|
|
||||||||


*
The Medical Research Council Immunochemistry Unit, Department of Biochemistry, University of Oxford, Oxford, United Kingdom, and
Department of Biochemistry, University of Toronto, Toronto, Ontario, Canada
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
2-macroglobulin
(
2M)4
constitute a family of proteins with a high degree of sequence
similarity and similar gene organization. They are thought to have
arisen by gene duplication from a common ancestor (1). C3, C4, and
2M share the unusual structural feature of an intrachain
thioester bond, which enables them to bind covalently to "target"
molecules (2, 3, 4). Upon proteolytic activation, the thioester becomes
"exposed" on the surface of the molecule and reactive with
nucleophiles. Ultimately, a covalent bond is formed between the acyl
group of the thioester and the amino or hydroxyl groups of the target.
The thioester proteins show significant differences in their binding
preferences. C3 and the C4B isotype of human C4 predominantly form
ester bonds with hydroxyl groups on carbohydrates or proteins, whereas
the C4A isotype and
2M primarily form amide bonds with
proteins (5, 6). An investigation of the binding specificities of the
two human C4 isotypes has revealed that a short sequence of 4 amino
acids, which lies
100 residues C-terminal of the thioester site,
accounts for the difference in binding specificity (7). Mutagenesis
studies have shown that a His residue at position 1106 in C4B catalyzes
the binding reaction of C4B with hydroxyl-bearing substrates, including
water (8, 9). We have demonstrated that the mechanism of the reaction
(Fig. 1
2M lack a His residue
at the relevant position, and their reaction with hydroxyls is not
catalyzed. In this case, the thioester is relatively long-lived and is
attacked directly in a noncatalytic way, thereby explaining the
preference for the more nucleophilic amino groups (9). Like C4B, C3 has
a His at the equivalent position (1126 in human C3, prepro-C3
numbering), reacts with hydroxyl groups, and has an activated
t1/2 that is too short to be determined.
However, the covalent binding of C3 to amino groups is significantly
less efficient than C4B. Thus, although it is tempting to extrapolate
the same chemical mechanism for C4B to account for the covalent binding
activities of C3, it is necessary to obtain independent experimental
evidence for C3 itself.
|
| Materials and Methods |
|---|
|
|
|---|
Mutation of the human C3 cDNA was performed by the gapped plasmid method (12) in the vector pSVC3 (13) and inserted into the expression vector pEE6.HCMV.GS (Celltech, Slough, U.K.) (6). Stable Chinese hamster ovary K1 cell lines expressing human C3 and its variants (C3-H1126A, C3-H1126D, and C3-H1126K) were established as previously described for C4 (9, 10). The expression of C3 was assayed by an inhibition ELISA. A total of 100 µl of C3-containing tissue culture supernatant was added to 100 µl of rabbit anti-human C3 (diluted 1:60,000 in PBS, 0.1% BSA) and incubated at 37°C for 1 h. The mixture was transferred to microtiter plate wells that had been precoated with 100 ng of C3 per well (Polysorb, Nunc, Life Technologies, Paisley, U.K.). The binding of the Ab to the plate was detected using alkaline phosphatase-conjugated goat anti-rabbit IgG and was quantitated by comparison with a standard containing a known amount of C3.
Purification of proteins
Using ion-exchange chromatography on a Q-Sepharose Fast Flow
column (16 mm x 200 mm) (Pharmacia, Uppsala, Sweden) equilibrated
with 20 mm Tris/HCl, 50 mM
-aminocaproic acid, 5 mM EDTA, 0.1 mM
Pefablock, and 50 mM NaCl (pH 7.4), expressed C3 variants were purified
from the tissue culture supernatants of stable transfectants grown for
10 days after reaching confluence. The protein was eluted with a 220 ml
gradient to 500 mM NaCl in the same buffer. Fractions containing C3
were loaded onto a Mono-S column (Pharmacia) (5 mm x 50 mm) that
had been equilibrated with 20 mM
2-[N-morpholino]ethanesulfonic acid and 50 mM NaCl (pH
6.0) and eluted with a 20 ml linear gradient to 500 mM NaCl. The
C3-containing fractions were run on a Mono-Q column (Pharmacia) using
the same buffer system described for the Q-Sepharose column; elution
was with a linear gradient from 50 to 500 mM NaCl over 20 ml. Plasma
C3, C4, and C1s were prepared as described previously (14, 15).
Binding reactions
[2-3H]glycerol (1 Ci/mmol), [2-3H]glycine (17.9 Ci/mmol), and [methyl-3H]methylamine hydrochloride (75 Ci/mmol) were obtained from Amersham (Little Chalfont, U.K.). The concentration of active protein was determined by incorporating [3H]methylamine (200 mCi/mmol) into the intact thioester bond of C3 or C4 (16). The covalent binding of C3 and C4 to [3H]glycerol (10 mM; 200 mCi/mmol) and [3H]glycine (0.15 mM; 200 mCi/mmol) was determined in 10 mM sodium phosphate, 140 mM NaCl, and 1 mM EDTA (pH 7.2) using human C1s (0.1% w/w) to activate C4 and trypsin (1% w/w) to activate C3; incubation was for 15 min at 37°C.
C3-H1126K characterization
Plasma C3 and C3-H1126K were activated with trypsin (1% w/w) in 10 mM sodium phosphate, 140 mM NaCl, and 1 mM EDTA (pH 7.2) in the presence of [2-3H]iodoacetic acid (70 µM; 145 mCi/mmol) (Amersham) for 15 min at 37°C. Cold iodoacetic acid (final concentration of 10 mM) was added, and the material was dialyzed extensively against 100 mM Tris (pH 8.0) to remove excess [3H]iodoacetic acid. The resulting material was reduced, alkylated, and digested with trypsin (2% w/w) for 16 h at 37°C. The tryptic peptides were fractionated by reverse phase HPLC on an Aquapore OD300(C18) column (Applied Biosystems, Warrington, U.K.) (100 mm x 2 mm) in 0.2% trifluoroacetic acid over a linear gradient of acetonitrile to 60% at a flow rate of 0.2 ml/min. Peaks were collected manually, and those containing counts were sequenced on an Applied Biosystems 473A gas-phase sequencer.
| Results |
|---|
|
|
|---|
Three C3 variants in which His1126 was
substituted with Ala, Asp, or Lys were created by site-directed
mutagenesis at the cDNA level. The cDNA of the variant C3
molecules and that of wild-type (wt) C3 was introduced into the
expression vector pEE6.HCMV.GS, and the plasmids were transfected
into Chinese hamster ovary cells. The recombinant C3 proteins were
recovered from the supernatants and were purified by conventional
chromatographic techniques. The yields of active proteins were in the
range of 3 to 7 mg/l of tissue culture supernatant. As judged by
SDS-PAGE, the proteins were correctly processed, having the same chain
structure as plasma C3. [3H]methylamine could be
incorporated into the C3
-chains, indicating the presence of an
intact thioester.
The binding properties of the recombinant C3 molecules with glycine and
glycerol as representative small molecules with amino and hydroxyl
groups, respectively, were studied and compared with C3 and C4A that
had been purified from plasma (Table I
).
The results are expressed as
k2/k0, which is the ratio
of the rate constants governing transacylation to the model substrate
and water, respectively. Plasma C3 and recombinant wt C3 bound to
glycerol, but their reaction with glycine was very low, even at a
glycine concentration of 5 mM. When the two mutant C3 molecules in
which the His residue was replaced either by Ala or Asp were tested,
glycerol binding was undetectable; however, close to 100% binding was
observed at a glycine concentration of 5 mM. The
k2/k0 values could be
established using lower concentrations of glycine. In both cases, the
k2/k0 values with glycine
were lower than that seen with C4A but significantly higher than that
observed with the plasma C3 or recombinant wt C3. Thus, replacing
His1126 with either Ala or Asp converted the binding
characteristics of the C3 from a hydroxyl group (glycerol) binding
molecule to an amino group (glycine) binding molecule.
|
In our previous study on C4A and C4B, we demonstrated that the
residue at position 1106, if it was nucleophilic, attacks the thioester
to form an intermediate. In the case of His in C4B, this intermediate
was so short-lived that it was impossible to observe directly. However,
a mutant in which His was replaced by Lys formed a stable amide bond
during activation, and we were able to purify and characterize the
cross-linked peptide (Fig. 2
). Therefore,
we created a C3 mutant in which His was replaced with Lys. The
C3-H1126K mutant was activated with trypsin in the
presence of [3H]iodoacetic acid, which labeled the free
sulfhydryl group released from the thioester upon activation. After the
removal of excess [3H]iodoacetic acid by dialysis,
C3-H1126K was fully digested with trypsin, and the peptides
were isolated by reverse phase HPLC. The peptide containing the
thioester sequence was followed by radioactivity, which eluted from the
column at 54% acetonitrile. Four sequences (Table II
) were obtained in this fraction;
peptides 1 through 3 can be identified as having originated from C3.
Peptide 4 was from bovine fetuin, which was most likely a
contaminant from the FCS in the tissue culture medium. Peptide 1
contains the thioester sequence, but the glutamyl residue of the
thioester at position 12 was undetectable. Peptide 2 contains
Lys1126 at position 15, which was also undetected. This
observation is consistent with our conjecture that Lys1126
was covalently linked to the Gln1013 of the thioester upon
activation. In a similar experiment, plasma C3 was treated under
identical conditions; the radioactivity associated with the
thioester-containing peptide was eluted at an acetonitrile
concentration of 50%, i.e., at a different position in the gradient
from the C3-H1126K thioester peptide. Sequence analysis
revealed that this peptide from plasma C3 had the sequence
HLIVTPSGCGEENMIGMTP, where Gln1013 was
identified as the residue Glu (underlined) at position 12. This
sequence is to be expected when the thioester is hydrolyzed upon
activation.
|
|
| Discussion |
|---|
|
|
|---|
The x-ray crystal structure of a C3-C1010A derivative of
the human C3d fragment has recently been determined (44), which shows
that the alignment of the side chains of Ala1010
(Cys1010 in the native molecule), Gln1013, and
His1126 is fully consistent with the catalyzed
transacylation mechanism initially proposed for C4B (10) and
demonstrated here for C3. In the model of thioester-intact C3d
calculated from the structure of the modified protein, little movement
of the peptide backbone is required for thioester formation. The
His1126 ring nitrogen is positioned
4 Å from the
carbonyl carbon of the thioester bond. Therefore, some local
conformational change is needed to permit the nucleophilic attack
required to form the acyl-imidazole intermediate.
Phylogenetic analysis of the sequence data indicates that the
thioester-containing complement proteins and the protease inhibitor
2M have evolved from a common ancestor (1). Most
2M sequences, including the recently published
Limulus
2M (17), have an Asn at the position
equivalent to His1126 of C3 (Table III
) and are consequently predicted to
exhibit a C4A-like binding specificity. However, the role of the
covalent binding reaction in the
2M proteins is unclear.
Only in the case of the monomeric
-macroglobulin, such as
1I3 of the rat, is covalent binding necessary for
protease inhibitory activities. The polymeric
2M
proteins rely on the physical entrapment of proteases by a Venus
fly-trap-like mechanism for protease inhibition (41).
|
2M is an unusual protease inhibitor, in that it is
active against a very wide range of proteases. This activity is
possible because the protein possesses a stretch of
30 amino acids
that includes cleavage sites for proteases of all classes and with many
specificities. Cleavage of any bond in this area, which is called the
bait region, causes a conformational change in the protein and
activation of the thioester, leading to entrapment and inhibition. The
physiologic roles of
2M are manyfold, but they include
the clearance of both the intracellular proteases released by tissue
damage and the proteases produced by invading microorganisms. The
multiplicity of cleavage sites in the bait region allows proteases to
be inhibited that have never before been encountered. At some time in
evolution, it is possible that the ancestor of present day
2M could have become bound not to the activating
bacterial proteases but also to the protease-producing organisms, and
that this protein could then have acted as an opsonin. Indeed, present
day
2M proteins, together with their receptors, have an
opsonic function in the clearance of proteases and cytokines, as well
as in the delivery of Ags to APCs (41). What then was the advantage of the mutation that led to the substitution of His for Asn and the acquisition of the catalyzed binding reaction, which has been conserved in all of the C3 proteins for which data are currently available? The answer is probably twofold: the increased efficiency of the binding to hydroxyl substrates and the increased rate of the hydrolysis of the activated thioester. Efficient binding to hydroxyl substrates has an obvious advantage in the immune system, since most pathogenic microorganisms are coated with a carbohydrate-rich cell wall. In addition, the thioester-binding reaction is promiscuous in the sense that, depending upon whether the catalyzed or noncatalyzed transacylation mechanism is followed, any available hydroxyl or amino group, including those on the tissues of the host, can act as acceptors. A complement component with a noncatalyzed binding reaction and, consequently, an extended t1/2 for the reactive thioester would be able to diffuse away from the site of activation and opsonize the tissues of the host. The catalyzed binding reaction ensures that binding occurs only in close proximity to the initial site of activation. It is interesting to note that in the monomeric macroglobulins, which evolved as a branch from the tetrameric macroglobulins, both Asn and His are found at the site of interest (23, 24, 25). It may not be incidental that the monomeric macroglobulins rely on covalent binding to inhibit protease activities; for the same reasons, His may have evolved as a completely distinct event from that seen in the complements (25).
To date, only primates, cattle, and sheep have been shown to possess an isotype of C4 (i.e., C4A) displaying a noncatalytic covalent binding reaction (27, 42). C4 is not as potentially hazardous to the host as is C3. For classical pathway C3-convertase formation to occur, C4 must be deposited close to the activating C1, as C2 must bind to the C4 before it is in turn cleaved by C1 (43). Bound C3, on the other hand, can form an active convertase anywhere, because once factor B binds to C3, it is activated by factor D, which is present in solution and is not localized to the activating surface. A C3 with a noncatalyzed binding reaction could consequently result in considerable damage to surrounding tissues. The presence of a His residue in C3 and C4 allows the regulation of covalent bond formation, drastically reducing the t1/2 of the activated form and thus favoring the binding to hydroxyl groups in the vicinity of the C3 and C4 convertase enzymes that have been recruited to the surface of pathogenic organisms. Thus, the biologic significance of the His is to enable the functioning of both C3 and C4. In many species, the C4 gene has become duplicated, probably independently in different species (27). In some species, such as primates and bovidae, His has been replaced by Asp in one of the duplicated C4 genes and the A form of C4 has emerged. Although this replacement may facilitate C4 transacylation to hydroxyl-poor target molecules, it is noteworthy that at least one gene expressing the B form of C4 has been retained in these species, perhaps to ensure the presence of a C4 with carbohydrate-binding specificity.
| Acknowledgments |
|---|
| Footnotes |
|---|
2 Current address: Department of Biochemistry, Sofia University, 8 Dragan Tzankov, 1421 Sofia, Bulgaria. ![]()
3 Address correspondence and reprint requests to Dr. S. K. Alex Law, The Medical Research Council Immunochemistry Unit, Department of Biochemistry, University of Oxford, South Parks Road, Oxford, OX1 3QU, U.K. E-mail address: ![]()
4 Abbreviations used in this paper:
2M,
2-macroglobulin; wt, wild-type. ![]()
Received for publication January 29, 1998. Accepted for publication March 23, 1998.
| References |
|---|
|
|
|---|
2-macroglobulin and complement components C3 and C4. Proc. Natl. Acad. Sci. USA 82:9.
2-macroglobulin. FEBS Lett. 123:145.[Medline]
2-macroglobulin. Complement 5:89.[Medline]
-macroglobulin
1I3. FEBS Lett. 368:87.[Medline]
2-macroglobulin. Eur. J. Biochem. 242:822.[Medline]
2-macroglobulin and assignment of the chromosomal locus. Proc. Natl. Acad. Sci. USA 82:2282.
2-macroglobulin. Proc. Natl. Acad. Sci. USA 81:7353.
1-macroglobulin. Biochemistry 31:2346.[Medline]
2-macroglobulin mRNA. J. Biol. Chem. 260:14224.
-Glu thiol ester. Biochim. Biophys. Acta 1162:230.[Medline]
1-inhibitor 3. Eur. J. Biochem. 164:375.[Medline]
-macroglobulin family. J. Biochem. 120:1167.
-chain. Immunol. Invest. 15:365.[Medline]
2-macroglobulin family. J. Immunol. 154:2164.[Abstract]
2-macroglobulin, complement, and biologic defense: antigens, growth factors, microbial proteases, and receptor ligation. Lab. Invest. 71:792.[Medline]
This article has been cited by other articles:
![]() |
J. Mutsuro, N. Tanaka, Y. Kato, A. W. Dodds, T. Yano, and M. Nakao Two Divergent Isotypes of the Fourth Complement Component from a Bony Fish, the Common Carp (Cyprinus carpio) J. Immunol., October 1, 2005; 175(7): 4508 - 4517. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Boshra, A. E. Gelman, and J. O. Sunyer Structural and Functional Characterization of Complement C4 and C1s-Like Molecules in Teleost Fish: Insights into the Evolution of Classical and Alternative Pathways J. Immunol., July 1, 2004; 173(1): 349 - 359. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. Jelezarova, A. Luginbuehl, and H. U. Lutz C3b2-IgG Complexes Retain Dimeric C3 Fragments at All Levels of Inactivation J. Biol. Chem., December 19, 2003; 278(51): 51806 - 51812. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Verschoor, M. A. Brockman, M. Gadjeva, D. M. Knipe, and M. C. Carroll Myeloid C3 Determines Induction of Humoral Responses to Peripheral Herpes Simplex Virus Infection J. Immunol., November 15, 2003; 171(10): 5363 - 5371. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Lin, D. R. Sutherland, W. Horsfall, N. Totty, E. Yeo, R. Nayar, X.-F. Wu, and A. C. Schuh Cell surface antigen CD109 is a novel member of the alpha 2 macroglobulin/C3, C4, C5 family of thioester-containing proteins Blood, March 1, 2002; 99(5): 1683 - 1691. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |