No job name

Biochemistry 2003, 42, 257-264
LytG of Bacillus subtilis Is a Novel Peptidoglycan Hydrolase: The Major Active Gavin J. Horsburgh, Abdelmadjid Atrih, Michael P. Williamson, and Simon J. Foster* Department of Molecular Biology and Biotechnology, UniVersity of Sheffield, Sheffield, S10 2TN, United Kingdom ReceiVed July 29, 2002; ReVised Manuscript ReceiVed October 21, 2002 ABSTRACT: LytG (YubE) of Bacillus subtilis is a novel 32 kDa autolysin produced during vegetativegrowth under the control of EσA RNA polymerase. Muropeptide analysis of vegetative cells of B. subtilisrevealed LytG to be the major glucosaminidase responsible for peptidoglycan structural determinationduring vegetative growth. Overexpression and purification of LytG allowed its biochemical characterization.
Despite sequence homology suggesting muramidase activity, LytG is a novel glucosaminidase withexoenzyme activity and may form part of a novel family of autolysins. It is involved in cell division,lysis, and motility on swarm plates.
Peptidoglycan is essential for the maintenance of cellular no change in growth, division or sporulation (20). Although viability and shape determination for most eubacteria. It is Ortiz et al. (19) suggest the production of a similar exo- - a dynamic structure continually being synthesized, modified, N-acetylglucosaminidase at low levels in B. subtilis 168, no and hydrolyzed to allow for cell growth and division.
further evidence of its existence has been shown.
Bacteria produce a complement of autolysins capable of Renaturing SDS-PAGE analysis and the genome se- hydrolyzing the peptidoglycan structure of their own cell wall quence of B. subtilis have revealed the presence of multiple (1). These autolysins have been implicated in several putative novel autolysins, which have begun to be character- important cellular functions, including differentiation, cell ized (21-25). The complex compensatory nature of the lysis, cell wall growth and turnover, cell separation, com- autolysins means that it is important to identify and analyze petence, motility, and antibiotic-induced lysis (2-6). Bacillus the total complement of these enzymes to determine their subtilis 168 has two major autolysins expressed during the individual and combined roles. One of the putative novel vegetative phase of growth, a 50 kDa N-acetylmuramyl-L- autolysins identified by sequence homology is a 32kDa alanine amidase (LytC) and a 90 kDa N-acetylglucosamini- possible muramidase (25), encoded by the lytG (formerly dase (LytD) (7, 8). Both enzymes are dispensable for growth yubE) gene, and shows greatest homology to the autolysins but have often mutually compensatory roles in cell wall AcmA of Lactococcus lactis and AlyS of Enterococcus hirae.
turnover, motility, and cell separation (9-13). Expression The lytG gene is situated in a possible bi-cistronic operon of lytC and lytD is mainly under the control of the sigma with the downstream gene yubF, which encodes a putative factor, σD,1 which controls the flagellar, chemotaxis and protein of unknown function. In this study, we have motility regulon (12, 14, 15).
characterized LytG as a novel exoglucosaminidase with a As well as the discovery of the endo- -N-acetylglu- cosaminidase (LytD) in B. subtilis 168 an exo- -N-acetyl-glucosaminidase has been characterized from B. subtilis B(16-19). This exo- -N-acetylglucosaminidase has a molec- EXPERIMENTAL PROCEDURES
ular weight of about 75 kDa and an optimum pH of 5.9. Amutant lacking the exo- -N-acetylglucosaminidase showed Bacterial Strains, Plasmids, and Growth Conditions. All B. subtilis strains and plasmids used in this study are shown in Table 1. Unless otherwise stated B. subtilis strains were This research was funded by the BBSRC and the Royal Society.
* Corresponding author. Mailing address: Department of Molecular grown in nutrient broth (Oxoid) at 37 °C with shaking (250 Biology and Biotechnology, University of Sheffield, Firth Court, rpm) or on nutrient agar plates at 37 °C. Plasmids were Western Bank, Sheffield, S10 2TN, United Kingdom. Phone: 44 114 constructed in, and prepared from, Escherichia coli strain 222 4411. Fax: 44 114 272 8697. E-mail: s.foster@sheffield.ac.uk.
1 Abbreviations, RP-HPLC, reverse phase high-pressure liquid chro- DH5R grown in Luria-Bertani (LB) broth or on LB agar at matography; σ, sigma factor; COSY, correlated spectroscopy; TOCSY, 37 °C. When appropriate, chromosomal drug resistance total correlated spectroscopy; ROESY, rotating frame nuclear Over- markers in B. subtilis were selected with kanamycin (10 µg hauser effect spectroscopy; EDTA, ethylenediaminetetraacetic acid; WT, mL-1), erythromycin (1 µg mL-1), lincomycin (25 µg mL-1), wild type; CWBP, cell wall binding protein; MurNAc, N-acetylmuramicacid; GlcNAc, N-acetylglucosamine; A2pm, meso-diaminopimelic acid.
spectinomycin (100 µg mL-1), phleomycin (0.3 µg mL-1), 258 Biochemistry, Vol. 42, No. 2, 2003 trpC2 metB5 xin-1 SP (s) lytG::pMUTIN4, Emr trpC2 metB5 xin-1 SP (s) lytC::ble trpC2 metB5 xin-1 SP (s) lytD::spc trpC2 metB5 xin-1 SP (s) sigD::pLM5 Cmr trpC2 metB5 xin-1 SP (s) lytC::ble lytD::spc trpC2 metB5 xin-1 SP (s) lytC::ble lytD::spc sigD::pLM5 Cmr trpC2 metB5 xin-1 SP (s) lytG::pMUTIN4, Emr lytC::ble trpC2 metB5 xin-1 SP (s) lytG::pMUTIN4, Emr lytD::spc trpC2 metB5 xin-1 SP (s) lytG::pMUTIN4, Emr lytC::ble lytD::spc trpC2 metB5 xin-1 SP (s) lytG::pMUTIN4, Emr lytC::ble lytD::spc sigD::pLM5 Cmr trpC2 metB5 xin-1 SP (s) yubF::kan trpC2 metB5 xin-1 SP (s) yubF::kan lytC::ble trpC2 metB5 xin-1 SP (s) yubF::kan lytC::ble lytD::spc supE44∆ lacU169 (Φ80 lacZ ∆M15) hsdR17 recA1 endA1 gyrA96 thi-1 relA1 pUC119 with 2.25 kb BamHI-SacI fragment containing yubF gene with an pUC119 with 2.25 kb BamHI-SacI fragment containing yubF gene with a kanamycin cassette inserted in the engineered XhoI site pMUTIN4 containing a 300 bp HindIII-BamHI insert carrying part of the yubE gene pET24d containing lytG gene with its signal sequence removed and C-terminal a Arrows (>) indicate transformation of recipient strain with donor chromosomal DNA. b Bacillus Genetic Stock Center, Ohio State University, or chloramphenicol (5 µg mL-1). In E. coli, plasmids were and also 33 bp into the yubF gene. A kanamycin cassette selected with ampicillin (50 µg mL-1).
was excised from pDG782 (29) by digestion with XhoI and Construction of Mutants Insertionally InactiVated in lytG SalI and then ligated with pGJH170. The resultant plasmid, and yubF. (i) Plasmid Construction. Nucleotide sequence pGJHy2k, therefore contained the yubF gene inactivated by from the B. subtilis complete genome sequence (26) was used the insertion of a kanamycin cassette. Multiple mutant strains to clone an internal fragment of the lytG gene into pMUTIN4 were made by transformation with the appropriate chromo- (27) to create plasmid pGJH100. Using genomic DNA as a template a 300 bp fragment was amplified by PCR using (ii) Transformation of E. coli and B. subtilis. Transforma- the forward primer (5’-GCCGAAGCTT64CCGCTAGCTC- tion of E. coli was performed as described by Hanahan (30).
TGTTTGTTT83) and the reverse primer (5’-CGCGGATCC368 Transformation of B. subtilis 168 with pGJH100 and CGAATGGTTTTTCTTTTCC349), where the restriction sites pGJHy2k was performed by the competent cell method (31).
are underlined and the internal sequence of the gene is Confirmation of lytG gene disruption by means of Campbell- italicized (numbering is with respect to the A of the type recombination was achieved by Southern blot analysis, translational start codon of the gene). Nucleotide sequence using pMUTIN4 HpaI digested fragment containing the from the B. subtilis complete genome sequence (26) was used multiple cloning site as a probe. Southern blot analysis, using to amplify, by PCR, two fragments either side of the yubF the yubF containing 2.25 kbp fragment from pGJH170 as a gene and overlapping. One stretching from 982 bp before probe, showed that the kanamycin resistance cassette had the start of, to 50 bp into the yubF gene was amplified using been inserted into the yubF gene by a double crossover event the forward primer (5’-CGCGCGGATCCCGAAGGAAG- creating the strain GJH145. Hybridization, probe labeling, GTT-958) and the reverse primer (5’-TAAGCTAGTACTCTC- and detection were done using the Boehringer Mannheim GAGCAACCGTGTTTCTGCGTC13). The other fragment nonradioactive DNA labeling and detection kit.
stretches from 50 bp in, to 1002 bp after the end of, the Analysis of Gene Expression. To measure gene expression gene and was amplified using the forward primer (5’- during sporulation, synchronous sporulation was performed GACGCAGAAACACGGTTGCCTCGAGAGTACTAGCT- using the resuspension method of Sterlini & Mandelstam TA50) and the reverse primer (5’-GCGAGCTCGCCGACA- (32). After the initiation of sporulation (t0), samples were ATCGGCGG1002). The two fragments were placed in a PCR harvested every hour for 9 h and sporulation morphology reaction with the two external primers to amplify the whole monitored by microscopy. Levels of -galactosidase activity region. The product was cleaned by gel extraction and were measured as described by Horsburgh and Moir (33).
digested with BamHI and SacI alongside pUC119. The Cell Separation. The measurement of filamentation and digested products were ligated by the method of Sambrook macrofilamentation was performed as described by Blackman et al. (28). The resultant plasmid pGJH170 was digested with et al. (13), with the exception of shaking at 40 rpm instead XhoI that cuts the ligated fragment at the engineered site of 45 rpm for liquid cultures grown at 25 °C.
Biochemistry, Vol. 42, No. 2, 2003 259 Swarm Plate Assay. Swarming motility of strains was measured on nutrient agar plates (0.3% w/v) as describedby Blackman et al. (13).
Cell Autolysis and Preparation of Cell-Wall-Binding Protein (CWBP) Extracts. Lysis of cells of B. subtilis parentaland mutant was performed and cell wall binding proteinsprepared as described by Blackman et al., (13).
OVerexpression and Purification of LytG Protein. Nucle- otide sequence from the B. subtilis complete genomesequence (26) was used to amplify the lytG gene with bothits putative signal sequence and stop codon removed. The801 bp fragment of lytG was amplified using the forwardprimer (5’-GTTTCCATGGCAACTTTATCAAAACCGAT-TGA110) and the reverse primer (5’-TAAACTCGAGGGT-TGCCTCCTTTATTTCA827). The fragment was cleaned bygel extraction and restricted with XhoI and NcoI alongsidepET24d. The restricted products were ligated using themethod of Sambrook et al. (28), creating the plasmidpETGJH6. Transformation of E. coli BL21(DE3) withpGJH6 was as described by Hanahan (30). The cloned lytGgene was sequenced which determined that no errors in itssequence had occurred during amplification (results notshown). Cultures of BL21(DE3) containing pETGJH6 in LB + glucose (0.5% w/v) + Kanamycin (30 µg mL-1) weregrown overnight (25 °C, 250 r.p.m) and inoculated 1 in 20into LB with glucose and kanamycin. Cells were grown at25 °C, 250 rpm until OD600 reached 0.6-0.8 at which point(IPTG) was added to a final concentration of 0.4 mM.
Samples were taken before the addition of IPTG and athourly intervals for 4 h for SDS-PAGE and renaturingSDS-PAGE analysis. Cells were harvested at 4 h and FIGURE 1: The lytG gene is vegetatively expressed. (A) Expression sonicated, and His-tagged LytG protein was purified using of lytG::lacZ was measured in GJH100 during growth in NB at 37 HisTrap metal chelate affinity chromatography as described °C. OD600 (closed symbols), -galactosidase activity (open sym- in the manufacturers instructions (Amersham Pharmacia bols). (B) Analysis of the 5′ end of the lytG transcript by primerextension. The sequencing ladder is in lanes A, C, G, and T. Lane 1 is the primer extension of RNA from strain 1A304. (C) SDS-PAGE and Renaturing SDS-PAGE. Protein samples Chromosomal DNA sequence proximal to the lytG transcriptional from overexpression studies and CWBP samples were start site. The transcriptional and translational starts are shown in analyzed by 12% (w/v) SDS-PAGE (34) and Coomassie bold and italics, respectively. The putative -10 and -35 regions blue staining to visualize proteins. Renaturing gel electro- phoresis using purified B. subtilis vegetative cell walls as intervals. Some samples were boiled and treated with substrate was used to detect autolysin activity as described Cellosyl as described by Atrih et al. (36). Before purification, samples were reduced using borohydride, which converts the Analysis of LytG ActiVity. All assays were carried out at reducing sugar to its corresponding alcohol. Muropeptides 37 °C in a Victor2 (Wallac), in 96 well plates with B. subtilis were analyzed by mass spectrometry (MS) measurements cell walls (0.19 mg mL-1) and purified LytG (20 µg mL-1).
using MALDI-TOF (36). NMR analysis was carried out on Activity was measured as loss of OD450, with 1 unit of Bruker DRX-500 and DRX-600 spectrometers. Samples were enzyme activity being defined as that, which will result in analyzed using two-dimensional correlated spectroscopy loss of 0.001 OD450 units/min-1. To determine the optimum (COSY), total correlated spectroscopy (TOCSY), and rotating pH, buffer and cation for LytG activity sodium citrate, frame nuclear Overhauser effect spectroscopy (ROESY). For phosphate, and Tris/HCl buffers with pH’s ranging from 3 TOCSY, a spin-lock field of 9 kHz was used over a mixing to 6.3, 5.8 to 8.0, and 7 to 9, respectively were used. The time of 95 ms, while for ROESY, a spin-lock field of 3 kHz presence of 20mM of the cations, MgCl2, CaCl2, CuCl2, was used over a mixing time of 150 ms. Spectra were HgCl2, or ZnCl2, the chelator EDTA, or a range of MgCl2 processed using FELIX (Accelrys Inc., San Diego, CA).
and CaCl2 concentrations (0-100mM) were also tested.
Analysis of lytG Promoters. Primer extension was carried out as described previously (34).
Analysis of Enzymatic Hydrolysis of Peptidoglycan by RP- Analysis of lytG Expression. Expression of lytG was HPLC. Preparation of cell wall peptidoglycan and separation measured using a lacZ reporter fusion in strain GJH100.
of soluble muropeptides were carried out as previously Maximal expression was observed during the mid-exponen- described (35) but with the following modifications. Pepti- tial phase of vegetative growth (Figure 1A), with none during doglycan (2 mg mL-1) was hydrolyzed with purified LytG sporulation (results not shown). Total RNA was purified from (50-200 µg mL-1) at 37 °C with samples taken at defined the wild-type strain at the estimated time of maximal 260 Biochemistry, Vol. 42, No. 2, 2003 FIGURE 2: Role of autolysins in cell separation. Formation of boliin liquid cultures occurs due to the increased filamentation in strainGJH110 (lytC lytD lytG) mutant (B) compared to strain SH128(lytC lytD) mutant (A). Liquid cultures (10 mL nutrient broth) weregrown overnight at 25 °C with gentle shaking (40 rpm) and pouredinto Petri dishes for photography.
expression (1.5-2 h) of lytG. The transcriptional site for lytGwas determined by primer extension to be located 48/49 bpupstream of the lytG translational start site (Figure 1B). Thepromoter has a -35 and -10 region of TTGAAA andTAAAAT, respectively.
Role of LytG. Cell Growth and DiVision. In nutrient broth (37 °C, 250 rpm) all B. subtilis (Table 1) strains grew atequivalent rates and reached the same final OD600 (resultsnot shown). Although accurate readings could not be obtainedfor strains SH131(lytC lytD sigD) and GJH110 (lytC lytDlytG) due to hyperfilamentation of cells. When strains weregrown at 25 °C with shaking at 40 rpm clumps of cells (boli)became visible in the medium in strain GJH110 (lytC lytD FIGURE 3: Role of LytD and LytG in B. subtilis peptidoglycanstructural determination. RP-HPLC muropeptide elution patterns lytG) but not in strain SH128 (lytC lytD) (Figure 2).
of peptidoglycan from B. subtilis 168 vegetative cells of 1A304 However, SH115 (lytC), SH119 (lytD), and GJH100 (lytG) (wild type) (A), GJH100 (lytG) (B), and SH119 (lytD) (C).
showed no such degree of boli formation. After growth at25 °C, 40 rpm, strains GJH110 (lytC lytD lytG) and SH128(lytC lytD) were both observed to be highly filamentous by was exacerbated (16% lysis after 90 min treatment). A similar compensatory role for LytD and LytG in lysis in response Swarming Motility. Strains SH115 (lytC),GJH100 (lytG), to sodium azide was also noted (results not shown).
and GJH104 (lytC lytG) were motile in stationary phase when Role of LytG in in ViVo Peptidoglycan Structural Deter- viewed by phase-contrast microscopy. These strains were mination. B. subtilis has a characteristic muropeptide profile analyzed on swarm plates, which measures chemotaxis as during vegetative growth (Figure 3A) (35). Analysis by RP- well as bacterial motility. Motile strains form a halo of diffuse HPLC of purified cell walls of wild type, strain SH119 (lytD), growth round a tightly packed central colony on swarm and strain GJH100 (lytG) showed a substantial reduction in plates. Strain SH115 (lytC) showed a reduction in swarming muropeptide 17 and a complete loss of muropeptide 29 (see motility compared to its parent 1A304, the halo diameter Figure 3; muropeptide designation from ref 35). Muropeptide being 44 (SD ( 0.7) versus 74 (SD ( 0.5) mm. This result 17 was previously shown to contain two products named is consistent with the findings of Blackman et al. (13). Strain 17a and 17b, corresponding to disaccharide tripeptide dis- GJH100 (lytG) also showed a reduction in swarming motility, accharide tetrapeptide with 1 phosphate and two amidations the halo diameter being 48 (SD ( 1.3) mm. GJH104 (lytC and disaccharide tripeptide disaccharide tetrapeptide with two lytG) demonstrated an even greater reduction in swarming amidations and missing a glucosamine (35). Muropeptide (32 (SD ( 0.4) mm). No swarming motility defect was noted 29 was also previously identified as disaccharide tripeptide for strain GJH145 (yubF). SH128 (lytC lytD) and GJH110 disaccharide tetrapeptide disaccharide tetrapeptide missing (lytC lytD lytG) do not swarm at all (as previously noted for a glucosamine (35). In contrast SH119 (lytD) (Figure 3C) showed no alteration in muropeptide profile compared to Cell Autolysis. Inactivation of lytG (strain GJH100) alone had no significant effect on lysis by cloxacillin (results not Analysis of Cell-Wall-Binding Protein Extracts. Renaturing shown). Also, in combination with LytC, there was no defect SDS-PAGE analysis of cell-wall-binding protein (CWBP) over and above LytC. However, although SH119 (lytD) only extracts of wild type and mutant strains revealed that LytG showed a modest reduction in lysis rate in the presence of was not a major CWBP and that its action could not be cloxacillin compared to 1A304 (79% vs 90% lysis after 90 demonstrated on zymograms from whole cell extracts (results min treatment, respectively), in GJH144 (lytG lytD) the effect Biochemistry, Vol. 42, No. 2, 2003 261 FIGURE 4: LytG has glucosaminidase activity. RP-HPLC muropep-tide elution patterns of soluble peptidoglycan fragments from B.
subtilis 168 (HR) vegetative cells hydrolyzed with: (A) Cellosyl,(B) LytG, or (C) LytG and Cellosyl.
Role of LytF. GJH145 (yubF) had no phenotypic differ- ences from the wild type (1A304) in any of the above assays(results not shown).
Analysis of LytG ActiVity. LytG protein was overexpressed in E. coli and purified by virtue of a C-terminal His tag.
The purified protein showed maximal activity in sodiumcitrate buffer (10 mM, pH 5.7). Activity was greatest in the FIGURE 5: 500 MHz NMR spectra of muropeptide 5(1). Spectra were acquired in water containing 10% D2O at 35 °C. (a) 1D spectrum. The water resonance was suppressed by presaturation EDTA (results not shown). The inhibition of LytG activity and has been removed by a low-frequency digital convolution filter.
by EDTA suggests that LytG was purified with chelated (b) Part of the 2D TOCSY spectrum, indicating spin system Mg2+ and that the presence of that cation is essential for assignments. The A2pm signals correspond to correlations with both LytG activity. Under optimal conditions hydrolysis (as the R and δ amides. Correlations involving the sugar resonances measured by loss of optical density) was observed with are boxed. (c) Part of the 2D ROESY spectrum, same spectral regionas part (b). The cross-peaks marked 1 and 2 are discussed in the purified M. luteus and B. subtilis cell walls but not with S. aureus, S. mutans, E. faecalis, or L. arabinosus (results notshown). The optimal specific activity of the purified enzyme ever, the muropeptide elution pattern from HPLC shows that was 450 U mg-1 using B. subtilis vegetative cell walls as these muropeptides have different retention times than their equivalents observed by Atrih et al. (35) after muramidase Determination of LytG Hydrolytic Bond Specificity. B. (Cellosyl) digestion. This difference could be accounted for subtilis 1A304 purified peptidoglycan was hydrolyzed with by the relative position of glucosamine residues. To deter- purified LytG under conditions which resulted in <5% lossof OD mine the structures of these muropeptides, NMR analysis 450 and the soluble material was analyzed (Figure 4B).
Two major peaks were observed (peaks 4 and 5, Figure 4B) was performed. One-dimensional spectra were typical of which did not correspond to any of the previously character- muropeptides. In particular, the spectrum of muropeptide 5(1) ized major Cellosyl derived muropeptides (Figure 4A). The (Figure 5a) contained two signals between 7.0 and 7.5 ppm, novel products were collected, desalted, and analyzed by with intensities corresponding to single protons, indicating mass spectroscopy. On desalting it was observed that peak that 5(1) has a single amidation on one of the two available 5 consisted of two muropeptides. Both of these (called A2pm residues. By contrast, the spectrum of 5(2) (data not muropeptides 5 (1) and 5 (2)) were collected and analyzed shown) indicates two amidations, one on each A2pm residue.
by mass spectroscopy and amino acid analysis (Table 2).
The spectrum of muropeptide 5(1) (Figure 5a) has two This identified these three muropeptides as disaccharide prominent intense peaks at 2.04 and 2.11 ppm. We have tripeptide (muropeptide 4) and disaccharide tripeptide dis- previously shown (35) that these can be assigned to the acetyl accharide tetrapeptide (muropeptides 5(1) and 5(2)). How- methyls of MurNAc and GlcNAc, respectively, and do not 262 Biochemistry, Vol. 42, No. 2, 2003 change position significantly on reduction of the reducing not sufficiently diagnostic to allow us to say with confidence which resonances come from MurNAc and which from The structures of the muropeptides were determined using GlcNAc. This assignment is made most unambiguously from 2D COSY, TOCSY, and rotating frame NOE (ROESY) the ROESY spectrum (Figure 5c) which contains intense spectra, and the method is illustrated using spectra of cross-peaks (labeled 1 and 2), respectively from the MurNAc muropeptide 5(1) (Figure 5). The spin systems are clearly and GlcNAc 2′-amide proton to its adjacent acetyl methyl.
identifiable from TOCSY spectra (Figure 5b), which indicate This demonstrates that the reducing end is GlcNAc. Other for example two closely similar sets of sugar resonances sequential NOEs confirm the expected disaccharide tripeptide coupled to 2′-amide protons at around 8.35 ppm (left-hand disaccharide tetrapeptide structure.
box in Figure 5b), and two more sets coupled to amide After partial hydrolysis of B. subtilis 1A304 peptidoglycan protons at around 7.9 ppm (right-hand box in Figure 5b).
with LytG as above, total material was further hydrolyzed The chemical shifts indicate that the sets around 8.35 ppm by Cellosyl, which resulted in >95% loss of OD450, and the retain characteristic pyranose sugar frequencies (in particular, soluble products were analyzed by HPLC. Two apparently the anomeric proton at around 4.7 ppm), while the sets novel muropeptides (2 and 3) were observed in the LytG around 7.9 ppm are from the reduced alcohol, i.e., this sugar and Cellosyl digested material (Figure 4C) compared with is the reducing end. However, the chemical shift values are the Cellosyl control (Figure 4A), and a large increase in peak Biochemistry, Vol. 42, No. 2, 2003 263 1 was also noted (Figure 4C). Muropeptides 1, 2, and 3 change, or steric hindrance, due to the presence of two (Figure 4C) were collected, desalted, and analyzed by mass amidations, may prevent the action of Cellosyl on both spectroscopy and amino acid analysis (Table 2). Muropep- potential cleavage sites in 5(2). Thus, the amidation state tides 1, 2, and 3 (Figure 4C) are muramic acid tripeptide, effects peptidoglycan hydrolase activity. This may form a monosaccharide tripeptide monosaccharide tetrapeptide, and novel mode of posttranslational regulation of peptidoglycan disaccharide tripeptide disaccharide tetrapeptide missing one hydrolase activity via this subtle modification of the stem glucosamine, respectively (Table 2). For muropeptide 3 the monosaccharide may be substituted with either the tri- or Hydrolysis of peptidoglycan with LytG alone resulted in the appearance of only disaccharide containing muropeptides A digestion time course also failed to reveal any muropep- (Figure 4B). These were monomer and dimer muropeptides.
tides other than disaccharides (muropeptides 4 and 5). Also All three contained only disaccharide substitutions. Larger Cellosyl digestion of soluble LytG products did not give any muropeptides were never recovered in significant quantities.
products apart from muropeptides 1, 2, and 3 (results not Thus LytG is, at least primarily, an exoenzyme acting processively from the ends of the glycan strands. This is thefirst autolytic exo-glucosaminidase to be described in B. DISCUSSION
subtilis 168. The enzyme was able to efficiently hydrolyzeonly cell wall peptidoglycan from B. subtilis and M. luteus Sequence homology had suggested that LytG may be a but not S. aureus and other species, which may be due to muramidase (25). Thus, it was surprising that lack of LytG different substrate chemical structures. O-Acetylation of resulted in the disappearance of muropeptides from the B. muramic residues in S. aureus may be responsible for this subtilis 168 peptidoglycan profile that corresponded to those lack of activity, which has also been shown to prevent the which had been previously predicted to be due to glu- hydrolysis of S. aureus peptidoglycan by lysozyme (38).
cosaminidase activity, as a result of the lack of an N- Also, HF treatment of B. subtilis cell walls led to an increase acetylglucosamine residue at the nonreducing terminus (35).
in peptidoglycan hydrolysis rate by LytG. Teichoic acid and In contrast, missing the major, well characterized glu- membrane phospholipids have previously been proposed to cosaminidase, LytD, did not affect the muropeptide profile.
have moderating effects on LytD activity (8).
Thus, it was proposed that LytG may be a glucosaminidase LytG is a new member of the family of vegetative and constitutes the major activity of this type involved in autolysins of B. subtilis (25). These enzymes perform, in peptidoglycan hydrolysis in the intact cell, even though it is many cases, overlapping and mutually compensatory roles not a major cell wall protein. The lack of LytD products in multiple physiological functions. A number of autolysins may be due to the absence of appropriate substrate for this have been shown to be involved in separation of daughter enzyme and stringent posttranslational control of activity.
cells after septation (6, 39). LytC, LytD, and LytG are all Despite intense speculation the biochemical regulation of involved in boli formation (probably due to hyperfilamen- autolysins in situ which prevents spontaneous autolysis has tation), as only the triple mutant forms such large clumps.
remained elusive. Very recently it has been shown that the Such levels of functional redundancy may be the result of cell wall of B. subtilis is protonated during growth (37). Both the importance of cell separation in the natural environment LytD and LytG have slightly acid pH optima, which suggests and the presence of the autolysins in the exposed location they may well be active in the cell wall environment.
of the cell wall open to attack by proteases etc. LytG also To verify the hydrolytic bond specificity of LytG, the plays a role in motility and chemotaxis, as evidenced by a protein was overexpressed and purified. Muropeptide analysis defect in swarming ability. This could be due to slight of LytG hydrolyzed peptidoglycan revealed unequivocally changes in filamentation status, which would impact on that LytG is a glucosaminidase (Table 2 and Figure 4).
chemotactic ability, or flagellar extrusion may be effected Cellosyl (muramidase) digestion of LytG hydrolyzed pep- by aberrant cell wall hydrolysis. Lysis due to azide or cell tidoglycan results in the loss of muropeptides 4 and 5 (Figure wall antibiotics occurs as a post-mortem event. In this case 4B) and the appearance of 1, 2, and 3 (Figure 4C). Thus LytG and LytD showed mutually compensatory roles.
muropeptides 4 and 5 are glucosaminidase-derived substrates Transcriptional fusion and primer extension analysis for the muramidase. Muropeptide 4 is hydrolyzed by Cellosyl revealed lytG transcription to occur from a single promoter, to produce muropeptide 1. Muropeptides 5(1) and 5(2) are similar to the consensus sequence typical of σA-dependent muramidase hydrolyzed to make muropeptides 2 and 3, promoters (40). This is in contrast to lytC and lytD, which respectively, as revealed by their relative amidation status are completely or in part under the control of the alternative (Table 2). This confirms that LytG is a novel glucosamini- sigma factor σD. Thus, LytG may have its primary function dase unrelated to previous enzymes of this class. The during exponential growth. The roles of the large complement homology of LytG to AcmA suggests that either related of seemingly compensatory autolysins during growth and the enzymes may have different activities or that previous mechanisms whereby all these components combine to create methodologies used to determine hydrolytic bond specificity a functional cell wall architecture have remained obscure.
have not been stringent enough. Thus, it is important to verify The classical representation of peptidoglycan is as long activity using the sensitive RP-HPLC assay.
glycan strands cross-linked periodically by peptide side The presence of muropeptide 3 (which retains an N- chains. However evidence from Escherichia coli (41) and acetylglucosamine residue) suggests that Cellosyl is unable recently from S. aureus (42) have revealed this convention to fully hydrolyze muropeptide 5(2). Muropeptide 2 was to be misleading. The average glycan strand length in S. shown to have only one amidation by NMR, while muropep- aureus is in fact 6 disaccharides. Thus, the bulk of the tide 3 had two amidations (Table 2). A conformational peptidoglycan is made up of short chains, which must be 264 Biochemistry, Vol. 42, No. 2, 2003 highly cross-linked in order to maintain integrity. The 18. Brewer, S. J., and Berkeley, R. C. W. (1973) Biochem. J. 134, presence of oligomer muropeptides linking multiple chains 19. Ortiz, J. M., Berkeley, R. C. W., and Brewer, S. J. (1973) J. Gen. alludes to a specific organization within the peptidoglycan.
The accurate chain length of peptidoglycan of B. subtilis has 20. Ortiz, J. M. (1974) J. Bacteriol. 117, 909-910.
not yet been determined. The activity of glucosaminidases, 21. Foster, S. J. (1992) J. Bacteriol. 174, 464-470.
such as LytG and LytD, will be to modify glycan strand 22. Margot, P., Wahlen, M., Gholamhuseinian, A., Piggot, P., and Karamata, D. (1998) J. Bacteriol. 180, 749-752.
length. Their role in this crucial parameter for peptidoglycan 23. Ishikawa, S., Hara, Y., Ohnishi, R., and Sekiguchi, J. (1998) J. structure is currently under investigation.
24. Margot, P., Pagni, M., and Karamata, D. (1999) Microbiology 145, REFERENCES
25. Smith, T. J., Blackman, S. A., and Foster, S. J. (2000) Microbiology 1. Ghuysen, J.-M., Tipper, D. J., and Strominger, J. L. (1966) 26. Kunst, F., Ogasawara, N., Moszer, I., and 148 other authors (1997) 2. Rogers, H. J., Thurman, P. F., and Burdett, I. D. J. (1983) J Gen. 27. Vagner, V., Dervyn, E., and Ehrlich, S. D. (1998) Microbiology 3. Fein, J. E., and Rogers, H. J. (1976) J. Bacteriol. 127, 1427- 28. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular 4. Fein, J. E. (1979) J. Bacteriol. 137, 933-946.
Cloning: a Laboratory Manual, 2nd ed., Cold Spring Harbor 5. Pooley, H. M., and Karamata, D. (1984) J. Bacteriol. 160, 1123- 29. Gue´rout-Fleury, A.-M., Shazand, K., Frandsen, N., and Stragier, 6. Ward, J. B., and Williamson, R. (1984) in Microbial Cell Wall Synthesis and Autolysis (Nombela, C., Ed.) pp 159-166, Elsevier, 30. Hanahan, D. (1983) Mol. Biol. 166, 557-580.
31. Anagnostopoulos, C, and Spizizen, J. (1961) J. Bacteriol. 81, 741- 7. Herbold, D. R., and Glaser, L. (1975) J. Biol. Chem. 250, 1676- 32. Sterlini, J. M., and Mandelstam, J. (1969) Biochem. J. 113, 29- 8. Rogers, H. J., Taylor, C., Rayter, S., and Ward, J. B. (1984) J. Gen. Microbiol. 130, 2395-2402.
33. Laemmli, U. K. (1970) Nature 227, 680-685.
9. Kuroda, A., and Sekiguchi J. (1991) J. Bacteriol. 173, 7304- 34. Horsburgh, M. J., and Moir, A. (1999) Mol. Microbiol. 32, 41- 10. Lazarevic, V., Margot, P., Soldo, B., and Karamata, D. (1992) J. 35. Atrih, A., Bacher, G., Allmaier, G., Williamson, M. P., and Foster, Gen. Microbiol. 138, 1949-1961.
S. J. (1999) J. Bacteriol. 181, 3956-3966.
11. Margot, P., and Karamata, D. (1992) Mol. Gen. Genet. 232, 359- 36. Atrih, A., Zo¨llner, P., Allmaier, G., Williamson, M. P., and Foster, S. J. (1998) J Bacteriol. 180, 4603-4612.
12. Margot, P., Maue¨l, C., and Karamata, D. (1994) Mol. Microbiol. 37. Calamati, H. G., Ehringer, W. D., Koch, A. L., and Doyle, R. J.
(2001) Proc. Natl. Acad. Sci U.S.A. 98, 15260-15263.
13. Blackman, S. A., Smith, T. J., and Foster, S. J. (1998) Microbiol. 38. Snowden, M. A., Perkins, H. R., Wyke, A. W., Hayes, M. V., and Ward, J. B. (1989) Microbiology 135, 3015-3022.
14. Helmann, J. D., Marquez, L. M., and Chamberlin, M. J. (1988) J. 39. Forsberg, C., and Rogers, H. J. (1971) Nature 229, 272-273.
40. Haldenwang, W. G. (1995) Microbiol. ReV. 59, 1-30.
15. Rashid, M. H. Mori, M., & Sekiguchi, J. (1995) Microbiology 41. Harz, H., Burgdorf, K., and Holtje, J.-V. (1990) Anal. Biochem. 16. Ortiz, J. M., Gillespie, J. B., and Berkeley, R. C. W. (1972) 42. Boneca, I. G., Huang, Z.-H., Gage, D., and Tomasz, A. (2000) J. Biochim. Biophys. Acta 289, 174-186.
17. Berkeley, R. C. W., Brewer, S. J., Ortiz, J. M., and Gillespie, J.
B. (1973) Biochim. Biophys. Acta 309, 157-168.

Source: http://nmr.group.shef.ac.uk/MPW/125.pdf