вход по аккаунту



код для вставкиСкачать
PROTEINS: Structure, Function, and Genetics 31:391–405 (1998)
Structural Investigation of C4b-Binding Protein
by Molecular Modeling: Localization
of Putative Binding Sites
Bruno O. Villoutreix,1* Ylva Härdig,1 Anders Wallqvist,2 David G. Covell,3 Pablo Garcı́a de Frutos,1
and Björn Dahlbäck1
1Department of Clinical Chemistry, The Wallenberg Laboratory, University Hospital, Lund University,
Malmö, Sweden
2Department of Chemistry, Rutgers University, Piscataway, New Jersey
3Frederick Cancer Research and Development Center, National Cancer Institute, Science Applications International
Corporation, Frederick, Maryland
C4b-binding protein (C4BP)
contributes to the regulation of the classical
pathway of the complement system and plays
an important role in blood coagulation. The
main human C4BP isoform is composed of one
b-chain and seven a-chains essentially built
from three and eight complement control
protein (CCP) modules, respectively, followed
by a nonrepeat carboxy-terminal region involved in polymerization of the chains. C4BP is
known to interact with heparin, C4b, complement factor I, serum amyloid P component,
streptococcal Arp and Sir proteins, and factor
VIII/VIIIa via its a-chains and with protein S
through its b-chain. The principal aim of the
present study was to localize regions of C4BP
involved in the interaction with C4b, Arp, and
heparin. For this purpose, a computer model of
the 8 CCP modules of C4BP a-chain was constructed, taking into account data from
previous electron microscopy (EM) studies.
This structure was investigated in the context of known and/or new experimental data.
Analysis of the a-chain model, together with
monoclonal antibody studies and heparin binding experiments, suggests that a patch of positively charged residues, at the interface between the first and second CCP modules, plays
an important role in the interaction between
C4BP and C4b/Arp/Sir/heparin. Putative binding sites, secondary-structure prediction for
the central core, and an overall reevaluation of
the size of the C4BP molecule are also presented. An understanding of these intermolecular interactions should contribute to the rational design of potential therapeutic agents
aiming at interfering specifically some of these
protein–protein interactions. Proteins 31:391–
405, 1998. r 1998 Wiley-Liss, Inc.
Key words: complement control protein; protein modeling; blood coagulation;
C4b-binding protein
Human C4b-binding protein (C4BP) is a highmolecular-weight glycoprotein (570 kDa) composed,
for the main isoform, of seven identical a-chains (549
residues each with three potential glycosylation
sites) and a single b-chain (235 residues with five
potential glycosylation sites).1,2 Like many other
complement regulators (e.g., factor H), C4BP contains several 60 amino acid long repeating modules
(59 repeats for human C4BP) termed complement
control protein (CCP) modules, also called short
consensus repeats (SCRs) or Sushi domains.3,4 The
C4BP a- and b-chains are made of eight and three
CCP modules, respectively, further continued by a
nonrepeat carboxy-terminal region of about 60 residues involved in polymerization of the chains. C4BP
has been studied by EM and X-ray scattering and
has been shown to have a spider-like organization.5–7
C4BP was first identified as a regulator of the
complement system8,9 and has been shown to be an
acute phase protein.10,11 Upon C4b binding, C4BP
interferes with the formation and decay of the classical pathway C3-convertase (C4b2a) and functions as
a cofactor to factor I in the proteolytic degradation of
C4b. In addition to C4b, C4BP is known to bind two
Grant sponsor: The Swedish Medical Research Council;
Grant numbers: 07143 and 11793; Grant sponsor: The Alfred
Österlund Trust; Grant sponsor: The Albert Påhlsson Trust;
Grant sponsor: The Göran Gustafsson Trust; Grant sponsor:
The King Gustav V and Queen Victoria Trust; Grant sponsor:
Ax:Son Johnsons Trust; Grant sponsor: The University Hospital, Malmö; Grant sponsor: The Louis Jeantet Foundation of
*Correspondence to: Bruno O. Villoutreix, Department of
Clinical Chemistry, University Hospital, Lund University, S205 02 Malmö, Sweden. E-mail: bruno.villoutreix@klkemi.mas.
Received 19 September 1997; Accepted 22 December 1997
plasma proteins: vitamin K-dependent protein S
(PS), an anticoagulant cofactor to activated protein C
(APC),2,12 and serum amyloid P component (SAP), a
protein found in all types of amyloid deposits, including those present in Alzheimer disease.13–15
The PS–C4BP interaction is of significant importance in coagulation because about 70% of PS is
noncovalently bound to C4BP and when complexed,
PS loses its APC cofactor activity. PS deficiencies
have been reported in patients suffering from thrombosis, and in this situation only the concentration of
free PS can be linked unambiguously to thrombotic
diseases.16 It is possible that C4BP binds to another
coagulation plasma protein, factor VIII/VIIIa.17 It
was recently shown that C4BP interacts with proteins Arp and Sir, two Ig-binding cell surface proteins from the Gram-positive bacterium Streptococcus pyogenes, both belonging to the antiphagocytic M
protein family.18 Such pathogens can cause suppurative and inflammatory infections of the skin and
throat and postinfection sequelae such as acute
glomerulonephritis. The C4BP–Arp or –Sir interaction has been suggested to increase virulence by
protection from complement attack.19 Finally, the
negatively charged heparin molecule binds to C4BP
and it was shown that this interaction competes with
the binding of C4b but not of PS.20,21
Several studies have aimed at localizing the binding sites for C4BP ligands. PS, via its sex hormone
binding globulin domain (SHBG),22 is the only known
protein that binds to the C4BP b-chain23,24 and this
interaction involves only the most amino-terminal
CCP module of the b-chain.25 The C4b-binding site
was first suggested to be in the middle of the a-chain
(6th and/or 7th CCP).26,27 However, more recent
studies have shown the importance of the Nterminal part (CCPs 1 to 3).28–30 The C4b- and
Arp/Sir-binding sites are partially overlapping and
involve CCP modules 1 to 3 of the a-chain.29 SAP has
been shown to bind to the central core fragment of
C4BP.15 The C4BP core fragment is obtained after
digestion of the molecule by chymotrypsin, which
cleaves the a-chain within CCP7, thereby liberating
most of the chain (CCPs 1 to 7) from the remaining
part of the protein.23 In contrast, the b-chain, also
digested by chymostrypsin,23 is not involved in the
C4BP–SAP interaction. The stoichiometry of the
SAP–C4BP and PS–C4BP interactions are 1:1,
whereas each a-chain, most likely, contains binding
sites for C4b, Arp/Sir, and heparin. However, at least
in the case of the C4BP–C4b interaction, only two to
four C4b molecules bind to C4BP, presumably due to
steric hindrance.31
Only limited information concerning the tertiary
structure of complement control proteins is available. However, the three-dimensional (3D) structure
of several CCP modules has been determined by
NMR spectroscopy.32–34 These modules share important structural similarity, with local differences,
mainly within loops, confirming that the consensus
sequence forming their hydrophobic core is of major
importance in conferring topological similarities. In
consequence, these NMR structures are valid templates for comparative modeling studies.35 CCP modules are characterized by the conservation of several
key residues, among which four cysteines disulfidebridged in a 1–3, 2–4 fashion. The consensus residues form a compact hydrophobic core that is surrounded by five b-strands.33
The study of C4BP using X-ray crystallography or
NMR spectroscopy has not been possible because of
the size of the molecule, the presence of several
isoforms, and heterogeneous glycosylation. Therefore molecular modeling methods seem very appropriate to study the 3D structure of this protein. In the
present investigation, comparative model building of
the eight CCP modules from the C4BP a-chain was
carried out, taking into account electron micrograph
data with the goal of identifying potential binding
sites, and with a special emphasis on the Arp/C4b/
heparin–C4BP interactions. For this purpose, interactive structural analysis of the C4BP a-chain model,
together with computation of some of its electrostatic
properties and theoretical enumeration of possible
binding sites, was undertaken. The interaction between Arp/C4b/heparin and C4BP was studied, experimentally guided by the analysis of the model
structure. The human C4BP a-chain model was also
analyzed in conjunction with previously reported
interspecies multiple sequence alignment.36,37 These
results allow us to propose molecular mechanisms
for some of the C4BP interactions.
Homology Modeling
The modeling of CCP modules from the C4BP
b-chain has been previously reported.35,38 The same
approach was applied for building the CCPs from the
a-chain and is briefly described below. From the 20
CCP modules of factor H, the 3D structures of CCPs
5, 16, and a 15–16 pair have been investigated by
NMR spectroscopy.32–34 The coordinates of these
modules were obtained from the Protein Data Bank
(PDB).39,40 A Silicon Graphics Indigo2 R10000 workstation was employed for the construction of the
models, together with the molecular modeling programs InsightII, Biopolymer, Homology, Discover,
and Delphi (Biosym-MSI, San Diego, CA, USA).
Model building
The sequence of each CCP module of the C4BP
a-chain was aligned to the factor H sequence (repeat
15th or 16th) according to Figure 1. The eight CCP
modules were first built independently. The NMR
coordinates from either H15 or H16 were used as
initial templates. The deletions in C4BP relative to
the reference proteins were refined by 200 cycles of
energy minimization, keeping the rest of the struc-
Fig. 1. Sequence alignment of the eight CCP modules from the
a-chain of human C4BP to the 15th or 16th factor H modules and
secondary-structure prediction for the nonrepeat central core
region. The dashed lines indicate the disulfide bridging pattern of
the C4BP a-chain CCP modules. The sixth CCP module has an
extra disulfide shown by a solid black line (see text). Cysteine
residues belonging to the CCP modules are numbered above the
sequences. The ‘$‘ symbol depicts consensus sequence for
N-glycosylation. C4BP modules were built using the presented
alignment, i.e., H15 was used as template for C4BP CCPs 1, 5,
and 6, while H16 was used for C4BP CCPs 2, 3, 4, 7, and 8. Black
lines above and below the C4BP nonrepeat sequences correspond to the PHD secondary structure prediction for the a- and
b-chains, respectively, and indicate a-helices.
ture fixed. The insertion regions were built from a
search41 among some of the high-resolution protein
structures present at the PDB. The minor steric
clashes and bond strains due to the introduction of
the new sidechains and loop building were regularized by a short energy minimization. Then, the first
and second CCP modules of the C4BP a-chain were
superimposed onto the H15–16 pair. H15 of a second
H15–16 pair was superimposed on the H16 component of the first pair. The third CCP of the a-chain of
C4BP was then superimposed on the H16 structure
of the second H15–16 pair. The process was repeated
until the a-chain was built. The model was then
briefly energy minimized.
Fig. 2. Validation of the model structure. a (left): Ribbon representation of the C4BP aCCP1 model structure. Richardson rendering for
the theoretical model of the first module of the a-chain. The disulfide
bridges are in blue, while some key residues conserved in the CCP
modules (hydrophobic core) are shown in magenta. The secondarystructure elements (b-sheets, in yellow) and other structural features
(turns, in green) were assigned according to the NMR template. The
remaining part of the structure is in white. b (above): Profile plot for
the entire a-chain model. The compatibility score for residues in a
21-amino acid sliding window is presented. The horizontal black line
represents the 0 value. Such profile indicates that the protein is well
Figure 3.
Fig. 4. C4BP a-chain model structure. a (left): Solid ribbon
representation of the a-chain. Overall representation of the model
structure with the cysteine residues shown in blue and asparagines, part of a consensus sequence for N-glycosylation, in
magenta. These asparagine sidechains are accessible to the
solvent in this model. The two chymotrypsin cleavage sites are
highlighted in red. b (right): The a-chain CCP1 and CCP2 model
structures. Some residues accessible to the solvent and forming
patches (hydrophobic, polar, positively charged) of potential impor-
tance for intermolecular interactions are presented (see text). The
NMR structure of a dodecamere heparin molecule65 was docked
manually onto the positive cluster in order to probe further the
possible C4BP–heparin interaction. K30, strictly conserved in the
sequences,37 could also be involved in the interaction with heparin. The distance between K30 and K79 is about 40 Å and the
length of the heparin molecule is about 50 Å. The insertion region
when compared to the NMR template (between C106 and C122) is
painted in green for orientation.
Fig. 3. Schematic diagram of human C4BP. Model structures
for the a- and b-chains were organized according to electron
micrograph images.5,6 The previously reported model for the three
CCP modules of the b-chain35 was rebuilt with the intermodule
angle stretched in a similar fashion as one of the a-chain. The
interaction between human C4BP and PS has been recently
narrowed to the first CCP module of the C4BP b-chain25 and to the
PS SHBG-like region.22,62 The PS binding site, by means of
synthetic peptides competition assays, was suggested to involve a
linear epitope (residues 31 to 45) on the first CCP of the C4BP
b-chain.63 Indeed, a cluster of solvent-exposed hydrophobic residues and two lysines at the surface of the first CCP module are
likely to play important roles in the PS–C4BP interaction.35
Monoclonal antibody 6F6, shown to interact with a linear epitope
involving C4BP b-chain residues 51–65, is also known to inhibit
the binding of PS to C4BP.64 Binding sites at the surface of the
a-chain discussed in this article are presented. The distances
reported are in Angstroms.
Theoretical Enumeration of Potential
Binding Sites
The method for identifying potential binding sites
is described in Young et al.,42 but a brief summary
will be presented here.
A simplified model consisting of only the Ca coordinates was used to represent the geometry of each
residue.42–46 The lattice model of Ca atoms for the
protein studied here fit the molecule coordinates
within 1-Å RMS deviation.47 Lattice positions exterior to the molecular surface can be examined for
their interactions with residues of the target molecule. This method can be thought as similar to
calculating molecular surfaces based on their accessibility to a water molecule probe.48 The probable
binding strength of each exterior position is determined by simply counting the number and type of
target Ca positions within 7.5 Å of any exterior
points. The amino acid composition of the defined
clusters thus scores the relative strength for which a
ligand might bind at that site.
Scoring binding strength
The binding strength for each exterior position is
the sum of scores for its constituent residues:
fcluster 5 S i51
where N is the number of neighboring (d , 7.5 Å)
residues. Parameters for each residue type were
based on assigning a hydrophobic value* determined
from residue-based contact energies as calculated by
Miyazawa and Jernigan.49 These contact energies
can be understood in terms of the hydrophobic–
hydrophilic designations of amino acids and the
pairings that contribute most to protein stability.44,46,49 Hydrophobicity assignments using this procedure show a strong correlation with the TanfordNozaki scale50 and others, as shown by Cornette et
al.51 Determination of scores for all surface regions
can be completed in less than one CPU minute on a
typical Silicon Graphics Workstation.
Electrostatic Potential
Electrostatic potential calculations were carried
out for the C4BP a-chain model structure. The
Delphi package, which solves the Poisson-Boltzmann equation by a finite difference method,52,53 was
used for the calculation of the 3D distribution of the
electrostatic potential at physiological ionic strength
and pH. The atomic coordinates of the model structure were mapped into a 3D grid with a resolution of
about one grid point/2 Å. The grid was chosen to
leave a 10-Å border between the protein and the grid
edge. The dielectric constant was set to 9 for the
protein interior and 80 for the surrounding solvent.
Atomic radii definitions were taken from the Delphi
default parameters with the radii of hydrogen atoms
set to zero. A standard set of formal charges was
assigned to the titratable residues (e.g., Arg, 10.5 e
at the Nh1 and Nh2 nucleus). The N-terminal amino
group was given a charge of 11 e while the Cterminal carboxy group was considered neutral. The
resulting 3D isopotential contours were analyzed
interactively within InsightII.
Secondary-Structure Prediction for Central
Core Nonrepeat Region
Secondary structure prediction was performed with
the Profile-fed neural network systems from HeiDelberg (PHD)54,55 for the nonrepeat region of the human C4BP a- and b-chains. This method was selected since it has an average cross-validated
accuracy of around 72% on a set of 250 sequences
with known structure. The entire human sequence of
the mature chains was provided as query input.
Experimental Design
C4BP5 and C456 were purified from human plasma
according to the procedure described in the respective references. C4(H2O) was used instead of C4b
and was prepared by repeated freezing and thawing
of purified C4.57,58 Protein Arp was a kind gift from
Dr. G. Lindahl. C4BP was radiolabeled using Iodobeads (Pierce Chemical Co., Rockford, IL, USA). The
monoclonal antibody against C4BP (mAb104) used
in this study has been described previously.30
Inhibition experiments
Microtiterplates were coated with C4(H2O) or Arp,
50 µl/well at concentrations of 10 µg/ml in coating
buffer. Increasing amounts of standard heparin,
low-molecular-weight heparin, or mAb104 in dilution buffer were then added, together with a trace
amount of 125I-labeled C4BP (20,000 cpm/well). The
final volume was 50 µl/well. After an overnight
incubation, the plates were washed and the amount
of bound C4BP detected using a g-counter.
Comparative Model Building
Overall validation of the C4BP model structure
*The values of ei are: F 25.12, M 24.91, I 24.88, L 24.65, W
24.36, V 24.17, C 24.00, Y 23.24, A 22.82, H 22.75, G 22.34,
T 22.30, P 22.22, R 22.18, S 22.07, Q 21.98, E 21.94, N
21.90, D 21.81, K 21.50.
All eight individual CCP models from the C4BP
a-chain were screened interactively and their 3D
structure as well as the distribution of the hydropho-
Fig. 5. Theoretical enumeration of potential binding sites.
Clusters of residues identified by theoretical mean as potential
binding site area as reported in Table I are mapped in a simplified
manner onto the C4BP model structure. The Ca atom of the potentially
N-glycosylated asparagines is shown in a CPK representation.
bic, polar, and charged residues were found in accordance with the NMR templates. The distribution of
polar and apolar atoms of a protein is a wellestablished way to assess the quality of a 3D structure.59 The compact hydrophobic core of the CCP
modules34 is found in all eight C4BP individual
models, indicating that, despite a relatively low
sequence identity between the NMR template and
C4BP, our structural prediction is accurate. As an
example, some key sidechains contributing to this
hydrophobic core are shown for the first CCP of the
Fig. 6. Electrostatic surface potential of the a-chain model.
The model structure is shown with the same orientation as the one
used for Figure 5. The electrostatic isosurfaces at a level of -0.6
(red) and 10.6 (blue) kcal/mol/e are shown. A positive region, at
the interface between aCCP1 and aCCP2, that play a role in the
interaction between C4BP and C4b/Arp/heparin is highlighted.
Other positively charged residues of possible interest are also
listed (see text).
a-chain (Fig. 2a). The accuracy of a protein structure
can also be assessed by inspection of its 3D profile.60
When the compatibility scores of a misfolded protein
are plotted as a function of the amino acid sequence,
the overall profile is often below 0.1 and can dip
several times below zero.60 We have run a similar
calculation for the entire C4BP a-chain model (Fig.
2b). The profile is always above zero and the selfcompatibility scores are essentially above 0.1, supporting the quality of our model structure. The
backbone angle values were investigated using
ProStat (Biosym-MSI) and were found compatible
with those of experimentally determined structures.
The relative orientation of the modules, however,
is difficult to predict, since the NMR data for the
H15–16 pair indicate flexibility between the two
repeats.32 A variable intermodule angle is further
supported after observations of the electron micrographs obtained for factor H61 and C4BP,6 where
some regions of the molecules appear twisted. However, it is not known if these bends are specific to
these proteins or induced by experimental procedures. We assumed that the overall orientation of
one CCP module relative to the next would be similar
to the one observed for the H15–16 NMR pair but
with the intermodule angle stretched in order to
match the overall length of the a-chain as deduced
from previous EM studies.5,6 One CCP module has
an ellipsoid shape, with a long axis of about 32 Å,
thus eight CCP modules should have an approximate theoretical length of 250 Å, to which about
10–15 Å could be added due to the presence of the
intermodule linker regions. Thus one would expect a
theoretical length of about 260–270 Å for an extended tentacle of eight CCPs. The a-chain homology
model proposed here has an extended conformation
with a length of about 260 Å (Fig. 3). An overall
length of 330 Å between the N-terminus of the
a-chain and the center of the central core has been
reported.5 The diameter of the core has been estimated to be around 60 Å [5].
Structural Analysis and Theoretical
Enumeration of Potential Binding Sites
The interactive structural analysis of the a-chain
(Fig. 4a,b) models, in conjunction with investigation
of the a-chain sequences (rabbit,36 rat,37 bovine,66,67
human,68 mouse69), as described by Hillarp et al.,37 is
reported below. Potential binding sites on the achains are listed in Table I and are graphically
represented in Figure 5.
CCP1 (E residues 1 to 63)
Three patches of exposed hydrophobic residues are
observed and are generally well conserved in the
C4BP sequences from different species. The first
patch involves residues P4, A12, P13, M14, I16, and
L18. The second displays F10, L34, Y37, and Y62.
These residues could shield another hydrophobic
patch located on aCCP2 involving F84, V108, V113,
and the C65–C106 disulfide. The third hydrophobic
patch contains V55, Y56, F59, I61, and aCCP2 F84
(Fig. 4b). This surface forms part of a theoretically
identified binding site (cluster h, Table I and Fig. 5),
together with four polar residues: S42, T43, T45, and
T47. The latter residues are included in an area
presenting numerous short and polar sidechains,
generally conserved in the sequences (residues T25,
T27, T28, S40, S50, T58). Charged residues are
unevenly distributed on this module, with D15, E20,
D51, E53, R22, and K24 at the N-terminal pole, K30
in the middle, and R39 and H41 at the C-terminal
pole. Among them, E20, R22, K24, K30, and R39 are
generally conserved in the interspecies sequence
analysis, or replaced by a long polar residue or an
amino acid of opposite charge.37
a CCP2 (E residues 64 to 125)
At the surface of aCCP2 a patch of positively
charged residues well conserved within the different
species 36,37 can be observed. This involves K63, R64,
R66, H67, and K7g with contributions from aCCP1
R39 and possibly H41 and K30 (Fig. 4b).Only E70
and D81 somewhat counterbalance these positive
charges. This area is part of a theoretically identified
binding site (cluster b, Table I and Figs. 5 and 6).
Close to this positive surface, a loop presenting with
a short insertion when compared to the NMR template forms also part of cluster b (Table I and Fig. 5).
Several short and polar sidechains, conserved in the
species, are also noted on one face of aCCP2. The
residues involved are T80, S83, S86, S91, S101,
T102, and T103 (Fig. 4b).
a CCP3 (E residues 126 to 189)
An insertion comprising residues 174–178 with
respect to the template is found at the interface with
aCCP2 (Fig. 1). A consensus sequence for N-glycosylation is present at residue N173 (Figs. 1 and 4a).
This residue is solvent-accessible in the model and is
located at the interface with aCCP2 in a region rich
in polar and charged residues (aCCP2 R72, N73,
Q121, E123, and aCCP3 E172, E174). One face of
this module displays several charged residues (K126,
K128, D132, R134, R137, H138, E141, E142, D156),
most often conserved in the C4BP sequences. Another region, at the interface with aCCP2, presents
several solvent-accessible hydrophobic residues,
aCCP2 F97, L98, I99, P120, V125, and aCCP3 K126
(carbon sidechain), F144, A146, Y147, F149. This
last set of residues is part of a potential binding site
(Table I, cluster d, and Fig. 5). A patch of Ser/Thr
residues is also noted on aCCP3 (conserved in the
species) and involves S139, S150, T152, S154, S166,
S168, and S182.
a CCP4 (E residues 190 to 249)
One face presents several charged residues not
well conserved between the species (R192, K193,
D195, E200, D236, K238), with possible contribution
from aCCP3 R158 (Table I, cluster c, and Fig. 5). A
face with several hydrophobic sidechains, which
tends to be conserved in the sequences, was also
TABLE I. Theoretical Enumeration of Potential Binding Sites for the a-Chain of Human C4BP
E405, C406, D407, Y410, I411,
L412, V413, Q415, A416, P430,
Q431, and C432.
G36, Y62, K63, R64, C65, R66, F84,
V108, Q109, D110, G112, V113,
and G114.
I189, T190, C191, R192, K193,
P194, D195, V196, S197, H198,
G199, and E200.
L98, I99, P120, Q121, C122, E123,
I124, V125, and Y147.
C351, H352, E353, C374, G375,
D376, I377, C378, Y422, and
R134, N135, D156, P157, R158,
F159, I189, T190, A235, and
S154, C155, L161, L162, G163,
H164, A165, and C186.
S42, T43, Q44, T45, L46, T47, V55,
Y56, N57, and T58.
F159, S160, K188, I189, T190, I208,
Y209, and N210.
On aCCP7, close to aCCP8
This area could interact with SAP
The second best ranking score was
mainly on aCCP2 and at the
interface with aCCP1
On aCCP4 with contribution of the
linker region between the third
and fourth repeats
At the aCCP2–aCCP3 interface
This region is nearby and part of a
patch of positively charged residues
Interface between aCCP6 and 7
Possibly nearby residues F97; 144
and 149 also play a role
This forms a small groove at the
surface of the molecule
Interface between aCCP3 and 4
This region is close from the third
best scoring cluster (c)
On aCCP3
Nearby the cluster f
On aCCP1
Nearby the second best scoring
cluster (b)
This area is nearby clusters f, g,
and c
Interface between aCCP3 and 4
observed. This surface involves mainly M201, F205,
P207, I208, V216, K218 (carbon sidechain), V224,
L225, and V230. However, these residues are not all
close in space but are distributed over one entire face
of the module.
a CCP5 (E residues 250 to 315)
This module is not present in mouse C4BP.37,68 The
only striking solvent-exposed patches involve a set of
charged residues and, on the opposite side of the
repeat, five threonines. These include residues E263,
K270, E271, D272, R281, R283, D293, E294, T291,
T292, T296, T297, and T307. Overall, these residues
tend to be conserved in the C4BP sequences.
a CCP6 (E residues 316 to 376)
This module is not present in mouse C4BP.37,68
aCCP6 has most likely an extra disulfide bridge
involving C316 with C339 (Figs. 1 and 4a), which
was modeled by assuming the bridging pattern
suggested by Norman et al.34 An interesting surface
involves charged residues E326, H330, R331, K332,
R334, H338, D345, H352, and E353. A small set of
Ser/Thr residues is also noted nearby this patch.
This surface contains T328, S348, S350, T354, S355,
and S358. These two patches are partially conserved
in the sequences. The only hydrophobic patch noted
at the surface of this module contains residues
V340, Y341, F342, and Y343 at the interface with
a CCP7 (E residues 377 to 434)
Two patches of charged residues are noted here,
one involving residues K399, E400, E401, and K417,
and on another face K383, H386, H388, K390, E405,
D407, K408, and K433, further continued by aCCP8
R437 and R481. A solvent-accessible hydrophobic
patch was found on this module and involves mainly
I403, C406, I411, L412, V413, and A416, further
continued on aCCP8 L435 and V454 (Table I, cluster
a, and Fig. 5). Chymotrypsin cleaves this module at
residues Tyr395 and Trp42523 (Fig. 4a).
a CCP8 (E residues 436 to 490)
Two N-glycosylation sites are predicted at N458
and N480 (Figs. 1 and 4a). N480 is somewhat more
shielded from the solvent than N458 but is still
accessible. N480 is at the interface with aCCP7 and
surrounded by five positively charged residues
(aCCP8 R437, R481, and aCCP7 H386, K408, K433,
further extended by K383, H388, K390). N458 has
also several polar and charged residues in its direct
vicinity (K450, E455, E457, T460, S474, T476). One
patch of charged amino acids is noted and involves
R437, K438, E440, R445, D449, D451, and D464.
These charged residues tend to be conserved. One
solvent-accessible hydrophobic patch is found and
involves residues V469, V470, P472, Y484, and P485,
which are close from three charged amino acids
generally conserved in the sequences, E486, K489,
and E491.
Fig. 7. Inhibition experiments. a: Increasing concentrations of
heparin was allowed to compete with 125I C4BP for binding to
immobilized C4(H2O) (U), protein Arp (N), and mAb104 (M).
100% binding was estimated in the absence of fluid-phase
competitor. b: Increasing concentrations of mAb 104 was allowed
to compete with 125I C4BP for binding to immobilized Arp. The data
were obtained after three different experiments.
Electrostatic Potential and Heparin Effects on
C4BP Interactions
tration of heparin was required to displace the
binding of Arp and mAb 104 as compared to C4(H2O).
Similar results were obtained with lowmolecular-weight heparin (data not shown). The
ability of mAb 104 to displace the binding of C4BP to
immobilized Arp was tested in a similar assay (Fig.
7). It was found that this antibody was an efficient
inhibitor of the Arp–C4BP interaction. Since the
epitope for mAb 104 is located either on the first
aCCP or at the interface between aCCP1 and
aCCP2,30 the above data suggests that C4b, Arp,
and mAb 104 are interacting in part with the positive cluster found on the first and second CCP
Three other positive patches that could be potential binding sites for heparin were found in the
a-chain. One is located on aCCP5 and involves K270,
R281, and R283. A site on aCCP6 could involve six
charged residues, H330, R331, K332, R334, H338,
and H352. The last most likely site for heparin
binding is at the interface between aCCP7 and
aCCP8 and could involve eight positively charged
residues, K383, H386, K390, K408, K433, R437, and
R481 (Fig. 6).
The overall electrostatic potential of the a-chain is
mosaic, but some modules tend to be covered either
by positive or negative surface. The most striking
surfaces are reported below. The interface between
the first and the second repeats is mainly positive at
the energy level computed (Fig. 6), as well as at -1 or
-2 kcal/mol/e (data not shown). This region corresponds to the conserved patch of positively charged
residues discussed above (Fig. 4b). Another electropositive region of possible interest is at the interface
between aCCP7 and aCCP8 and could be of importance for interaction with the mainly electronegative
b-chain35 or with other molecules such as SAP. The
negative surfaces are mainly at the N-terminal pole
of aCCP3, at the interface between aCCP5 and
aCCP6, and at the C-terminal pole of aCCP8. This
last region on aCCP8 could be of importance for
interactions with the nonrepeat segment of C4BP or
with SAP.
Areas presenting a richer content in positively
charged residues could be involved in specific or
nonspecific heparin binding. One site, well conserved
in the C4BP sequences (6 to 8 positively charged
residues), was discussed above and is located at the
interface between aCCP1 and aCCP2 (Figs. 4b and
6). A series of experiments were designed to evaluate
the role of this positively charged cluster in the
C4BP interaction with heparin and with Arp/C4b.
Increasing concentration of heparin was added,
together with a trace amount of 125I-labeled C4BP,
to microtiter plates coated with C4(H2O), Arp, or
mAb 104 (a monoclonal antibody known to compete with the C4BP–C4b interaction). As can be seen
in Figure 7, heparin was able to interfere with all
three interactions, although a higher concen-
Secondary-Structure Prediction for Central
Core Nonrepeat Region
The C4BP b- and a-chains contained only bstrands, as expected, for the region encompassing
the three and eight CCP modules, respectively. The
nonrepeat areas were found essentially helical (74%
for the b-chain and 53% for the a-chain) (Fig. 1). This
last point will be analyzed in the Discussion section.
C4BP is a complex and important regulatory protein of the complement cascade. In addition, its role
in the coagulation system is certainly of importance
since it is known that patients with low free PS level
are at higher risk of thromboembolic disease.16,70
The potential role of C4BP in some bacterial infectious processes has been pointed out recently.18,19 It
is known that many pathogenic microorganisms
have evolved mechanisms to avoid complement attack, in some cases by expressing a protein with
functional similarity to a complement regulatory
molecule or by acquiring such a protein from the
host.71 It has also been reported that C4BP exacerbates the inflammatory response to Escherichia coli
infusions in baboons.72 For these reasons, the study
of the 3D structure and mechanisms of actions of
C4BP are of importance.
In the present investigation we have developed a
model for C4BP using as template previously reported NMR structures of CCP modules from factor
H. Comparative model building73,74 is the best method
presently available to predict the 3D structure of a
protein75 and has already been used successfully to
investigate CCP modules.35,38,76–79 These modules
are found in several families of proteins and seem to
be well adapted for protein–protein interaction. The
interface area between these modules is nicely designed to ‘hook‘ approaching molecules.
Intermolecular associations involve generally interactions between the surface-accessible portions of
molecules and seem to be mainly determined by the
details of geometric and chemical complementarity.45 Recently, Young et al.42 have described a method
for finding regions on protein surfaces where a
residue-based hydrophobicity potential is high. These
regions generally coincide with binding sites of ligands. In addition, the graphical analysis of the
three-dimensional distribution of the electrostatic
potential can help in pointing toward regions of
functional importance.80–82
Analysis of a-Chain
Our investigation aimed at the localization of
residues possibly involved in heparin, C4b, and
Arp/Sir binding. Several studies support the importance of the amino-terminal part (CCPs 1 to 3) for
the binding of C4b, including EM analysis,6,30 binding assays with chimeric constructs,28–30 and sequence comparisons.36,37 A recent report indicates
that the C4b- and Arp/Sir-binding sites are partially
overlapping18 and involve CCP modules 1 to 3 of the
a-chain.29 Furthermore, it was shown that the interactions C4BP–C4b31 and C4BP–Arp29 are ionic
strength-dependent. These suggest that electrostatic forces could play a role in the complex formation, since at this concentration salts should essentially shield charges. This hypothesis is further
supported by our study, which shows that heparin
competes with the C4BP–Arp interaction, highlighting the importance of positively charged residues. It
is known that heparin competes also with the C4BP–
C4b interaction.20 Monoclonal antibody 10430 and
mAb G11H1129 seem to interact between aCCP1 and
aCCP2. These two antibodies inhibit the contact
C4BP–C4b. Moreover, the mAb 104–C4BP interaction is also displaced by heparin (Fig. 7). Taken
together, these data suggest that an important site of
interaction for C4b involves at least the positive
cluster at the interface between aCCP1 and aCCP2
(Fig. 4b). However, since a higher concentration of
heparin was necessary to displace the C4BP–Arp
interaction when compared to C4BP–C4b, it is expected that the binding sites are not entirely equivalent. This is supported by the fact that the positive
cluster between CCPs 1 and 2 is very conserved in
C4BP sequences, while only C4BP from human (or
closely related primates) interacts with Arp.29
We have compared the sequences of C4BP CCPs 1
and 2 with the ones of other C4b-binding proteins,
complement receptor 1 (CR1), and decay accelerating factor (DAF). There is an important amino acid
conservation in the proposed C4b-binding site. Interestingly, C4BP, CR1, and DAF possess a positively
charged cluster at the interface between two CCPs.
This cluster is located between CCPs 1 and 2 for
C4BP and CR1, and between CCPs 2 and 3 for DAF.
In addition, several residues forming human C4BP
cluster b (Table I) are conserved also in CR1 and
DAF. These residues are in the C4BP, numbering
G36, K63, C65, R66, F84, I108 (or V), and V113.*
Furthermore, in CR1, residues G35, R64, and N65
(C4BP G36, R66, and H67, respectively) have been
shown to be important for C4b recognition by sitedirected mutagenesis.83 Also, C4BP R66 and H67
have been found to be important in Arp and C4b
binding by mutagenesis.29 A model for the four CCPs
of DAF also suggests that a positive surface, together
with a hydrophobic patch mainly at the interface
between CCPs 2 and 3, is of importance for the
interaction with the C3 convertase.78 All these data
support that cluster b is a binding site for C4b and
The sequence expected to be responsible for the
interaction with C4BP on the Arp protein contains
several conserved negatively charged residues.19 This
Arp sequence could thus interact with the positive
cluster found between C4BP aCCP1 and aCCP2.
C4b residues 738 to 826 have been suggested to play
a role in the interaction with C4BP.27 The region
*The sequences compared were human, orangutan, mouse,
and guinea pig DAF; human, rabbit, rat, mouse, and bovine
C4BP; human, chimpanzee, and hamadryas baboon CR1; and
the mouse CRRY and CRY proteins.
encompassing C4b residues 740–757 contains a stretch
of negatively charged residues (740EILQEEDLIDEDDIPVRS757)27 that would seem well adapted for the
interaction with the aCCP1–aCCP2 positive cluster.
Theoretical analysis of the a-chain also suggests
that CCP3 and CCP4 contain potential binding sites
(Fig. 5). These regions could play some role in the
recognition of factor I, C4b, or Arp. Interestingly, no
binding sites were found on CCP 5 and CCP6, but
only at the interface between CCP 6 and CCP7 (Fig.
5). The fifth and sixth repeats are indeed missing in
the mouse sequence,69 suggesting that these two
modules may not be important for the function of the
molecule. This is further supported because a recombinant C4BP molecule with the three amino-terminal CCPs from mouse did bind human C4b.28 Monoclonal antibody 7F227 has been reported to interact
with aCCP6 and to inhibit the C4b–C4BP interaction. It seems likely that CCP6 does not contain the
binding site for C4b and that the competition observed is due to steric hindrance. The top ranking
score on the a-chain (Fig. 5 and Table I) could be
involved in an interaction with SAP since this cluster (cluster a) is nearby the central core (Figs. 3 and
4) and two potential glycosylation sites. It is known
that deglycosylation lowers the affinity for the C4BP–
SAP interaction.15
Basic Structure and Central Core of C4BP
The predominant form of human C4BP consists of
seven identical a-chains and a single b-chain linked
together by disulfide bonds within the C-terminal
nonrepeat region. Covalent bonds between the achains, however, are not necessary for association
since the cysteine residues in this region of the
molecule are not present in mouse C4BP, which also
lacks the b-chain.69,84 This indicates that noncovalent forces play an important role in the polymerization of the chains. Secondary-structure prediction for
the nonrepeat regions suggested this area to be
predominantly helical (Fig. 1). The central core was
shown via EM experiments to have a ring-like
structure with an inner diameter of 13 Å and an
outer diameter of 60 Å.5 Building part of the nonrepeat core as a-helices allows for the formation of a
structure presenting mainly one face with hydrophobic sidechains and another with hydrophilic amino
acids (Fig. 8). Taken together, these data suggest
that the central core nonrepeat region could be
arranged as a barrel with the hydrophobic face of the
helices packing against each other and the hydrophilic surfaces having contact with the solvent. Such
hydrophobic surfaces could explain why polymerization of the chains is independent of disulfide bridging.
Fig. 8. Amphipathic a-helix for regions of the nonrepeat central
core. Several segments from the nonrepeat central core area are
predicted to adopt an a-helical conformation (see Fig. 1). The
region of residues 515 to 533 from the a-chain and of residues 200
to 210 from the b-chain were built by assuming a standard helical
geometry. It can be seen clearly that one face of the helix displays
essentially hydrophobic residues while the other contains essentially hydrophilic amino acids. Additional information is needed,
however, to build the entire central core region.
The lack of 3D information for the nonrepeat core
and a-chain of C4BP prompted us to undertake the
present molecular modeling study. Most of the putative binding sites described here are in agreement
with the experimental data on record, validating the
overall accuracy of our model structure. A region
involved in the interaction of Arp/C4b/heparin with
C4BP, presenting positively charged residues, should
be located at the interface between the first and
second CCP modules of the a-chain. Heparin-like
molecules on the surface of cells could play a role in
the functioning of C4BP, such as local downregulation of the complement system. Polysulfated compounds and chemically modified peptides could be
used to inhibit undesired macromolecular interactions. Our results will be useful for further characterization of these protein assemblages and for investigation of the central core region. Site-directed
mutagenesis is presently undergoing in our laboratory to assess the role of the predicted binding sites.
We are grateful to Dr. G. Lindahl for providing Arp
1. Hillarp, A., Hessing, M., Dahlbäck, B. Protein S binding in
relation to the subunit composition of human C4b-binding
protein. FEBS Lett. 259:53–56, 1989.
2. Dahlbäck, B. Protein S and C4b-binding protein: Compo-
nents involved in the regulation of the protein C anticoagulant system. Thromb. Haemost. 66:49–58, 1991.
Reid, K.B.M., Bentley, D.R., Campbell, I.D., et al. Complement system proteins which interacts with C3b or C4b: A
superfamily of structurally related proteins. Immunol.
Today 7:230–234, 1986.
Janatova, J., Reid, K.B.M., Willis, A.C. Disulfide bonds are
localized within the short consensus repeat units of complement regulatory proteins: C4b-binding protein. Biochemistry 28:4754–4761, 1989.
Dahlbäck, B., Müller-Eberhard, H.J. Ultrastructure of
C4b-binding protein fragments formed by limited proteolysis using chymotrypsin. J. Biol. Chem. 259:11631–11634,
Dahlbäck, B., Smith, C.A., Müller-Eberhardt, H.J. Visualization of human C4b-binding protein and its complexes
with vitamin K-dependent protein S and complement
protein C4b. Proc. Natl. Acad. Sci. USA 80:3461–3465,
Perkins, S.J., Chung, L.B., Reid, K.B.M. Unusual ultrastructure of complement component C4b-binding protein of
human complement by synchrotron X-ray scattering and
hydrodynamic analysis. Biochem. J. 233:799–807, 1986.
Cooper, N.R. Isolation and analysis of the mechanism of
action of an inactivator of C4b in normal human serum. J.
Exp. Med. 141:890–903, 1975.
Shiraishi, S., Stroud, R.M. Cleavage product of C4b produced by enzymes in human serum. Immunochemistry
12:935–939, 1975.
Barnum, S.R., Dahlbäck, B. C4b-binding protein, a regulatory component of the classical pathway of complement, is
an acute-phase protein and is elevated in systemic lupus
erythematosus. Compl. Inflamm. 7:71–77, 1990.
Garcı́a de Frutos, P., Mohammed Alim, R.I., Härdig, Y.,
Zöller, B., Dahlbäck, B. Differential regulation of a- and
b-chains of C4b-binding protein during acute-phase response resulting in stable plasma levels of free anticoagulant protein S. Blood 84:815–822, 1994.
Dahlbäck, B., Stenflo, J. High molecular weight complex in
human plasma between vitamin K-dependent protein S
and complement component C4b-binding protein. Proc.
Natl. Acad. Sci. USA 78:2512–2516, 1981.
de Beer, F.C., Baltz, M.L., Holford, S., Feinstein, A., Pepys,
M.B. Fibronectin and C4-binding protein are selectively
bound by aggregated amyloid P component. J. Exp. Med.
154:1134–1139, 1981.
Schwalbe, R.A., Dahlbäck, B., Hillarp, A., Nelsestuen, G.L.
Independent association of serum amyloid P component,
protein S, and complement C4b with complement C4bbinding protein and subsequent association of the complex
with membranes. J. Biol. Chem. 265:21749–21757, 1990.
Garcı́a de Frutos, P., Härdig, Y., Dahlbäck, B. Serum
amyloid P component binding to C4b-binding protein. J.
Biol. Chem. 270:26950–26955, 1995.
Simmonds, R.E., Zöller, B., Ireland, H., et al. Genetic and
phenotypic analysis of a large (122 member) protein S
deficient knindred provides an explanation for the familial
coexistence of type I and III plasma phenotypes. Blood
89:4364–4370, 1997.
Koppelmann, S.J., van’t Veer, C., Sixma, J.J., Bouma, B.N.
Synergistic inhibition of the intrinsic factor X activation by
protein S and C4b-binding protein. Blood 86:2653–2660,
Thern, A., Stenberg, L., Dahlbäck, B., Lindahl, G. Igbinding surface proteins of Streptococcus pyogenes also
bind human C4b-binding protein (C4BP), a regulatory
component of the complement system. J. Immunol. 154:375–
386, 1995.
Johnsson, E., Thern, A., Dahlbäck, B., Hedén, L.-O., Wikström, M., Lindahl, G. A highly variable region in members
of the streptococcal M protein family binds the human
complement regulator C4BP. J. Immunol. 157:3021–3029,
Hessing, M., Vlooswijk, R.A., Hackeng, T.M., Kanters, D.,
Bouma, B. The localization of heparin-binding fragments
on human C4b-binding protein. J. Immunol. 144:204–208,
Garcı́a de Frutos, P., Dahlbäck, B. Interaction between
serum amyloid component and C4b-binding protein associated with inhibition of factor I-mediated C4b degradation.
J. Immunol. 152:2430–2437, 1994.
He, X., Shen, L., Malmborg, A., Smith, K.J., Dahlbäck, B.,
Linse, S. Binding site for C4b-binding protein in vitamin
K-dependent protein S fully contained in the carboxyterminal laminin-G-type repeats. A study using recombinant factor IX-protein S chimeras and surface plasmon
resonance. Biochemistry 36:3745–3754, 1997.
Hillarp, A., Dahlbäck, B. The protein S-binding site localized to the central core of C4b-binding protein. J. Biol.
Chem. 262:11300–11307, 1987.
Hillarp, A., Dahlbäck, B. Novel subunit in C4b-binding
protein required for protein S binding. J. Biol. Chem.
263:12759–12764, 1988.
Härdig, Y., Dahlbäck, B. The amino-terminal module of the
C4b-binding protein b-chain contains the protein Sbinding site. J. Biol. Chem. 271:20861–20864, 1996.
Chung, L.P., Reid, K.B.M. Structural and functional studies on C4b-binding protein, a regulatory component of the
human complement system. Biosci. Rep. 5:855–865, 1985.
Hessing, M., Kanters, D., Takeya, H., et al. The region
Ser333–Arg356 of the a-chain of human C4b-binding protein is involved in the binding of complement C4b. FEBS
Lett. 317:228–232, 1993.
Ogata, R.T., Mathias, P., Bradt, B.M., Cooper, N.J. Murine
C4b-binding protein. Mapping of the ligand-binding site
and the N-terminus of the pre-protein. J. Immunol. 150:
2273–2280, 1993.
Accardo, P., Sánchez-Corral, P., Criado, O., Garcı́a, E.,
Rodrı́guez de Córdoba, S. Binding of human complement
component C4b-binding protein (C4BP) to Streptococcus
pyogenes involves the C4b-binding site. J. Immunol. 157:
4935–4939, 1996.
Härdig, Y., Hillarp, A., Dahlbäck, B. The amino-terminal
module of the C4b-binding protein a-chain is crucial for the
C4b binding and Factor I-cofactor function. Biochem. J.
323:469–475, 1997.
Ziccardi, R.J., Dahlbäck, B., Müller-Eberhard, H.J. Characterization of the interaction of human C4b-binding protein
(C4BP) with physiological ligands. J. Biol. Chem. 259:
13674–13679, 1984.
Barlow, P.N., Steinkasserer, A., Norman, D.G., et al. Solution structure of a pair of complement modules by nuclear
magnetic resonance. J. Mol. Biol. 232:268–284, 1993.
Barlow, P.N., Norman, D.G., Steinkasserer, A., et al. Solution structure of the fifth repeat of factor H: A second
example of the complement control protein module. Biochemistry 31:3626–3634, 1992.
Norman, D.G., Barlow, P.N., Baron, M., Day, A.J., Sim,
R.B., Campbell, I.D. Three-dimensional structure of a
complement control protein module in solution. J. Mol.
Biol. 219:717–725, 1991.
Villoutreix, B.O., Fernández, J.A., Teleman, O., Griffin,
J.H. Comparative modeling of the three SCR domains of
the b-chain of C4b-binding and evaluation of potential sites
of interaction with protein S. Protein Eng. 8:1253–1258,
Garcı́a de Frutos, P., Dahlbäck, B. cDNA structure of rabbit
C4b-binding protein a-chain: Preserved sequence motive in
complement regulatory protein modules which bind C4b.
Biochim. Biophys. Acta 1261:285–289, 1995.
Hillarp, A., Wiklund, H., Thern, A., Dahlbäck, B. Molecular
cloning of rat C4b-binding protein a- and b-chains. J.
Immunol. 158:1315–1323, 1997.
38. Fernández, J.A., Villoutreix, B.O., Hackeng, T.M., Griffin,
J.H., Bouma, B.N. Analysis of protein S C4b-binding
protein interactions by homology modeling and inhibitory
antibodies. Biochemistry 33:11073–11078, 1994.
39. Abola, E.E., Bernstein, F.C., Bryant, S.H., Koetzle, T.F.,
Weng, J. Information content, software systems, scientific
applications. In: ‘Crystallographic Databases.‘ Allen, F.H.,
Bergerhoff, G., Sievers, R., (eds.). Bonn: Data Commission
of the International Union of Crystallography, 1987:107–
40. Bernstein, F.C., Koetzle, T.F., Williams, G.J.B., et al. The
Protein Data Bank: a computer-based archival file for
macromolecular structures. J. Mol. Biol. 112:535–542, 1977.
41. Jones, T., Thirup, S. Using known substructures in protein
model building and crystallography. EMBO J. 5:819–822,
42. Young, L., Jernigan, R.L., Covell, D.G. A role for hydrophobicity in protein–protein interactions. Protein Sci. 3:717–
729, 1994.
43. Levitt, M. A simplified representation of protein conformations for rapid simulation of protein folding. J. Mol. Biol.
104:59–107, 1976.
44. Covell, D., Jernigan, R. Conformations of folded proteins in
restricted space. Biochemistry 29:3287–3294, 1990.
45. Dill, K.A. Dominant forces in protein folding. Biochemistry
29:7133–7155, 1990.
46. Covell, D.G. Lattice model simulations of polypeptide
chain folding. J. Mol. Biol. 235:1032–1043, 1994.
47. Covell, D.G. Folding protein alpha-carbon chains into
compact forms by Monte Carlo methods. Proteins 14:192–
204, 1992.
48. Christenson, H.K., Claesson, P.M. Cavitation and the
interaction between macroscopic surfaces. Science 239:390–
392, 1988.
49. Miyazawa, S., Jernigan, R.L. Estimation of effective interresidue contact energies from protein crystal structures:
Quasi-chemical approximation. Macromolecules 18:534–
552, 1985.
50. Nozaki, Y., Tanford, C. The solubility of amino acids and
two glycine peptides in aqueous ethanol and dioxane
solutions. J. Biol. Chem. 246:2211–2217, 1971.
51. Cornette, J.L., Cease, K.B., Margalit, H., Spouge, J.L.,
Berzofsky, J.A., DeLisi, C. Hydrophobicity scales and computational techniques for detecting amphipathic structures in proteins. J. Mol. Biol. 195:659–685, 1987.
52. Gilson, M., Sharp, K., Honig, B. Calculating the electrostatic potential for molecules in solution. J. Comp. Chem.
9:327–335, 1987.
53. Klapper, I., Hagstrom, R., Fine, R., Sharp, K., Honig, B.
Focusing of electric fields in the active site of Cu–Zn
superoxide dismutase: Effects of ionic strength and amino
acid modification. Proteins 1:47–59, 1986.
54. Rost, B. PHD: Predicting one-dimensional protein structure by profile based neural networks. Methods Enzymol.
266:525–539, 1996.
55. Rost, B., Sander, C. Prediction of protein secondary structure at better than 70% accuracy. J. Mol. Biol. 232:584–
599, 1993.
56. Andersson, M., Hansson, A., Englund, G., Dahlbäck, B.
Inhibition of complement components C3 and C4 by cadralazine and its active metabolite. Eur. J. Clin. Pharmacol.
40:261–265, 1991.
57. von Zabern, I., Gigli, I. Conformational changes in complement component C4 induced by activation, treatment with
amines, chaotropes, or freezing–thawing, detectable by
radioiodination using lactoperoxidase. J. Immunol. 128:
1439–1442, 1982.
58. von Zabern, I., Bloom, E.L., Chu, V., Gigli, I. The fourth
component of human complement treated with amines or
chaotropes or frozen–thawed (C4b-like C4): Interaction
with C4-binding protein and cleavage by C3b/C4b inactivator. J. Immunol. 128:1433–1438, 1982.
59. Novotny, J., Bruccoleri, R., Karplus, M. An analysis of
incorrectly folded protein models. J. Mol. Biol. 177:787–
818, 1984.
Lüthy, R., Bowie, J.U., Eisenberg, D. Assessment of protein
models with three-dimensional profiles. Nature 356:83–85,
DiScipio, R.G. Ultrastructures and interactions of complement factors H and I. J. Immunol. 149:2592–2599, 1992.
Villoutreix, B.O., Garcı́a de Frutos, P., Lövenklev, M.,
Linse, S., Fernlund, P., Dahlbäck, B. The SHBG region of
the anticoagulant cofactor protein S: Secondary-structure
prediction, circular dichroism and analysis of naturally
occuring mutations. Proteins 29:478–491, 1997.
Fernández, J.A., Griffin, J.H. A protein S binding site on
C4b-binding protein involves b-chain residues 31–45. J.
Biol. Chem. 269:2535–2540, 1994.
Hessing, M., Kanters, D., Heijnen, H.F.G., Hackeng, T.M.,
Sixma, J.J., Bouma, BN. Structure function studies on
human C4b-binding protein using monoclonal antibodies.
Eur. J. Immunol. 21:2077–2085, 1991.
Mulloy, B., Forster, M.J., Jones, C., Davies, D.B. NMR and
molecular modeling studies of the solution conformation of
heparin. Biochem. J. 293:849–858, 1993.
Hillarp, A., Thern, A, Dahlbäck, B. Bovine C4b-binding
protein: Molecular cloning of the alpha- and beta-chains
provides structural background for lack of complex formation with protein S. J. Immunol. 153:4109–4199, 1994.
Hillarp, A., Dahlbäck, B. Cloning of cDNA coding for the
b-chain of human complement component C4b-binding
protein: Sequence homology with the a-chain. Proc. Natl.
Acad. Sci. USA 87:1183–1187, 1990.
Chung, L.P., Bentley, D.R., Reid, K.B.M. Molecular cloning
and characterization of the cDNA coding for C4b-binding
protrein, a regulatory protein of the classical pathway of
the human complement system. Biochem. J. 230:133–141,
Kristensen, T., Ogata, R.T., Chung, L.P., Reid, K.B.M.,
Tack, B.F. cDNA structure of murine C4b-binding protein,
a regulatory component of the serum complement system.
Biochemistry 26:4668–4674, 1987.
Comp, P., Esmon, C. Recurrent venous thromboembolism
in patients with a partial deficiency of protein S. N. Engl. J.
Med. 311:1525–1528, 1984.
Cooper, N.R. Complement evasion strategies of microorganisms. Immunol. Today 12:327–331, 1991.
Taylor, F.B., Dahlbäck, B, Chang, A.C., et al. Role of free
protein S and C4b-binding protein in regulating the coagulant response to Escherichia coli. Blood 86:2642–2652,
Blundell, T.L., Sibanda, B.L., Sternberg, M.J.E., Thornton,
J.M. Knowledge-based prediction of protein structures and
design of novel molecule. Nature 326:347–352, 1987.
Greer, J. Comparative modeling methods: Application to
the family of the mammalian serine proteases. Proteins 7:
317–334, 1990.
Sánchez, R., Sali, A. Advances in comparative protein
structure modeling. Curr. Opin. Struct. Biol. 7:206–214,
Rossi, V., Gaboriaud, C., Lacroix, M., et al. Structure of the
catalytic region of human complement protease C1s: Study
by chemical cross-linking and three-dimensional homology
modeling. Biochemistry 34:7311–7321, 1995.
Molina, H., Perkins, S.J., Guthridge, J., Gorka, J., Kinoshita, T., Holers, V.M. Characterization of a complement
receptor 2 (CDR2, CD21) ligand-binding site for C3. J.
Immunol. 154:5426–5435, 1995.
Kuttner-Kondo, L., Medof, M.E., Brodbeck, W., Shoham,
M. Molecular modeling and mechanism of action of human
decay-accelerating factor. Protein Eng. 9:1143–1149, 1996.
Mumenthaler C, Schneider U, Buchholz CJ, Koller D,
Braun W, Cattaneo R. A 3D model for the measles virus
receptor CD46 based on homology modeling, Monte Carlo
simulations and hemagglutinin binding studies. Protein
Sci. 6:588–597, 1997.
80. Honig, B., Nicholls, A. Classical electrostatics in biology
and chemistry. Science 268:1144–1149, 1995.
81. Villoutreix, B.O., Spassov, V.Z., Atanasov, B.P., Herve, G.,
Ladjimi, M.M. Structural modeling and electrostatic properties of aspartate transcarbamylase from Saccharomyces
cerevisiae. Proteins 19:230–243, 1994.
82. Karshikoff, A., Bode, W., Tulinsky, A., Stone, S.R. Electrostatic interactions in the association of proteins: An analy-
sis of thrombin–hirudin complex. Protein Sci. 1:727–735,
83. Krych, M., Hourcade, D., Atkinson, J.P. Sites within the
complement C3b/C4b receptor important for the specificity
of ligand binding. Proc. Natl. Acad. Sci. USA 88:4353–4357,
84. Rodrı́guez de Cordoba, S., Pérez-Blas, M., Ramos-Ruiz, P.,
Sánchez-Corral, P., Pardo-Manuel de Villena, F., ReyCampos J. The gene coding for the beta-chain of C4bbinding protein (C4BPB) has become a pseudogene in the
mouse. Genomics 21:501–509, 1994.
Без категории
Размер файла
866 Кб
Пожаловаться на содержимое документа