Báo cáo Y học: GNA33 from Neisseria meningitidis serogroup B encodes a membrane-bound lytic transglycosylase (MltA) pdf

GNA33 from
Neisseria meningitidis
serogroup B encodes
a membrane-bound lytic transglycosylase (MltA)
Gary T. Jennings
1
*, Silvana Savino
1
*, Elisa Marchetti
1
, Beatrice Arico
`
1
, Thomas Kast
2
, Lucia Baldi
1
,
Astrid Ursinus
2
, Joachim-Volker Ho¨ ltje
2
, Robert A. Nicholas
3
, Rino Rappuoli
1
and Guido Grandi
1
1
I.R.I.S., Chiron S.p.A., Siena, Italy;
2

Max Planck Institute fur Entwicklungsbiologie, Abteilung Biochemie, Tubingen, Germany;
3
Department of Pharmacology, University of North Carolina at Chapel Hill, NC, USA
In a previous study, we used the genome of serogroup B
Meningococcus to identify novel vaccine candidates. One of
these molecules, GNA33, is well conserved among Men-
ingococcus B strains, other Meningococcus serogroups and
Gonococcus and induces bactericidal antibodies as a result of
being a mimetic antigen of the PorA epitope P1.2. GNA33
encodes a 48-kDa lipoprotein that is 34.5% identical with
membrane-bound lytic transglycosylase A (MltA) from
Escherichia coli. In this study, we expressed GNA33, i.e.
Meningococcus MltA, as a lipoprotein in E. coli. The lipo-
protein nature of recombinant MltA was demonstrated by
incorporation of [
3
H]palmitate. MltA lipoprotein was
purified to homogeneity from E. coli membranes by cation-
exchange chromatography. Muramidase activity was con-
firmed when MltA was shown to degrade insoluble murein
sacculi and unsubstituted glycan strands. HPLC analysis
demonstrated the formation of 1,6-anhydrodisaccharide
tripeptide and tetrapeptide reaction products, confirming
that the protein is a lytic transglycosylase. Optimal
muramidase activity was observed at pH 5.5 and 37 °Cand
enhanced by Mg
2+
,Mn
2+
and Ca

2+
. The addition of Ni
2+
and EDTA had no significant effect on activity, whereas
Zn
2+
inhibited activity. Triton X-100 stimulated activity
5.1-fold. Affinity chromatography indicated that MltA
interacts with penicillin-binding protein 2 from Meningo-
coccus B, and, like MltA from E. coli, may form part of a
multienzyme complex.
Neisseria meningitidis is a Gram-negative, capsulated
b-proteobacterium capable of causing severe meningitis
and septicemia with a fatality rate of  10% [2]. The
complete 2 272 351-bp genomic sequence of Meningococcus
serogroup B (strain MC58) has been determined and used
by us to identify novel vaccine candidates against this
pathogenic organism [1,2]. We amplified, cloned and
expressed in Escherichia coli selected ORFs encoding pro-
teins with predicted surface exposure. Recombinant pro-
teins were purified, immunized in mice, and the resultant
sera analysed by FACS, ELISA, and bactericidal assay.
GNA33 was positive in all three analyses and highly
conserved (99.2± 0.7%) among 22 strains of Meningococ-
cus B, nine strains from Meningococcus serogroups A, C, Y,
X, Z, W135, and 95.8% conserved in Neisseria gonorrhoeae
[1]. Further study revealed that GNA33 elicits protective
antibodies to meningococci by mimicking a surface-exposed
epitope on loop 4 of porin A in strains with serosubtype
P1.2 [3].

The ORF of GNA33 encodes a protein 441 amino acids
in length with an N-terminal 20-amino-acid lipopolypep-
tide signal sequence (LPSS) with a consensus lipoprotein-
processing site, LAAC [4]. Sequence comparison showed
that GNA33 is 34.5% identical and 41.3% homologous
with the 38-kDa membrane-bound lytic transglycosylase A
(MltA) from E. coli (Fig. 1). In E. coli, four additional
exo-specific lytic transglycoylases (MltB, MltC, MltD, and
Slt70) have been identified and/or characterized [5–8].
These lytic transglycosylases exhibit no significant
sequence homology with each other. With the exception
of Slt70 (soluble lytic transglycosylase), they are all
lipoproteins that attach to the outer membrane [7–10].
Homologues of all these lytic transglycosylases have been
identified in Meningococcus B [2], which, like their E. coli
counterparts, also exhibit little sequence conservation with
each other.
Lytic transglycosylases are a unique class of lysozyme-like
enzymes that catalyze cleavage of the b-1,4-glycosidic bond
between N-acetylmuramic acid (MurNAc) and N-acetyl
glucosamine (GlcNAc). However, unlike lysozyme where
the glycosyl moiety is transferred to H
2
O, cleavage by lytic
transglycosylases is followed by an intramolecular transgly-
cosylation [10]. In this reaction, the glycosidic linkage
between the muramyl and glucosaminyl residues is trans-
ferred to the C6 position of the muramyl residue to form
terminal 1,6-anhydromuramic acid-containing products
[10]. By virtue of their ability to cleave the polysaccharide

backbone of the peptidoglycan layer, lytic transglycosylases
are thought to play a role in synthesis and degradation of
the murein sacculus. It has been proposed that lytic
transglycosylases play important roles in cellular elongation,
Correspondence to G. T. Jennings, Cytos Biotechnology AG,
Wagistrasse 25, CH-8952 Zu
¨
rich-Schlieren, Switzerland.
Fax: + 41 1 733 4659, Tel.: + 41 1 733 4642,
E-mail:
Abbreviations: GNA, genome-derived Neisseria antigen; LPSS, lipo-
polypeptide signal sequence; MipA, MltA-interacting protein; MltA,
membrane-bound lytic transglycosylase A; PBP, penicillin-binding
protein.
*Note: these authors contributed equally to this work.
(Received 29 March 2002, revised 14 June 2002,
accepted 20 June 2002)
Eur. J. Biochem. 269, 3722–3731 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03064.x
septation, recycling of muropeptides, and pore formation
[7,10,11].
Current models of cell wall synthesis in Gram-negative
bacteria predict the necessity for murein synthases and
lytic enzymes to interact in a co-ordinated and controlled
manner [10]. Indeed, interactions between lytic transgly-
cosylases (MltA, MltB and Slt70), bifunctional transgly-
cosylase-transpeptidases (PBP1A, PBP1B, PBP1C),
transpeptidases (PBP2, PBP3), and endopeptidases
(PBP4 and PBP7) of E. coli have been reported [12,13].
In particular, affinity chromatography and/or surface
plasmon resonance have shown interactions between

MltA, PBP1B, PBP1C, PBP2, PBP3 and a newly identi-
fied scaffolding protein, MipA [14]. It is thought that,
through such interactions, the enzymes required for cell
wall metabolism associate and form a multienzyme
complex [10,14]. An enzyme complex would not only
provide a means for regulating peptidoglycan synthesis but
would also provide a way to control the potentially
autolytic activity of proteins such as MltA. To date, no
evidence of these associations in Neisseria species has been
reported.
In this study we cloned GNA33 (MltA) from Meningo-
coccus serogroup B. The recombinant lipoprotein was
expressed in E. coli, purified, and assayed for its muram-
idase and lytic transglycosylase activity. In addition, we used
affinity chromatography to investigate the hypothesis that
MltA associates with other enzymes involved in peptido-
glycan metabolism and thus may be part of a multienzyme
complex.
EXPERIMENTAL PROCEDURES
Vector construction
Three versions of meningococcal mltA were amplified by
PCR and cloned into the expression vector pET21b+
(Novagen) via 5¢ NdeIand3¢ XhoI restriction sites. These
included a full-length form incorporating its endogenous
20-amino-acid LPSS, a form containing a 19-amino-acid
LPSS from an unrelated Meningococcus B lipoprotein,
GNA1946 [1], and a truncated form lacking any leader
sequence (Fig. 1).
Full-length mltA was amplified using a forward primer
containing an NdeI restriction site (5¢-CGCGGATCCCA

TATGAAAAAATACCTATTCCGC-3¢) incorporating
the ATG start codon. The reverse primer (5¢-CCCGCTC
GAGTTACGGGCGGTATTCGG-3¢) contained a XhoI
restriction site and was used for all three constructs. The
construct containing the GNA1946 LPSS was made using a
forward primer (5¢-GGGAATTCCATATGAAAACCTT
CTTCAAAACCCTTTCCGCCGCCGC
GCTAGCGCT
CATCCTCGCCGCCTGCCAAAGCAAGAGCATC-3¢)
spanning the entire leader of GNA1946 and containing 18
nucleotides overlapping the mltA sequence. A conservative
double nucleotide substitution (underlined) was made in a
region of the primer encoding the GNA1946 LPSS. This
substitution introduced an NheI restriction and was
designed to allow the GNA1946 LPSS to be ligated into
any of the meningococcus genes that we have previously
expressed in pET-21b [1]. This restriction site was not used
Fig. 1. Amino acid sequence of MltA from
N. meningitidis serogroup B: comparison with
MltA from E. coli. The amino-acid sequence
of MltA from Meningococcus B (strain 2996)
(NmMltA) was compared with MltA from
E. coli (EcMltA) using the
GAP
program
included in the Genetics Computer Group
(GCG) Wisconsin Package version 10.0. The
20-amino-acid LPSS is underlined. The LPSS
was identified using the program
PSORT

avail-
able at . The 19-amino-
acid LPSS from the Meningococcus Bgene
GNA1946 (GNA1946L), was used to replace
the MltA leader peptide and is shown above
the meningococcal sequence. Amino acids are
identified by the standard single letter code.
Ó FEBS 2002 MltA from N. meningitidis serogroup B (Eur. J. Biochem. 269) 3723
in this study. The truncated gene lacking the 20-amino-acid
leader peptide was amplifed using the forward primer,
5¢-CGCGGATCCCATATGCAAAGCAAGAGCATCC
AAA-3¢.
PCR was performed in a reaction volume of 100 lL
comprising 10 m
M
Tris/HCl (pH 8.3), 50 m
M
NaCl,
1.5 m
M
MgCl
2
,0.8 m
M
dNTPs, 40 l
M
each oligonucleotide
primer, and 2.5 U TaqI DNA polymerase (PerkinElmer,
Boston, MA, USA). Template DNA for the reaction was
200 ng genomic DNA from Neisseria meningitidis B 2996.

The primary denaturation step was performed at 94 °Cfor
3 min and the remainder of the first five cycles with
denaturation, annealing and polymerization conditions of
94 °Cfor40s,52°C for 40 s and 72 °Cfor1min,
respectively. The annealing temperature was increased to
65 °C for the next 30 cycles, and a final 7 min extension at
72 °C completed the reaction. PCR products were purified
using the Qiagen Gel Extraction Kit. Ligations and
transformations into E. coli DH5 were performed as
described by Sambrook et al. [15]. After selection, amplifi-
cation and purification, the plasmids were used to transform
E. coli BL21(DE3) (Novagen, Madison, WI, USA). The
genomic sequence of Meningococcus B is known for the
strain MC58 [2]. The nucleotide sequence of mltA from
strain 2996 has 17 nucleotide substitutions (of which 16 are
silent) with respect to mltA from strain MC58. Only one of
these base changes results in an amino-acid substitution,
Ser312 to Ala.
Expression and purification of recombinant MltA
E. coli BL21(DE3) cells harboring the three versions of
pET21b-MltA (see above) were grown at 30 °CinLuria–
Bertani medium containing 100 lgÆmL
)1
ampicillin until
the D
550
reached 0.6–0.8. Isopropyl thio-b-
D
-galactoside
was added to a final concentration of 1.0 m

M
,andthe
culture shaken for an additional 3 h. Cells were collected by
centrifugation at 8000 g for 15 min at 4 °C. All subsequent
procedures were performed at 4 °C.
For purification of lipidated MltA, cells were resuspend-
ed in 25 mL 50 m
M
phosphate/300 m
M
NaCl, pH 8.0,
containing complete protease inhibitor (Roche, Basel,
Switzerland). Bacteria were disrupted by osmotic shock
with two or three passages through a French Press (SLM
Aminco). Unbroken cells were removed by centrifugation
at 5000 g for 15 min, and membranes sedimented by
centrifugation at 100 000 g for 45 min. The pellet was
resuspended in 20 m
M
Tris/HCl (pH 8.0)/1.0
M
NaCl
containing complete protease inhibitor, and the suspension
mixed for 2 h. After centrifugation at 100 000 g for
45 min, the pellet was resuspended in 20 m
M
Tris/HCl
(pH 8.0) containing 1.0
M
NaCl, 5.0 mgÆmL

)1
Chaps, 10%
(v/v) glycerol and complete protease inhibitor. The solution
was stirred overnight, centrifuged at 100 000 g for 45 min,
and the supernatant dialysed for 6 h against 20 m
M
Bicine
(pH 8.5)/120 m
M
NaCl/5.0 mgÆmL
)1
Chaps/10% (v/v)
glycerol. The dialysate was cleared by centrifugation at
13 000 g for 20 min and applied to a Mono S FPLC ion-
exchange column (Pharmacia, Uppsala, Sweden) at a flow
rate of 0.5 mLÆmin
)1
. Elution was performed using a
stepwise NaCl gradient.
The protein was also expressed and purified in a form
lacking the LPSS. After expression and harvesting, cells
were resuspended in 20 m
M
Bicine (pH 8.5)/20 m
M
NaCl/
10% (v/v) glycerol containing complete protease inhibitor
and disrupted with a Branson Sonifier 450. The sonicate was
centrifuged at 8000 g for 30 min to remove unbroken cells,
and MltA was precipitated from the supernatant by the

addition of saturated (NH
4
)
2
SO
4
solution. MltA was
precipitated between 35% and 70% saturation and was
collected by centrifugation at 8000 g for 30 min. The pellet
was dissolved in 20 m
M
Bicine (pH 8.5)/20 m
M
NaCl/10%
(v/v) glycerol and dialysed against this buffer overnight. The
dialysate was centrifuged at 13 000 g for 20 min, and the
supernatant was applied to an FPLC Mono S ion-exchange
column at a flow rate of 0.5 mLÆmin
)1
. The protein was
eluted from the column with a stepwise NaCl gradient.
Purifications were analysed by SDS/PAGE [16], and protein
concentration determined by the Bradford method. West-
ern-blot analysis was performed using polyclonal antisera as
described previously [1].
Palmitate labelling
Palmitate incorporation by recombinant MltA was con-
firmed as described by Kraft et al. [17]. Briefly, E. coli
BL21(DE3) harbouring one of the three pET21b-MltA
constructs were grown at 30 °C in Luria–Bertani medium

containing 100 lgÆmL
)1
ampicillin and 5 lCiÆmL
)1
[
3
H]palmitate (Amersham) until the D
550nm
reached
0.4–0.8. Expression of recombinant protein was induced
for 1 h by the addition of isopropyl b-
D
-thiogalactoside
(final concentration 1 m
M
), and the bacteria harvested by
centrifugation at 3000 g for 15 min. Cells were washed twice
with cold NaCl/P
i
, suspended in 20 m
M
Tris/HCl (pH 8.0)/
1m
M
EDTA/1.0% (w/v) SDS, lysed by boiling for 10 min,
and centrifuged for 10 min at 13 000 g. Cold acetone was
added to the supernatant, and, after 1 h at )20 °C, protein
was collected at 13 000 g for 10 min. Protein was resus-
pended in 1.0% (w/v) SDS, boiled with SDS/PAGE sample
buffer, and subjected to SDS/PAGE using a 12.5%

separating gel. Gels were fixed for 1 h in 10% (v/v) acetic
acid, and soaked for 30 min in Amplify solution (Amer-
sham). The gel was vacuum-dried under heat and exposed
to Hyperfilm (Kodak) overnight at )80 °C.
Assay for muramidase activity
Purified, recombinant MltAs expressed with the GNA1946
LPSS or without an LPSS were assessed for their ability to
degrade insoluble murein sacculi into soluble muropeptides
by the method of Ursinus & Holtje [18]. Murein lysis
activity was determined using peptidoglycan radiolabelled
with meso-2,6-diamino-3,4,5-[
3
H]pimelic acid as substrate.
Enzyme (3–10 lg total) was incubated for 45 min at 37 °C
in a total volume of 100 lL comprising 10 m
M
Tris/maleate
(pH 5.5), 10 m
M
MgCl
2
, 0.2% (v/v) Triton X-100
and [
3
H]diaminopimelic acid-labelled murein sacculi
( 10 000 c.p.m.). The assay mixture was placed on ice
for 15 min with 100 lL1.0%(w/v)N-cetyl-N,N,N-trime-
thylammonium bromide, and the precipitated material
separated by centrifugation at 10 000 g. The radioactivity
in the supernatant was measured by liquid-scintillation

counting. The E. coli lytic transglycosylase Slt70 was used as
a positive control for the assay, and the negative control
comprised the above assay solution without enzyme.
3724 G. T. Jennings et al.(Eur. J. Biochem. 269) Ó FEBS 2002
Assay for lysis of poly(MurNAc-GlcNAc) glycan strands
The ability of MltA to utilize purified glycan strands as
substrate was determined by the method described by
Ursinus & Holtje [18]. Poly(MurNAc-GlcNAc)
n>30
,
labelled with N-acetyl-
D
-1-[
3
H]glucosamine, was incubated
with 3 lgMltAin10m
M
Tris/maleate (pH 5.5)/10 m
M
MgCl
2
/0.2% (v/v) Triton X-100 for 30 min at 37 °C. The
reaction was stopped by boiling for 5 min, and the pH of
the sample adjusted to 3.5 by addition of 10 lL 20% (v/v)
phosphoric acid. The components of the assay were then
separated by RP-HPLC on a Nucleosil 300 C
18
column as
described by Harz et al. [19]. The E. coli lytic transglycos-
ylase MltA was used as a positive control in the assay. A

negative control was performed in the absence of enzyme.
Analysis of reaction products
The nature of the reaction products resulting from the
digestion of unlabelled E. coli murein sacculus were deter-
mined by RP-HPLC as described by Glauner [20]. Murein
sacculi digested with the muramidase Cellosyl were used to
calibrate and standardize the Hypersil ODS column.
Gel filtration
The molecular masses of the recombinant proteins were
estimated using either FPLC Superose 12 (H/R 10/30) or
Superdex 75 gel-filtration columns (Pharmacia). The buf-
fers were 20 m
M
Bicine (pH 8.5) with and without
5.0 mgÆmL
)1
Chaps, respectively. In addition, each buffer
contained 150–200 m
M
NaCl and 10% (v/v) glycerol.
Proteins were dialysed against the appropriate buffer and
applied in a volume of 200 lL. Gel filtration was performed
with a flow rate of 0.5–2.0 mLÆmin
)1
and the eluate
monitored at 280 nm. Fractions were collected and
analysed by SDS/PAGE. Blue Dextran 2000 and the
molecular-mass standards ribonuclease A, chymotryp-
sin A, ovalbumin A, and BSA (Pharmacia) were used to
calibrate the columns. The molecular mass of the sample

was estimated from a calibration curve of K
av
vs. log
(molecular mass) of the standards.
Preparation of membrane extracts for affinity
chromatography
A detergent-solubilized membrane extract was prepared
from an acapsulated N. meningitidis strain, M7. An over-
night culture of strain M7 was inoculated into 2 L Muller-
Hinton broth containing 0.25% (w/v) glucose, and grown at
37 °C in an atmosphere of 5.0% CO
2
. When the D
550
reached 0.6, the culture was cooled on ice and harvested by
centrifugation at 8000 g; all the following steps were
performed at 4 °C. The pellet was resuspended in 10 m
M
Tris/HCl (pH 8.0) containing complete protease inhibitor
and DNase (10 lgÆmL
)1
), and the cells were disrupted with
a French Press. Membranes were spun down at 100 000 g
for 45 min and resuspended in 10 m
M
Tris/maleate
(pH 6.8) containing 2.0% (v/v) Triton X-100, 10 m
M
MgCl
2

, 150 m
M
NaCl and EDTA-free complete protease
(buffer I). After stirring overnight, membrane debris was
removed by centrifugation (100 000 g for 45 min), and the
supernatant containing solubilized protein stored at )20 °C.
Affinity chromatography
Purified leaderless MltA (10 mgÆmL
)1
gel) was coupled to
CNBr-activated Sepharose 4B (Pharmacia) according to
the manufacturer’s protocol. CNBr-activated Sepharose 4B
prepared without protein and where the functional groups
were neutralized with Tris was used as a control for
nonspecific binding to the resin. Disposable columns
containing either control or MltA-coupled resin were
prepared and equilibrated with 20 col. vol. buffer I. Solu-
bilized membrane extract was applied to both columns at a
flow rate of 0.25 mLÆmin
)1
, then washed with 5 · 1.0 mL
buffer I. Retained proteins were eluted by increasing the
NaCl concentration in a stepwise fashion. Salt concentra-
tions of 300 m
M
, 600 m
M
and 1.0
M
in buffer I were applied

in 5 · 1.0 mL aliquots, and the eluates retained for analysis
by SDS/PAGE, penicillin-binding assay, and Western blot.
Penicillin-binding assay
Penicillin-binding proteins (PBPs) were identified using the
125
I-labelled Bolton–Hunter derivative of ampicillin pre-
pared as described previously [21]. Briefly, 4 lL(2.4lg
total) of the labelled ampicillin derivative was incubated for
30 min at 37 °Cwith40lL of the fractions eluted from
control and MltA-coupled affinity columns. The reaction
was stopped by the addition of 4 lL penicillin G
(60 mgÆmL
)1
), and the reaction complexes separated by
SDS/PAGE and visualized by autoradiography.
Preparation of antisera to PBP2
Recombinant PBP2 from N. gonorrhoeae was purified as a
soluble, active form. PBP2 was expressed in the cytoplasm
of E. coli as a fusion protein to maltose-binding protein
(MBP) with a His
6
tag at its N-terminus. Codons 44–581,
which encode the entire periplasmic domain of PBP2, were
fused in-frame to the C-terminus of MBP via an interven-
ing tobacco etch virus (TEV) protease site. The fusion
protein was overexpressed in E. coli, purified on a Ni
2+
/
nitrilotriacetate column, and cleaved with His
6

–TEV
protease (fusion protein/TEV protease, 20 : 1, w/w) in
50 m
M
Tris/HCl (pH 8.0)/500 m
M
NaCl/10% glycerol.
After digestion, PBP2 was again purified by metal chelate
affinity chromatography to remove uncut fusion protein,
His
6
–MBP and the protease. PBP2 was not eluted in the
flow through, which contained unrelated contaminant
proteins, but was eluted from the column with 10 m
M
imidazole. Purified PBP2 was judged to be at least 95%
pure by SDS/PAGE. The protein was concentrated to
6mgÆmL
)1
and stored at )80 °C. Purified PBP2 was used
to immunize mice, and antisera were collected as described
byPizzaet al.[1].
Western blot
Fractions eluted from the MltA-coupled affinity column
were separated by discontinuous SDSPAGE using a 12.5%
separating gel [15]. Proteins were electroblotted onto a
nitrocellulose membrane and probed with antisera to PBP2
diluted 1 : 1000. Immunoreactive proteins were detected
using the enhanced chemiluminescent method (Amersham,
Chicago, IL, USA) and fluorography.

Ó FEBS 2002 MltA from N. meningitidis serogroup B (Eur. J. Biochem. 269) 3725
RESULTS
Cloning and expression in
E. coli
Expression of MltA in E. coli was observed when the
gene was cloned with either its own 20-amino-acid LPSS
or the 19-amino-acid LPSS from an unrelated Meningo-
coccus lipoprotein, GNA1946. However, the level of
expression was much lower when the native leader
peptide was used (result not shown). Hence, for purposes
of purification and characterization, we used the clone
incorporating the LPSS from GNA1946. MltA cloned
without a leader peptide was expressed very efficiently
and represented about 20% of total cellular protein as
judged by densitometry. This truncated, soluble form of
the protein was used for affinity chromatography (see
below).
MltA incorporating the LPSS from GNA1946 was
routinely expressed at 30 °C because expression of the
recombinant protein at 37 °C resulted in lysis of host
cells. Lysis at 37 °C was observed within 60 min of
induction of expression and could be prevented by the
addition of 12% (w/v) sucrose and 10 m
M
MgSO
4
.
Overexpression of E. coli MltA also results in formation
of spheroplasts and cell lysis [9]. However, in contrast
with our results, lysis due to overexpression of E. coli

MltA occurs at 30 °C, but not at 37 °C. With E. coli
MltA, this effect is due to the temperature sensitivity of
its muramidase activity, which exhibits maximum activity
at 30 °C and a 93% reduction in activity at 37 °C. It also
has been reported that a 55-fold overexpression of E.coli
lytic transglycosylase MltB resulted in rapid cell lysis at
37 °C [8]. Similar to our observation with Meningococcus
MltA, autolysis induced by overexpression of E. coli
MltB was also prevented by osmotic protection during
growth.
Purification of recombinant proteins
Recombinant MltA lipoprotein was purified from the
membrane fraction of E. coli as described in Experimen-
tal Procedures. Analysis of the purification by SDS/
PAGE showed that MltA lipoprotein was localized in
the membrane fraction (Fig. 2, lane 2). Western-blot
analysis with polyclonal sera raised against MltA failed
to detect MltA in any of the soluble fractions obtained
before Chaps extraction, demonstrating exclusive local-
ization of the lipoprotein to the membrane fraction
(result not shown). After solubilization of MltA with
Chaps, it was necessary to maintain NaCl at a minimum
concentration of 120 m
M
to prevent the lipoprotein from
precipitating. The predicted pI for MltA is 10.5. The
basic nature of the protein enabled FPLC cation-
exchange chromatography to be performed under condi-
tions that allowed almost complete removal of contam-
inating proteins in a single step (Fig. 2, lane 4). Similarly,

this property was exploited to perform a simple two-step
procedure for the purification of the truncated version of
MltA, which involved salting out and cation exchange.
The leaderless form is found exclusively in the cytosolic
fraction of E. coli and was purified to homogeneity as
judged by SDS/PAGE with Coomassie blue staining
(Fig. 2, lane 5).
Molecular mass
The molecular masses of the lipoprotein and truncated
forms of MltA were determined under denaturing condi-
tions by SDS/PAGE (Fig. 2). The two forms of the protein
migrate to the same position in the gel (Fig. 2), and, from a
calibration plot of log mass vs. relative mobility of protein
standards, the masses of both forms of MltA were
calculated to be 44.5 kDa. This is in agreement with the
molecular mass of 45 869 Da predicted from the amino-
acid composition of the protein excluding the first
19-amino-acids of the leader peptide. As the lipoprotein
expressed with its 2138-Da leader sequence migrates to the
same position as leaderless MltA, it is reasonable to
conclude that the signal peptide is cleaved when this clone
is expressed. The presence of detergent in the purification
prevented an accurate estimation of molecular mass for
MltA lipoprotein using molecular exclusion chromatogra-
phy. As truncated MltA lacking its LPSS was purified in the
absence of detergent, we determined the native molecular
mass using this form of the protein (see Experimental
Fig. 2. SDS/polyacrylamide gel showing the purification and molecular
mass of recombinant forms of MltA. Proteins were separated by SDS/
PAGE on a 12.5% separating gel and stained with Coomassie Brilliant

Blue. Lane M, molecular-mass standards; lane 1, bacterial lysate after
expression; lane 2, membrane fraction after 100 000 g centrifugation;
lane 3, soluble fraction after extraction of membrane fraction with
0.5% CHAPS; lane 4, an aliquot from the peak fraction from Mono S
FPLC ion-exchange chromatography; lane 5, truncated MltA
(expressed without the LPSS) after Mono S FPLC ion-exchange
chromatography.
3726 G. T. Jennings et al.(Eur. J. Biochem. 269) Ó FEBS 2002
procedures). Truncated MltA was eluted with a K
av
corresponding to a molecular mass of 31 600 Da. This
value is low compared with that of the denatured protein
suggesting either an interaction with the column or a smaller
than expected Stokes’ radius. Nevertheless the native
molecular mass of the truncated form of MltA is more
indicative of a monomer than a dimer.
Confirmation that MltA is a lipoprotein
To test if recombinant MltA expressed with either its
endogenous LPSS or GNA1946 LPSS was lipidated, the
ability of the proteins to incorporate [
3
H]palmitate was
examined. Proteins extracted from cells grown in the
presence of the radiolabel were examined by SDS/PAGE
and autoradiography (Fig. 3). A labelled band with a
molecular mass of 44 kDa was observed for MltA cloned
with either its own leader or the leader from GNA1946. The
radiolabel was not incorporated when MltA lacking an
LPSS was expressed.
MltA is a muramidase

Both the purified lipoprotein and truncated form of MltA
showed muramidase activity when assayed for their ability
to degrade murein sacculi to soluble muropeptides. How-
ever, the activity observed with the lipoprotein form was
21.6-fold higher than the activity of the truncated form. For
this reason, the lipoprotein was chosen for further kinetic
analyses. The activity of MltA lipoprotein was enhanced
5.1-fold by the addition of 0.2% (v/v) Triton X-100 to the
assay, whereas Triton X-100 had no measurable effect on
the activity of the truncated soluble form of the protein.
Biochemical and kinetic properties of the enzyme
The effect of pH on muramidase activity was determined in
Tris/maleate buffer over the pH range 5.0–8.0. The optimal
pH for the reaction was determined to be 5.5 (data not
shown). The optimum pH for lytic transglycosylase activity
byMltAfromE. coli is 4.5 [18]. Enzyme activity was
measured over the temperature range 18–42 °C. Maximum
activity was observed at 37 °C (data not shown). As we
observed that MltA has 77% of the activity at 30 °Casit
does at 37 °C, the stability of cells expressing Meningococcus
MltA at 30 °C is unlikely to be due solely to a temperature-
dependent decrease in murein lytic activity as previously
described for E. coli MltA (see above).
The effect of ions on muramidase activity was determined
by performing the reaction with a variety of bivalent
cations, at a final concentration of 10 m
M
.Maximum
activity was found with Mg
2+

, which stimulated activity
2.1-fold. Mn
2+
and Ca
2+
stimulated enzyme activity to a
similar extent, whereas Ni
2+
and EDTA had no significant
effect on activity. In contrast, Zn
2+
significantly inhibited
enzyme activity (data not shown).
Initial-rate kinetic analyses were performed with sub-
strate concentrations ranging from 2.6 to 52.0 mgÆL
)1
.An
analysis of the Michaelis–Menten curve (data not shown)
showed that the enzyme exhibits typical first-order and zero-
order kinetics. As the substrate for the reaction is insoluble,
it is not possible to determine the K
m
for the reaction in
molar terms [18]. However, the apparent K
m
of 8.2 mgÆL
)1
determined from a double-reciprocal Lineweaver–Burk plot
is slightly lower than the value (52.6 mgÆL
)1

)obtained
previously for MltA from E. coli [18].
Substrate specificity and reaction product
The ability of MltA to lyse isolated glycan strands
comprising poly(MurNAc-GlcNAc)
n>30
was demonstrated
when we separated 1,6-anhydrodisaccharide subunit reac-
tion products from the oligosaccharide substrate by HPLC
(Fig. 4). The same elution profile was observed when we
assayed E. coli MltA in a control experiment (result not
shown). The use of isolated glycan strands as a substrate
further demonstrates homology with E. coli MltA, which is
also capable of utilizing both murein sacculi and isolated
glycan strands [18,22].
HPLC analysis of the digestion products after incubation
of MltA with murein sacculi showed two major peaks eluted
with retention times of 52.4 and 68.9 min (Fig. 5). By
comparing the elution profile of the calibration standard,
it was determined that these major reaction products
Fig. 3. Demonstration that MltA is a lipoprotein. E. coli BL21(DE3)
harbouring pET21b-MltA cloned with its own LPSS, with the LPSS
from GNA1946, or without a leader sequence were grown in the
presence of [
3
H]palmitate. Expression of recombinant protein was
induced for 1 h at 30 °C by the addition of 1 m
M
isopropyl b-
D

-
thiogalactoside. Cells were then washed, lysed and protein precipitated
as described in Experimental Procedures. Proteins were separated by
SDS/PAGE using a 12.5% separating gel, and the labelled proteins
were visualized by autoradiography. Lane 1, MltA cloned without an
LPSS; lane 2, MltA cloned with its own LPSS; lane 3, MltA cloned
with the LPSS from GNA1946. Molecular masses of marker proteins
are indicated on the left and the position of MltA is indicated by an
arrow.
Ó FEBS 2002 MltA from N. meningitidis serogroup B (Eur. J. Biochem. 269) 3727
corresponded to 1,6-anhydrodisaccharide tripeptide and
tetrapeptide, respectively. The formation of the 1,6-anhydro
intramolecular bond within the muramic acid moiety
confirms that the enzyme is indeed a lytic transglycosylase
(Fig. 5).
MltA–Sepharose affinity chromatography of membrane
proteins
The leaderless form of MltA was expressed, purified and
covalently bound to CNBr-activated Sepharose. This col-
umn was used to isolate MltA-interacting proteins from a
membrane fraction of Meningococcus B. Proteins were
eluted with a stepwise NaCl gradient and assayed for
penicillin-binding activity by incubation with
125
I-labelled
ampicillin. PBPs were visualized by SDS/PAGE and auto-
radiography (Fig. 6). A control column prepared without
MltA was used to assess the specificity of binding. The most
intensely labelled band at 62 kDa observed in the starting
material was retained by MltA–Sepharose during loading

and washing, but was completely eluted with 300 m
M
NaCl.
In contrast, the intensely labelled 46-kDa band observed in
the starting material was not retained by the column and was
eluted in the flow through. Vollmer et al. [14] reported that
400 m
M
NaCl was sufficient to completely disrupt binding of
PBPs to E. coli MltA. When the autoradiograph and
Coomassie blue-stained gel were overlaid, it was not possible
to see a protein band corresponding to the 62-kDa
radioactive band. This is characteristic of PBPs, which are
typically of low abundance; for example, E. coli PBP2 is
present at only  50 copies per cell [23,24]. To date, four
PBPs have been identified in Meningococcus B: PBP1, PBP2,
PBP3 and PBP4. These proteins have predicted molecular
masses of 88.9 kDa, 63.6 kDa 50.5 kDa and 34.1 kDa,
respectively [25,26]. Hence we reasoned that the 62-kDa PBP
specifically retained during affinity chromatography is
PBP2. To confirm this hypothesis, we analysed affinity-
chromatography fractions by Western-blot analysis using
polyclonal antisera raised against PBP2 from Gonococcus
(Fig. 7). Gonococcal PBP2 has 98% sequence identity with
PBP2 from Meningococcus serogroup B. Immunoblots
showed an immunoreactive band with a molecular mass of
62 kDa in the starting material and in the fraction obtained
after elution with 300 m
M
NaCl. Moreover, this band

migrated to the same position as purified gonococcal PBP2.
The 88-kDa immunoreactive band observed in the starting
material was not retained by the MltA affinity column.
Taken together these results demonstrate an interaction
involving MltA and PBP2.
DISCUSSION
A genomics-based approach to vaccine discovery previ-
ously identified GNA33 as a potential vaccine candidate
Fig. 5. HPLC analysis of muropeptides after digestion of murein sacculi
with MltA. Isolated murein sacculi were digested with purified MltA
and reduced with sodium borohydride. The resulting muropeptides
were separated by RP-HPLC on a Hypersil ODS column. Elution was
performed with a linear gradient from 50 m
M
sodium phosphate
(pH 4.32) to 50% methanol in 50 m
M
sodium phosphate (pH 4.95).
The column was calibrated and standardized with murein sacculi
digested with the muramidase Cellosyl.
Fig. 4. HPLC analysis demonstrating hydrolysis of isolated glycan
strands. Poly(MurNAc-GlcNAc)
n>30
was incubated without (A) or
with (B) MltA as described in Experimental procedures. At the com-
pletion of the incubation, the sample was passed over a Nucleosil 300
C
18
column to which was applied 0.1 m
M

sodium phosphate buffer
(pH 2), 5% acetonitrile for 5 min, 100% methanol for 5 min, and
again starting buffer. The radioactivity of the eluate was monitored.
The peak eluted between 20.2 and 22.2 min corresponds to intact
glycan strands. In a control assay in which MltA was replaced with
E. coli MltA, the same elution profile as seen in (B) was observed (data
not shown).
3728 G. T. Jennings et al.(Eur. J. Biochem. 269) Ó FEBS 2002
against meningococcal infection. Sequence comparison
predicted that GNA33 encodes a lipoprotein homologous
to the lytic transglycosylase MltA from E. coli.To
definitively identify and characterize GNA33, we cloned
and expressed the ORF of GNA33 in E. coli with and
without an LPSS. Although the level of expression of the
truncated form was 20-fold higher than of the lipoprotein
form, incorporation of an LPSS in MltA increased
specific activity by 22-fold. Incorporation of
3
[H]palmitate,
cleavage of the leader peptide, and localization of the
protein to the membrane fraction all suggest that recom-
binant MltA is correctly processed as a lipoprotein in
E. coli. Moreover, purification of enzymatically active
protein and lysis of host cells during expression confirmed
the fidelity of the heterologous expression system. When
MltA was expressed with its own LPSS, the level of
expression was low. The level of expression was increased
significantly by fusing codons 21–441 of MltA to an LPSS
from an unrelated Meningococcus B lipoprotein,
GNA1946. This LPSS in combination with MltA is

obviously efficiently processed by the lipoprotein-process-
ing machinery of E. coli.
We demonstrated that recombinant MltA is capable of
lysing murein sacculi, confirming that the protein is a
muramidase. The lipoprotein produced two major reac-
tion products, 1,6-anhydrodisaccharide tripeptide and
tetrapeptide, confirming that the protein is indeed a lytic
transglycosylase. Of the four exo-specific lytic transglycos-
ylases in E. coli studied to date, only MltA is capable of
utilizing unsubstituted murein glycan strands as substrate
[18]. The ability of meningococcal MltA to also utilize the
unsubstituted substrate shows a functional similarity
between the two homologues. Furthermore, in many of
the biochemical parameters assessed, such as pH opti-
mum, K
m
and requirement for bivalent cations, the
N. meningitidis and E.coli enzymes are similar [9,18].
These results confirm the sequence-based prediction that
GNA33 is a homologue of E. coli lytic transglycosylase
MltA. For these reasons, we assigned the name MltA to
GNA33.
In this study, we used affinity chromatography to
demonstrate an association between meningococcal MltA
and PBP2. The ability to interact with a PBP is a
characteristic common to MltA from N. meningitidis and
E. coli and is the first description of such an association
beyond that reported for E. coli. E. coli MltA is thought
to form part of an enzyme complex composed of murein
synthases and muramidases. This association is believed

to facilitate the co-ordinated action of different enzymes
involved in enlargement and septation of the murein
sacculus [10]. Reconstitution experiments with E. coli
Fig. 6. PBP assay of proteins fractionated by affinity chromatography
on MltA-sepharose. Aliquots of fractions obtained from the elution of
the meningococcal membrane extract from either a MltA–Sepharose
or control column were assayed for the presence of PBPs with
125
I-labelled ampicillin as detailed in Experimental Procedures. The
labelled fractions were subjected to SDS/PAGE on a 10% separating
gel and visualized by autoradiography after 100 h exposure. SM is
membrane extract before addition to the column. C indicates eluates
obtained from the control column, and T represents eluates from the
MltA–Sepharose column. Shown are the first two fractions from the
wash with buffer I (150 m
M
NaCl) and each of the elution steps in
300 m
M
NaCl, 600 m
M
NaCl and 1
M
NaCl. The position of molec-
ular-mass markers is indicated.
Fig. 7. Western blot of proteins fractionated by affinity chromatography
on MltA–Sepharose. Aliquots of fractions obtained after elution of the
meningococcal membrane extract from the MltA–Sepharose column
were analysed by immunoblotting with anti-PBP2 sera. Immunore-
active bands were detected by enhanced chemiluminesence as

described in Experimental procedures. Lane 1, purified gonococcal
PBP2; lane 2, membrane extract from meningococcus before addition
to the column; lane 3, fraction obtained after elution with 300 m
M
NaCl. The positions of molecular-mass markers are shown.
Ó FEBS 2002 MltA from N. meningitidis serogroup B (Eur. J. Biochem. 269) 3729
MltA and PBP1B demonstrated the necessity for the
structural protein MipA, and it has been proposed that
this enzyme serves as a scaffold for assembly of the
multienzyme complex [14]. We performed an extended
homology search of the Meningococcus B genome but
failed to identify a homologue of MipA. A similar
situation exists for Haemophilus influenzae, which contains
a homologue of MltA but not MipA [14]. A
BLAST
search showed that N. meningitidis PBP2 and E. coli
PBP3 have 39% identity and 59% homology over a 541-
amino-acid overlap and revealed that meningococcal
PBP2 is more homologous to E. coli PBP3 than PBP2.
In fact, Meningococcus does not have a homologue of
E. coli PBP2, which is involved in maintaining the
characteristic rod shape of the bacterium. Interestingly,
it is the presence of either PBP2 or PBP3 in the enzyme
complex of E. coli that confers a specific function to the
complex [14]. In E. coli, PBP2 is known to be responsible
for cell elongation, whereas PBP3 is involved in septum
formation [27,28]. It will be interesting to determine if
such an enzyme complex exists in Meningococcus,the
nature and composition of the protein components, and
in particular the function of the association between

MltA and PBP2.
We initially reported that antibodies raised against
GNA33 are bactericidal, a property known to correlate
with protective effects in humans [1]. It was subsequently
discovered that antibodies elicited by vaccination with
GNA33 are bactericidal because MltA is an effective
mimetic antigen of the PorA epitope P1.2 [3]. In its own
right, MltA may be a useful vaccine for the prevention of
disease caused by P1.2 strains. Furthermore, it has been
suggested that substituting strain specific PorA loops into
MltA or its subdomains may generate immunogenic
mimetics of other serotype PorA epitopes [3]. The ease of
expression and purification demonstrated in this work
further suggests the great potential that MltA offers as a
recombinant vaccine candidate against meningococcal
infection. A direct role for lytic transglycosylases in
meningococcal disease is suggested by an investigation of
genes required for bacteraemic disease. In an infant rat
model of N. meningitidis infection, Sun and co-workers
[29] used insertional mutagenesis to identify genes essen-
tial for pathogenesis, one of which was the gene encoding
MltB. A further role for lytic transglycosylases in disease
may be associated with their reaction products. The 1,6-
anhydrodisaccharide-containing metabolites, such as
those shown here to be produced by meningococcal
MltA, have been shown to have diverse biological
activities. For instance, the cytopathology of respiratory
epithelium that is characteristic of Bordetella pertussis
infection is caused by 1,6-anhydromuramic acid-contain-
ing products [30]. The same compounds are also capable

of inducing sleep and arthritis [31,32]. Perhaps most
importantly is the potential of 1,6-anhydromuramyl
peptides to induce meningeal inflammation [33]. Hence
lytic transglycosylases such as MltA may be directly
involved in the pathogenesis associated with meningoc-
cocal infection. The potential that lytic transglycosylases
offer as targets for disease intervention combined with
their importance in growth, septation, recycling of
peptidoglycan, and pore formation makes them worthy
of further investigation.
ACKNOWLEDGEMENTS
We would like to thank Vega Masignani and Maria Scarselli for
sequence comparisons and database searches, and Mariagrazia Pizza
for many helpful discussions. We are also grateful to Giorgio Corsi for
preparing the figures and to Catherine Mallia for formatting and
submitting the manuscript.
REFERENCES
1. Pizza, M., Scarlato, V., Masignani, V., Giuliani, M.M., Arico, B.,
Comanducci, M., Jennings, G.T., Baldi, L., Bartolini, E.,
Capecchi, B. et al. (2000) Identification of vaccine candidates
against serogroup B meningococcus by whole-genome sequencing.
Science 287, 1816–1820.
2. Tettelin, H., Saunders, N.J., Heidelberg, J., Jeffries, A.C., Nelson,
K.E.,Eisen,J.A.et al. (2000) Complete genome sequence of
Neisseria meningitidis serogroup B strain MC58. Science 287,
1809–1815.
3. Granoff, D.M., Moe, G.R., Giuliani, M.M., Adu-Bobie, J.,
Santini, L., Brunelli, B., Piccinetti, F., Zuno-Mitchell, P., Lee, S.S.,
Neri, P. et al. (2001) A novel mimetic antigen eliciting protective
antibody to Neisseria meningitidis. J. Immunol. 167, 6487–6496.

4. Hayashi, S. & Wu, H.C. (1990) Lipoproteins in bacteria.
J. Bioenerg. Biomemb. 22, 451–469.
5. Betzner, A.S. & Keck, W. (1989) Molecular cloning, over-
expression and mapping of the slt gene encoding the soluble lytic
transglycosylase of Escherichia coli. Mol. Gen. Genet. 219, 489–491.
6. Dijkstra, A.J., Hermann, F. & Keck, W. (1995) Cloning and con-
trolled overexpression of the gene encoding the 35 kDa soluble lytic
transglycosylase from Escherichia coli. FEBS Lett. 366, 115–118.
7. Dijkstra, A.J. & Keck, W. (1996) Identification of new members of
the lytic transglycosylase family in Haemophilus influenzae and
Escherichia coli. Microb. Drug Resist. 2, 141–145.
8. Ehlert, K., Holtje, J V. & Templin, M.F. (1995) Cloning and
expression of a murein hydrolase lipoprotein from Escherichia coli.
Mol. Microbiol. 16, 761–768.
9. Lommatzsch, J., Templin, M.F., Kraft, A.R., Vollmer, W. &
Holtje, J V. (1997) Outer membrane localization of murein
hydrolases: MltA, a third lipoprotein lytic transglycosylase in
Escherichia coli. J. Bacteriol. 179, 5465–5470.
10. Holtje, J V. (1998) Growth of the stress-bearing and shape-
maintaining murein sacculus of Escherichia coli. Microbiol. Mol.
Biol. Rev. 62, 181–203.
11. Holtje, J V. & Tuomanen, E.I. (1991) The murein hydrolases of
Escherichia coli: properties, functions and impact on the course of
infections in vivo. J. Gen. Microbiol. 137, 441–454.
12. vonRechenberg, M., Ursinus, A. & Holtje, J V., (1996) Affinity
chromatography as a means to study multienzyme complexes
involved in murein synthesis. Microb. Drug Resist. 2, 155–157.
13. Romeis, T. & Holtje, J V. (1994) Specific interaction of penicillin-
binding proteins 3 and 7/8 with soluble lytic transglycosylase in
Escherichia coli. J. Biol. Chem. 269, 21603–21607.

14. Vollmer, W., von Rechenberg, M. & Holtje, J V. (1999)
Demonstration of molecular interactions between the murein
polymerase PBP1B, the lytic transglycosylase MltA, and the
scaffolding protein MipA of Escherichia coli. J. Biol. Chem. 274,
6726–6734.
15. Sambrook, J., Fritsch, E.F. & Maniatis, T. (1989) Molecular
Cloning: A Laboratory Manual, 2nd edn. Cold Spring Harbor
Laboratory Press, Cold Spring Harbor, NY.
16. Laemmli, U.K. & Favre, M. (1973) Maturation of the head of
bacteriophage T4. I. DNA packaging events. J. Mol. Biol. 80,575–
599.
17. Kraft, A.R., Templin, M.F. & Holtje, J V. (1998) Membrane-
bound lytic endotransglycosylase in Escherichia coli. J. Bacteriol.
180, 3441–3447.
3730 G. T. Jennings et al.(Eur. J. Biochem. 269) Ó FEBS 2002
18. Ursinus, A. & Holtje, J V. (1994) Purification and properties of a
membrane-bound lytic transglycosylase from Escherichia coli.
J. Bacteriol. 176, 338–343.
19. Harz, H., Burgdorf, K. & Holtje, J V. (1990) Isolation and
separation of the glycan strands from murein of Escherichia coli by
reversed-phase high-performance liquid chromatography. Anal.
Biochem. 190, 120–128.
20. Glauner, B. (1988) Separation and quantification of muropeptides
with high-performance liquid chromatography. Anal. Biochem. 17,
451–464.
21. Schwarz, U., Seeger, K., Wengenmayer, F. & Strecker, H. (1981)
Penicillin-binding proteins of Eschericia coli identified with a
125
I-labelled derivative of ampicillin. FEMS Microbiol. Lett. 10,
107–109.

22. Romeis, T., Vollmer, W. & Holtje, J.V. (1993) Characterization of
three different lytic transglycosylases in Escherichia coli. FEMS
Microbiol. Lett. 111, 141–146.
23. Dougherty, T.J., Kennedy, K., Kessler, R.E. & Pucci, M.E. (1996)
Direct quantitation of the number of individual penicillin-
binding proteins per cell in Escherichia coli. J. Bacteriol. 178, 6110–
6115.
24. Spratt, B.G. (1977) Properties of the penicillin-binding proteins of
Escherichia coli K12. Eur. J. Biochem. 72, 341–352.
25. Ropp, P.A. & Nicholas, R.A. (1997) Cloning and characterization
of the ponA gene encoding penicillin-binding protein 1 from
Neisseria gonorrhoeae and Neisseria meningitidis. J. Bacteriol. 179,
2783–2787.
26. Zhang, Q.Y. & Spratt, B.G. (1989) Nucleotide sequence of the
penicillin-binding protein 2 gene of Neisseria meningitides. Nucleic
Acids Res. 17, 5383.
27. Park, J.T. & Burman, L. (1973) FL-1060: a new penicillin with a
unique mode of action. Biochem. Biophys. Res. Commun. 51,
863–868.
28. Schmidt, L.S., Botta, G. & Park, J.T. (1981) Effects of
furazlocillin, a beta-lactam antibiotic which binds selectively
to penicillin-binding protein 3, on Escherichia coli mutants
deficient in other penicillin-binding proteins. J. Bacteriol. 145,
632–637.
29. Sun, Y.H., Bakshi, S., Chalmers, R. & Tang, C.M. (2000) Func-
tional genomics of Neisseria meningitidis pathogenesis. Nat. Med.
6, 1269–1273.
30. Cookson, B.T., Tyler, A.N. & Goldman, W.E. (1989) Primary
structure of the peptidoglycan-derived tracheal cytotoxin of Bor-
detella pertussis. Biochemistry 28, 1744–1749.

31. Krueger, J.M., Pappenheimer, J.R. & Karnovsky, M.L. (1982)
Sleep-promoting effects of muramyl peptides. Proc. Natl. Acad.
Sci. USA 79, 6102–6106.
32. Fleming, T.J., Wallsmith, D.E. & Rosenthal, R.S. (1986)
Arthropathic properties of gonococcal peptidoglycan fragments:
implications for the pathogenesis of disseminated gonococcal
disease. Infect. Immun. 52, 600–608.
33. Tuomanen, E., Hengstler, B., Zak, O. & Tomasz, A. (1986)
Induction of meningeal inflammation by diverse bacterial cell
walls. Eur. J. Clin. Microbiol. 5, 682–684.
Ó FEBS 2002 MltA from N. meningitidis serogroup B (Eur. J. Biochem. 269) 3731