Báo cáo Y học: Conformational analysis of opacity proteins from Neisseria meningitidis ppt

Conformational analysis of opacity proteins from
Neisseria
meningitidis
Marien I. de Jonge
1,2
, Martine P. Bos
3
, Hendrik J. Hamstra
1
, Wim Jiskoot
4
, Peter van Ulsen
4
,
Jan Tommassen
4
, Loek van Alphen
1
and Peter van der Ley
1
1
Laboratory of Vaccine Research, National Institute of Public Health and the Environment, RIVM Bilthoven, the Netherlands;
2
Department of Medical Microbiology, University of Amsterdam/AMC, Amsterdam, the Netherlands;
3
Department of Molecular
Microbiology and
4
Department of Pharmaceutics, Utrecht University, Utrecht, the Netherlands
Opacity-associated (Opa) proteins are outer membrane
proteins which play a critical role in the adhesion of patho-

genic Neisseria spp. to epithelial and endothelial cells and
polymorphonuclear neutrophils. The adherence is mainly
mediated by the CD66-epitope-containing members of the
carcinoembryonic-antigen family of human cell-adhesion
molecules (CEACAM). For the analysis of the specific in-
teractions of individual Opa proteins with their receptors,
pure protein is needed in its native conformation. In this
study, we describe the isolation and structural analysis of
opacity proteins OpaJ129 and OpaB128 derived from
Neisseria meningitidis strain H44/76. When the Opa proteins
were produced with the phoE signal sequence in Escherichia
coli, they were localized at the cell surface and the recom-
binant bacteria were found to specifically interact with
CEACAM1. For refolding and purification, the proteins
were overproduced without their signal sequences in E. coli,
resulting in its cytoplasmic accumulation in the form of
inclusion bodies. After solubilization of the inclusion bodies
in urea, the proteins could be folded efficiently in vitro, under
alkaline conditions by dilution in ethanolamine and the
detergent n-dodecyl-N,N-dimethyl-1-ammonio-3-propane-
sulfonate (SB12). The structure of the refolded and purified
proteins, determined by circular dichroism, indicated a high
content of b-sheet conformation, which is consistent with
previously proposed topology models for Opa proteins. A
clear difference was found between the binding of refolded
vs. denatured OpaJ protein to the N-A1 domain of CEA-
CAM1. Almost no binding was found with the denatured
Opa protein, showing that the Opa–receptor interaction is
conformation-dependent.
Keywords: Opa protein; Neisseria meningitidis;CEACAM

receptor; in vitro folding; conformation.
The pathogenic bacteria Neisseria meningitidis and Neisseria
gonorrhoeae express a family of genes encoding outer
membrane proteins that are structurally related but highly
polymorphic. These proteins were originally identified as
colony-opacity-associated (Opa) proteins [1]. Opa proteins
appear to play a critical role in the ÔintimateÕ adhesion of the
bacteria to epithelial and endothelial cells and to polymor-
phonuclear neutrophils [2,3]. The majority of Opa proteins
bind to carcinoembryonic antigen cell-adhesion molecules
(CEACAM, formerly called CD66) [4]. CEACAM proteins
are expressed on various epithelial and endothelial cells as
well as on some lymphoid and myeloid cells [5]. A minority of
the Opa proteins target heparan sulfate proteoglycans
(HSPG) [6,7]. Recently, several Opa proteins were found
that did not bind to any of these human receptors, suggesting
that these Opa proteins haveadditional functions or that they
recognize additional receptors [8]. Opa-receptor-mediated
adhesion can lead to invasion of the bacteria into thedifferent
cell types expressing CEACAM proteins [8,9]. Opa expres-
sion was found in mucosal as well as disease isolates, 87.5%
of meningococcal strains tested bind to CEACAM1 [10].
Although the binding specificity of the variable Opa
proteins to the conserved human receptors has been studied
extensively [11], not much is known about the binding sites
present in the Opa proteins. A two-dimensional topology
model has been proposed, in which the Opa proteins form
eight-stranded b-barrels, exposing four loops at the cell
surface [12]. The variability of the Opa proteins is mainly
concentrated in surface-exposed loops two and three. An

intriguing question is how the binding function of the Opa
proteins can be conserved despite the hypervariability
observed.
Pure protein is needed for detailed structure–function
relationship studies addressing this question. Due to phase-
variable expression of the Opa proteins in Neisseria spp.
[13], it is difficult to purify individual Opa proteins without
contamination of other Opa proteins from neisserial cells.
To overcome this problem, the invasion-associated men-
ingococcal OpaJ129 and OpaB128 proteins, both of which
bind to CEACAM1, were produced in this study in E. coli.
We developed a method to produce pure, folded and
functional Opa protein. The Opa proteins were isolated in
the form of inclusion bodies and subsequently folded in vitro.
Correspondence to M. de Jonge, Laboratory of Vaccine Research,
National Institute of Public Health and the Environment,
RIVM Bilthoven, the Netherlands.
Fax: + 31 030 2744429, Tel.: + 31 030 2743999,
E-mail:
Abbreviations: CEACAM, carcinoembryonic-antigen cell adhesion
molecule; HSPG, heparan sulfate proteoglycans; IPTG, isopropyl
thio-b-
D
-galactoside; OMC, outer membrane complex; Opa proteins,
opacity associated proteins; SB12, n-dodecyl-N,N-dimethyl-
1-ammonio-3-propanesulfonate.
(Received 12 July 2002, accepted 5 September 2002)
Eur. J. Biochem. 269, 5215–5223 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03228.x
The structure of refolded Opa protein was studied by
circular dichroism (CD) spectroscopy. The spectra are

indicative of a high content of b-strands, consistent with the
previously proposed structural models. Refolded Opa
protein was shown to be functional by specific binding to
the N-A1 part of CEACAM1.
MATERIALS AND METHODS
Construction of the expression systems
The genes encoding OpaJ129 and OpaB128 were isolated
from H44/76 using Taq polymerase (Amersham, Piscata-
way, NJ, USA) and general opa primers (5¢-CTTCT
CTTCTCTTCCGCAGC-3¢ and 5¢-TCGGTATCGGGG
AATCAGAA-3¢), cloned into plasmid pCR2.1 (Topo TA
cloning kit, Invitrogen, Carlsbad, CA, USA) and subse-
quently sequenced using M13-forward and M13-reverse
primers (Invitrogen). Plasmids pCR2.1 containing opaJ129
and opaB128 were used to amplify the DNA sequences
encoding the mature OpaB128 and OpaJ129 proteins with
Taq polymerase. The primers used (5¢-AGCGC
CCA
TGGCAAGTGAAG-3¢ and 5¢-GGCATCGGGATCCG
GGAATCAG-3¢) were based on the DNA sequence of
opaB128 and opaJ129 of N. meningitidis strains H44/76
(unpublished) and 190/87 (GenBank accession no.
AF016285) [12]. The primers contained base substitutions
(underlined) to introduce NcoIandBamHI cleavage sites,
respectively. The PCR product was cloned in plasmid
pCR2.1. The NcoI–BamHI fragment was isolated from the
resulting plasmid and ligated into the NcoI–BamHI digested
expression vector pET11d (New England Biolabs, Inc.,
Beverly, MA, USA) downstream of the inducible T7
promoter. In the resulting construct, the codon for the first

amino acid residue of the mature Opa protein was situated
directly downstream of the ATG start codon. The sequences
of the inserts were checked by DNA sequencing, using the
DNA sequencing kit and the ABI Prism 310 genetic
analyser, according to the instructions of the manufacturer
(ABI Prism, Perkin Elmer Applied Biosystems, Warrington,
UK).
Plasmids pET11d-opaB128 and pET11d-opaJ129 were
used to transform the E. coli strain BL21 (DE3) containing
a chromosomal copy of the T7 RNA polymerase gene
under control of the lac promotor [14]. Plasmid-containing
derivatives of this strain were grown at 37 °CinLuria-
Bertani (LB) medium (BIO 101, Inc., Carlsbad, CA,
USA) supplemented with 100 lgÆmL
)1
ampicillin (Sigma,
St. Louis, MO, USA).
The OpaB128 and OpaJ129 expression at the cell surface
of E. coli strain CE1265 was realized using the expression
vector pMR05, containing the complete phoE gene [15].
PCR amplifications were performed on pCR2.1 containing
either opaB128 or opaJ129 using Taq polymerase and
mutagenic primers (5¢-ATAGATCTCGGGGAATCAG
AAGCG-3¢ and 5¢-CTTCTCTTCTCTTCTGCAGC-3¢)
to generate a PstI site between the signal sequence and the
mature portion and a BglII site behind the stopcodon of
opaB128 and opaJ129.ThePstI–BglII fragments of
opaB128 and opaJ129 were used to replace a PstI–BglII
fragment of the phoE gene in pMR05, resulting in an in-
frame fusion of opa to the signal peptide of phoE and

expression from the phoE promotor. The resulting plasmids
were used for transformation of strain CE1265, which
expresses the pho regulon constitutively due to a phoR
mutation [16]. Expression of OpaB128 and OpaJ129 was
determined by assaying the binding of monoclonal anti-
bodies MN20E12.70 (M. de Jonge, G. Vidarson, H. H. van
Dijken, P. Hoogerhout, L. van Alphen, J. Dankert & P. van
der Ley, unpublished results) and 15-1-P5.5 [18] in a colony
blotting experiment [19]. The bla-opaB fusion construction,
whichresultedinE. coli surface expression of gonococcal
OpaB, is described by Belland et al.[3].Thesurface
expression was confirmed with immunofluorescence. Cells
were washed with NaCl/P
i
and after blocking overnight in
NaCl/P
i
with 3% BSA, incubated with 15-1-P5.5 [18]
(diluted 1 : 100) for 1 h, followed by an incubation with
Alexa-conjugated goat anti-(mouse IgG) (Molecular Probes
Inc., Eugene, OR, USA) (diluted 1 : 300) for 1 h. After
washing three times, cells were again fixated in NaCl/P
i
with
2% formaldehyde (Merck, Darmstadt, Germany). The
construction of the recombinant N-domains of the different
CEACAM proteins is described by Bos et al.[20].
Binding of His-tagged CEACAM fragments
to bacterial cells
The binding of His-tagged CEACAM fragments to bacter-

ial cells was measured as described previously [20]. The
expression of the N-terminal domains of the CEACAM
proteins was regulated by the inducible T7 promoter. The
Opa-expressing bacteria (3 · 10
8
) in 200 lL Hepes buffer
(10 m
M
Hepes, pH 7.4, 145 m
M
NaCl, 5 m
M
KCl, 0.5 m
M
MgCl
2
and 1 m
M
CaCl
2
) were incubated with 10 lLof
cleared bacterial cell lysate containing the His-tagged
N-terminal domains of either CEACAM1 or CEACAM8,
for 20 min at 37 °C. Bacteria were collected by centrifuga-
tion (5 min at 2000 g) and washed with 1 mL of Hepes
buffer. The pellet was resuspended, and processed for SDS/
PAGE and Western blotting, with monoclonal antibody
4B12 (1 : 5000) [21] for the detection of OpaJ129 and anti-
His
6

monoclonal Ig (1 : 10 000) (Amersham Pharmacia
Biotech GmbH, Freiburg, Germany) for the detection of
the N-terminal domains of the CEACAM proteins.
Production and purification of inclusion bodies
Cultures of the E. coli strain BL21 (DE3) containing either
pET11d-opaB128 or pET11d-opaJ129, grown overnight at
37 °C, were diluted 1 : 10 into fresh LB medium supple-
mented with 0.5% glucose (Fluka, Buchs, Switzerland) and
100 lgÆmL
)1
ampicillin. When the culture reached an
optical density of 660 nm (D
660
) of 0.6, isopropyl thio-b-
D
-galactoside (IPTG) (Boehringer Mannheim, Germany)
was added to a final concentration of 1 m
M
.After3hof
incubation at 37 °C, the cells were harvested by centrifuga-
tion at 4500 r.p.m. for 15 min at 4 °C (Centrikon T324,
Rotor A6.9, Kontron Instruments, Milan, Italy). The pellet
waswashedwith10m
M
Tris/HCl (pH 8) and centrifuged at
4500 r.p.m. for 15 min at 4 °C in the same rotor. After
resuspension in the same buffer, cells were disrupted using a
French Press (SLM-Aminco) at 9000 p.s.i. three times. The
inclusion bodies were collected by a low-speed centrifuga-
tion step at 2800 g for 10 min at 4 °C (Megafuge 1.0 R,

Hereaus sepatech, Germany). The pellet was resuspended in
8
M
urea, 50 m
M
glycine (pH 8.0). Ultracentrifugation at
5216 M. I. de Jonge et al.(Eur. J. Biochem. 269) Ó FEBS 2002
100 000 g for 2.5 h at 4 °C was used to remove residual
membrane fragments and the supernatant was stored at
4 °C. The protein concentration was determined with the
Pierce protein assay (Pierce, Rockford, IL, USA) using BSA
as a standard. Proteins were separated by SDS/PAGE
with 0.4 m
M
thioglycolic acid (Sigma-Aldrich, Steinheim,
Germany), included in the separating gel. After blotting on
poly(vinylidene difluoride) (Millipore, Bedford, MA, USA)
membranes and staining of the blots with Coomassie
Brilliant Blue, protein bands were cut from the membranes
and used for N-terminal sequencing.
Semi-native-polyacrylamide gel electrophoresis
To determine the heat-modifiability of wild type OpaB128
and OpaJ129, outer membrane complexes (OMCs) were
isolated from N. meningtidis strain H44/76 according to the
protocol described by Davies et al. [22]. The expression of
wild-type OpaJ129 was determined by Western blotting
using monoclonal antibody 15-1-P5.5 [18]. The expression
of wild-type OpaB128 was determined by Western blotting
using monoclonal antibody MN20E12.70 (M. de Jonge, G.
Vidarson, H. H. van Dijken, P. Hoogerhout, L. van Alphen,

J. Dankert & P. van der Ley, unpublished results).
Semi-native polyacrylamide gel electrophoresis was per-
formed by using SDS-free 11% polyacrylamide gels.
Loading buffer containing either 0.1% or 2.0% SDS
(Fluka, Buchs, Switzerland) was added to the samples
which were subsequently incubated at room temperature
and 100 °C, respectively. After electrophoresis, protein
bands were visualized with Coomassie Brilliant Blue (Fluka,
Buchs, Switzerland).
Refolding and purification
To find the optimal folding conditions, buffers with
different NaCl concentrations (100–300 m
M
) and final urea
concentrations (125 m
M
)1
M
) were tested at different pH
values ranging from pH 7.2–12.0. Furthermore, different
protein dilutions (1 : 20 to 1 : 200) and n-dodecyl-N,
N-dimethyl-1-ammonio-3-propanesulfonate (SB-12, Fluka,
Buchs, Switzerland) concentrations were tested. All refold-
ing experiments were performed overnight at 4 °C.
In the optimal folding procedure, Opa (10 mgÆmL
)1
)
dissolved in 8
M
urea and 50 m

M
glycine (pH 8.0) was
diluted 100-fold in refolding buffer containing 328 m
M
ethanolamine (pH 12), 0.5% SB12 (i.e. approximately
5 · critical micelle concentration). Prior to use, SB12 was
purified by passing a solution of the detergent in methanol/
chloroform (1 : 1, v/v) over an Al
2
O
3
column, to remove all
acidic impurities present in the commercial preparation [23].
After incubation overnight at 4 °C, the refolding mix was
neutralized with HCl to pH 7.5 and 10 m
M
Tris was added
to buffer the solution. Subsequently, to remove ethanol-
amine the solution was washed in a concentrator (Schleicher
and Schuell, Dassel, Germany) with 10 m
M
Tris/HCl, 0.5%
SB12 (pH 7.5) (buffer A).
An SP-Sepharose-HP column (volume 15 mL) (Amer-
sham Pharmacia Biotech Europe GmbH, Freiburg, Ger-
many) was equilibrated with buffer A, loaded with
approximately 10 mg refolded OpaJ129 and washed twice
with buffer A with pH 7.5 and pH 8.5. The proteins were
eluted with a linear gradient of NaCl from 0–1
M

in
120 mL. To check folding and purification, SDS/PAGE
was performed under seminative and denaturing conditions.
The folded and purified proteins were stored at )20 °C.
Circular dichroism spectroscopy
Circular dichroism (CD) spectra were recorded at 25 °Cwith
a dual-beam DSM 1000 CD spectrophotometer (On-Line
Instrument Systems, Bogart, GA, USA). The subtractive
double-grating monochromator was equipped with a fixed
disk, holographic gratings, and 1.24-mm slits. For far-UV
and near-UV measurements, gratings with 2400 lines per mm
(blaze wavelength 230 nm) and 600 lines per mm (blaze
wavelength 300 nm), respectively, were used. Far-UV spec-
tra were recorded from 250 to 200 nm (cell-path length
0.5 mm). For near-UV measurements (320–250 nm), cells
with a path length of 1 cm were used. The Opa protein
concentration was 0.54 mgÆmL
)1
. The results depicted
represent the average of at least six repeated scans (step
resolution 1 nm, 1 s each step), from which the correspond-
ing buffer spectrum was subtracted. The measured CD
signals were converted to molar ellipticity [h],basedona
mean residual weight of 112 (OpaB128) or 111.5 (OpaJ129).
For the comparison between folded and denatured
protein, folded protein in buffer A was incubated for
20 min at 100 °C with 1.85% SDS.
Immunodotblotting
Opa proteins were diluted to 1 lg per 100 lLin10m
M

Tris/HCl pH 7.5, 0.2% SB-12 (native) or in 2% SDS and
incubated at 100 °C (denatured) and spotted on to nitro-
cellulose (1 lg per spot). Filters were blocked with NaCl/P
i
/
T/P (NaCl/P
i
, 0.01% Tween-20, 0.5% Protifar) and subse-
quently incubated for 1 h with 1 mL of receptor sample
in 10 mL NaCl/P
i
/T/P. The blots were incubated with
1 : 10.000 diluted anti-His Ig (Amersham), followed by
1 : 10.000 dilution of peroxidase-conjugated goat anti-
mouse Ig (BioSource, Camarillo, CA, USA) and ECL
detection (Pierce).
OpaD protein is a purified Opa preparation from
meningococcal membranes [21] (generously donated by
M. Achtman, Max-Planck Institute, Berlin, Germany).
This preparation was successfully used previously to detect
CEACAM binding in a dot-blot assay [10].
The N- and A1 domains of CEACAM1 were amplified
from CEACAM cDNA (gift from M. Kuroki, Fukuoka
University, Fukuoka, Japan) with primer pair 5¢-ATCATA
TGCAGCTCACTACTGAATCCATGCC-3¢ and 5¢-AT
CGGGATCCCTAACTCACTGGGT TCTGTATTTC-3¢
and cloned into pET15a (Invitrogen) using NdeIandBamHI
sites included in the primers. This results in an N-terminal
6·His-tag addition to the CEACAM N-domain. The two
domains with the His-tag were subcloned into pET26b

(Invitrogen) using NcoIandBamHI restriction, resulting in
plasmid pVB1. This vector adds the pelB signal sequence to
the CEACAM domains, which allows secretion of the
protein into the periplasm. BL21 cells containing pVB1 were
grown in LB containing 50 lgÆmL
)1
kanamycin to a D
600
of
0.6. Cells were induced with 0.2 m
M
IPTG and grown
overnight at room temperature. The induced cell pellet was
resuspended in 200 m
M
Tris/HCl pH 8.0, 0.5 m
M
EDTA,
0.5
M
sucrose. Lysozyme was added to 60 lgÆmL
)1
and
Ó FEBS 2002 Conformation of meningococcal Opa proteins (Eur. J. Biochem. 269) 5217
the suspension was diluted 2 · with 0.5 m
M
EDTA and
incubated for 10 min at room temperature. Cells were
collected by centrifugation for 2 min at 8000 g. (Eppendorf
centrifuge) and the supernatant was collected as the

periplasmic fraction containing the Opa receptor.
RESULTS
Binding of Opa proteins to CEACAM1
N. meningitidis strain H44/76 can make four different Opa
proteins, some of which appear to be associated with the
ability to invade human nasopharyngeal cells [24]. Among
those, major differences in the hypervariable regions were
found between OpaB128 and OpaJ129; the other two Opa
proteins have sequences closely resembling either one (M. de
Jonge, G. Vidarson, H. H. van Dijken, P. Hoogerhout, L.
van Alphen, J. Dankert & P. van der Ley, unpublished
results). The majority of Opa proteins bind to CEACAM
proteins, reviewed by Billker et al. [11]. The binding of
OpaB128 and OpaJ129 to CEACAM1 and CEACAM8
was determined, using E. coli cells expressing this Opa
protein at the cell surface. For this purpose we cloned the
fragment of opaB128 and opaJ129 encoding the mature
domain of the protein downstream of the promoter and the
signal sequence-encoding part of the phosphate-limitation
inducible phoE gene of E. coli. The resulting plasmids were
used to transform E. coli strain CE1265, which expresses the
pho regulon constitutively due to a phoR mutation [16].
Surface expression of OpaB 128 and OpaJ129 was con-
firmed by immunoblotting outer membrane complexes
(OMCs) (Fig. 1) with monoclonal antibodies MN20E12.70
and 15-1-P5.5. We determined binding of soluble his-tagged
N-terminal domains of CEACAM1 or CEACAM8 to
recombinant OpaB128 or OpaJ129 expressed at the cell
surface with an anti-His monoclonal antibody, as previously
described by Bos et al. [20]. The E. coli bacteria expressing

either OpaB128 or OpaJ129 at the cell-surface were
incubated with N-domains of the two different CEACAM
proteins (Fig. 2, lanes 5–8). Surface expressed recombinant
gonococcal OpaB protein was included in these experiments
as a positive control (Fig. 2, lanes 1 and 2). After
incubation, the bacterial cells were collected by centrifuga-
tion and the proteins were separated by SDS/PAGE. After
blotting to nitrocellulose filters, the presence of CEACAM
in the bacterial cell pellets was evaluated with an anti-His Ig.
Like the bacteria expressing the gonococcal OpaB protein
(lane 1 and 2) the bacteria expressing OpaB128 or OpaJ129
bound to CEACAM1 (lane 5 and 7) while no binding was
found with CEACAM8 (lanes 6 and 8).
Expression system for Opa proteins
To obtain large quantities of OpaB128 and OpaJ129
protein, part of the opa sequence encoding the mature Opa
protein without the signal sequence was cloned into pET11d
under the control of the inducible T7 promoter. The
recombinant genes were expressed in the E. coli strain
BL21 (DE3) upon addition of IPTG. The Opa proteins
accumulated in the cytoplasm as inclusion bodies, which
could be separated from the other cell components by
centrifugation. After dissolving these inclusion bodies in 8
M
urea followed by an ultracentrifugation step to remove
residual membrane fragments, the Opa protein in the
supernatant was approximately 90% pure as determined
by SDS/PAGE (Fig. 3A). N-terminal amino acid sequen-
cing of the purified proteins revealed the sequence ASEDG,
Fig. 1. Western blots showing the heat-modifiability of OpaB128 and OpaJ129 expressed in N. meningitidis and E. coli. OMCs of H44/76 expres-

sing either OpaB128 or OpaJ129 and OMCs of E. coli strain CE1265 containing the phoE-opa fusion plasmid pMR05-opaB128 or pMR05-opaJ129
were separated by seminative-PAGE and analysed by Western blotting, using either the OpaB128 or OpaJ129 specific monoclonal antibody
(MN20E12.70 or 15-1-P5.5, respectively). Samples were treated in sample buffer containing 0.1% SDS at room temperature (RT) or 2.0% SDS at
100 °C, prior to electrophoresis.
Fig. 2. Binding of His-tagged N-domains of CEACAM1 and 8 by
OpaJ129- and OpaB128-expressing E. coli cells. The binding of the
N-terminal fragments of CEACAM1 and CEACAM8 by MS11-
OpaB-expressing E. coli (lanes 1 and 2), by E. coli not expressing an
Opa protein (lanes 3 and 4), binding of OpaJ129- and OpaB128-
expressing E. coli to CEACAM1 (lane 5 and 7, respectively) and to
CEACAM 8 (lane 6 and 8, respectively) was studied. The bacteria were
incubated with cleared lysates of E. coli containing the N-domains of
the CEACAM proteins and were processed for immunoblotting.
Bound N-domain was detected with anti-His Ig. Opa protein expres-
sion of the variants was evaluated with mAb 4B12.
5218 M. I. de Jonge et al.(Eur. J. Biochem. 269) Ó FEBS 2002
exactly corresponding to the predicted N-terminus of the
mature Opa. This indicated that the N-terminal methionine
encoded by the ATG initiation codon was efficiently
removed in vivo by the methionine endopeptidase [25].
Folding experiments with Opa proteins
Correct folding of recombinant Opa protein was evaluated
using the property that native Opa proteins migrate faster
in semi-native PAGE than heat-denatured forms [1]. To
check whether this heat-modifiability character also applies
to OpaB128 and OpaJ129, OMCs were isolated from
N. meningitidis strain H44/76 expressing them, as deter-
mined by Western blotting using monoclonal antibodies
MN20E12.70 (M. de Jonge, G. Vidarson, H. H. van Dijken,
P. Hoogerhout, L. van Alphen, J. Dankert & P. van der Ley,

unpublished results) or 15-1-P5.5 [18], respectively. Semi-
native PAGE, followed by Western blotting, confirmed that
both OpaB128 and OpaJ129 migrated with an apparent
molecularmassof 23 kDa, whereas completely unfolded
OpaB128 or OpaJ129 migrates as a protein of  27 kDa
(Fig. 1). The heat-modifiability of wild type OpaB128 and
OpaJ129 was taken as marker for correct folding of the
proteins purified from the inclusion bodies. The correct
folding of OpaB128 and OpaJ129 expressed at the surface of
E. coli strain CE1265 was confirmed in the same assay
(Fig. 1).
We diluted the urea-solubilized protein solution
(10 mgÆmL
)1
) 100-fold in various buffers with different
pHs, all containing 0.5% SB12 (w/v) and incubated the
samples overnight at 4 °C. When the pH of the refolding
buffer was below 10, no or hardly any refolding was
observed. However, in 328 m
M
ethanolamine buffer with
pH 10.5 (i.e. just above the calculated pI of OpaJ129 and
OpaB128, 10.3 and 10.4, respectively) almost 50% of
OpaJ129 and > 50% of OpaB proved to be refolded
according to semi-native PAGE analysis (data not shown).
To increase the folding efficiency, several buffering
substances and final protein concentrations were tested
at different pH values, and different NaCl and urea
concentrations. Although inclusion of 200 m
M

NaCl in
the refolding buffer improved refolding considerably this
condition was not applied further, because salt interferes
with the subsequent protein purification by ion-exchange
chromatography. The variation in protein and urea con-
centrations had almost no effect (data not shown). How-
ever, at pHs further above the calculated pI of OpaB128 and
OpaJ129 the folding appeared very efficient (Fig. 3A).
Summarizing, efficient refolding was achieved by a 100-fold
dilution of 10 mgÆmL
)1
Opa protein solubilized in 8
M
urea
in folding buffer containing 328 m
M
ethanolamine and
0.5% (w/v) SB12. The optimal pH for efficient refolding was
11 for OpaB128 and 12 for OpaJ129. More than 95% of the
protein proved to adopt a folded state under these
conditions, as shown by semi-native PAGE (Fig. 3).
Purification of folded Opa proteins
To remove unfolded protein and other contaminants, the
in vitro folded OpaJ129 proteins were purified by ion-
exchange chromatography. We observed that during the
purification by anion-exchange chromatography (Q-Seph-
arose HP at pH 12) a substantial proportion of the refolded
OpaJ129 protein eluted in the denatured state. Probably, the
protein was unstable in the alkaline conditions applied
during purification. Therefore, the pH was reduced from 12

to 7.5 after refolding. This procedure did not affect the
folding state of either OpaB128 or OpaJ129 as determined
by seminative PAGE (data not shown). The neutralized
protein solution was applied to a cation-exchange column
(SP-Sepharose HP at pH 7.5). Protein was eluted from the
column with a linear salt gradient, resulting in the elution of
either folded OpaB128 or folded OpaJ129 as a single peak.
Apparently, due to a difference in affinity, the folded protein
Fig. 3. Semi-native PAGE analysis of in vitr o folding of OpaB128 (A) and OpaJ129 (B). (A) Semi-native-PAGE analysis of in vitro folding of
unpurified OpaB128 (Coomassie stained). Lane 1 isolated inclusion bodies. Lane 2 in vitro folded protein. Lane 3 denatured protein. Lane 4 and 5,
in vitro folded and denatured OpaB after additional purification. Lane 6, molecular mass marker. Samples 2 and 4 were incubated at room
temperature in loading buffer containing 0.1% SDS, samples 1, 3 and 5 were incubated at 100 °C in loading buffer containing 2.0% SDS prior to
electrophoresis. (B) Semi-native PAGE analysis of in vitro folding of unpurified OpaJ129 (Coomassie stained). Lane 1, isolated inclusion bodies.
Lane 2, in vitro folded protein. Lane 3, denatured protein. Samples 1 and 3 were incubated as samples 1, 3 and 5 (Fig. 3A) and sample 2 was treated
as sample 2 and 4 (Fig. 3A). (C) Coomassie stained polyacrylamide gel showing in vitro folded OpaJ129, after additional purification. Purified
protein samples 1 and 2 were treated as samples 2 and 3 (Fig. 3A), respectively.
Ó FEBS 2002 Conformation of meningococcal Opa proteins (Eur. J. Biochem. 269) 5219
was purified from the residual unfolded protein as well as
from other contaminants.
Analysis of the Opa protein conformation
by circular dichroism
To test whether the in vitro folded OpaB128 and OpaJ129
had adopted the expected b-sheet conformation, CD spectra
were recorded for folded Opa protein and Opa protein that
was denatured by boiling in 1.85% SDS. The far-UV spectra
revealed a clear difference between the secondary structures
of the folded and denatured proteins (Fig. 4A). The charac-
teristic feature of the spectrum, recorded for folded OpaB128
was a minimum at 217 nm. Characteristic features of the
spectrum, recorded for folded OpaJ129, were a maximum at

approximately 232 nm and a minimum at 215 nm. The
minimum negative peaks in this range are indicative for the
content of b-sheet. The characteristic features of folded
OpaB128 and OpaJ129 disappeared upon denaturation,
with the minimum shifting to approximately 209 nm and the
maximum disappearing. This spectrum suggested the pres-
ence of a considerable proportion of a-helix. Apparently,
boiling in SDS induces a non-native structure.
Near-UV CD permits assessment of the differences
between the tertiary structure of folded and denatured
OpaJ129. Figure 4B shows that a less pronounced peak at
approximately 293 nm characterized the near-UV CD
spectrum of folded OpaB128, while folded OpaJ129 was
characterized by two peaks at approximately 265 nm and
293 nm. After denaturation, this characteristic feature of the
folded OpaJ129 protein changed into a spectrum with a
broad positive ellipticity and a maximum at around 270 nm,
while the major difference between refolded and denatured
OpaB128 was measured between 250 and 265 nm. The
differences in the spectra between refolded and denatured
Opa protein are indicative for a major structural change
after denaturation.
Functional analysis of purified refolded
and denatured Opa protein
In a receptor overlay experiment, equal amounts of refolded
anddenaturedOpaJandOpaDwereappliedtonitrocel-
lulose and incubated with bacterial lysates containing the
CEACAM1-N-A1 domain. Binding of CEACAM1-N-A1
was determined by monoclonal anti-His Ig reacting with the
His-tagged CEACAM1-N-A1 protein. Refolded OpaJ129

bound to CEACAM1-N-A1, consistent with the binding
experiments with the OpaJ129-expressing E. coli bacteria
(Fig. 5). The binding of refolded Opa appeared to be
conformation-dependent, as no binding was found with
denatured OpaJ129.
DISCUSSION
The majority of Opa proteins have been shown to speci-
fically target members of the CEACAM receptor family
[10,26]. How this binding function can be conserved
Fig. 4. Far-UV (A) and near-UV (B) circular dichroism spectra of refolded and heat-denatured OpaB128. (A) Far-UV circular dichroism spectra of
refolded OpaB128 (interrupted line) and heat-denatured OpaB128 in 1.85% SDS containing buffer (solid line). (1) Far-UV circular dichroism
spectra of refolded OpaJ129 (interrupted line) and heat-denatured OpaJ129 in 1.85% SDS containing buffer (solid line) (2). (B) Near-UV circular
dichroism spectra of refolded OpaB128 (interrupted line) and heat-denatured OpaB128 in 1.85% SDS containing buffer (solid line). (1) Near-UV
circular dichroism spectra of refolded OpaJ129 (interrupted line) and heat-denatured OpaJ129 in 1.85% SDS containing buffer (solid line) (2).
5220 M. I. de Jonge et al.(Eur. J. Biochem. 269) Ó FEBS 2002
despite the hypervariability of the surface-exposed regions
of the Opa proteins is still an enigma. The detailed
identification of the receptor-binding Opa regions would
aid greatly in the development of new vaccines or
antimicrobials specifically targeted at blocking this essen-
tial adhesion process. The study of the molecular inter-
actions between the CEACAM receptors and Opa
proteins would be facilitated greatly by the availability
of large quantities of pure Opa proteins. This was
achieved in the present study for OpaB128 and OpaJ129,
two representative Opa proteins present in invasive
variants of N. meningitidis strain H44/76.
Previously, the isolation and purification of Opa proteins
from meningococcal strains has been described [21,27].
However, translation of the constitutively transcribed opa

genes depends on the expression status of the individual opa
loci, which are subject to high-frequency phase variation [28].
Due to this phase variation it is difficult to express and purify
a single individual Opa protein without significant contam-
ination from different Opa proteins expressed from other
loci. However, for detailed structure–function analysis pure
protein is needed and by cytoplasmic expression of OpaB128
and OpaJ129 in E. coli, we were able to isolate highly pure
protein, not contaminated with other Opa proteins.
In order to determine their binding to CEACAM1, we
assessed the binding of surface-expressed OpaB128 and
OpaJ129 in an E. coli background to the N-terminus of
CEACAM1. CEACAM8 was taken as a negative control
because it does not function as an Opa receptor as has been
shown for the gonococcal and meningococcal Opa proteins
analyzed to date. Both OpaB128 and OpaJ129 bound
specifically to the N-terminal domain of CEACAM1. As
the majority of Opa proteins recognize this receptor,
OpaB128 and OpaJ129 seem to be typical members of this
protein adhesin family [10].
OpaB128 as well as OpaJ129 was expressed cytoplasmi-
cally in E. coli in the form of inclusion bodies and
subsequently refolded and purified. The characteristic heat
modifiability of Opa proteins was used to monitor their
refolding. Similar to other b-barrel outer membrane
proteins such as OmpA (E. coli)andP5(Haemophilus
influenzae), this heat-modifiable characteristic of Opa cor-
relates with folding into the native structure [29,30]. Both
OmpA and P5 are integrated into the membrane as eight-
stranded b-barrels [31,32] and the same structure has been

predicted for Opa proteins [12,33].
The CD measurements showed a clear difference
between the structure of folded Opa and Opa denatured
by boiling in SDS. The far-UV spectrum we recorded for
folded Opa resembles that of folded OmpA from E. coli
[34] and purified P5 from H. influenzae [32]. The spectra
are indicative of a high content of b-strands, consistent
with the (proposed) structure of these outer membrane
proteins (Fig. 4A). The difference between the near-UV
CD spectra of folded and denatured Opa supports the
conclusion that denatured protein has undergone a major
conformational change. In the proposed topology model
for Opa proteins, 31% of the amino acid chain is
predicted to form a transmembrane b-barrel. The high
content of b-strands reflected in the CD spectra reported
here suggests that a significant part of the extracellular
loops may also adopt this secondary structure. It is thus
conceivable that Opa proteins form a more extended
b-barrel structure that protrudes from the outer mem-
brane into the extracellular space, similar to what was
described recently for the OmpT outer membrane prote-
ase from E. coli [35].
The pH and the salt concentration are the most critical
factors in the folding efficiency of Opa. It appeared that a
pH above the calculated pI is needed for efficient folding, as
has also been found for the OmpA protein from E. coli and
the PorA protein from N. meningitidis [36,37]. The present
study demonstrates how two different Opa proteins, with
approximately 70% homology, can be folded in vitro under
similar conditions. This method will allow us to establish a

collection of different Opa proteins, suitable for studying
the interactions with CEACAM receptors.
In the receptor binding experiments OpaD was used as a
positive control, as was also done in similar experiments by
Virji et al. [10]. In earlier experiments using only the
N-terminus of CEACAM1 we could not find reproducible
binding to Opa. However, when the N-A1 domain of
CEACAM1 was used instead, it became clear that refolded
OpaJ129 is functional in receptor binding. The binding
between Opa and CEACAM1 seemed to be conformation-
dependent since almost no binding was found with dena-
tured Opa protein.
To conclude, with our purification and folding proce-
dures, we were able to isolate pure and native OpaB128 and
OpaJ129, both adhesins binding to the CEACAM1 recep-
tor. Conformational analysis of the purified, refolded
proteins provided the first experimental evidence for a
secondary structure dominated by b-strands, confirming
previously proposed topology models. Purified and refolded
Opa proteins will be used for detailed structural and
functional analysis.
ACKNOWLEDGEMENTS
We would like to thank F. van der Lecq at the sequencing centre of
the Centre for Biomembranes and Lipid Enzymology at Utrecht
University for N-terminal protein sequencing. We are grateful to
M. Kuroki at the Fukuoka University, for the generous gift of
Fig. 5. Binding of native and denatured OpaD and OpaJ to its receptor
was determined by immunodotblotting.
Ó FEBS 2002 Conformation of meningococcal Opa proteins (Eur. J. Biochem. 269) 5221
cDNA from CEACAM, to M. Achtman at the Max-Planck

Institute in Berlin, for the generous gift of purified OpaD and to
B. Kuipers at the RIVM in Bilthoven and W. Zollinger at the
Walter Reed Army Institute of Research in Washington for
providing us with monoclonal antibodies. We also thank
W. van Noppen at the University of Amsterdam/AMC for critically
reading the manuscript.
REFERENCES
1. Swanson, J. (1978) Studies on gonococcus infection. XII. Colony
color and opacity variants of gonococci. Infect. Immun. 19,320–
331.
2. Virji, M., Makepeace, K., Ferguson, D.J., Achtman, M. &
Moxon, E.R. (1993) Meningococcal Opa and Opc proteins: their
role in colonization and invasion of human epithelial and
endothelial cells. Mol. Microbiol. 10, 499–510.
3. Belland, R.J., Chen, T., Swanson, J. & Fischer, S.H. (1992)
Human neutrophil response to recombinant neisserial Opa
proteins. Mol. Microbiol. 6, 1729–1737.
4. Dehio, C., Gray-Owen, S.D. & Meyer, T.F. (1998) The role
of neisserial Opa proteins in interactions with host cells. Trends
Microbiol. 6, 489–495.
5. Benchimol, S., Fuks, A., Jothy, S., Beauchemin, N., Shirota, K. &
Stanners, C.P. (1989) Carcinoembryonic antigen, a human tumor
marker, functions as an intercellular adhesion molecule. Cell 57,
327–334.
6. van Putten, J.P. & Paul, S.M. (1995) Binding of syndecan-like
cell surface proteoglycan receptors is required for Neisseria
gonorrhoeae entry into human mucosal cells. EMBO J. 14, 2144–
2154.
7. Chen, T., Belland, R.J., Wilson, J. & Swanson, J. (1995)
Adherence of pilus– Opa+ gonococci to epithelial cells in vitro

involves heparan sulfate. J. Exp. Med. 182, 511–517.
8. Muenzner, P., Dehio, C., Fujiwara, T., Achtman, M., Meyer, T.F.
& Gray-Owen, S.D. (2000) Carcinoembryonic antigen family
receptor specificity of Neisseria meningitidis Opa variants
influences adherence to and invasion of proinflammatory cyto-
kine-activated endothelial cells. Infect. Immun. 68, 3601–3607.
9. Makino, S., van Putten, J.P. & Meyer, T.F. (1991) Phase variation
of the opacity outer membrane protein controls invasion by
Neisseria gonorrhoeae into human epithelial cells. EMBO J. 10,
1307–1315.
10. Virji, M., Watt, S.M., Barker, S., Makepeace, K. & Doyonnas, R.
(1996) The N-domain of the human CD66a adhesion molecule is a
target for Opa proteins of Neisseria meningitidis and Neisseria
gonorrhoeae. Mol. Microbiol. 22, 929–939.
11. Billker, O., Popp, A., Gray-Owen, S.D. & Meyer, T.F. (2000) The
structural basis of CEACAM-receptor targeting by neisserial Opa
proteins. Trends Microbiol. 8, 258–260.
12. Malorny, B., Morelli, G., Kusecek, B., Kolberg, J. & Achtman,
M. (1998) Sequence diversity, predicted two-dimensional protein
structure, and epitope mapping of neisserial Opa proteins.
J. Bacteriol. 180, 1323–1330.
13. Stern, A. & Meyer, T.F. (1987) Common mechanism controlling
phase and antigenic variation in pathogenic neisseriae. Mol.
Microbiol. 1, 5–12.
14. Studier, F.W. & Moffatt, B.A. (1986) Use of bacteriophage T7
RNA polymerase to direct selective high-level expression of cloned
genes. J. Mol. Biol. 189, 113–130.
15. Agterberg, M., Adriaanse, H. & Tommassen, J. (1987) Use of
outer membrane protein PhoE as a carrier for the transport of a
foreign antigenic determinant to the cell surface of Escherichia coli

K-12. Gene 59, 145–150.
16. Korteland, J., Overbeeke, N., de Graaff, P., Overduin, P. &
Lugtenberg, B. (1985) Role of the Arg158 residue of the outer
membrane PhoE pore protein of Escherichia coli K12inbacter-
iophage TC45 recognition and in channel characteristics. Eur. J.
Biochem. 152, 691–697.
17. Reference withdrawn.
18. Wang, J.F., Caugant, D.A., Morelli, G., Koumare, B. &
Achtman, M. (1993) Antigenic and epidemiologic properties of the
ET-37 complex of Neisseria meningitidis. J. Infect. Dis. 167,
1320–1329.
19. van der Ley, P. & Poolman, J.T. (1992) Construction of a multi-
valent meningococcal vaccine strain based on the class 1 outer
membrane protein. Infect. Immun. 60, 3156–3161.
20. Bos,M.P.,Kuroki,M.,Krop-Watorek,A.,Hogan,D.&Belland,
R.J. (1998) CD66 receptor specificity exhibited by neisserial Opa
variants is controlled by protein determinants in CD66 N-
domains. Proc. Natl Acad. Sci. USA 95, 9584–9589.
21. Achtman, M., Neibert, M., Crowe, B.A., Strittmatter, W.,
Kusecek, B., Weyse, E., Walsh, M.J., Slawig, B., Morelli, G. &
Moll, A. (1988) Purification and characterization of eight class 5
outer membrane protein variants from a clone of Neisseria
meningitidis serogroup A. J. Exp. Med. 168, 507–525.
22. Davies, R.L., Wall, R.A. & Borriello, S.P. (1990) Comparison of
methods for the analysis of outer membrane antigens of Neisseria
meningitidis by western blotting. J. Immunol. Methods 134,215–
225.
23. Dekker, N., Merck, K., Tommassen, J. & Verheij, H.M. (1995)
In vitro folding of Escherichia coli outer-membrane phospholipase
A. Eur. J. Biochem. 232, 214–219.

24. deVries,F.P.,VanderEnde,A.,VanPutten,J.P.M.&Dankert,
J. (1996) Invasion of primary nasopharyngeal epithelial cells by
Neisseria meningitidis is controlled by phase variation of multiple
surface antigens. Infect Immun. 64, 2998–3006.
25. Ben-Bassat, A., Bauer, K., Chang, S.Y., Myambo, K., Boosman,
A. & Chang, S. (1987) Processing of the initiation methionine from
proteins: properties of the Escherichia coli methionine amino-
peptidase and its gene structure. J. Bacteriol. 169, 751–757.
26. Bos, M.P., Grunert, F. & Belland, R.J. (1997) Differential
recognition of members of the carcinoembryonic antigen family
by Opa variants of Neisseria gonorrhoeae. Infect. Immun. 65,
2353–2361.
27. Blake, M.S. & Gotschlich, E.C. (1984) Purification and partial
characterization of the opacity-associated proteins of Neisseria
gonorrhoeae. J. Exp. Med. 159, 452–462.
28. Stern, A., Brown, M., Nickel, P. & Meyer, T.F. (1986) Opacity
genes in Neisseria gonorrhoeae: control of phase and antigenic
variation. Cell 47, 61–71.
29. Vogel, H. & Jahnig, F. (1986) Models for the structure of
outer-membrane proteins of Escherichia coli derived from
raman spectroscopy and prediction methods. J. Mol. Biol. 190,
191–199.
30. Webb, D.C. & Cripps, A.W. (1999) A method for the purification
and refolding of a recombinant form of the nontypeable Haemo-
philus influenzae P5 outer membrane protein fused to poly-
histidine. Protein Expr. Purif. 15, 1–7.
31. Pautsch, A. &Schulz, G.E. (1998) Structure of the outer membrane
protein A transmembrane domain. Nat. Struct. Biol. 5, 1013–1017.
32. Webb, D.C. & Cripps, A.W. (1998) Secondary structure
and molecular analysis of interstrain variability in the P5 outer-

membrane protein of non-typable Haemophilus influenzae
isolated from diverse anatomical sites. J. Med. Microbiol. 47,
1059–1067.
33. van der Ley, P. (1988) Three copies of a single protein II-
encoding sequence in the genome of Neisseria gonorrhoeae
JS3: evidence for gene conversion and gene duplication. Mol.
Microbiol. 2, 797–806.
34. Kleinschmidt, J.H., den Blaauwen, T., Driessen, A.J. & Tamm,
L.K. (1999) Outer membrane protein A of Escherichia coli inserts
and folds into lipid bilayers by a concerted mechanism.
Biochemistry 38, 5006–5016.
5222 M. I. de Jonge et al.(Eur. J. Biochem. 269) Ó FEBS 2002
35. Vandeputte-Rutten, L., Kramer, R.A., Kroon, J., Dekker, N.,
Egmond, M.R. & Gros, P. (2001) Crystal structure of the outer
membrane protease OmpT from Escherichia coli suggests a novel
catalytic site. EMBO J. 20, 5033–5039.
36. Surrey, T. & Jahnig, F. (1992) Refolding and oriented insertion of
a membrane protein into a lipid bilayer. Proc.NatlAcad.Sci.USA
89, 7457–7461.
37. Jansen, C., Wiese, A., Reubsaet, L., Dekker, N., de Cock, H.,
Seydel, U. & Tommassen, J. (2000) Biochemical and bio-
physical characterization of in vitro folded outer membrane
porin PorA of Neisseria meningitidis. Biochim. Biophys. Acta 1464,
284–298.
Ó FEBS 2002 Conformation of meningococcal Opa proteins (Eur. J. Biochem. 269) 5223