| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Microbiology and Infectious Diseases, Flinders Medical Center, Flinders University, Adelaide, Australia;
Immunity et Infections Virales, IVMC, Lyon, France;
The Rockefeller University, New York, NY 10021;
Department of Immunology, Institute of Medical and Veterinary Sciences, Adelaide, Australia; and
¶ Austin Research Institute, Heidelberg, Australia
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
GAS possess numerous virulence factors but their M protein is regarded
as one of the most important (5). Streptococcal M proteins
appear as long hairlike filaments on the bacterial surface and are
composed of two 
helical chains which are predominantly arranged in
a coiled-coil conformation (6, 7). All of the M proteins
contain a conserved C-terminal region near the cell surface and a
hypervariable N-terminal region which provides the basis for
serological typing (5).
M protein is involved in the attachment of GAS to skin epithelial cells, which is a vital first step in the establishment of impetigo and cellulitis. The M protein receptor on keratinocytes is membrane cofactor protein (CD46) (8), which is a regulator of the complement system that functions as a cofactor for factor I-mediated inactivation of the opsonins C3b and C4b. M protein is also critical for protecting organisms from phagocytosis. In the absence of type-specific Abs, strains of GAS expressing M protein fail to activate either the classical or alternative pathways of complement. M-negative organisms, however, activate complement and are thus efficiently opsonized and phagocytosed (9, 10, 11, 12). M6 and other M serotypes bind three plasma complement regulators: C4b-binding protein (C4BP) and the C3b-binding proteins, factor H (fH) and factor H-like protein 1 (fHL-1). These interactions are thought to contribute to the antiphagocytic properties of M protein by regulating complement activation on the GAS surface (13, 14, 15).
CD46, C4BP, fH, and fHL-1 belong to a family of complement regulatory
proteins that share a structurally conserved motif termed the short
consensus repeat (SCR). Each SCR is composed of 
60 aa, contains four
invariant cysteines forming two intradomain disulfide bonds, and a
linker region of three to eight residues joining the adjacent domains
(16). CD46 possesses four SCR domains and, in addition, a
region rich in serine, threonine, and proline, a region of 12 aa of
unknown function, a transmembrane region, and a short cytoplasmic tail
(17). C4BP contains - and 
-chain subunits of eight
and three SCRs, respectively. The most common isoform of C4BP consists
of seven -chains and one -chain linked by disulfide bonds
(18). It has been demonstrated that the M protein
binding site within C4BP resides in SCRs 1 and 2 of the -chain
(19, 20). fH is composed of 20 SCRs (21) and
fHL-1, an alternatively spliced variant of fH, consists of fH SCRs
1–7, with four additional hydrophobic amino acids at the carboxyl
terminus (22). Binding of M protein to fH and fHL-1 is
mediated by SCR 7 (23, 24, 25) and recently an additional
hydrophobic binding site at the carboxyl terminus of fHL-1 has also
been suggested (26).
In this study, the adherence of M6 GAS to the human keratinocyte cell
line HaCat was confirmed as being M protein dependent. A major M
protein binding site was localized to SCRs 3 and 4 by use of CD46 SCR
deletion and substitution mutants. A molecular model of SCRs 3 and 4
revealed similarity to the M protein binding sites identified on fH SCR
7 (27) and C4BP -chain SCRs 1 and 2 (28),
allowing prediction of potential residues in CD46 involved in M protein
binding. Furthermore, the M protein- and C3b-binding sites were
identified as being distinct since CD46 cofactor activity was not
inhibited by excess M protein, and the presence of excess C3b did not
inhibit the ability of CD46 to bind M protein.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
M6 protein was purified from the periplasm of transformed Escherichia coli as previously described (29). The final preparation was shown to be homogenous based on Coomassie blue-stained SDS-polyacrylamide gels, Western blotting using a monoclonal anti-M6 protein Ab, and N-terminal amino acid sequence analysis. The GAS strains examined in this study included JRS4, an M6 protein-expressing organism, and JRS145, an M6 protein-deficient derivative of JRS4 (30, 31).
Cloning and expression of soluble CD46 and CD46/CD55 chimeras
Soluble recombinant CD46 protein and soluble chimeric proteins
(isoform STP BC), lacking the transmembrane and cytoplasmic sequence,
were expressed in transfected human kidney fibroblasts (293 cell line)
(32). Constructs encoding the chimeric proteins, in which
the third and/or fourth SCR domains were replaced with the
corresponding SCRs from CD55, were generated from cDNA constructs by
splice-overlap extension PCR and mutagenesis to insert a stop codon
before the transmembrane sequence (33). The proteins
made were soluble CD461–4,
CD461–2CD553–4,
CD461–3CD554, and
CD461–2,4CD553 (Fig. 1
).
|
Truncated CD46 SCR deletion mutants consisting only of SCRs 2–4
and SCRs 3 and 4, labeled CD462–4 and
CD463–4 respectively (Fig. 1
), were prepared by
PCR using CD46 cDNA template (kindly supplied by Dr. D. Lublin,
Washington University, St. Louis, MO). Forward primers contained
EcoRI restriction sites and the common reverse primer
contained a stop codon and XbaI restriction site. Primer
sequences, with restriction sites shown in lower case, were SCR 2
forward, TCCgaattcAGAGAAACATGTCCATATAT; SCR 3 forward,
TCCgaattcAAGGTTTTGTGTACACCAC; and SCR 4 reverse,
GCtctagaTCAAAGACACTTTGGAACTG. The resulting PCR products were cloned
into the pPICZA yeast expression vector (Invitrogen, Carlsbad, CA),
downstream of the AOX1 methanol-inducible gene promoter. DNA sequencing
on an automated Applied Biosystems 373A fluorescent sequencing system
using the PRISM Ready Reaction dideoxy dye terminator cycle sequencing
kit (PerkinElmer Applied Biosystems, Foster City, CA) was
performed across the entire CD462–4 and
CD463–4 inserts. The plasmids were then
electroporated into the methylotrophic yeast, Pichia
pastoris (strain GS115) according to the manufacturer’s protocol.
Expression of recombinant proteins was induced for 4 days by the
presence of 1–2% methanol.
Purification of recombinant CD46 proteins
Recombinant CD46 proteins were purified on affinity columns made by coupling either the mouse anti-CD46 mAb E4.3 (34) or the IgG fraction of polyclonal rabbit anti-CD46 antiserum to cyanogen bromide-activated Sepharose (Amersham Pharmacia Biotech, Amersham, U.K.) using the manufacturer’s protocol. Supernatants (100–500 ml) from the transfected 293 cell line and P. pastoris expression cultures were passed over the affinity columns. After extensive washings with PBS, bound proteins were eluted, respectively, with diethylamine (34) or 3 M potassium thiocyanate and dialyzed immediately into 50 mM phosphate buffer. The P. pastoris eluates were concentrated in a stirred ultrafiltration cell (Amicon, Beverly, MA). The concentration of soluble CD461–4 was determined by absorbance at 280 nm using an extinction coefficient value of 0.92 (34). The specific concentration of the CD46/CD55 chimeras and the truncated CD46 SCR deletion mutants was determined by serial dilutions against a standard concentration of soluble CD461–4 in a dot blot assay or ELISA.
Western blotting
The identity of recombinant CD46 proteins was confirmed by Western blotting. Proteins were separated on a nonreducing 12.5% SDS-polyacrylamide gel and transferred to Hybond C+ nitrocellulose (Amersham Pharmacia Biotech). Membranes were incubated with rabbit polyclonal anti-CD46 Ab (1:10,000 v/v), followed by HRP-conjugated anti-rabbit IgG (1:2000 v/v; Silenus, Hawthorn, Australia) and washed between incubation steps with 50 mM phosphate buffer. Labeled proteins were finally detected by ECL (Amersham Pharmacia Biotech).
Adherence assay
The binding of M6-positive GAS and M6-negative GAS to the human
keratinocyte cell line HaCat was performed as previously described
(35). Briefly, HaCat cells were cultured on glass
coverslips. The M6-positive or M6-negative GAS were grown overnight at
37°C under 5% CO2 in brain-heart
infusion medium (Oxoid, Basingstoke, U.K.), washed in PBS, and
resuspended in serum-free DMEM (HyClone Laboratories, Logan, UT) at an
A600 = 0.05 (4 x
107 bacteria/ml). The HaCat-covered glass
coverslips were incubated for 2 h with 500 µl of the bacterial
suspension, and nonadherent organisms were removed by washing with PBS.
The coverslips were immersed in ethanol, glued to microscope slides,
Gram stained, and finally examined under oil immersion. Five random
fields containing >100 HaCat cells were examined and the number of
bacteria bound per HaCat cell was determined. Experiments were repeated
six times in duplicate and the means were calculated.
Dot blot analysis of the interaction between recombinant CD46 and M6 protein
Purified M6 protein (5 µg) and bovine albumin (5 µg; CSL, Victoria, Australia) were dried onto Hybond C+ nitrocellulose membranes. After blocking with 5% skim milk for 1 h, the membranes were sequentially incubated with recombinant CD46 (0.5 µg/ml) for 2 h, rabbit polyclonal anti-CD46 Ab (1:10,000 v/v) for 1 h, and then HRP-conjugated anti-rabbit IgG (1:2000 v/v) for 1 h. CD46 and Abs were all diluted in 1% skim milk in 50 mM phosphate buffer. Membranes were washed between incubation steps with 50 mM phosphate buffer and bound proteins were detected by ECL.
ELISA analysis of the interaction between CD46 and M6 protein
Purified M6 protein (1 µg) or bovine albumin (1 µg) in 100 µl of 100 mM bicarbonate buffer (pH 9.5) was coated overnight to wells of Maxisorb ELISA plates (Nunc, Roskilde, Denmark). Nonspecific sites were blocked with 200 µl of 5% skim milk for 1 h. Plates were then sequentially incubated with 100 µl of recombinant CD46 at 10 µg/ml for 2 h, rabbit polyclonal anti-CD46 Ab (1:4000 v/v) for 1 h, and HRP-conjugated anti-rabbit IgG (1:2000 v/v) for 1 h. CD46 and Abs were all diluted in 1% skim milk in 50 mM phosphate buffer. Wells were washed between each incubation with 50 mM phosphate buffer. Substrate was added and the A490 was determined. Experiments were repeated four times in duplicate and the means and SDs were calculated.
The possible effect of C3b on the CD46-M6 protein interaction was studied in a similar assay except that the binding of M6 protein (1 µg) to immobilized CD46 (1 µg) was determined in the presence or absence of 500 µg of C3b (kindly supplied by Dr. T. Seya, Osaka, Japan). M6 protein was detected by sequential incubations with monoclonal anti-M6 protein Ab (10 µg/ml) for 1 h and HRP-conjugated anti-mouse IgG (1:1000 v/v) for 1 h. M6 protein and Abs were all diluted in 1% skim milk in 50 mM phosphate buffer.
ELISA analysis of the interaction between CD46 and GAS
M6-positive and M6-negative GAS were grown overnight in brain-heart infusion medium without agitation at 37°C under 5% CO2. Cells were harvested by centrifugation, washed twice with PBS, and resuspended in 100 mM bicarbonate buffer (pH 9.5) to achieve 108 bacteria/ml. Subsequently, 100 µl of each suspension was aliquoted into wells of an ELISA plate and dried overnight at 37°C. Wells were blocked with 5% skim milk, washed with 50 mM phosphate buffer, and then sequentially incubated with 100 µl of recombinant CD46 (2.5 µg/ml) for 1 h, rabbit polyclonal anti-CD46 Ab (1:4000 v/v) for 30 min, and HRP-conjugated protein A (1:2000 v/v; Amersham Pharmacia Biotech) for 30 min. CD46 and Abs were all diluted in 1% skim milk in 50 mM phosphate buffer. Wells were washed between each incubation with 50 mM phosphate buffer. Substrate was added and the A490 was determined. Experiments were repeated four times in duplicate and the means and SDs were calculated.
Molecular modeling of SCR domains 3 and 4 of CD46
Homology-based molecular modeling of CD46 SCRs 3 and 4 was
performed using the experimentally determined tertiary structure of
CD46 SCRs 1 and 2 as template (Brookhaven Protein Data Bank, PDB, entry
1CKL) (37) and the Insight II program package (Molecular
Simulations, San Diego, CA) and Indigo 2 work station (Silicon
Graphics, Mountain View, CA). Nonhomologous residues were replaced and
a total of three loops was needed to equalize the number of amino acids
in the model and the template. Loop 1 was constructed for residues
40–47 of the model (spliced from PDB entry 1AP8), loop 2 for residues
62–64 (from 1ACC), and loop 3 for residues 113–117 (from 1AZO). The
interdomain angles were adapted from the CD46 SCRs 1 and 2 template
structure. The preliminary model structure was subjected to energy
minimizations by gradually relaxing the molecule: first, only hydrogens
were allowed to move; second, only side chains were allowed to move;
and third, all of the atoms were allowed to move, also allowing the
interdomain angle to change. In all energy minimizations, a conjugate
gradient algorithm with CFF91 force field was used until the maximal
derivative was <4.2 J/Å. Subsequently, the structure was soaked in a
waterbox of 2091 water molecules (box dimensions, 60 x 46 x
31 Å) to achieve an 6-Å-thick layer of water around the model. The
soaked model was again subjected to energy minimizations by gradually
relaxing the structure. Thereafter, the assembly of the model and the
water molecules was subjected to 10 picoseconds of molecular dynamics
simulation at 100,000 and 50 picoseconds at 300,000 under periodic
boundary conditions as described previously (38). A low
potential energy structure was chosen and processed as described
previously (39).
The achieved model of CD46 SCRs 3 and 4 was then compared with the
previously published molecular models of two M protein binding human
proteins, fH SCRs 6 and 7 (27, 40, 41) and C4BP -chain
SCRs 1 and 2 (20). In the comparison, the known M
protein-binding residues of fH SCRs 6 and 7 and C4BP -chain SCRs 1
and 2 and the corresponding positively charged residues on CD46 SCRs 3
and 4 model were highlighted. The illustrations were designed with
WebLab ViewerPro 3.5 (Molecular Simulations). The model structure of
CD46 SCRs 3 and 4 has been deposited in the Brookhaven PDB (entry
1HR4). The model structure of C4BP -chain SCRs 1 and 2 was kindly
provided by Dr. A. Blom (Mälmo, Sweden).
Cofactor assay
The cofactor activity of recombinant CD46 was assayed in 20 mM phosphate buffer (pH 6.0) in a final volume of 30 µl containing 0.25 µg of C3b, 1 µg of factor I (Calbiochem-Novabiochem, La Jolla, CA), 40 ng of soluble CD461–4, and either buffer or 2.5 µg of M6 protein (final concentration of 83.5 µg/ml). Reactions that did not contain factor I or soluble CD461–4 were used as negative controls. The reactions were incubated at 37°C for 3 h. Samples were separated under reducing conditions in a 7.5% polyacrylamide gel, transferred to a nitrocelluose membrane, and sequentially incubated with 5% skim milk for 30 min, goat anti-C3c Ab (1:1000 v/v; Silenus) for 30 min, and HRP-conjugated anti-goat IgG (1:2000 v/v; Calbiochem-Novabiochem) for 30 min. The membrane was washed between incubations with PBS and protein was detected by ECL.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To confirm the role of M protein in the adherence of M6 GAS to host cells, we examined the binding of M6-positive and M6-negative GAS strains to the human keratinocyte cell line HaCat. M6-positive organisms adhered to HaCat cells with an average of 60 ± 23 organisms/cell, whereas the M6-negative organisms failed to bind in significant numbers (average, 6 ± 2 organisms/cell). This indicates that adherence of M6 GAS to HaCat cells is M protein dependent.
Recombinant CD46 mutants
It has been previously shown that CD46 is the receptor on
keratinocytes responsible for M protein binding (8). To
map the M6 protein binding site, we used recombinant soluble CD46 and
soluble CD46/CD55 chimeras, in which the transmembrane and cytoplasmic
domains of CD46 were deleted, and for the chimeras, SCRs 3 and/or 4 of
CD46 were replaced with the corresponding SCR domains from CD55
(CD461–4,
CD461–2CD553–4,
CD461–3CD554, and
CD461–2, 4CD553). In
addition, truncated CD46 SCR deletion mutants,
CD462–4 and CD463–4, were
constructed (Fig. 1). Each protein
migrated on SDS-PAGE according to its predicted molecular mass,
allowing for N- and O-linked glycosylation. The
P. pastoris produced CD463–4 SCR
deletion mutant migrated as two bands (25 and 30 kDa) under nonreducing
conditions (Fig. 2), but as a single band
under reducing conditions. This may be due to misfolding of a portion
of the protein, resulting in aberrant migration of one band under
nonreducing conditions.
|
The M protein binding site within CD46 was mapped by a number of
different assays. We initially used soluble
CD461–4 and the CD46/CD55 chimeras and
determined their M protein-binding characteristics by dot blot. In this
assay, M6 protein or albumin was immobilized onto nitrocellulose before
incubation with the test proteins. Soluble
CD461–4 bound to M6 protein but not to the
negative control, albumin. The chimeras,
CD461–2CD553–4,
CD461–3CD554, and
CD461–2,4CD553, exhibited
no consistent binding to M6 protein or albumin (Fig. 3). These preliminary data indicated that
CD46 SCRs 3 and 4 are responsible for M protein binding.
|
). These data indicate that CD46
SCRs 3 and 4 alone, in the absence of adjacent SCRs, can mediate M
protein binding.
|
).
|
M protein binding sites have been mapped for C4BP (19, 20), fH (23, 24, 25), and in this report for CD46 to
-chain SCRs 1 and 2, SCRs 6 and 7, and SCRs 3 and 4, respectively.
The sequence similarity between these M protein-binding SCR was
determined by sequence alignments using the ClustalW alignment program
(Genetics Computing Group, Madison, WI) with manual corrections
to align structurally important cysteine residues. The overall sequence
identity of the domains ranged from 35 to 46%. The highest similarity
was between CD46 SCR 4 and fH SCR 7 (46%), followed by CD46 SCR 3 and
fH SCR 7 with 41% identity. The identity and homology between C4BP
-chain SCRs 1 and 2 and fH SCRs 6 and 7 or CD46 SCRs 3 and 4 was
lower (<35% and <39%, respectively).
In a previous study, we used site-directed mutagenesis to identify four
residues within fH SCR 7 and the linker region between fH SCRs 6 and 7
(Arg369, Lys370,
Arg386, Lys387) to be
critical for M protein binding (41). Blom et al.,
(20) used site-directed mutagenesis to identify four
residues within C4BP -chain SCRs 1 and 2 to be involved in M protein
binding (Lys63, Arg64,
Arg66, and His67). An amino
acid sequence alignment of the M protein-binding domains of fH, C4BP,
and CD46 revealed one common feature: two or three positively charged
residues within or next to the linker region between the two domains.
In addition, SCR 4 of CD46 contained two positively charged residues
(Lys210, Lys211) in the
same location as the M protein-binding residues on SCR 7 of fH
(Arg404, Lys405). In the
primary structures of fH SCRs 6 and 7 or CD46 SCRs 3 and 4, there were
phenylalanine residues corresponding to the position of the M protein
binding His67 of C4BP (Fig. 6).
|
To model the tertiary location of the positively charged residues
that were suggested by the sequence alignments to be involved in M6
protein binding, we constructed a molecular model of CD46 SCRs 3 and 4.
The experimentally determined tertiary structure of CD46 SCRs 1 and 2
was used as a template. The locations of the suggested M6
protein-binding residues and other positively charged residues were
visualized using a structure that had been subjected to energy
minimizations and a total of 60 picoseconds of molecular dynamics
simulations. Four positively charged residues
(Lys190, Lys193,
Arg195, and Lys210) were
found to form a cluster of positive charges at the interface of CD46
SCRs 3 and 4 (Fig. 7), similar to that
observed at the interface of fH SCRs 6 and 7 (27) and C4BP
-chain SCRs 1 and 2 (20, 28). In all three M
protein-binding complement regulators, the positively charged residues
formed a semicircle, which contained one phenylalanine residue.
|
The primary function of CD46 is cofactor activity for proteolytic
cleavage of C3b and C4b, with SCR domains 3 and 4 shown to bind C3b
(42). As SCRs 3 and 4 have been identified to bind both
C3b and M protein, we assessed whether these ligand binding sites
within SCRs 3 and 4 are distinct or overlapping. The effect of M
protein binding on cofactor activity of CD46 was determined with a C3b
fluid-phase cofactor assay in the presence of a 30-fold molar excess of
M protein, using conditions in which limited C3b cleavage occurred. C3b
is composed of an '-chain (107 kDa) and -chain (75 kDa) and, as
expected, in the presence of factor I and CD46 the '-chain was
cleaved at two sites to generate fragments of 67, 42, and 40 kDa (Fig. 8). In the presence of excess M protein,
no reduction of CD46 cofactor activity was detected, suggesting that
the binding sites on CD46 for M protein and C3b are distinct (Fig. 8).
To further confirm this finding, we performed an ELISA in which CD46
was immobilized to the wells of an ELISA tray and then incubated with M
protein in the presence or absence of a 30-fold molar excess of C3b. No
reduction in binding of M protein to CD46 in the presence of C3b was
observed (mean binding of M6 in the presence of C3b was 103% relative
to control).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The M6 protein receptor on keratinocytes has been reported to be CD46 (8), a widely expressed complement regulator (45, 46). A number of other pathogens have also been identified which utilize CD46 as a receptor to initiate infections. These include measles virus (47), which binds CD46 SCRs 1 and 2 (48), human herpesvirus 6 (49), Neisseria gonorrhoeae, and Neisseria meningitidis (50). The broad tissue expression of CD46 provides opportunities for wide tissue tropism of disease; however, it is not clear whether this is a factor in pathogen-host receptor selection.
To further examine the interaction between M protein and CD46, we used a number of CD46/CD55 chimeric SCR replacement mutants (CD461–2CD553–4, CD461–3CD554, and CD461–2, 4CD553) and examined their ability to interact with M protein. Replacement of either CD46 SCR 3 and/or 4 with the corresponding SCR domains of CD55 largely diminished binding to M6 protein, indicating that both of these SCR domains are essential for full binding. As previous mutagenesis studies have revealed that some SCRs involved in binding interactions require flanking SCRs to allow full functional activity (51), we therefore determined the minimum number of SCRs that are required for M protein recognition. Truncated CD46 SCR deletion mutants consisting of SCRs 2 and 4 and 3 and 4 were therefore constructed and both were shown to mediate binding, indicating that SCRs 3 and 4 contain a major binding site for M protein without a need for adjacent domains.
M proteins of GAS strains also bind other structurally related
complement regulators including fH and C4BP via SCRs 6 and 7 and
-chain SCRs 1 and 2, respectively (19, 20, 23, 24, 25).
Previously, molecular models of both fH SCRs 6 and 7 and C4BP -chain
SCRs 1 and 2 had been constructed to define M protein binding sites.
Both models defined a cluster of positively charged residues at the
interface of the SCR domains, which by site-directed mutagenesis have
been identified to be involved in M protein binding (20, 27, 41). His67 of C4BP has also been
identified to be involved in M protein binding (20). To
further delineate the residues of CD46 involved in M protein binding,
we constructed a homology-based molecular model of CD46 SCRs 3 and 4
and compared it with the models of fH SCRs 6 and 7 and C4BP -chain
SCRs 1 and 2. The model of CD46 SCRs 3 and 4 similarly contained
positively charged residues in an equivalent position (Fig. 7). Thus,
it is probable that these residues are directly involved in the binding
of M6 protein to CD46. Our finding that both the SCR 3 and SCR 4
domains of CD46 are needed for full M6 binding is in agreement with
this hypothesis, since the chimeric
CD461–3CD554 protein lacks
three of these four positively charged residues, whereas the
CD461–2,4CD553 chimera
introduces a negatively charged (Glu191) residue
between two positively charged residues (Lys190
and Lys193).
In addition to the positively charged semicircle found on all three models of the M protein-binding domains, we identified one phenylalanine residue within the semicircle of each model. To ascertain the role of these phenylalanine residues in M protein binding, mutagenesis studies are needed. Since His67 of C4BP has been shown to be necessary for efficient M protein binding by C4BP and since the M protein-C4BP interaction is relatively insensitive to high salt concentration (up to 1.5 M), it seems that hydrogen bonds and hydrophobic interactions are involved in this interaction (20). Combined with the fact that phenylalanine is the most frequently found residue involved in protein-protein interactions in general (52), it is relatively likely that the structurally conserved phenylalanine in C4BP, fH, and CD46 is involved in M protein binding.
Our predictions regarding the residues in CD46 involved in M protein binding do, however, require some caution. Although in general each single domain of a multidomain protein can be accurately modeled if good quality homologous templates are used, the prediction of the rotation and torsion angles between two SCR domains can be problematic. The possible rotation and torsion angles vary considerably in different SCR-containing proteins. In our CD46 SCRs 3 and 4 model, we have used a relatively long molecular dynamics simulation at 300,000 to allow possible movement of the SCR domains to find one possible conformation for the SCR pair. Our model shows one of several possible interdomain torsion and rotation angles and the distribution of surface charge associated with it. Additional experimental mutagenesis studies are required to confirm that the residues, Lys190, Lys193, Arg195, Phe196, and Lys210 are involved in M protein binding.
In addition to M protein binding, SCRs 3 and 4 of CD46 are also involved in the C3b/C4b binding and cofactor activity (42). In this study, we demonstrate that 1) CD46 maintains its C3b cofactor activity in the presence of excess M protein and 2) excess C3b does not inhibit the ability of CD46 to bind M protein by ELISA, suggesting that the binding sites on CD46 for M protein and C3b are distinct. These results, however, need confirmation with specific mutagenesis studies but are similar to the finding with fH, in which cofactor and M protein binding sites are clearly distinct (24, 51). Assuming the fluid-phase assay utilizing soluble proteins predicts what occurs for cell surface receptor-ligand interaction, these results also suggest that wild-type CD46 would retain functional activity after binding to GAS via M6 protein. Thus, CD46 may not only act as a receptor on host cells for the establishment of streptococcal infections, but it also may be involved in protection of the organism from destruction by the host complement system.
The results suggesting the separate location of M protein and C3b binding sites on CD46 and our prediction of the CD46 residues involved in M protein binding (Lys190, Lys193, Arg195, Phe196, and Lys210) are in partial agreement with a recent study by Liszewski et al. (53). By alanine-scanning mutagenesis, it was shown that replacement of either Lys193, Arg195, Phe196, or Lys210 did not abolish C3b binding but decreased binding to 37, 50, 35, and 68%, respectively. A Lys190 mutant was not constructed. Of these residues, only the replacement of Lys193 or Phe196 with alanine abrogated the cofactor activity of CD46. However, the inhibition caused by alanine substitutions of Lys193, Arg195, or Phe196 might be due to an altered orientation between the SCR 3 and 4 domains, since these residues are located at the hinge region. In contrast and, more importantly, other substitutions in CD46 SCRs 3 and 4 that were outside our predicted M protein site resulted in the complete loss of C3b binding and cofactor activity. To determine whether the suggested four positively charged residues and one hydrophobic residue on CD46 are in fact important for M protein binding, further analyses are needed with amino acid substitution constructs.
In summary, we have shown that the binding of CD46 to serotype M6 GAS requires M protein. We have also shown that CD46 SCRs 3 and 4 contain a major M protein binding site and the M protein/CD46 interaction 1) does not measurably inhibit proteolytic cleavage of C3b and 2) is not inhibited by C3b. Our results therefore suggest that CD46 not only acts as a receptor on host cells for M6 GAS adherence but might also function by regulating complement activation on GAS and have a direct pathogenic role. On the basis of sequence comparisons and molecular modeling, we additionally propose that Lys190, Lys193, Arg195, Phe196, and Lys210 are involved in M protein binding. However, further studies are needed to locate the precise residues involved in the M protein-CD46 interaction.
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. David Gordon, Department of Microbiology and Infectious Diseases, Flinders Medical Center, Bedford Park, SA-5042, Australia. E-mail address: D.Gordon{at}flinders.edu.au
3 Abbreviations used in this paper: GAS, group A streptococcus; C4BP, C4b-binding protein; fH, factor H; fHL-1, factor H-like protein 1; SCR, short consensus repeat.
Received for publication February 2, 2001. Accepted for publication February 15, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-helical coiled-coil structure and arrangement on the cell surface. Proc. Natl. Acad. Sci. USA 78:4689.
light chain dimer: a unique human miniautoantibody against complement factor H. J. Immunol. 163:4590.This article has been cited by other articles:
|
|
 |
|
  A. H. Nobbs, R. J. Lamont, and H. F. Jenkinson Streptococcus Adherence and Colonization Microbiol. Mol. Biol. Rev., September 1, 2009; 73(3): 407 - 450. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Fisher, Y.-S. Huang, X. Li, K. S. McIver, C. Toukoki, and Z. Eichenbaum Shr Is a Broad-Spectrum Surface Receptor That Contributes to Adherence and Virulence in Group A Streptococcus Infect. Immun., November 1, 2008; 76(11): 5006 - 5015. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Lovkvist, H. Sjolinder, R. Wehelie, H. Aro, A. Norrby-Teglund, L. Plant, and A.-B. Jonsson CD46 Contributes to the Severity of Group A Streptococcal Infection Infect. Immun., September 1, 2008; 76(9): 3951 - 3958. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. K. Zaiss, M. J. Cotter, L. R. White, S. A. Clark, N. C. W. Wong, V. M. Holers, J. S. Bartlett, and D. A. Muruve Complement Is an Essential Component of the Immune Response to Adeno-Associated Virus Vectors J. Virol., March 15, 2008; 82(6): 2727 - 2740. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. S. Jokiranta, Z.-Z. Cheng, H. Seeberger, M. Jozsi, S. Heinen, M. Noris, G. Remuzzi, R. Ormsby, D. L. Gordon, S. Meri, et al. Binding of Complement Factor H to Endothelial Cells Is Mediated by the Carboxy-Terminal Glycosaminoglycan Binding Site Am. J. Pathol., October 1, 2005; 167(4): 1173 - 1181. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Fleischli, S. Verhaagh, M. Havenga, D. Sirena, W. Schaffner, R. Cattaneo, U. F. Greber, and S. Hemmi The Distal Short Consensus Repeats 1 and 2 of the Membrane Cofactor Protein CD46 and Their Distance from the Cell Membrane Determine Productive Entry of Species B Adenovirus Serotype 35 J. Virol., August 1, 2005; 79(15): 10013 - 10022. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. Price, J. Schaumburg, C. Sandin, J. P. Atkinson, G. Lindahl, and C. Kemper Induction of a Regulatory Phenotype in Human CD4+ T Cells by Streptococcal M Protein J. Immunol., July 15, 2005; 175(2): 677 - 684. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Gaggar, D. M. Shayakhmetov, M. K. Liszewski, J. P. Atkinson, and A. Lieber Localization of Regions in CD46 That Interact with Adenovirus J. Virol., June 15, 2005; 79(12): 7503 - 7513. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. McRae, T. G. Duthy, K. M. Griggs, R. J. Ormsby, P. J. Cowan, B. A. Cromer, W. J. McKinstry, M. W. Parker, B. F. Murphy, and D. L. Gordon Human Factor H-Related Protein 5 Has Cofactor Activity, Inhibits C3 Convertase Activity, Binds Heparin and C-Reactive Protein, and Associates with Lipoprotein J. Immunol., May 15, 2005; 174(10): 6250 - 6256. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Pandiripally, L. Wei, C. Skerka, P. F. Zipfel, and D. Cue Recruitment of Complement Factor H-Like Protein 1 Promotes Intracellular Invasion by Group A Streptococci Infect. Immun., December 1, 2003; 71(12): 7119 - 7128. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. L. Greenstone, F. Santoro, P. Lusso, and E. A. Berger Human Herpesvirus 6 and Measles Virus Employ Distinct CD46 Domains for Receptor Function J. Biol. Chem., October 11, 2002; 277(42): 39112 - 39118. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) or albumin
(
) was coated onto wells
of an ELISA plate. After blocking nonspecific sites with 5% skim milk,
the wells were sequentially incubated with the mutant CD46 proteins,
polyclonal rabbit antisera against human CD46, and HRP-conjugated
anti-rabbit IgG. Substrate was added and the OD was read at 490 nm.
Experiments were repeated four times in duplicate and the means and SD
were calculated.
