1

PEPTIDOGLYCAN N-ACETYLGLUCOSAMINE DEACETYLASES FROM BACILLUS CEREUS, HIGHLY CONSERVED PROTEINS IN BACILLUS

ANTHRACIS*

Emmanuel Psylinakis,2 Ivo G. Boneca,3 Konstantinos Mavromatis,4 Alexandra Deli,1,2 Emma Hayhurst,5 Simon J. Foster,5 Kjell M. Vårum,6 and

Vassilis Bouriotis1, 2 Department of Biology, Enzyme Biotechnology Group, University of Crete, P.O. Box 2208, Vasilika Vouton 714 09, Heraklion, Crete, Greece,1 Institute of Molecular Biology and Biotechnology, P.O. Box 1527, Vasilika Vouton 711 10, Heraklion, Crete, Greece,2 Unité de Pathogénie Bactérienne des Muqueuses, Institut Pasteur, 28 Rue du Dr. Roux, 75724 Paris, Cedex 15, France,3 JGI Production Genomics Facility 2800 Mitchell Drive Walnut Creek, CA 94598, USA,4 Department of Molecular Biology and Biotechnology, University of Sheffield, Sheffield, S10 2TN, United Kingdom,5 and Norwegian Biopolymer Laboratory (NOBIPOL), Department of Biotechnology, The Norwegian University of Science and Technology, 7034 Trondheim, Norway6.

Running title: Polysaccharide deacetylases in Bacillus cereus Address correspondence to: Vassilis Bouriotis, Department of Biology, University of Crete, P.O. Box 2208, Vasilika Vouton, 711 10, Heraklion, Crete, Greece, Tel/Fax: +30-281-0394375, E-mail: bouriotis@imbb.forth.gr

The genomes of Bacillus cereus and its closest relative Bacillus anthracis contain 10 polysaccharide deacetylase homologues. Six of these homologues have been proposed to be peptidoglycan N-acetylglucosamine deacetylases. Two of these genes namely bc1960 and bc3618 have been cloned and expressed in Escherichia coli and the recombinant enzymes have been purified to homogeneity and further characterized.

Both enzymes were effective in deacetylating cell wall peptidoglycan from the Gram(+) Bacillus cereus and Bacillus subtilis and the Gram(-) Helicobacter pylori as well as soluble chitin substrates and N-acetylchitooligomers. However the enzymes were not active on acetylated xylan. These results provide insight into the substrate specificity of Carbohydrate Esterase Family 4 Enzymes (CE4).

It was revealed that both enzymes deacetylated only the GlcNAc residue of the synthetic muropeptide N-acetyl-D-glucosamine-(ß-1,4)-N-acetylmuramyl-L-alanine-D-isoglutamine. Analysis of the constituent muropeptides of peptidoglycan from B. subtilis and H. pylori resulting from incubation of the enzymes BC1960 and BC3618 with these polymers and subsequent hydrolysis by Cellosyl and mutanolysin respectively similarly revealed that both

enzymes deacetylate GlcNAc residues of peptidoglycan.

Kinetic analysis towards GlcNAc2-6 revealed that GlcNAc4 was the favorable substrate for both enzymes. Identification of the sequence of N-acetychitooligosaccharides (GlcNAc2-4) following enzymatic deacetylation by using 1H-NMR revealed that both enzymes deacetylate all GlcNAc residues of the oligomers except the reducing end ones.

Enzymatic deacetylation of chemically acetylated vegetative peptidoglycan from B. cereus by BC1960 and BC3618 resulted in increased resistance to lysozyme digestion. This is the first biochemical study of bacterial peptidoglycan N-acetylglucosamine deacetylases.

Polysaccharide deacetylases belong to

Carbohydrate Esterase Family 4 (CE4) which includes chitin deacetylases, acetyl-xylan esterases, xylanases, rhizobial NodB chitooligosaccharide deacetylases and peptidoglycan deacetylases (1). All these enzymes share a universal conserved region called polysaccharide deacetylase domain (according to the Henrissat classification). All five members of this family catalyze the hydrolysis of either N-linked acetyl group from N-acetylglucosamine residues (chitin

JBC Papers in Press. Published on June 16, 2005 as Manuscript M407426200

Copyright 2005 by The American Society for Biochemistry and Molecular Biology, Inc.

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

2

deacetylase, NodB and peptidoglycan N-acetylglucosamine deacetylase), or O-linked acetyl groups from O-acetylxylose residues (acetyl xylan esterase, xylanase) (2-4).

A large number of open reading frames encoding for putative polysaccharide deacetylases have been identified in genomes of Gram-positive bacteria. It has been suggested that they correspond to peptidoglycan deacetylases, although this has been verified only in two cases. Two genes namely (pgdA) from Streptococcus pneumoniae and (pdaA) from Bacillus subtilis have been identified encoding for peptidoglycan N-acetylglucosamine deacetylase (5) and an enzyme required for the production of muramic δ-lactam in the spore cortex of B. subtilis respectively (6, 7). Recently the crystal structure of this peptidoglycan deacetylase was reported (8). However, the corresponding gene products have not been purified and characterized biochemically.

The recent sequencing of the Bacillus cereus (9) and Bacillus anthracis (10) genomes revealed in each a multiplicity (10 ORFs) of putative polysaccharide deacetylases. Six of these genes have been proposed to encode for putative peptidoglycan deacetylases in B. cereus and have almost identical amino acid sequence with the corresponding ones from B. anthracis implying similar functional roles of these proteins in the two bacteria.

Given the laboratory safety precautions necessary for working with highly infectious agents and the recent concerns and proscriptions related to B. anthracis as a potential bioweapon (class A agent, Center for Disease Control, USA) the B. cereus enzymes offer themselves as suitable models for studying the corresponding proteins of B. anthracis.

The objective of this study is to shed light on the role of bacterial GlcNAc polysaccharide deacetylases and furthermore based on the extensive homologies to contribute to our understanding of the physiology of B. anthracis, an interesting pathogenic microbe close relative to B. cereus.

This report describes the cloning and expression of two genes encoding for peptidoglycan GlcNAc deacetylases from B. cereus ATCC 14579 as well as the characterization of the purified recombinant enzymes.

EXPERIMENTAL PROCEDURES

Materials and Methods Primers were synthesized by the

Microchemistry Facility of IMBB. The expression plasmid pET-26b(+) and E. coli BL21-DE3 pLysS were from Novagen. cDNA clones from B. cereus ATCC14579 were from Integrated Genomics Inc. (Chicago, US). All chromatographic materials were from Pharmacia. Ni-NTA agarose, PCR, gel extraction and plasmid purification kits were from Qiagen. Enzymes and reagents for acetate determination were purchased from Roche Diagnostics. Substrates and common biochemicals were purchased from Sigma. Restriction enzymes as well as all DNA modifying enzymes were from MINOTECH Biotechnology. Cloning and expression of the bc1960 and bc3618 genes of B. cereus into pET expression Vectors- The genes were amplified from cDNA clones using DNA polymerase chain reaction. An Nde I and a XhoI site was incorporated at the start and the end of each gene in order to produce an in frame C-terminal 6xHis-tag fused construct in pET-26b(+) vector. The amplified genes were purified, digested with the corresponding enzymes and ligated into pET-26 vectors placing the polysaccharide deacetylase genes under the transcriptional control of the T7lac promoter. The resulting plasmids were transformed into BL21-DE pLysS, the transformants were screened for the gene of interest and the inserts were sequenced at the automatic sequencing facility of IMBB. The transformed strains containing each of the cloned genes were stored at -80 oC in 20% glycerol until further use. 20 ml of a saturated culture of each of the transformed polysaccharide deacetylases expression strains was inoculated into 500 ml of LB medium containing suitable antibiotics and incubated at 30 oC on a rotary shaker to an A600 of 0.6. The cultures were transferred to 25 oC, IPTG was added to a final concentration of 1mM and the cells were further incubated for 16 hours. Purification of BC1960 and BC3618-Cells were harvested by centrifugation and resuspended in 50mM Tris-HCl buffer pH 8.0, 1M NaCl, 5mM imidazole (Buffer A). After sonication, the soluble fraction was collected by centrifugation and loaded on Ni-NTA agarose, equilibrated

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

3

with buffer A. Proteins were eluted using a step gradient of imidazole (250mM). Active fractions were pooled, concentrated (3 ml) and subsequently applied onto a Sephacryl S200 HR column previously equilibrated in 20 mM Tris-HCl pH 7.4, 200mM NaCl. Fractions containing enzyme activity were pooled and concentrated (1mg/ml). Preparation of radiolabeled substrates-Preparation of cell wall peptidoglycan from vegetative B. cereus and B. subtilis was performed according to a previously described protocol (11). Labeling of glycol chitin and peptidoglycan was performed using [3H]acetic anhydride according to Araki and Ito (12). Enzyme Assays-Standard enzyme assays were performed in a mixture containing 20mM Tris-HCl pH 8.0 (for BC3618) or 20mM Mes-NaOH pH 6.0 (for BC1960), 1mM CoCl2 and 5 µl of the substrate (1mg/ml). Incubation time was 30 min at 50 oC (for BC1960), or 37 oC (for BC3618).

We have employed two different assays for determining polysaccharide deacetylase activity. (i) In a radiometric assay, deacetylase activity was estimated using as substrate partially O-hydroxyethylated chitin (glycol chitin) and peptidoglycan radiolabeled in N-acetyl groups (12). (ii) In a coupled assay, acetate released by the action of polysaccharide deacetylases on various polysaccharides and N-acetylchitooligomers was determined by the enzymatic method of Bergmeyer via three coupled enzyme reactions (13).

Kinetic properties of BC1960 and BC3618 towards N-acetylchitooligosaccharides were determined as below. The reactions were performed with various concentrations of GlcNAc2-6 in 20mM Tris-HCl pH 8.0 (for BC3618) or 20mM Mes-NaOH pH 6.0 (for BC1960), 1mM CoCl2 at 50 oC (for BC1960) or 37 oC (for BC3618) for 10 minutes. Reactions were terminated by heating the mixture at 100 oC for 5 min. Acetic acid released was determined by using an acetic acid determination Kit (Roche Diagnostics) (13). Analysis of reaction products- A commercially available muropeptide N-acetyl-D-glucosamine-(ß-1,4)-N-acetylmuramyl-L-alanine-D isoglutamine (GMDP; 40 µg/ml) was treated either with BC1960 or BC3618. Reaction products were separated by HPLC, desalted and further analyzed by MALDI-TOF and MALDI-PSD as previously described (14).

Purified peptidoglycan (200 µg) of Helicobacter pylori, was incubated with BC1960 or BC3618 for 18 hours at 50°C and 37°C respectively. Samples were boiled for 10 minutes and insoluble peptidoglycan was recovered by centrifugation and washed once with water. The resulting pellet was further digested with mutanolysin (500 µg/ml) in 12.5 mM sodium phosphate pH 5.8 for 18 hours at 37°C with stirring. Samples were boiled for 5 minutes, the supernatants were reduced with sodium borohydride and muropeptides were analyzed by HPLC as previously described (15). Individual peaks were also purified and analyzed by MALDI-TOF and MALDI-PSD as previously described (14).

Preparation and reversed phase-HPLC analysis of soluble and reduced muropeptides from B. subtilis peptidoglycan was carried out as previously reported (11). Reduced muropeptides were separated using a Waters HPLC system on a Hypersil octadecylsilane column (4.6 by 250 mm; particle size, 5 mm). Elution buffers were as follows: Buffer A, 40 mM sodium phosphate pH 4.5; Buffer B, 40 mM sodium phosphate pH 4.0 containing 20% (vol/vol) methanol. A small amount of sodium azide (142 µl from a 1% [wt/vol] solution) was added to 1 lt of buffer A to equalize its A202 with that of buffer B. The column was equilibrated at 52°C with buffer A at a flow rate of 0.5 ml /min for 20 min. Soluble- reduced muropeptides (60 µl) were injected, and a linear gradient of 0 to 100% buffer B over a period of 270 min was started 5 min after sample injection. The flow rate was constant at 0.5 ml /min over the course of the gradient, and the eluted compounds were detected by monitoring A202.

GlcNAc2-4 (2-4 mg) were incubated with 0.5 mg of each enzyme in 0.5 ml buffer (25 mM Mes-NaOH buffer pH 6.0 for BC1960 and 50 mM Tris-HCl buffer pH 8.0 for BC3618) at 37 oC for between 18 to 48 hours. The reaction was stopped by heating to 100 oC for 3 minutes, cooled and the pH adjusted to 4.5 by HCl before lyophilization. The samples were dissolved in 0.7 ml D2O and the pD adjusted to 4 using DCl and the proton NMR-spectra obtained as previously described (16). Polyacrylamide gel electrophoresis-Homogeneous polyacrylamide gel electrophoresis under denaturing and reducing conditions was performed according to Laemmli (17). Protein bands were visualized by staining with Coomassie Brilliant Blue R.

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

4

RESULTS

Comparison with other polysaccharide deacetylases. The deduced sequences of bc1960 (NP_831730) and bc3618 (NP_833348) consist of 275 and 213 amino acids respectively. BC1960 and BC3618 exhibit the highest identity to the peptidoglycan N-acetylglucosamine deacetylase from S. pneumoniae (38% and 36% identity respectively). Residues conserved in all members of CE4 family enzymes were also identified in the B. cereus deacetylases (Fig. 1).

Comparison with B. anthracis. Figure

2 lists the 10 open reading frames of B. cereus alongside the orthologous members of B. anthracis. These pairings represent the best sequence alignments determined by CLUSTAL analysis of the individual proteins for each organism. Ten ORFs found in B. anthracis were identical in size and most of them were within 90% identical to the corresponding ones from B. cereus.

The assignment of putative enzymatic activities used here were obtained from the annotated ERGO-light data base (Legend Fig 2, ref. 10).

Enzyme purification and

characterization. In addition to bc1960 and bc3618 genes we have also expressed the remaining four putative GlcNAc deacetylase genes. However we were unable to demonstrate enzymatic activity of these gene products towards various chitin and peptidoglycan substrates tested.

Both polysaccharide deacetylases were produced containing a C-terminal 6xHis-tag and isolated from the soluble fraction of the cell extract. Purification was achieved in one step using Ni-NTA affinity chromatography with an overall yield 5 mg of protein per liter of E. coli culture. The recombinant enzymes were purified to homogeneity and appear to be monomers as judged by SDS-PAGE and gel filtration chromatography. BC3618 and BC1960 exhibit molecular masses of ~28 kDa and ~ 30 kDa respectively in agreement with the masses (27,711 Da for BC3618 and 30,570 Da for BC1960) estimated from the nucleotide sequences (Fig. 3).

Both enzymes were active on radiolabeled glycol chitin and peptidoglycan from B. cereus vegetative cell walls but not on xylan. Chitin deacetylase from Mucor rouxii was not effective in deacetylating cell wall peptidoglycan from B. cereus (data not shown).

Both enzymes were inhibited by the presence of 1mM Cu2+ and Zn2+ tested as chlorides whereas they were not inhibited by other divalent or monovalent metals (Mg2+, Ca2+, Mn2+, K+, Na+) tested up to 50mM concentration. A maximum 10% increase in activity of both enzymes was observed by the addition of 1mM CoCl2 in the assay buffer. Both enzymes were not inhibited by acetate even at concentrations up to 50 mM. The purified recombinant enzymes exhibited different pH, temperature optima and thermal stability, as determined by using both radiolabeled glycol chitin and peptidoglycan as substrates. BC1960 exhibits a pH optimum at pH 6.0 and a remarkable thermal stability since it retains 95% of its activity after preincubation at 50 oC for 24 hours. On the other hand BC3618 exhibits a pH optimum at pH 8.0 and is inactivated after incubation at 50 oC for 1h. Optimum temperature for enzyme activity was determined to be 50 oC and 37 oC for BC1960 and BC3618 respectively (data not shown).

The specificity of the enzymes for various N-acetylchitooligomers was examined and the kinetic parameters were also calculated (table 1). Both enzymes did not deacetylate N-acetylglucosamine and required at least two sugar residues for catalysis. Kinetic parameters for GlcNAc2-6 were obtained from Lineweaver-Burk plot analysis and the enzyme reaction rates for these substrates seemed to follow Michaelis-Menten kinetics. Both enzymes exhibited the highest Vmax value toward GlcNAc4 and the lowest Km for GlcNAc6 among these substrates. The resulting Kcat/Km ratios indicated that GlcNAc4 was the favorable substrate for both enzymes. BC3618 exhibited higher Kcat/Km values compared to BC1960 toward GlcNAc3-6 substrates, while BC1960 exhibited higher Kcat/Km toward GlcNAc2.

Enzymatic deacetylation of N-acetyl-

D-glucosamine-(ß-1,4)-N-acetylmuramyl-L-alanine-D isoglutamine. The synthetic muropeptide N-acetyl-D-glucosamine-(ß-1,4)-N-acetylmuramyl-L-alanine-D isoglutamine (GMDP) was incubated with both deacetylases.

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

5

Both enzymes converted GMDP in two new muropeptides (peaks 1 and 2, Fig. 4A). The new species were purified, desalted and analyzed by MALDI-TOF. The two muropeptides had exactly the same molecular mass (m/z 678,2749 and 678,3139) and differed from GMDP (m/z 720,2915) by m/z 42 which corresponds to the loss of an acetyl group. Further, we performed MALDI-PSD on peaks 1 and 2 to determine which moiety of GMDP was modified by both deacetylases (Fig. 4B and data not shown). Interestingly PSD analysis of peaks 1 and 2 revealed that only the N-acetylglucosamine residue had lost the N-linked acetyl functional group and that no modification of the N-acetylmuramic acid residue had occurred. Indeed, a major fragment (m/z 517,4) corresponds to the loss of glucosamine (m/z 161). Fragmentation from the C-terminal end of the deacetylated GMDP generates a and b type ions. For example, a1-NH3 ion m/z 504.2 is further converted into m/z 343,2 corresponding to the loss of m/z 161 therefore a glucosamine residue. Finally, the presence of the ion m/z 318,2 which corresponds to the N-acetylmuramic acid alone clearly excludes deacetylation of the N-acetylmuramic acid residue. Furthermore, no fragments corresponding to the loss of N-acetylglucosamine were observed (loss of m/z 203). The same fragmentation patterns were observed for peak 2 (data not shown) excluding a minor deacetylation of GMDP at the N-acetylmuramic acid residue. These results also imply that despite the conversion of GMDP into two mew muropeptide peaks with different elution properties, the two species are structurally identical. This behavior might be due to different protonation states of the deacetylated GMDP in the elution buffers.

Enzymatic deacetylation of H. pylori

and B. subtilis peptidoglycans by BC1960 and BC3618. In order to further examine the substrate specificity of BC1960 and BC3618, we incubated insoluble peptidoglycan from H. pylori with each enzyme followed by mutanolysin digestion and analyzed the resulting muropeptides by RP-HPLC. Muropeptide analysis indicated that both deacetylases generated the same kind of muropeptides (BC3618 converted 57 % of muropeptides while BC1960 converted 90% of muropeptides into new species, data not shown). However, after 18 hours treatment, BC1960 seemed to convert

almost all H. pylori monomeric muropeptides into new muropeptide species, while treatment with BC3618 resulted in a mixture of normal H. pylori muropeptides and new species (Fig. 5). Analysis of the resulting muropeptides by MALDI-TOF showed that both enzymes were able to deacetylate monomeric muropeptides (GlcNAc-MurNAc) with varying peptide chain length (table 2). The appearance of the same muropeptide in two individual peaks (table 2, peaks 4, 4a* and peaks 5*, 5a*) might be due to different protonation states of the deacetylated muropeptides in the elution buffers. Interestingly, even anhydro muropeptides (retention time over 65 minutes) (18), were converted into new species (table 2, peaks 6* and 9*), which corresponded to the deacetylated ones. Fragmentation of the deacetylated muropeptides by MALDI-PSD confirmed that N-acetylglucosamine residue of muropeptides is deacetylated.

In order to examine the effect of these enzymes on Bacilli peptidoglycan we used as a model substrate peptidoglycan from vegetative B. subtilis (Table 3). Identification of muropeptides was based on the previously reported analysis of B. subtilis peptidoglycan (11). Incubation of B. subtilis vegetative peptidoglycan with BC1960 and BC3618 resulted in an increase of the deacetylated muropeptides from 23.6% (native peptidoglycan) to 36.1% and 26.6% respectively. Both enzymes catalyze the deacetylation of GlcNAc residues on peptidoglycan, as confirmed by the increase of peak 4 which corresponds to a disaccharide tripeptide with an amide group missing the N-acetyl-group from GlcNAc and the concomitant decrease of muropeptide 3 which corresponds to the acetylated form of this muropeptide (Table 3). GlcNAc deacetylase activity of BC1960 is also confirmed by the increase of peaks 18, 22, 23 and 32 which correspond to muropeptides de-acetylated on GlcNAc residues identified in B. subtilis peptidoglycan. BC3618 could also deacetylate to a lower extent certain muropeptides (Table 3, peaks 18 and 23) while it was ineffective in deacetylating other muropeptides (Table 3, peaks 22 and 32). Analysis of enzymatically deacetylated oligomers GlcNAc2-4. Figure 6 shows the 1H-NMR spectra (anomer region) of the chitin oligomers (dimer, trimer and tetramer) obtained after treatment with BC3618. The spectra of the

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

6

dimer (Fig. 6A) shows that the major resonances are the α-anomer reducing end of an acetylated unit (A) at 5.19 ppm and the β–anomer reducing end of an acetylated unit at 4.74 ppm (19). The α- and β-anomer reducing end resonances from a deacetylated unit (D), which would be expected to appear at 5.43 ppm and 4.92 ppm are only present in very low concentrations in the spectrum. The internal D- and A-units resonate at around 4.9 ppm and 4.6 ppm, and only the resonances from the internal D-units can be seen in the spectra (20). Thus, the major dimer contains an acetylated reducing end and a deacetylated non-reducing end (DA). Similarly, the trimer (Fig 6B) and tetramer (Fig. 6C) spectra were assigned, and it was found that the major trimer is DDA and the major tetramer DDDA. Thus, BC1960 and BC3618 were able to deacetylate all units on chitin oligomers (dimer to tetramer) except the reducing end unit. BC1960 and BC3618 behaved similarly with respect to the specificity on chitin oligomers, except that with BC1960 it was possible to deacetylate the reducing end unit of the trimer and the tetramer (but not with the dimer) as substrates upon extended incubation with the enzyme, while this was not the case with BC3618.

Lysozyme digestion of B. cereus peptidoglycan. In order to examine the effect of the degree of acetylation of vegetative peptidoglycan from B. cereus on the rate of lysozyme degradation four different peptidoglycan substrates from B. cereus were examined. a) Native vegetative peptidoglycan b) chemically N-acetylated peptidoglycan and chemically N-acetylated peptidoglycan incubated with either c) BC1960 or d) BC3618. We observed that although native peptidoglycan couldn’t be digested, 40% of the chemically acetylated peptidoglycan was solubilized by lysozyme (Fig. 7). Partial digestion of acetylated peptidoglycan by lysozyme might be due to incomplete chemical acetylation of the polymer. However, incubation of chemically acetylated peptidoglycan with either BC1960 or BC3618 resulted in modified substrates resistant to hydrolysis by lysozyme.

DISCUSSION

The genomes of B. cereus (9) and its

closest relative B. anthracis (10) contain each a set of 10 polysaccharide deacetylase homologues, six of which have been proposed to encode for putative peptidoglycan N-acetylglucosamine deacetylases and share almost identical amino acid sequences (Fig. 2).

In our effort to shed light on the role of polysaccharide deacetylases in bacteria and contribute to the understanding of the physiology of B. anthracis, we report for the first time the cloning and expression of two peptidoglycan GlcNAc deacetylase genes bc1960 and bc3618 from B. cereus as well as the purification and characterization of the recombinant enzymes. Previous reports on bacterial peptidoglycan deacetylases addressed either partially purified enzymes or not biochemically characterized gene products (12, 5-7).

BC1960 exhibits an overall 36% identity and 57% similarity with BC3618. The sequence similarity with other known polysaccharide deacetylases extends only to the C-terminal, which is the putative NodB domain (Fig. 1). The purified recombinant enzymes exhibit different pH, temperature optima and thermal stability, while they are not inhibited by acetate and they are both activated by CoCl2. Both enzymes were effective in deacetylating peptidoglycan from B. cereus, B. subtilis, H. pylori as well as glycol chitin, N-acetyl chitooligomers and the synthetic muropeptide GMPD. However, the enzymes were not active on acetylated xylan. This is the first example of CE4 family enzymes which are active on both peptidoglycans and chitin substrates. It has been previously reported that chitin deacetylase from M. rouxii and acetylxylan esterase from Streptomyces lividans were active on both soluble chitin and acetyl xylan substrates (2). However, both enzymes were inactive towards peptidoglycan substrates. These results provide insight into the substrate specificity of the CE4 family enzymes.

Considering specificity of GlcNAc deacetylases, the enzymes are active on both peptidoglycans and soluble muropeptides. It seems that although the enzymes exhibit a strict specificity for deacetylating GlcNAc residues of peptidoglycan substrates they are able to hydrolyze various muropeptides of different

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

7

peptide side chains as exemplified in the case of peptidoglycan from H. pylori and B. subtilis (Fig.5, table 2 and table 3). Furthermore, the enzymes are active on substrates whereby the neighboring residue to GlcNAc is not N-acetylmuramic acid (MurNAc) as demonstrated in the cases of GlcNAcn (Fig. 6) and anhydromuropeptides from peptidoglycan of H. pylori (table 2). In support of this observation PgdA from S. pneumoniae, the only peptidoglycan deacetylase identified in this bacterium, carries out the deacetylation of almost 80% of GlcNAc residues in cell wall peptidoglycan (5).

Considering enzyme specificity of MurNAc deacetylases, expression of pdaA gene from B. subtilis in E. coli had no effect on modification of E. coli peptidoglycan structure whereas joint expression of muramoyl-L-alanine amidase (cwlD) and pdaA genes resulted in the formation of muramic-δ-lactam. PdaA probably carries out both deacetylation and lactam ring formation and requires the product of CwlD activity as a substrate (7). The recently reported structure of PdaA complexed with the substrate analogue N-acetylglucosamine, although possibly representing a non-productive binding mode, revealed interactions of an aspartic acid and three histidines all conserved in the nodB homologous domain with the ligand (8). BC1960 and BC3618 exhibit significant sequence similarity with PdaA (68% and 67% respectively). On the basis of the sequence similarities similar structures are predicted between the two enzymes and PdaA.

Kinetic analysis of BC1960 and BC3618 toward N-acetylchitooligosaccharides GlcNAc2-6 revealed that the catalytic efficiency of the enzymes depended on the length of the oligomers (table 1). Both enzymes required at least two sugar residues for catalysis (although BC3618 showed very low efficiency toward GlcNAc2). Furthermore, GlcNAc4 proved to be the favorable substrate for both enzymes.

Identification of the sequence of chitin oligomers (GlcNAc2-4) following enzymatic deacetylation by 1H-NMR revealed that both enzymes deacetylate all GlcNAc residues of the oligomers except the reducing end ones (Fig. 6).

It has been previously reported that the resistance of pneumonococcal and B. cereus peptidoglycan to lysozyme degradation depends on the degree of acetylation of peptidoglycan (5, 21). Recent analysis of B. cereus peptidoglycan revealed an unusual high percentage of deacetylated GlcNAc residues (21). We have similarly observed that chemical acetylation of vegetative peptidoglycan isolated from B. cereus (ATCC 14579) resulted in enhanced degradation rates by lysozyme. However, enzymatic deacetylation of chemically acetylated peptidoglycan by BC1960 and BC3618 resulted in resistance of this macromolecule to lysozyme degradation (Fig. 7). It is conceivable that the BC1960 and BC3618 or the other orthologues may provide protection against host lysozyme. Further investigations are needed to establish the importance, if any of these enzymes (knockout studies).

In light of the unusual occurrence of multiple putative deacetylases in the Bacillus spp. genomes, especially B. cereus and B. anthracis, we speculate that these enzymes may play different roles in cellular, developmental or environmental biology of B. cereus (e.g. sporulation, spore germination, vegetative growth, host-microbe interactions). Clarification of these questions will require a combined biochemical and genetic (knockout) analysis. Considering the current interest on B. anthracis as potential bioweapon and the difficulties encountered working with such agents, these studies may set a basis for potential drug design applications targeting enzymes involved in biosynthesis/modification of B. anthracis peptidoglycan.

REFERENCES 1. Coutinho, P.M., Henrissat, B. (1999) In Carbohydrate-Active Enzymes: an integrated database

approach. Gilbert, H.J., Davies, G., Henrissat, B., Svensson, B. Eds. Recent advances in carbohydrate bioengineering, The Royal Society of Chemistry, Cambridge, , pp. 3-12

2. Caufrier, F., Martinou, A., Dupont, C., and Bouriotis, V. (2003) Carbohydr. Res. 338, 6870-693

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

8

3. Tsigos, I., Martinou, A., Kafetzopoulos, D., and Bouriotis, V. (2000) Trends Biotechnol. 18, 305-312

4. Kafetzopoulos, D., Martinou, A., and Bouriotis, V. (1993) Proc. Natl. Acad. Sci. USA. 90, 2564-2568

5. Volmer, W., and Tomasz, A. (2000) J. Biol. Chem. 275, 20496-20501 6. Fukushima, T., Yamamoto, H., Atrih, A., Foster, S. J., and Sekiguchi, J. (2002) J. Bacteriol.

184, 6007-6015 7. Gilmore, M. E., Bandyopadhyay, D., Dean, A. M., Linnstaedt, S. D., and Popham, D.L. (2004)

J. Bacteriol. 186, 80-89 8. Blair, D. E., Van Aalten, D. M. F. (2004) FEBS Letters. 570, 13-19 9. Read, T., Peterson, S., Tourasse, N., Baillie, L., Paulsen, I., Nelson, K., Tettelin, H., Fouts, D.,

Eisen, J., Gill, S., Holtzapple, E., Okstad, O., Helgason, E., Rilstone, J., Wu, M., Kolonay, J., Beanan, M., Dodson, R., Brinkac, L., Gwinn, M., DeBoy, R., Madupu, R., Daugherty, S., Durkin, A., Haft, D., Nelson, W., Peterson, J., Pop, M., Khouri, H., Radune, D., Benton, J., Mahamoud, Y., Jiang, L., Hance, I., Weidman, J., Berry, K., Plaut, R., Wolf, A., Watkins, K., Nierman, W., Hazen, A., Cline, R., Redmond, C., Thwaite, J., White, O., Salzberg, S., Thomason, B., Friedlander, A., Koehler, T., Hanna, P., Kolsto, A.-B., and Fraser, C. (2003) Nature 423, 81-86

10. Ivanova, N., Sorokin, A., Anderson, I., Galleron, N., Candelon, B., Kapatral, V., Bhattacharyya, A., Reznik, G., Mikhailova, N., Lapidus, A., Chu, L., Mazur, M., Goltsman, E., Larsen, N., D'Souza, M., Walunas, T., Grechkin, Y., Pusch, G., Haselkorn, R., Fonstein, M., Ehrlich, S. D., Overbeek, R., and Kyrpides, N. (2003) Nature. 423, 87-91

11. Atrih, A., Bacher, G., Allmaier, G., Williamson, M. P., and Foster, S. J. (1999) J. Bacteriol. 181, 3956-3996

12. Araki, Y., Oba, S., Araki, S., and Ito, E. (1980) J. Biochem. 88, 469-479 13. Bergmeyer, H. U. (1974) Methods Enzym. Anal. 1, 112-117 14. Antignac A., Rousselle Jean-Claude., Namane, A., Labigne, A., Taha, M. K., and Boneca I. G.

(2003) J. Biol. Chem. 278, 31521-31528 15. Glauner B. (1988) Anal. Biochem.172, 451-464 16. Sørbotten, A., Horn, S. J., Eijsink, V.G.H., and Vårum K.M. (2005) FEBS Journal. 272, 538-

549 17. Laemmli, U. K. (1970) Nature 227, 680-685 18. Costa K., Bacher, G., Allmaier,G., Dominguez-Bello, M. G., Engstrand, L., Falk, P., de Pedro,

M. A., and Garcia-del Portillo, F. (1999) J. Bacteriol. 181, 3710-2715 19. Ishiguro, K., Yoshie, N., Sakurai, M., and Inoue, Y. (1992) Carbohydr. Res. 237, 333-338 20. Vårum, K.M., Anthonsen, M.W., Grasdalen, H., and Smidsrød, O. (1991) Carbohydr. Res. 211,

17-23 21. Severin, A., Tabei, K., and Tomasz, A. (2004) Microb. Drug. Resist. 10, 77-82 22. Overbeek, R., Larsen, N., Walunas, T., D'Souza, M., Pusch, G., Selkov, E. Jr., Liolios, K.,

Joukov, V., Kaznadzey, D., Anderson, I., Bhattacharyya, A., Burd, H., Gardner, W., Hanke, P., Kapatral, V., Mikhailova, N., Vasieva, O., Osterman, A., Vonstein, V., Fonstein, M., Ivanova, N., and Kyrpides, N. (2003) Nucleic Acids Res. 31, 164-171

FOOTNOTES

*The authors work was funded by the 5th Framework Program CARAPAX of the European Union. I.G. Boneca is an INSERM Research Associate and was supported by an ACI Microbiologie (INSERM MIC 0321) Grant. We thank Dr. N. Kyrpides, (Integrated Genomics, Chicago, USA), for providing the B. cereus ATCC 14579 clones. We also thank Prof. N. Panopoulos for critical reading of the manuscript. 1The abbreviations used are: GlcNAc, N-acetylglucosamine; GlcNAc2, diacetylchitobiose; GlcNAc3, tri-N-acetylchitotriose; GlcNAc4, tetra-N-acetylchitotetraose; GlcNAc5, penta-N-acetylchitopentaose; GlcNAc6, hexa-N-acetylchitohexaose; PCR, polymerase chain reaction; IPTG, isopropyl-β-D-

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

9

thiogalactoside; Mes, 4-morpholineethanesulfonic acid; GMDP, N-acetyl-D-glucosamine-(ß-1,4)-N-acetylmuramyl-L-alanine-D isoglutamine; HPLC, high pressure liquid chromatography; MALDI, matrix-assisted laser desorption ionization; TOF, time of flight; PSD, post source decay; SDS-PAGE, SDS-polyacrylamide gel electrophoresis; MurNAc, N-acetylmuramic acid.

FIGURE LEGENDS Fig. 1. Multiple sequence alignment of characterized members of the CE4 family enzymes with the two putative polysaccharide deacetylases BC1960 and BC3618 from B. cereus. PGDA, peptidoglycan GlcNAc deacetylase from S. pneumoniae (NP_358926); PDAA, spore cortex peptidoglycan deacetylase from B. subtilis (NP_388679); CDA_1, Chitin deacetylase from S. cerevisiae (NP_013411); AXEA, acetyl-xylan esterase from S. lividans (Q54413); NodB, chitooligosaccharide deacetylase from S. meliloti (Q52477). Only the putative catalytic sites are aligned. The black regions indicate identical residues in all sequences. Shading indicates amino acids identical in at least six of the compared sequences. Fig. 2. The orthologous open reading frames from B. cereus and B. anthracis. Values in parenthesis refer to the number of amino acids in the respective open reading frame. The percent identity and similarity refer to the overlapping regions of the respective orthologues. The bolded entries are those identified with enzymatic activity in this study. Possible function has been assigned to these enzymes by ERGO-light database (http://www.ergo-light.com/ERGO) on the basis of sequence comparisons, pattern-based analysis, phylogenetic clusters, networked cellular pathways and chromosomal neighbourhoods of functionally related genes (10, 22). CDA, chitooligosaccharide deacetylase; PD, polysaccharide deacetylase; PDA, peptidoglycan deacetylase. Fig. 3. SDS-PAGE of the purified polysaccharide deacetylases BC3618 (A) and BC1960 (B).

A) Lane 1, molecular weight markers; Lane 2, crude extract (50 µg); Lane 3, 30mM imidazole Ni-NTA eluate (50 µg); Lane 4, 250 mM imidazole Ni-NTA eluate (60 µg).

B) Lane 1, molecular weight markers; Lane 2, 250 mM imidazole Ni-NTA eluate (50 µg). Samples were electrophoresed on a 12% polyacrylamide gel under denaturing and reducing conditions. Protein bands were visualized by staining with Coomassie Brilliant Blue R. Fig. 4. HPLC, MALDI-TOF and MALDI-PSD analysis of N-acetyl-D-glucosamine-(ß-1,4)-N-acetylmuramyl-L-alanine-D-isoglutamine (GMDP ). GMDP (control) was treated with BC1960 or BC3618 and the resulting muropeptides (peak 1 and peak 2) were separated by HPLC and analyzed by MALDI-TOF (A). MALDI-PSD analysis was performed following HPLC purification and desalting the muropeptide corresponding to peak 1 (B). b and a ions correspond to fragments released from C-terminal end and y ions correspond to fragments released from N-terminal end. Fig. 5. RP-HPLC muropeptide elution patterns of peptidoglycan from H. pylori (NCTC11367) treated with BC3618 or BC1960. Peptidoglycan from H. pylori (control) was initially incubated with either BC1960 or BC3618, subsequently digested with mutanolysin and subjected to HPLC analysis as described in Experimental Procedures. Fig. 6. 1H-NMR analysis of reaction products resulting from incubation of GlcNAc2 (A) GlcNAc3 (B) GlcNAc4 (C) with either BC3618 or BC1960. GlcNAc2-4 (2-4 mg) were incubated with 0.5 mg of each enzyme in 0.5 ml buffer (25 mM MES-NaOH buffer pH 6.0 for BC1960; 50 mM Tris-HCl buffer pH 8.0 for BC3618) at 37 0C for between 18 to 48 hours. The reaction products were analyzed by 1H-NMR as previously described (16).

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

10

Fig. 7. The effect of chemical acetylation and enzymatic deacetylation on the susceptibility of the B. cereus peptidoglycan to lysozyme digestion. Chemically acetylated peptidoglycan, either non modified (squares) or treated with BC1960 (triangles) and BC3618 (circles) was incubated with 80 µg of lysozyme for designated times. (x), indicates control (non acetylated) peptidoglycan from vegetative B. cereus. Susceptibility to lysozyme digestion was determined by measuring the absorbance at 600 nm.

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

11

Figure 1

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

12

Figure 2

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

13

Figure 3

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

14

Table 1 Substrate specificity and kinetic properties of BC1960 and BC3618 toward GlcNAc2-6

Enzyme assay was performed as described under Experimental Procedures.

BC1960

BC3618

Substrate Relative activitya

(%)

Km

(mM)

Vmax

(µmol/min/mg)

Kcatb/Km

(mM-1 s-1)

Relative activity

(%)

Km

(mM)

Vmax

(µmol/min/mg)

Kcatb/Km

(mM-1 s-1)

GlcNAc 0 - - - 0 - - -

GlcNAc2 49.3 4.1 24.3 3 4.8 3.9 8.7 0.7

GlcNAc3 69.2 2.46 30.6 6.3 29.8 2.2 50 10.4

GlcNAc4 100 1.18 47.3 24 100 1.5 97.3 29.9

GlcNAc5 86.4 0.37 15.5 21.3 44.6 0.5 24.2 22.3

GlcNAc6 87 0.3 13.4 22.6 48.1 0.45 25 25.6

a The concentration of substrates was adjusted with respect to their content in N-acetyl residues. b The Kcat values derived from the expression Vmax/molecular mass of the protein

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

15

Figure 4

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

16

Figure 5

0 5 10 15 20 25 30 35 40 45

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

Η. pylori

BC3618

BC1960

1

2* 4* 8*9*

5a*5*

6*

4a*

7*

23

4

5

3 4

5

2 45

2* 4*8* 9*5a*

5*

4a*

Retention time (min)

Vol

ts

2

0 5 10 15 20 25 30 35 40 45

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

Η. pylori

BC3618

BC1960

1

2* 4* 8*9*

5a*5*

6*

4a*

7*

23

4

5

3 4

5

2 45

2* 4*8* 9*5a*

5*

4a*

Retention time (min)

Vol

ts

2

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

17

Table 2 Identities and quantification of muropeptides from H. pylori peptidoglycan

Individual HPLC peaks were purified and analyzed by MALDI-TOF.

Peak Nob MALDI-TOF Proposed structurec Native

Mol% BC1960 Mol%

BC3618 Mol%

1 993.3622 GlcNAc-MurNAc-Ala-Glu-mDap 1.60 0.00 0.00

2 964.4113 GlcNAc-MurNAc-Ala-Glu-mDap-Ala 8.88 5.83 n.d

3 1021.4249 GlcNAc-MurNAc -Ala-Glu-mDap-ala-Gly 5.15 0.00 4.99

4 721.2785 GlcNAc-MurNAc -Ala-Glu 17.77 9.45 13.6

5 1035.4717 GlcNAc-MurNAc -Ala-Glu-mDap-Ala-Ala 66.60 15.90 47.55

2* 922.3962 GlcN-MurNAc -Ala-Glu-mDap-Ala 0.00 6.58 5.03

4* 4a*

679.2679 679.2822 GlcN-MurNAc -Ala-Glu 0.00 10.33

8.88 5.89 4.04

5* 5a*

993.4518 993.4518 GlcN-MurNAc -Ala-Glu-mDap-Ala-Ala 0.00 26.46

16.57 13.84 5.05

6* 902.3769 GlcN -[(an)MurNAc]-Ala-Glu-mDap-Ala 0.00 n.a. n.a.

7* 1071.4541 n.d. 0.00 n.a. n.a.

8* n.d n.d. 0.00 n.a. n.a.

9* 973.4146 GlcN -[(an)MurMAc]-Ala-Glu-mDap-Ala-Ala 0.00 n.a. n.a.

bNumbers correspond with peaks in Fig 5. cMuropeptides best fitting the relative molecular masses measured by MALDI-TOF. De-N-acetylation of the N-acetylglucosamine residue has been confirmed by MALDI-PSD analysis of muropeptides. Abbreviations: GlcNAc, N-acetylglucosamine; MurNAc, N-acetylmuramic acid; GlcN, glucosamine; (an)MurNAc, (1→6)anhydro-N-acetylmuramic acid; mDap, meso-diaminopimelic acid; n.d, not determined; n.a, not applicable.

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

18

Table 3 Identities and quantification of muropeptides from B. subtilis peptidoglycan

Peptidoglycan from vegetative B. subtilis (HR, wild type) was initially incubated with either BC1960 or BC3618, subsequently digested with Cellosyl and subjected to RP-HPLC analysis as described in Experimental Procedures. Peak Noa

Muropeptide identitya Native Mol%

BC1960 Mol%

BC3618 Mol%

3 Disaccharide tripeptide with 1 amidation 26 13.4 14.8

4 Disaccharide tripeptide 1 amidation missing an acetyl group

4.1 8.9 6.0

18 Disaccharide tripeptide disaccharide tetrapeptide with 1

amidation and missing an acetyl group

3.9 4.2 4.4

21 Disaccharide tripeptide disaccharide tetrapeptide with 2 amidations

31.8 20.2 21.9

22 Disaccharide tripeptide disaccharide tetrapeptide with 2 amidations and missing an acetyl group

12.3 16.3 12.2

23 Disaccharide tripeptide disaccharide tetrapeptide with 2 amidations and missing an acetyl group

1.5 4.3 2.3

31 Disaccharide tripeptide disaccharide tetrapeptide disaccharide tetrapeptide with 3 amidations

3.7 2.6 2.7

32 Disaccharide tripeptide disaccharide tetrapeptide disaccharide tetrapeptide with 3 amidations and missing an acetyl group

1.9 2.6 1.7

a Peak number and muropeptide identity have been assigned according to Atrih et al (11).

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

19

Figure 6

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

20

Figure 7

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from

Emma Hayhurst, Simon J. Foster, Kjell M. Vårum and Vassilis BouriotisEmmanuel Psylinakis, Ivo G. Boneca, Konstantinos Mavromatis, Alexandra Deli,

conserved proteins in bacillus anthracisPeptidoglycan N-acetylglucosamine deacetylases from bacillus cereus, highly

published online June 16, 2005J. Biol. Chem. 

  10.1074/jbc.M407426200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

by guest on January 11, 2021http://w

ww

.jbc.org/D

ownloaded from