1. The whole genome sequencing data were deposited at the European Nucleotide Archive (ENA) under accession number PRJEB55822.

Determination of resistance towards antimicrobials

For the disk diffusion assays, overnight cultures of the indicated L. monocytogenes strains were adjusted to an OD₆₀₀ of 0.1. 100 µl of cultures were spread on BHI agar plates using a cotton swap. Plates contained 1 mM IPTG were indicated. 6 mm filter disks were placed on top of the agar surface, soaked with 20 µl of the appropriate antibiotic stock solution and the plates were incubated at 37 °C. The diameter of the inhibition zone was measured the next day. The following stock solutions were used: 50 mg ml⁻¹ fosfomycin, 1 mg ml⁻¹ t-Cin, 5 mg ml⁻¹ tunicamycin, 15 mg ml⁻¹ d-cycloserine, 60 mg ml⁻¹ nisin, 30 mg ml⁻¹ vancomycin, 5 mg ml⁻¹ moenomycin, 1 mg ml⁻¹ penicillin, 10 mg ml⁻¹ ampicillin and 250 mg ml⁻¹ bacitracin.

Spot plating assays

Overnight cultures of L. monocytogenes strains were adjusted to an OD₆₀₀ of 1 and serially diluted to 10^-6. 5 µl of each dilution were spotted on BHI agar plates, BHI agar plates containing 0.5 M sucrose and 0.025 µg ml⁻¹ penicillin; 100 µg ml⁻¹ lysozyme; 0.025 µg/ml moenomycin, 500 µl DMSO (≥99.5 %), 0.05 µg ml⁻¹ or 0.5 µg ml⁻¹ tunicamycin and plates incubated at 37 °C unless otherwise stated. Where indicated, plates were supplemented with 1 mM IPTG or 20 mM MgCl₂. Images of plates were taken after 20–24 h of incubation.

Cell length analysis

Overnight cultures of the indicated L. monocytogenes strains were inoculated to an OD₆₀₀ of 0.1 and grown to an OD₆₀₀ of 0.3–0.6 at 37 °C with agitation in BHI medium or BHI medium containing 1 mM IPTG and the appropriate antibiotic. To stain the bacterial membranes, 800 µl of the bacterial cultures were mixed with 40 µl of 100 µg ml⁻¹ nile red and incubated for 20 min at 37 °C. The cells were washed twice in PBS buffer and subsequently re-suspended in the same buffer. Cells were then fixed in 1.12 % paraformaldehyde for 20 min at room temperature in the dark and 1–1.5 µl of the cell suspension was spotted on microscope slides covered with a thin agarose layer (1.5 % in ddH₂O). Phase contrast and fluorescence images were taken using a Zeiss Axioskop 40 microscope equipped with an EC Plan-NEOFLUAR 100X/1.3 objective (Carl Zeiss, Göttingen, Germany) and coupled to an AxioCam MRm camera. Filter set 43 was used for the detection of the nile red signal. The images were processed using the Axio Vision software (release 4.7). The length of 50 cells per replicate was measured for each strain and the mean calculated. Statistical analysis was performed using the software GraphPad Prism (version 8).

Isolation of cellular proteins and western blotting

Bacteria from a 20 ml culture were harvested by centrifugation and washed with ZAP buffer (10 mM Tris-HCl, pH 7.5, 200 mM NaCl). The cells were subsequently resuspended in 1 ml ZAP buffer containing 1 mM PMSF and disrupted by sonication. Cellular debris was removed by centrifugation, the resulting supernatant collected and separated by SDS polyacrylamide gel electrophoresis (PAGE). Proteins were transferred onto positively charged polyvinylidene fluoride (PVDF) membranes using a semi-dry transfer unit. MurA and DivIVA were detected using a polyclonal rabbit antiserum raised against the B. subtilis MurAA protein (Kock et al., 2004), which also cross reacts with the L. monocytogenes MurA protein, and DivIVA (Marston et al., 1998) as primary antibody and an anti-rabbit immunoglobulin G conjugated to horseradish peroxidase antibody as the secondary antibody. Blots were developed using ECL chemiluminescence reagents (Thermo Scientific) and imaged using a chemiluminescence imager (Vilber Lourmat).

Cytochrome C binding assay

Cytochrome C binding assays were performed as previously described (Kang et al., 2015). Briefly, overnight cultures of L. monocytogenes were used to inoculated fresh cultures to an OD₆₀₀ of 0.05 in 4 ml BHI medium and the cultures were grown to an OD₆₀₀ of 0.6–0.8 at 37 °C and 200 rpm. Bacteria from 2 ml of the culture were harvested by centrifugation at 16.200 × g for 1 min. Cells were then washed twice with 20 mM MOPS (3-(N morpholino) propanesulfonic acid) buffer (pH 7) and adjusted to an OD₆₀₀ of 0.25 in the same buffer. Cytochrome C was added at a final concentration of 50 µg ml⁻¹ and the suspension was incubated in the dark for 10 min at room temperature. The suspension was centrifuged for 5 min at 16.200 × g, the supernatant removed and the absorbance of the supernatant was measured at 410 nm (OD₄₁₀ + cells). A reaction without cells was used as blank (OD₄₁₀ – cells). The percentage of bound cytochrome C was calculated as follows:

$$\%cytochromeCbound=100-[\left(O{D}_{410}+cells\right)/\left(O{D}_{410}-cells\right)$$

β-galactosidase assay

To compare the dlt promoter activity of the L. monocytogenes wildtype and suppressor strain 10403SΔeslB₍₃₎ P_dlt* (ANG5746), β-galactosidase assays were performed. For this purpose, promoter-lacZ fusions were integrated into the amyE locus of B. subtilis 168. The resulting B. subtilis strains were grown in CSE-glucose minimal medium at 37 °C to an OD₆₀₀ of 0.5–0.8, bacteria from a culture aliquot harvested and the β-galactosidase activity determined as described previously (Miller, 1972).

Results

ΔeslB phenotypes are restored in the presence of excess Mg²⁺

The eslABCR operon encodes the putative ABC transporter EslABC and the RpiR transcriptional regulator EslR. Absence of the transmembrane component EslB leads to a multitude of phenotypes including a cell division defect, decreased lysozyme resistance, the production of a thinner cell wall and a growth defect in media containing high sugar concentrations (Rismondo et al., 2021b). In addition, we observed that the growth of the eslB mutant is severely affected when grown on BHI plates at 42 °C (Fig. 2A). So far, the cellular function of the transporter EslABC and how it is mechanistically linked to cell division and cell wall biosynthesis in L. monocytogenes is unknown. Generally, many cell wall defects can be rescued by the addition of Mg²⁺, however, the reason for this is still debated. Recently, it was speculated that binding of Mg²⁺ to the cell wall inhibits peptidoglycan hydrolases and thereby stabilizes the bacterial cell wall (Tesson et al., 2022). As the phenotypic defects of the eslB mutant suggest that peptidoglycan biosynthesis might be impaired, we speculated that the addition of Mg²⁺ should rescue the growth of this strain. To test this hypothesis, the L. monocytogenes wildtype 10403S, the ΔeslB mutant and the ΔeslB complementation strain were inoculated on BHI plates containing sucrose and penicillin, lysozyme or were incubated at 42 °C in the absence or presence of 20 mM MgCl_2. The addition of Mg²⁺ could restore the growth of the ΔeslB mutant under all conditions tested (Fig. S1), further supporting the hypothesis that peptidoglycan biosynthesis is impaired in L. monocytogenes cells lacking EslB.

Fig. 2.

Mutation P159L in prpC suppresses eslB phenotypes. (A) Drop dilution assay. Dilutions of L. monocytogenes strains 10403S (wt), ΔeslB₍₂₎, ΔeslB₍₂₎ compl., and ΔeslB₍₂₎ prpC^P159L were spotted on BHI plates, BHI plates containing 100 µg/ml lysozyme or containing 0.5 M sucrose and 0.025 µg/ml penicillin G and were incubated at 37 °C or on BHI plates and were incubated at 42 °C. (B) Western blot. Protein samples of strains 10403S, ΔeslB₍₂₎, ΔeslB₍₂₎ compl. and ΔeslB₍₂₎ prpC^P159L were separated by SDS PAGE and transferred to a PVDF membrane. MurA was detected using a MurAA-specific antibody as described in the methods section. Protein samples of L. monocytogenes wildtype strain EGD-e and a clpC mutant were used as controls. A western blot showing DivIVA levels serves as a loading control. MurA level in the protein samples were quantified and plotted. For statistical analysis, an unpaired two-tailed t-test was used (* p ≤ 0.05).

Resistance profile of the eslB mutant towards cell wall-targeting antibiotics

To narrow down which point of the peptidoglycan biosynthesis pathway is impaired in the eslB mutant, we performed disk diffusion assays with antibiotics that target different steps of this pathway (Fig. 1). Compared to the wildtype, the eslB mutant was more sensitive towards t-Cin and fosfomycin (Fig. 3), which target the UDP-GlcNAc biosynthetic pathway and MurA, respectively (Marquardt et al., 1994, Pensinger et al., 2021, Sun et al., 2021). In contrast, the resistance towards d-cycloserine, nisin, vancomycin, moenomycin, penicillin, ampicillin and bacitracin (Fig. 1), which target processes downstream of MurA, was not altered in the eslB mutant (Fig. S2). The eslB mutant was more resistant towards tunicamycin, whose main target at low concentrations is TarO, the first enzyme functioning in the WTA biosynthesis pathway (Fig. 3). As we have seen above, the inhibition of WTA biosynthesis seems to be beneficial for the eslB mutant likely due to an increased flux of UDP-GlcNAc towards peptidoglycan biosynthesis. This antibiotic screen therefore suggests that one of the limiting factors of the eslB mutant might be the synthesis or correct distribution of UDP-GlcNAc, a precursor, which is used for both peptidoglycan biosynthesis and the synthesis and modification of WTA. This hypothesis is supported by the observation that overproduction of GlmM and GlmR, two proteins involved in the production of UDP-GlcNAc, can partially suppress the eslB phenotypes (Fig. S3).

Fig. 3.

Alterations in the resistance of the eslB mutant towards cell wall-targeting antibiotics. Disk diffusion assay. L. monocytogenes strains 10403S (wt), ΔeslB₍₂₎ and ΔeslB₍₂₎ compl. were spotted on BHI plates. Antibiotic-soaked disks placed on top the agar surface and the plates incubated for 24 h at 37 °C. The inhibition zones for the indicated strains were measured and the average values and standard deviation of at least three independent experiments were plotted. For statistical analysis, a one-way ANOVA coupled with a Dunnett‘s multiple comparison test was used (** p ≤ 0.01, *** p ≤ 0.001, **** p ≤ 0.0001).

Isolation of eslB suppressor mutants

To gain insights into the function of EslABC, we took advantage of the observation that the eslB mutant forms suppressors when grown on BHI plates that contain either 100 µg ml⁻¹ lysozyme, or 0.5 M sucrose and 0.025 or 0.05 µg ml⁻¹ penicillin. Genomic alterations present in independently isolated eslB suppressors were determined by whole genome sequencing. A large subset of eslB suppressors had mutations in walR (lmo0287) or walK (lmo0288), which encode the WalRK two-component system that is involved in cell wall metabolism (Dubrac et al., 2008, Howell et al., 2003). In addition, we identified mutations that mapped to genes associated with peptidoglycan biosynthesis (murZ (lmo2552), reoM (lmo1503), prpC (lmo1821), pbpA1 (lmo1892)), peptidoglycan hydrolysis (cwlO (lmo2505, spl), ftsX (lmo2506), ftsE (lmo2507)) and wall teichoic acid (WTA) biosynthesis and modification (tarL (lmo1077), dltX (promoter region, lmrg_02074)) (Table 1).

Table 1.

Identified sequence alterations in L. monocytogenes eslB deletion strains and suppressors.

Strain number	Reference position1	Type2	Ref3	Allele4	Frequency5	Annotations6	AA change7	Condition8
ANG4275 10403SΔeslB(1)	2425786–2425875	DEL			100 %	lmo2396, internalin	30 aa deletion
ANG5386	2583855	DEL	T	–	100 %	murZ, UDP-N-acetylglucosamine 1-carboxyvinyltransferase	M240fs	P
ANG5479	310043	SNV	G	A	100 %	walK, histidine kinase of TCS WalRK	R553H	P
ANG5480	309234	SNV	C	G	100 %	walK, histidine kinase of TCS WalRK	N283E	P
ANG5488	2537835	SNV	C	G	100 %	ftsX, membrane component of ABC transporter FtsEX	G243R	P
ANG5489	310094	SNV	C	T	100 %	walK, histidine kinase of TCS WalRK	A570V	P
ANG5499	1491251	DEL	T	–	100 %	reoM, regulator of MurA degradation	L23fs	P
ANG5662 10403SΔeslB(2)	–
ANG5698	2582654	DEL	G	–	100 %	murZ, UDP-N-acetylglucosamine 1-carboxyvinyltransferase	Q307fs	P
ANG5699	310,133	SNV	T	C	97.4 %	walK, histidine kinase of TCS WalRK	I583T	P
ANG5708	1852252	SNV	G	A	100 %	prpC, serine threonine phosphatase	P159L	P
ANG5710	2537273	SNV	C	T	98 %	cwlO, peptidoglycan dl-endopeptidase	R106H	P
ANG5714	2538568	SNV	G	A	98.4 %	ftsE, ATP binding protein of ABC transporter FtsEX	Q220-	P
ANG5733	308133	SNV	A	G	100 %	walR, response regulator of TCS WalRK	S216G	L
ANG5734	309772	SNV	C	G	100 %	walK, histidine kinase of TCS WalRK	H463D	L
ANG5737	309085	SNV	G	T	100 %	walK, histidine kinase of TCS WalRK	V234L	L
ANG5741	309824	SNV	G	A	100 %	walK, histidine kinase of TCS WalRK	R480H	L
ANG5685 10403SΔeslB(3)	–
ANG5717	1921373	SNV	G	A	100 %	pbpA1, bifunctional penicillin binding protein	G125D	P
ANG5729	1090622	SNV	C	T	100 %	tarL, teichoic acid ribitol-phosphate polymerase	P282L	P
ANG5746	988580	SNV	T	C	100 %	Promoter of dlt operon, involved in d-alanylation of wall teichoic and lipoteichoic acids	–	L

¹Reference position is based on the position in the L. monocytogenes 10403S reference genome (NC_01744).

²Type of mutation: SNV = single nucleotide variant; DEL = nucleotide deletion.

³Ref indicates base in reference genome.

⁴Allele indicates base at the same position in the sequenced strain.

⁵Frequency at which the base change was found in the sequenced strain.

⁶TCS = Two-component system.

⁷AA change indicates the resulting amino acid change in the protein found in the reference strains as compared to the sequenced strain.

⁸Condition, which was used to raise suppressors. L = BHI plates containing 100 µg/ml lysozyme. P = BHI plates containing 0.5 M sucrose and 0.025 or 0.05 µg/ml penicillin.

Suppression of eslB phenotypes by increased MurA levels

Under sucrose penicillin stress, two eslB suppressors with mutations in murZ and one eslB suppressor with a mutation in reoM or prpC were isolated. The proteins encoded by all three of these genes affect MurA protein levels (Wamp et al., 2022, 2020). Drop dilution assays were performed to assess whether mutations in murZ, reoM and prpC suppress the growth defect of the eslB mutant in the presence of sucrose and penicillin as well as other conditions under which the growth of the eslB mutant is severely affected. As expected, the growth of the L. monocytogenes strains 10403SΔeslB₍₁₎ murZ^M240fs, 10403SΔeslB₍₂₎ murZ^Q307fs, 10403SΔeslB₍₁₎ reoM^K23fs and 10403SΔeslB₍₂₎ prpC^P159L on BHI plates containing sucrose and penicillin is comparable to the wildtype strain 10403S. The suppressors with mutations in murZ, reoM and prpC also grew like wild-type on BHI plates containing lysozyme as well as on BHI plates that were incubated at 42 °C (Fig. 2A for 10403SΔeslB₍₂₎ prpC^P159L, data not shown for 10403SΔeslB₍₁₎ murZ^M240fs, 10403SΔeslB₍₂₎ murZ^Q307fs, 10403SΔeslB₍₁₎ reoM^K23fs).

MurZ is a homolog of the UDP-N-acetylglucosamine 1-carboxyvinyltransferase MurA, which is required for the first step of peptidoglycan biosynthesis (Fig. 1)(Du et al., 2000, Kock et al., 2004). While MurA is essential for the growth of L. monocytogenes, the deletion of murZ is possible and results in the stabilization of MurA (Rismondo et al., 2016). MurA is a substrate of the ClpCP protease and a recent study showed that ReoM is required for the ClpCP-dependent proteolytic degradation of MurA in L. monocytogenes (Rismondo et al., 2016, Wamp et al., 2020). The activity of ReoM is controlled by the serine/threonine kinase PrkA and the cognate phosphatase PrpC. The phosphorylated form of ReoM stimulates peptidoglycan biosynthesis by preventing ClpCP-dependent degradation of MurA, while MurA degradation is enhanced in the presence of non-phosphorylated ReoM (Wamp et al., 2022, 2020). The identified mutations in murZ and reoM in our suppressor strains lead to frameshifts, and thus to the production of inactive MurZ and ReoM proteins. It has been previously shown that MurA protein levels are increased in murZ and reoM mutants (Rismondo et al., 2016, Wamp et al., 2020) and we thus assume that this is also the case for the eslB suppressor strains 10403SΔeslB₍₁₎ murZ^M240fs, 10403SΔeslB₍₂₎ murZ^Q307fs and 10403SΔeslB₍₁₎ reoM^K23fs. The identified mutation in prpC in the suppressor strain 10403SΔeslB₍₂₎ prpC^P159L leads to the production of a variant of PrpC, in which proline at position 159 is replaced by a leucine, however, it is currently not known whether this amino acid exchange leads to decreased or enhanced activity of PrpC. A decreased activity of PrpC would lead to accumulation of MurA (Wamp et al., 2022), while enhanced activity would result in reduced MurA levels. To assess which affect the P159L mutation in prpC has on MurA levels, western blots were performed using a MurAA-specific antibody. Protein samples of L. monocytogenes strains EGD-e and EGD-e ΔclpC were used as controls. In accordance with previous studies, MurA accumulated in the clpC mutant (Rismondo et al., 2016). The production of the mutated PrpC variant, PrpC^P159L, also led to a slight accumulation of MurA in the eslB mutant background (Fig. 2B). Additionally, no change in MurA levels could be observed for the eslB mutant compared to the wildtype strain 10403S (Fig. 2B). These results suggest that while the decrease in peptidoglycan production seen in the eslB mutant is not caused by decreased MurA levels, the growth deficit of the eslB mutant can be rescued by preventing MurA degradation.

To confirm that the suppression of the eslB phenotypes in the strains with mutations in murZ, reoM or prpC is indeed the result of increased MurA protein levels, we integrated a second, IPTG-inducible copy of murA into the genome of the eslB mutant. First, we determined the resistance towards fosfomycin, which is a known inhibitor of MurA (Fig. 4A-B) (Kahan et al., 1974), for the L. monocytogenes wildtype and the eslB mutant harboring the empty plasmid pIMK3 or pIMK3-murA in the presence of IPTG. Strain 10403SΔeslB pIMK3 is threefold more sensitive to fosfomycin as compared to the cognate wildtype strain. The induction of murA expression increases the resistance of strains 10403S pIMK3-murA and 10403SΔeslB pIMK3-murA (Fig. 4A-B), suggesting that MurA is indeed overproduced in these strains.

Fig. 4.

MurA overexpression leads to suppression of eslB phenotypes. (A-B) Fosfomycin disk diffusion assay. (A) L. monocytogenes strains wt pIMK3, ΔeslB pIMK3, wt pIMK3-murA and ΔeslB pIMK3-murA were plated on BHI plates containing 1 mM IPTG. Fosfomycin-soaked disks were placed on the agar surface and the plates incubated for 24 h at 37 °C. Yellow lines indicate the diameter of the zone of inhibition. (B) The inhibition zones for the indicated strains were measured and the average values and standard deviation of three independent experiments were plotted. (C) Drop dilution assay. Dilutions of L. monocytogenes strains 10403S pIMK3 (wt pIMK3), ΔeslB pIMK3, wt pIMK3-murA and ΔeslB pIMK3-murA were spotted on BHI plates, BHI plates containing 100 µg/ml lysozyme or containing 0.5 M sucrose and 0.025 µg/ml penicillin G, which were incubated at 37 °C or on BHI plates, which were incubated at 42 °C. (D) Microscopy images of L. monocytogenes strains. Bacterial membranes were stained with nile red as described in the methods section. Scale bar is 2 µm. (E) Cell length of L. monocytogenes strains shown in panel B. The cell length of 50 cells per strain was determined and the median cell length calculated. The average values and standard deviations of three independent experiments are plotted. For statistical analysis, a one-way ANOVA coupled with Tukey‘s multiple comparison test was used (** p ≤ 0.01, *** p ≤ 0.001, **** p ≤ 0.0001). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Next, we assessed whether the growth phenotypes of the eslB mutant can be suppressed by the overexpression of murA. In the presence of the inducer IPTG, the growth of strain 10403SΔeslB pIMK3-murA was comparable to the corresponding wildtype strain under all conditions tested (Fig. 4C). These results indicate that overproduction of MurA can compensate for the loss of eslB and that the suppression of the eslB phenotypes in the murZ, reoM and prpC suppressors is likely the result of increased MurA levels.

Interestingly, increased MurA production also leads to the suppression of the cell division defect of the eslB mutant. Cells of the eslB mutant carrying the empty pIMK3 plasmid have a cell length of 3.19 ± 0.29 µm. In contrast, strain 10403SΔeslB pIMK3-murA produces cells with a length of 2.27 ± 0.03 µm, a size that is comparable to the length of L. monocytogenes wildtype cells (Fig. 4D-E).

Suppression by reduction of the glycosyltransferase activity of PBP A1

Multiple enzymes are involved in the production of lipid II, the peptidoglycan precursor, within the cytoplasm. After its transport across the membrane via MurJ, lipid II is incorporated into the growing glycan strand by glycosyltransferases, which can either be class A penicillin binding proteins (PBPs), monofunctional glycosyltransferases or members of the SEDS (shape, elongation, division, sporulation) family, such as RodA and FtsW (Cho et al., 2016, Emami et al., 2017, Meeske et al., 2016). L. monocytogenes encodes two class A PBPs, PBP A1 and PBP A2, in its genome (Korsak et al., 2010, Rismondo et al., 2015). One of the eslB suppressor strains, ANG5717, carries a point mutation in pbpA1 leading to the substitution of glycine at position 125 by aspartate. Interestingly, this glycine residue is part of the catalytic site of the glycosyltransferase domain, suggesting that the glycosyltransferase activity of the PBP A1 variant PBP A1^G125D might be reduced. Drop dilution assays and microscopic analysis revealed that the phenotypes associated with deletion of eslB can be rescued by pbpA1^G125D (Fig. 5A-C). The glycosyltransferase activity of class A PBPs can be inhibited by moenomycin (Huber and Nesemann, 1968). We thus tested whether the presence of moenomycin would enable the growth of the eslB mutant under heat stress. As shown in Fig. 5D, the eslB mutant is unable to grow at 42 °C on BHI plates or BHI plates containing DMSO, which was used to dissolve moenomycin. In contrast, nearly wildtype-like growth of the eslB mutant was observed on BHI plates that were supplemented with moenomycin. These results suggest that growth deficits of the eslB mutant can be rescued by the reduction of the glycosyltransferase activity of class A PBPs.

Fig. 5.

Inactivation of the glycosyltransferase of penicillin binding proteins rescues the eslB mutant. (A) Drop dilution assay. Dilutions of L. monocytogenes strains 10403S (wt), ΔeslB₍₃₎, ΔeslB₍₃₎ compl. and ΔeslB₍₃₎ pbpA1^G125D were spotted on BHI plates, BHI plates containing 100 µg/ml lysozyme or containing 0.5 M sucrose and 0.025 µg/ml penicillin G, which were incubated at 37 °C or on BHI plates, which were incubated at 42 °C. (B) Microscopy images of L. monocytogenes strains. Bacterial membranes were stained with nile red as described in the methods section. Scale bar is 2 µm. (C) Cell length of L. monocytogenes strains shown in panel B. The cell length of 50 cells per strain was determined and the median cell length calculated. The average values and standard deviations of three independent experiments are plotted. (D) Drop dilution assay. Dilutions of L. monocytogenes strains 10403S (wt), ΔeslB₍₂₎ and ΔeslB₍₂₎ compl. were spotted on BHI plates, BHI plates containing DMSO or containing 0.025 µg/ml moenomycin and incubated at 37 °C or 42 °C. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Suppression by reducing cwlO expression or CwlO activity

In addition to suppressors associated with MurA and PBP A1, we isolated suppressors carrying mutations in genes that either affect the transcription of cwlO or the activity of CwlO. CwlO is a dl-endopeptidase, which opens the existing cell wall to allow for the insertion of new lipid II precursors during cell elongation of B. subtilis (Bisicchia et al., 2007, Hashimoto et al., 2012). In B. subtilis, transcription of cwlO is induced by WalRK in response to low dl-endopeptidase activity (Bisicchia et al., 2007, Dobihal et al., 2019, Dubrac et al., 2008) and CwlO activity is stimulated by a direct interaction with the ABC transporter FtsEX (Meisner et al., 2013). In our suppressor screen, we isolated eight eslB suppressor strains that carried mutations in walK, which were either isolated under lysozyme or sucrose penicillin stress, and four eslB suppressor strains that carried mutations in walR from which the majority was isolated in the presence of lysozyme stress. We selected eslB suppressors 10403SΔeslB₍₂₎ walR^S216G and 10403SΔeslB₍₂₎ walK^H463D, which were isolated under lysozyme stress, and 10403SΔeslB₍₂₎ walK^I583T, which was isolated under sucrose penicillin stress, for further analysis. In addition, three suppressors with mutations were isolated in ftsE, ftsX and cwlO under sucrose penicillin pressure (Table 1).

Spot plating assays showed that the eslB suppressors with mutations in either walR or walK could overcome the growth defect of the eslB mutant under sucrose penicillin, lysozyme and heat stress (Fig. 6). In contrast, eslB suppressors harboring mutations in either ftsE, ftsX or cwlO, grow comparable to the wildtype under sucrose penicillin and heat stress, but showed a similar growth defect as the eslB mutant on plates containing lysozyme (Fig. 6A). Microscopic analysis of the eslB suppressors with mutations in walR, walK, ftsE, ftsX and cwlO revealed that these cells have a similar cell length to the L. monocytogenes wildtype 10403S and thus, the expression of mutated WalR, WalK, FtsE and CwlO variants could restore the cell division defect of the eslB mutant (Fig. 6B-C). Additionally, we observed that the cells of suppressor strains 10403SΔeslB₍₂₎ cwlO^R106H and 10403SΔeslB₍₂₎ ftsE^Q220- were bent compared to the wildtype strain. This phenotype is characteristic for B. subtilis cwlO and ftsE mutants (Domínguez‐Cuevas et al., 2013; Meisner et al., 2013) suggesting that the acquired mutations in cwlO and ftsE led to inactivation or reduced activity of CwlO and FtsE in L. monocytogenes.

Fig. 6.

Mutations in walR, walK, cwlO, ftsE and ftsX suppress eslB phenotypes. (A) Drop dilution assay. Dilutions of L. monocytogenes strains 10403S (wt), ΔeslB₍₂₎, ΔeslB₍₁₎, ΔeslB₍₂₎ compl., ΔeslB₍₁₎ compl., ΔeslB₍₂₎ walR^S216G, ΔeslB₍₂₎ walK^H463D, ΔeslB₍₂₎ walK^I583T, ΔeslB₍₂₎ cwlO^R106H, ΔeslB₍₂₎ ftsE^Q220- and ΔeslB₍₁₎ ftsX^G253R were spotted on BHI plates, BHI plates containing 100 µg/ml lysozyme or containing 0.5 M sucrose and 0.025 µg/ml penicillin G and were incubated at 37 °C or on BHI plates that were incubated at 42 °C. (B) Microscopy images of L. monocytogenes strains. Bacterial membranes were stained with nile red as described in the methods section. Scale bar is 2 µm. (C) Cell length of L. monocytogenes strains shown in panel B. The cell length of 50 cells per strain was determined and the median cell length calculated. The average values and standard deviations of three independent experiments are plotted. For statistical analysis, a one-way ANOVA coupled with a Dunnett‘s multiple comparison test was used (* p ≤ 0.05, **** p ≤ 0.0001). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

The observation that mutations accumulate in genes associated with CwlO led to the hypothesis that wildtype CwlO activity is toxic, due to the reduced peptidoglycan levels produced by the eslB mutant (Rismondo et al., 2021b). To confirm this hypothesis, an eslB cwlO double mutant was constructed, in which the expression of an ectopic copy of cwlO was placed under the control of an IPTG-inducible promoter. As expected, in the absence of the inducer, the eslB cwlO double mutant was able to grow on BHI plates supplemented with sucrose and penicillin. In accordance to the results obtained for the suppressors, the growth of a strain lacking both, EslB and CwlO, was still impaired in the presence of lysozyme. Furthermore, individual deletion of either eslB or cwlO resulted in a severe growth defect at higher temperatures, while the eslB cwlO double mutant was able to grow in the absence of IPTG at 42 °C. Induction of cwlO expression in strain 10403SΔcwlO ΔeslB pIMK3-cwlO resulted in a growth defect under sucrose penicillin and under heat stress, which is comparable to that of the eslB mutant (Fig. 7A). Next, we studied the cell morphology of strains lacking CwlO, EslB or both. In accordance with previous studies investigating the function of CwlO in B. subtilis (Domínguez-Cuevas et al., 2013, Meisner et al., 2013), the L. moncytogenes cwlO mutant formed slightly smaller, bent cells (Fig. 7B-C). Cells lacking both, EslB and CwlO, were also shorter than cells of the L. monocytogenes wildtype 10403S, while induction of cwlO expression in strain 10403SΔcwlO ΔeslB pIMK3-cwlO resulted in the formation of elongated cells (Fig. 7B-C). Altogether, these findings support our hypothesis that the suppression of the eslB phenotypes in the cwlO, ftsE and ftsX suppressors is the result of an inactivation of the dl-endopeptidase CwlO.

Fig. 7.

Absence of CwlO suppresses most of the eslB phenotypes. (A) Drop dilution assay. Dilutions of L. monocytogenes strains 10403S (wt), ΔeslB₍₂₎, ΔcwlO and ΔcwlO ΔeslB pIMK3-cwlO were spotted on BHI plates, BHI plates containing 100 µg/ml lysozyme or containing 0.5 M sucrose and 0.025 µg/ml penicillin G and were incubated at 37 °C or on BHI plates that were incubated at 42 °C. For induction of cwlO expression, the indicated plates were supplemented with 1 mM IPTG. (B) Microscopy images of L. monocytogenes strains. Bacterial membranes were stained with nile red as described in the methods section. Scale bar is 2 µm. (C) Cell length of L. monocytogenes strains shown in panel B. The cell length of 50 cells per strain was determined and the median cell length calculated. The average values and standard deviations of three independent experiments are plotted. For statistical analysis, a one-way ANOVA coupled with a Dunnett‘s multiple comparison test was used (* p ≤ 0.05, **** p ≤ 0.0001). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Suppression by inhibition of WTA biosynthesis

In B. subtilis, WalRK does not only stimulate transcription of cwlO, but also of the tagAB and tagDEFGH operons, which encode proteins required for WTA biosynthesis and export (Howell et al., 2003). In L. monocytogenes 10403S, WTA is composed of a ribitol phosphate backbone, which is attached to a Glc-Glc-glycerol phosphate-GlcNAc-ManNAc linker unit and modified with GlcNAc, rhamnose and d-alanine residues (Brown et al., 2013, Shen et al., 2017). Thus, UDP-GlcNAc is consumed by several enzymes during WTA biosynthesis and modification. As mentioned above, we isolated suppressors containing mutations in lmo1077, encoding a TarL homolog, and the promoter region of the dlt operon (Table 1). TarL is a teichoic acid ribitol phosphate polymerase required for the polymerization of the ribitol phosphate backbone of WTA (Fig. 1)(Brown et al., 2013). The dlt operon codes for proteins that are essential for the d-alanylation of WTA (Neuhaus and Baddiley, 2003, Perego et al., 1995, Rismondo et al., 2021a). Fig. 8A shows that the eslB mutant, which harbors a mutation in tarL, could grow again under sucrose penicillin and lysozyme stress, likely due to a reduction in WTA biosynthesis and potentially leading to an increase in the cellular UDP-GlcNAc pool, which can then be used by MurA. However, the tarL suppressor is unable to grow at 42 °C. A similar phenotype could be observed when the L. monocytogenes wildtype strain 10403S is grown at 42 °C on a plate containing high concentrations of tunicamycin, which inhibits TarO, the first enzyme required for WTA biosynthesis (data not shown). This indicates that the mutated TarL variant produced by the eslB suppressor strain 10403SΔeslB₍₃₎ tarL^P282L has a reduced activity. To further test if inhibition of WTA biosynthesis is indeed beneficial for the eslB mutant, we grew the eslB mutant at 42 °C on BHI plates containing different concentrations of tunicamycin. We could see partial suppression of the heat sensitivity at a tunicamycin concentration of 0.05 µg/ml and the eslB mutant grew comparable to the L. monocytogenes wildtype strain 10403S and the complementation strain on plates containing 0.5 µg/ml tunicamycin (Fig. S4). These results demonstrate that reduction of WTA biosynthesis leads to the suppression of the heat phenotype associated of the eslB mutant.

Fig. 8.

Suppression of eslB phenotypes by alterations in WTA biosynthesis and modification. (A) Drop dilution assay. Dilutions of L. monocytogenes strains 10403S (wt), ΔeslB₍₃₎, ΔeslB₍₃₎ compl., ΔeslB₍₃₎ tarL^P282L and ΔeslB₍₃₎ P_dlt* were spotted on BHI plates, BHI plates containing 100 µg/ml lysozyme or containing 0.5 M sucrose and 0.025 µg/ml penicillin G and were incubated at 37 °C or on BHI plates that were incubated at 42 °C. (B) β-galactosidase assay. The dlt promoter of L. monocytogenes wildtype 10403S (P_dlt) and suppressor strain ΔeslB₍₃₎ P_dlt* (P_dlt*) were fused with lacZ and integrated into the amyE locus of Bacillus subtilis. The promoter activity was determined in Miller units as described in the methods section and the average and standard deviation of three independent experiments were plotted. An unpaired t-test was used for statistical analysis (**** p ≤ 0.0001). (C) Cytochrome C assay. Cells of L. monocytogenes strains 10403S, ΔeslB₍₃₎, ΔeslB₍₃₎ compl. and ΔeslB₍₃₎ P_dlt* were incubated with cytochrome C as described in the methods section. The percentage of cytochrome C, which is bound by the cell surface, was calculated for three independent experiments and plotted. A strain lacking DltA was used as control. For statistical analysis, a one-way ANOVA coupled with a Dunnett‘s multiple comparison test was used (** p ≤ 0.01, **** p ≤ 0.0001).

Fig. 8A further shows that the eslB suppressor strain, which harbors a mutation in the dlt promoter region, could grow again under all conditions tested. To assess what consequence the point mutation in the promoter region of the dlt operon has on gene expression, we place the promoterless lacZ gene under the control of the dlt promoter or the mutated dlt promoter found in suppressor strain 10403SΔeslB₍₃₎ P_dlt*. The lacZ promotor gene fusions were introduced into B. subtilis and β-galactosidase activities determined. This analysis showed that the mutation in the dlt promoter resulted in increased promoter activity and likely increased d-alanylation of teichoic acids in the suppressor strain (Fig. 8B). The d-alanylation state of teichoic acids is an important factor that impacts the bacterial cell surface charge (Vadyvaloo et al., 2004). As we isolated a suppressor that increased the expression of the dlt operon, we wondered, whether the surface charge of eslB mutant cells is altered. To test this, we determined the binding capability of positively charged cytochrome C to the cell surface of different L. monocytogenes strains, which serves as a readout of the bacterial surface charge (Peschel et al., 1999, Wecke et al., 1997). A strain lacking d-alanine residues on LTA and WTA due to a deletion of dltA was used as a control. As expected, the cell surface of the dltA mutant had a higher negative charge as compared to the L. monocytogenes wildtype strain 10403S. A similar result was observed for the eslB mutant. The high negative surface charge of the eslB mutant could be partially rescued by the mutation in the dlt promoter, which is present in the eslB suppressor strain 10403SΔeslB P_dlt* (Fig. 8C). This result suggests that either the d-alanylation state of WTAs is altered in the absence of EslB or that the surface presentation of WTA is changed due to the production of a thinner peptidoglycan layer.

Discussion

In this study, we aimed to provide further insight into the connection between the predicted ABC transporter EslABC and peptidoglycan biosynthesis. Previous studies have shown that the deletion of eslB, coding for one of transmembrane components of the transporter, leads to a growth defect on sucrose containing media, the formation of elongated cells, the production of a thinner peptidoglycan layer, as well as sensitivity towards the natural antibiotic lysozyme and cationic antimicrobial peptides (Burke et al., 2014, Durack et al., 2015, Rismondo et al., 2021b), suggesting its involvement in peptidoglycan biosynthesis and cell division. It was proposed that the addition of Mg²⁺ rescues mutants with a defect in peptidoglycan biosynthesis by reducing the activity of peptidoglycan hydrolases (Tesson et al., 2022). In accordance with this, we observed a suppression of the eslB growth deficits under all conditions tested when an excess of Mg²⁺ was added. L. monocytogenes encodes several peptidoglycan hydrolases, including the dl-endopeptidase CwlO (Spl), the LytM-domain containing protein Lmo2504 and the two cell wall hydrolases NamA and CwhA (p60), which are required for daughter cell separation (Carroll et al., 2003, Pilgrim et al., 2003). We have isolated several eslB suppressors, which carry mutations in cwlO or mutations in other genes, e.g. ftsE or ftsX, which affect CwlO activity. After depletion of CwlO, the eslB mutant was able to grow in otherwise non-permissive conditions, suggesting that the activity of CwlO might be deregulated in the eslB deletion strain or that the peptidoglycan of the eslB mutant is more sensitive to hydrolysis by CwlO. CwlO activity is controlled by a direct protein–protein interaction with the ABC transporter FtsEX (Meisner et al., 2013). It is thus tempting to speculate that the ABC transporter EslABC might also affect CwlO activity or localization of proteins involved in peptidoglycan biosynthesis and/or degradation. Interestingly, a strain lacking both EslB and CwlO is able to grow at elevated temperatures, while both single mutants are not able to grow under this condition.

We could isolate eslB suppressor strains with mutations that are associated with stabilization of MurA, the UDP-GlcNAc 1-carboxyvinyltransferase responsible for the first step of peptidoglycan biosynthesis. Elevated levels of MurA, and thus, enhanced peptidoglycan biosynthesis could fully restore the phenotypic defects of the 10403SΔeslB strain. To identify the exact process of peptidoglycan biosynthesis, which is impaired in the eslB mutant, we performed a screen with antibiotics targeting different steps of this process. The absence of eslB only affected the resistance towards t-Cin and fosfomycin, which reduce the synthesis of UDP-GlcNAc and inhibit MurA, respectively (Marquardt et al., 1994, Pensinger et al., 2021, Sun et al., 2021). Surprisingly, we observed an increased resistance of the eslB mutant towards tunicamycin. The primary target of tunicamycin is TarO, the first enzyme of WTA biosynthesis, and, at high concentrations, also MraY, which is responsible for the production of lipid I during peptidoglycan biosynthesis (Fig. 1)(Campbell et al., 2011, Hakulinen et al., 2017, Price and Tsvetanova, 2007, Watkinson et al., 1971). During WTA biosynthesis, UDP-GlcNAc is used for the synthesis of the linker unit and for the modification of the WTA backbone (Eugster et al., 2015, Rismondo et al., 2018, Shen et al., 2017). Thus, reducing the production of WTA could increase the availability of UDP-GlcNAc for peptidoglycan biosynthesis and therefore support the growth of the eslB mutant under non-permissive conditions. In accordance with this, we observed a partial suppression of the eslB phenotypes by the overproduction of GlmM and GlmR, two enzymes required for the synthesis of UDP-GlcNAc (Pensinger et al., 2021). These results suggest that either the production or distribution of UDP-GlcNAc as a substrate between different pathways might be disturbed in the eslB mutant.

Cell wall stability depends on the balance between peptidoglycan biosynthesis and hydrolysis. An imbalance of one of these processes results in rapid cell lysis (Sassine et al., 2020). Recent studies suggest that new peptidoglycan precursors are inserted into the growing glycan chains by the Rod system, which is composed of several enzymes. In contrast, class A PBPs are thought to fill gaps and/or repair cell wall defects (Cho et al., 2016, Dion et al., 2019, Vigouroux et al., 2020). It was recently shown that enhanced endopeptidase activity leads to the activation of class A PBPs in E. coli (Lai et al., 2017). A similar mechanism seems to exist in B. subtilis, as either inactivation of PBP1 or inhibition of peptidoglycan hydrolases by the addition of Mg²⁺ suppresses growth and morphological defects of an mreB mutant (Tesson et al., 2022). In accordance with this, we also observed suppression of the heat sensitivity of the eslB mutant in presence of moenomycin, which specifically inhibits the glycosyltransferase activity of class A PBPs (Ostash and Walker, 2010, Van Heijenoort et al., 1978).

In our suppressor screen, we identified a strain carrying a mutation in the dlt promoter region, which leads to an overproduction of the Dlt enzymes. These enzymes are required for the modification of teichoic acids (TAs) with d-alanines (Neuhaus and Baddiley, 2003, Percy and Gründling, 2014, Rismondo et al., 2021a). The modification of TAs with d-alanine residues leads to a reduction in the negative surface charge and affects the activity of peptidoglycan hydrolases (Brown et al., 2013, Tesson et al., 2022, Vadyvaloo et al., 2004). Furthermore, lack of d-alanine modifications of TAs was shown to increase lysozyme sensitivity in Staphylococcus aureus (Herbert et al., 2007). Interestingly, the cell surface of the eslB mutant is more negatively charged, similar to that of a strain lacking d-alanine modifications on TAs. Based on our results, we propose the following model: The absence of EslB seems to affect the production and/or distribution of UDP-GlcNAc between different pathways, leading to the production of a thinner peptidoglycan layer. The reduced peptidoglycan thickness could subsequently result in the presentation of a larger portion of WTA on the bacterial cell surface, which would explain the higher negative surface charge (Fig. 9). This higher negative surface charge would enhance the binding capability and/or activity of cationic antimicrobial peptides, lysozyme as well as peptidoglycan hydrolases (Low et al., 2011, Neuhaus and Baddiley, 2003, Ragland et al., 2017, Steudle and Pleiss, 2011, Weidenmaier et al., 2003) and furthermore explain the sensitivity of the eslB mutant towards CwlO activity, lysozyme and cationic antimicrobial peptides (Burke et al., 2014, Rismondo et al., 2021b).

Fig. 9.

Model of altered WTA presentation on the cell surface of an eslB mutant. The Gram-positive cell wall is composed of a thick layer of peptidoglycan (PG) and wall teichoic acids (WTA), which are attached to the MurNAc-moiety of the PG backbone (Brown et al., 2013). WTA of L. monocytogenes 10403S are composed of a negatively charged ribitol phosphate backbone (indicated by -), which can be modified with GlcNAc residues (blue hexagons) and positively charged d-alanine residues (indicated by + ). The absence of EslB seems to affect the production and/or distribution of the PG precursor UDP-GlcNAc between different pathways, which results in the production of a thinner PG layer. WTAs are thought to partially stick out of the PG layer and we propose that due to the thinner PG layer produced by the eslB mutant, a larger portion of the WTA backbone, which is negatively charged, could be presented on the bacterial cell surface. This would result in a higher negative cell surface charge of the eslB mutant, which would affect the binding capability and/or activity of PG hydrolases, lysozyme and cationic antimicrobial peptides (Low et al., 2011, Ragland et al., 2017, Steudle and Pleiss, 2011, Weidenmaier et al., 2003). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

The essential two-component system WalRK stimulates the transcription of several peptidoglycan hydrolases in Gram-positive bacteria (Delaune et al., 2011, Delauné et al., 2012, Dubrac et al., 2008, Dubrac and Msadek, 2004, Howell et al., 2003). In B. subtilis, WalRK also regulates the expression of genes involved in WTA biosynthesis and export (Howell et al., 2003). The regulon of the WalRK system has not yet been determined for L. monocytogenes, however, it was shown that the system is essential (Fischer et al., 2022). Inactivation of WalRK usually leads to cell death of wildtype cells due to loss of peptidoglycan hydrolase activity, however, cell death could be prevented by the inhibition of peptidoglycan biosynthesis (Salamaga et al., 2021). We have isolated several eslB suppressors with mutations in walR and walK. As it is unlikely that all of these mutations are gain-of-function mutations, we hypothesize that these mutations result in a reduced activity of the WalRK system. This would lead to a reduced peptidoglycan hydrolase activity as well as a reduction in WTA content, and reduction of the latter would at least partially restore the bacterial cell surface charge of the eslB mutant.

Taken together, we could show that the lack of EslB results in a defect in peptidoglycan biosynthesis, which can be suppressed by modulating the activity of enzymes involved in either peptidoglycan biosynthesis or hydrolysis. Our results suggest that the production or distribution of the peptidoglycan precursor UDP-GlcNAc between different pathways might be disturbed in the eslB mutant. Further studies are required to prove this hypothesis and to determine the function of the ABC transporter EslABC.

CRediT authorship contribution statement

Lisa Maria Schulz: Conceptualization, Funding acquisition, Investigation, Visualization, Formal analysis, Writing – review & editing. Patricia Rothe: Investigation, Formal analysis. Sven Halbedel: Supervision, Writing – review & editing. Angelika Gründling: Conceptualization, Funding acquisition, Writing – review & editing. Jeanine Rismondo: Conceptualization, Funding acquisition, Investigation, Visualization, Formal analysis, Writing – original draft.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article: