| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Institute of Enzymology, Biological Research Center, Hungarian Academy of Sciences, Budapest, Hungary;
Department of Biochemistry, Eötvös Loránd University, Budapest, Hungary; and
Laboratoire d’Enzymologie Moléculaire, Institut de Biologie Structurale Jean-Pierre Ebel, Grenoble, France
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
region
interacting with other subcomponents and C-terminal 
B region
mediating proteolytic activity. The B region consists of two
complement control protein modules (CCP1, CCP2) and a serine protease
domain (SP). To clarify the role of the individual domains in the
structural and functional properties of the B region we produced the
CCP1-CCP2-SP (B), the CCP2-SP, and the SP fragments in recombinant
form in Escherichia coli. We successfully renatured the
inclusion body proteins. After renaturation all three fragments were
obtained in activated form and showed esterolytic activity on synthetic
substrates similar to each other. To study the self-activation process
in detail zymogen mutant forms of the three fragments were constructed
and expressed. Our major statement is that the ability of
autoactivation and C1s cleavage is an inherent property of the SP
domain. We observed that the CCP2 module significantly increases
proteolytic activity of the SP domain on natural substrate, C1s.
Therefore, we propose that CCP2 module provides accessory
binding sites. Differential scanning calorimetric measurements
demonstrated that CCP2 domain greatly stabilizes the structure of SP
domain. Deletion of CCP1 domain from the CCP1-CCP2-SP fragment results
in the loss of the dimeric structure. Our experiments also provided
evidence that dimerization of C1r is not a prerequisite for
autoactivation. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B catalytic
region of the molecule, whereas the N-terminal interacting region
of C1r (CUB-EGF) binds the C1s subunit in a
Ca2+-dependent manner.
The catalytic C-terminal B fragment of C1r, consisting of the two
CCP domains followed by the activation peptide of the protease and the
serine protease domain (B chain), can be obtained by autolysis or by
limited proteolysis of extrinsic proteases (e.g., thermolysin)
(6). The CCP repeat is 
60 residues in length and is
widespread among complement proteins. It is likely that the CCP domains
significantly contribute to the specificity of the interaction and
catalytic properties of the B fragment. The B fragment is a
dimer, like the intact C1r molecule, and is capable of autoactivation
and can cleave and activate zymogen C1s (7). A recent
structural model of (B)2 suggests a loose
head-to-tail assembly of the monomers, where the -chain (the two CCP
modules and the activation peptide) of one monomer interacts with
the serine protease module of the other monomer (8).
To gain detailed information about the role of the individual domains
in the structure and function of the catalytic region of human C1r we
expressed and characterized three different recombinant fragments in
Escherichia coli: the B (CCP1-CCP2-SP), the catalytic
domain with one CCP module (CCP2-SP), and the serine protease module
including the activation peptide (SP), respectively. Our aim was
to determine which domains are responsible for the dimerization and how
the CCP domains modulate the stability and the proteolytic action of
the serine protease domain. Another objective was to clarify the
mechanism of autoactivation of the proenzyme C1r. For this purpose we
expressed mutant forms of the recombinant fragments, which were
stabilized in the zymogen state. Using these mutants we could obtain
new information about the details of the autoactivation process.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The cDNA fragments corresponding to the amino acids 309–705 (CCP1-CCP2-SP), 376–705 (CCP2-SP), and 451–705 (SP) amino acids of human C1r were amplified by PCR using the proofreading Pfu DNA polymerase enzyme (Stratagene, La Jolla, CA) and the full-length cDNA template (9). For the amplification procedure the following forward primers were obtained from East Port Scientific (Budapest, Hungary): 5'-GCGAAGCTTGCCCCCAGCCCAAGACCCTA-3', 5'-GCGAAGCTTGTGGGCAGCCCCGAAACCTG-3', and 5'-GCGAAGCTTGTGGGAAGCCCGTGAACC-3' in the case of CCP1-CCP2-SP, CCP2-SP, and SP, respectively. The reverse primer was 5'-GCGGTCGACTCAGTCCTCCTCCTCCATCT-3' for each fragment. The PCR products were digested with HindIII and SalI (cleavage sites are underlined) and were ligated into the HindIII/XhoI site of the pET-17b expression vector (Novagen, Madison, WI) in frame with the following Tag sequence 5'-ATGTGCACCCAAGCT-3'. As a result of this the recombinant proteins contained four extra amino acids (Ser-Thr-Gln-Ala) at their N terminus. The constructs were verified by DNA sequencing. In vitro mutagenesis experiments were conducted by means of the QuickChange site-directed mutagenesis kit (Stratagene). The primer pairs (only the sense sequence is shown) were: GTGGAACAGAGGCAGCAGATAATCGGAGGGCAAAAAG for the Arg463Gln and GCCTGCCAGGGGGATGCTGGGGGCGTTTTTGCA for the Ser654Ala mutations.
Production, renaturation, and purification of the recombinant proteins
The expression plasmids were transformed into the BL21(DE3)
pLysS host strain, and the transformants were selected on Luria-Bertani
medium plates containing ampicillin and chloramphenicol. The
expression was conducted according to the manufacturer’s instructions
(10). After induction with isopropyl
-D-thiogalactoside, the cells were collected in a 1/10 volume
Tris-EDTA buffer and frozen at -20°C. The cells were then
thawed and incubated for 30 min at room temperature in the presence of
0.5% Nonidet P-40. The viscous solution was sonicated to shear the
DNA, and the inclusion bodies were collected by centrifugation
(12,000 x g, 15 min, 4°C). The supernatant was
discarded and the pellet was washed three times with Tris-EDTA buffer
(1/10 of the culture volume). The inclusion bodies were solubilized in
6 M GuHCl, 0.1 M Tris-HCl (pH 8.3), 100 mM DTT for 3 h at room
temperature. The solution contained 10 mg/ml protein. The
solubilized proteins were diluted to 400-fold into the refolding
buffers. The refolding buffers contained 50 mM Tris-HCl (pH 8.3), 3 mM
reduced glutathione, 1 mM oxidized glutathione, 5 mM EDTA, and 2 M
GuHCl in the case of the CCP1-CCP2-SP fragment, or 0.5 M arginine in
the case of the CCP-SP and SP fragments. The renaturation process was
conducted at 15°C overnight. The renatured protein solutions were
then dialyzed against 50 mM Tris-HCl pH 7.4, 145 mM NaCl, filtrated on
a 0.45-µm nitrocellulose membrane and concentrated.
The renatured proteins were purified on Q-Sepharose-Fast Flow column (Pharmacia Biotech, Uppsala, Sweden). The samples were dialyzed against and the column was equilibrated in a buffer containing 20 mM NaCl and 20 mM Tris-HCl, pH 8.3, for the CCP1-CCP2-SP and CCP2-SP or pH 7.4 for the SP fragments. The samples were loaded onto the column and the elution was conducted with a linear NaCl gradient from 20 to 400 mM. Fractions were identified by SDS-PAGE. The recombinant proteins were further purified by gel filtration using a Superose-12 fast protein liquid chromatography column (Pharmacia Biotech) in 50 mM NaCl, 20 mM Tris-HCl (pH 8.3) for CCP1-CCP2-SP and CCP2-SP, or pH 7.4 for SP. The concentration of the recombinant proteins was determined by absorbance at 280 nm using the absorption coefficients 15.2, 15.8, and 15.4 (1%, 1 cm) for the CCP1-CCP2-SP, CCP2-SP, and SP fragments, respectively. For calculation of absorption coefficients we used the method of Gill et al. (11) taking disulfide bridges into account. The molecular masses calculated from the amino acid sequences were 45,532, 37,670, and 28,976 Da for CCP1-CCP2-SP, CCP2-SP, and SP fragments, respectively.
N-terminal sequencing
After SDS-PAGE and blotting to polyvinylidene difluoride membrane, the N-terminal amino acid sequences of the recombinant proteins were determined by a pulsed-liquid phase protein sequencer ABI 471A.
Gel filtration chromatography
Gel filtration experiments were conducted using a Superose-12 FPLC column (Pharmacia Biotech). In the 4.0–6.0 pH range the column was equilibrated with 20 mM citrate, 100 mM NaCl buffer. At pH 7.0 and 8.0 buffers containing 20 mM Tris and 100 mM NaCl were used. Cytochrome c, carbonic anhydrase, albumin, and human IgG were used as m.w. standards. Relative m.w. of the C1r fragments were calculated using linear fitting of the elution volume vs m.w. of the standard proteins in logarithmic scale at the appropriate pHs.
Enzymatic assays
Esterolytic activity. The rates of hydrolysis were measured on the Z-Lys-S-Bzl and Z-Gly-Arg-S-Bzl thioesters. The release rate of HS-Bzl was measured through its reaction with 4,4'-dithiodipyridine (12), and was followed with a Jasco V-550 spectrophotometer at a wavelength of 324 nm. Assays were conducted following the method of McRae et al. (13) at 30°C in 20 mM Tris buffer at pH 7.5 containing 145 mM NaCl. Values of kcat/KM were directly determined from the catalytic rate at low substrate concentrations (10–30 µM).
C1s cleavage.
Proenzyme C1s was expressed in baculovirus expression system using High
Five insect cell culture (Invitrogen, Carlsbad, CA).
Functionally pure (80%) C1s proenzyme was obtained by purifying the
cell culture supernatant on a DEAE Sepharose FF column (Amersham
Pharmacia Biotech, Piscataway, NJ) as described in Ref.
(14). C1s preserved its proenzyme state during the
purification and storage. Proenzyme C1s cleavage ability of the C1r
fragments was tested by means of the esterolytic activity of the
generated active C1s molecules on the Z-Lys-S-Bzl thioester substrate.
An enzyme/C1s molar ratio of 1:50 was used for most of the experiments.
C1s proenzyme solution (10–15 ml) with a protein concentration of
0.1 µM in 20 mM Tris, 145 mM NaCl buffer pH 7.4 was thermostated
at 30°C. The C1r SP, CCP2-SP, and CCP1-CCP2-SP fragments were added
to it at a final concentration of 1–3 x
10-9 M. At 1-min intervals, 1 ml of the mixture
was withdrawn and the esterolytic activity was measured by the addition
of the Z-Lys-S-Bzl substrate at a final concentration of 100 µM. The
maximal specific activity value of totally activated C1s of 182
s-1 at 100 µM Z-Lys-S-Bzl concentration was
used for the calculation of the actual concentration of C1s. Values of
kcat/KM
were calculated by linear fitting for the first five to eight points
where the amount of the cleaved, active C1s was <10% of the total
proenzyme concentration.
C1r autoactivation experiments. The S654A mutant proenzyme SP, CCP2-SP, and CCP1-CCP2-SP fragments were used as substrates to investigate the autoactivation ability of the wild-type active C1r fragments. Measurements were conducted in 20 mM Tris, 145 mM NaCl buffer at pH 8.3. Reaction was started by the addition of the active enzyme to the S654A mutant proenzyme solution thermostated at 37°C. An enzyme/zymogen ratio of 1:10–1:50 was used. Ten-microliter aliquots were removed at 10–15 time points in the range of 0.5 min to 1 h and added to 10 µl of 5% SDS sample buffer (15) containing 5% 2-ME and were immediately incubated for 3 min at 100°C. The cleavage at the activation site of the proenzyme molecules was followed by reducing SDS-PAGE. The acrylamide gels were stained with Coomassie brilliant blue. The concentration of the uncleaved proenzyme vs time was calculated from the density of its bands recorded by a Bio-Rad GelDoc2000 imaging system (Hercules, CA). After curve fitting and derivation in Origin 5.0 data analyzing software (MicroCal, Northampton, MA) the cleavage rate vs proenzyme concentration was calculated and proved to be linear in the concentration range used; therefore, kcat/KM values could be obtained from the slope of the curve.
Differential scanning calorimetry (DSC). Calorimetric measurements were performed on a VP-DSC (MicroCal) differential scanning calorimeter. Denaturation curves were recorded between 10 and 80°C at a pressure of 2.5 atm, using a scanning rate of 1°C/min. The protein concentration was set to 0.1 mg/ml. Samples were dialyzed against 20 mM Tris pH 8.3, 145 mM NaCl, and the dialysis buffer was used as a reference. Heat capacities were calculated as outlined by Privalov (16).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Three cDNA fragments from the catalytic region of C1r (Fig. 1
) have been cloned into a modified
pET-17b vector in fusion with the (Met)-Ser-Thr-Gln-Ala sequence. Each
insert begins with a Cys (Cys309,
Cys376, Cys451 in the case
of CCP1-CCP2-SP (B), CCP2-SP, and SP, respectively) at the N
terminus and ends with Asp705 at the C terminus.
The mature proteins have an N-terminal sequence
Ser-Thr-Gln-Ala-(Cys)... as verified by protein sequencing. To
prevent autoactivation, stabilized mutant constructs were also
expressed. In one series we introduced the
Arg463Gln mutation into the cDNAs, whereas in
another series we changed the active site Ser654
to Ala for all the three fragments. The expression plasmids were
transformed into the E. coli BL21(DE3) pLysS strain, and the
recombinant protein expression was induced by adding isopropyl
-D-thiogalactoside. After induction, the cells were lysed and the
soluble and insoluble fractions were separated by centrifugation and
analyzed on SDS-PAGE (data not shown). In the soluble fraction we could
not detect recombinant proteins using Coomasie blue staining, whereas
the pellet contained almost exclusively the recombinant C1r fragments
(purity 80%). Because the recombinant proteins were present as
inclusion bodies, renaturation procedures were needed to generate the
native, folded structure. The inclusion bodies were solubilized in 6 M
guanidine-HCl solution, which contained 100 mM DTT to reduce all the
disulfide bridges. The solubilized recombinant proteins (10 mg/ml)
were then diluted 400-fold (final concentration 25 µg/ml) using
different refolding buffers and incubated at 15°C overnight. Many
different refolding solutions containing various additives and
different oxido shuffling reagents were tested for the three fragments
and the best ones were selected for large scale renaturation (17, 18). We found the highest renaturation yield using 2 M
guanidine-HCl in the case of CCP1-CCP2-SP fragment and 0.5 M
L-arginine in the case of CCP2-SP and SP
fragments. The optimal oxido shuffling system was the mixture of
reduced and oxidized glutathione in a ratio of 3 mM reduced
glutathione/1 mM oxidized glutathione at pH 8.3 in all experiments. The
efficiency of the folding process could be estimated by reducing
SDS-PAGE, because native, functionally active C1r can cleave itself
into two chains ( 18-kDa and B 30-kDa chains in the case of the
CCP1-CCP2-SP fragment). Because the denatured recombinant proteins have
single-chain structure in the inclusion bodies, the appearance of two
chains on the reducing gel is a good indicator of autoactivation and of
successful renaturation of the wild-type fragments (Fig. 2A). Edman degradation of the
large (30-kDa) chains yielded the Ile-Ile-Gly-Gly-Gln sequence in all
cases, indicating that the correct autolytic cleavage at the
Arg463-Ile464 bond between
the - and B chains had occurred during the activation process. The
efficiency of the renaturation was 10–20%, allowing us to obtain
enough material for all subsequent physicochemical and functional
studies. After the renaturation process the aggregated material was
removed by filtration on a 0.45-µm nitrocellulose membrane, and the
refolded recombinant proteins were purified by anion-exchange and
gel-filtration chromatography as described in Materials and
Methods. On the Q-Sepharose Fast Flow column most of the
contaminants did not bind to the resin at low ionic strength (20 mM
NaCl and 20 mM Tris-HCl) and could be removed by washing the column
with the low salt buffer. The correctly folded, native recombinant
fragments eluted as single peaks detected at 280 nm during the
ascending salt gradient. There was no difference between the elution
volume of the wild-type (activated) and the R463Q or S654A mutant
(zymogen) fragments, although the recombinant fragments were
essentially pure after the ion-exchange chromatography to remove the
traces contaminants we performed a gel filtration chromatography on a
Superose-12 FPLC column. After this step, the protein solutions were
concentrated and the concentrations of the recombinant proteins were
measured from the absorbance at 280 nm. The final yields for the
CCP1-CCP2-SP, CCP2-SP, and SP fragments were 2, 5, and 2 mg/l of
culture, respectively. Both the wild-type and the zymogen mutant
fragments yielded a single band on nonreducing SDS-PAGE analysis,
although with different apparent molecular mass (45 and 39 kDa
for wild-type CCP1-CCP2-SP and zymogen CCP1-CCP2-SP, respectively;
38 and 34 kDa for wild-type and zymogen CCP2-SP; and 31 and
27 kDa for the wild-type and zymogen SP fragment). However, on the
reducing gel the wild-type fragments exhibited the activated two-chain
forms (30 kDa for the B chain and 18 kDa for the -chain),
whereas the zymogen mutants retained a single-chain structure. To prove
the correct folding of the zymogen mutants, the renatured proteins were
converted into the two-chain form. The Arg463Gln
mutants could be specifically cleaved at the
Gln463-Ile464 and activated
by thermolysin (Fig. 2B). After thermolysin treatment all
the three R463Q fragments showed proteolytic and esterolytic activity
similar to that of the wild-type autoactivated proteases (Table I). The Ser654Ala
mutants cannot autoactivate themselves, but wild-type C1r fragments
could cleave them at
Arg463-Ile464 activation
site, as verified by protein sequencing.
|
|
|
Gel permeation chromatography: relative molecular mass.
To determine the relative molecular mass and investigate the
dimerization properties of the expressed and renaturated C1r fragments,
the relative molecular mass of the three C1r fragments vs pH were
analyzed by gel permeation chromatography. The molecular mass of the
three C1r fragments were determined relative to the standards at every
pH (Fig. 3).
|
90 kDa at neutral or alkaline pH. Below pH 6.0 it showed a
sigmoidal shaped decrease to 44–47 kDa, which is the molecular mass of
the monomer CCP1-CCP2-SP, and is in accord with the SDS-PAGE. DSC. DSC measurements were performed, on the one hand, to check the native structure of the fragments and, on the other hand, to investigate the role of the individual domains in the conformational stability of the catalytic region of C1r.
The SP fragment showed a sharp, cooperative melting transition at a
relatively low temperature (47.5°C), indicating a compact, stable
structure of the molecule (Fig. 4).
|
The CCP1-CCP2-SP fragment showed an unfolding transition at 57.5°C, which is in good agreement with that of the equivalent CCP1-CCP2-SP produced in baculovirus expression system (58.3°C) (19). These data prove that renatured CCP1-CCP2-SP is in a native form and that its conformational stability is similar to that of a CCP1-CCP2-SP with a somewhat larger N-terminal and with carbohydrate side-chains. The stability of CCP1-CCP2-SP is somewhat higher than that of CCP2-SP. The presence of the CCP1 module and the dimerization of the CCP1-CCP2-SP fragment exert less effect on the stability of the protein as compared with the significant stabilizing effect of the CCP2 module in the interaction with SP domain in the CCP2-SP construct.
Functional characterization of the recombinant proteins
Esterolytic activity on synthetic substrates.
The values of the catalytic efficiency
(kcat/KM)
for the reaction of the C1r fragments with the Z-Lys-S-Bzl and
Z-Gly-Arg-S-Bzl thioesters are presented in Table I
. Z-Lys-S-Bzl is not
a "good" substrate for C1r, but its spontaneous hydrolysis rate is
very low; therefore, measurements of low catalytic activity was
possible. Z-Gly-Arg-S-Bzl, a more sensitive thioester substrate of C1r,
was hydrolyzed at a high rate by the fragments. The three C1r fragments
showed similar esterolytic activities on the thioester substrates
indicating similar active site conformations in the SP, CCP2-SP, and
CCP1-CCP2-SP fragments. However, the CCP2-SP fragment proved to be
slightly more potent as compared with the others. The observed
kcat/KM
values on the Z-Gly-Arg-S-Bzl substrate are four times higher than
those described previously for the entire C1r molecule isolated from
human serum (13).
The catalytic efficiency of C1s cleavage by the recombinant C1r
fragments.
The ability of the C1r fragments to cleave proenzyme C1s was tested
through the esterolytic activity of the activated C1s molecules on the
Z-Lys-S-Bzl thioester substrate. Although C1r also cleaves this
substrate (see Table I), its catalytic efficiency is about two orders
of magnitude less than that of C1s. An enzyme-substrate ratio of 1:50
assures that the activity of C1r on the Z-Lys-S-Bzl was negligible
compared with that of C1s. The proenzyme C1s concentration (0.04–0.1
µM) was orders of magnitude below the
KM value, and this allowed direct
calculation of the
kcat/KM
values from the linear part of the C1s activation curve. Kinetic
analysis of the activation of C1s by CCP1-CCP2-SP is shown on Fig. 5. The slope of the line is proportional
to the catalytic efficiency. The results with the three C1r fragments
are summarized in Table II. All the three
fragments efficiently cleaved proenzyme C1s. The CCP2-SP fragments
exhibited an exceptionally high
kcat/KM
value.
|
|
. In all enzyme-substrate pairs a high
catalytic efficiency, comparable to the activities on synthetic ester
substrates, could be detected. SP fragment activated the three S654A
mutants at a similar rate. The highest
kcat/KM
values were obtained for the self-activation of the CCP2-SP fragment
(i.e., its activity on the S654A mutant CCP2-SP) (Fig. 6).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-hydroxylation of the Asn residue in the EGF domain
(Asn167 in C1r and Asn149
in C1s) is very low (14, 22) and the glycosylation
patterns differ significantly from the complex type glycosylation found
in the case of serum proteins. However, these differences in the
posttranslational modifications do not affect the functional properties
of the recombinant proteins, because even aglycosylated proteins retain
their biological activity (14, 21). It has been reported
that simple (single-domain) serine proteases (trypsin, chymotrypsin)
can be produced in E. coli and the inclusion bodies can be
successfully renatured to obtain natively folded, functionally active
recombinant proteins with a high yield (17, 23, 24, 25, 26). Based
on these results we chose to express fragments from human C1r in
bacteria. The B fragment of C1r, which can be obtained from the
intact molecule by limited proteolysis retains the proteolytic activity
of the molecule. It consists of three domains, two CCP modules and the
serine protease domain, and it is a dimer at physiological pH, like the
entire molecule. It is also capable of autoactivation. Because this
fragment preserves several basic functions of C1r, it is an attractive
candidate to study the role and significance of the CCP modules in the
function and regulation of C1r. To explore the contribution of the
individual domains to the above-mentioned properties of C1r we
successively deleted the CCP domains preceding the SP domain from the
cDNA. As a result, we made and expressed three cDNA constructs. The
recombinant proteins accumulated as inclusion bodies inside the
E. coli cells. After disruption of the cells we purified and
renatured these proteins. The renatured fragments were purified to
homogeneity by ion-exchange chromatography and gel filtration. After
these procedures all the three fragments were in a correctly folded,
functionally active form as confirmed by subsequent physicochemical and
enzymatic measurements. After renaturation the wild-type
fragments were present in the activated, two-chain form.
Autoactivation is a good indicator of correct folding. Indeed, by the
end of the renaturation procedure the activation was complete. To study
the zymogen form of the enzymes, as well as the autoactivation
procedure, we constructed and expressed mutant C1r fragments, i.e.,
stable proenzymes. In one set of experiments we mutated the
Arg463-Ile464 cleavage site
to Gln-Ile (27), whereas in another set we changed the
catalytic site Ser654 to Ala. In the first case
the zymogen cannot be activated by a trypsin-like serine protease;
however, it could be cleaved and converted to an active enzyme by
thermolysin. In the case of the Ser654Ala mutants
the Arg-Ile bond can be cleaved by wild-type C1r, but the mutant itself
cannot function as an active protease. All of these C1r fragments were
produced with a similar yield except the CCP2-SP fragments where the
renaturation efficiency was significantly higher. This can be
interpreted as the CCP2 domain closely associates with the SP domain
forming a compact cooperative folding unit. The SP domain alone is less
stable as it is indicated by the DSC measurements. The CCP1-CCP2-SP
fragment contains the CCP1 module, which associates loosely to the
CCP2-SP. The larger, more complex structure of CCP1-CCP2-SP can account
for its lower renaturation yield.
The DSC measurements indicate that all of the renatured fragments
have compact, folded structures. The calorimetric heat denaturation
curves (excess heat capacity against temperature) show cooperative
unfolding of the native structures in the case of the three fragments.
The heat capacity curve of the recombinant CCP1-CCP2-SP is essentially
identical with the curve of CCP1-CCP2-SP expressed in the baculovirus
insect cell system. According to the DSC curves the SP fragment has the
least stable structure. The CCP2 module exerts a dramatic stabilizing
effect, as the midpoint of the heat denaturation peak of SP (47.5°C)
is being shifted to 55.4°C in the case of the CCP2-SP fragment (Fig. 4). The presence of the CCP1 module stabilizes the structure further,
but this effect is much less significant than that of CCP2. These
results are in agreement with the homology model of the catalytic
region of C1r (8) and with the crystal structure of the
CCP2-SP fragment of C1s (28). According to the crystal
structure the CCP2 module associates to SP domain through a rigid
module-domain interface involving intertwined proline and tyrosine-rich
polypeptide segments. Such a rigid CCP-SP assembly is conserved in
other extracellular proteases (29).
Gel filtration experiments show that the SP and CCP2-SP fragments are
monomers at all pH values, whereas the CCP1-CCP2-SP fragment is a dimer
at neutral and basic pH (Fig. 3). Like serum C1r (30), our
recombinant CCP1-CCP2-SP dimers dissociate under slightly acidic
conditions. These results provide straightforward experimental evidence
that the CCP1 domain is involved in the dimerization of C1r. Previously
a three-dimensional model of activated (B)2
has been constructed, which was based on chemical cross-linking and
homology modeling (8). Chemical cross-linking of the
(B)2 fragment produced by autolytic cleavage
of the active serum C1r indicated the existence of converse salt
bridges between Lys299 in the N-terminal region
of the segment of one monomer, and Glu510 of
the serine protease domain of the other monomer. However, our
CCP1-CCP2-SP construct begins with the Cys309 of
C1r preceded by the Ser-Thr-Gln-Ala N-terminal fusion peptide and
therefore lacks Lys299. The fact that our
recombinant CCP1-CCP2-SP is a stable dimer indicates that formation of
salt bridges between Lys299 and
Glu510 is not a prerequisite of dimerization.
Future site-directed mutagenesis experiments could reveal the amino
acid residues of the CCP1 module that are involved in C1r
dimerization.
Regarding enzymatic activity it is somewhat surprising that the serine protease domain of C1r alone can cleave C1s with a rate comparable with that of the activation by the CCP1-CCP2-SP fragment. This indicates that the SP domain contains all the structural elements necessary for C1s binding and cleavage. However, the presence of the CCP2 domain alone in the CCP2-SP fragment causes a dramatic increase (one order of magnitude) of the kcat/Km value of the reaction. We can conclude that the CCP2 domain is responsible for the enhancement of the efficiency of the C1s cleaving activity of C1r. We suggest that although the SP domain alone can cleave C1s, the CCP2 domain provides with additional contact surfaces for binding and orienting the substrate. As we mentioned above, our results obtained by DSC indicate that the CCP2 domain strongly stabilizes the structure of the SP domain. The increase of the kcat/Km value could be explained in principle by this domain-domain interaction. However, the similar esterolytic activity of the recombinant C1r fragments shows that this is not the case. It is obvious that all the three recombinant C1r fragments contain a fully functional active site. The changes in stability caused by the addition of CCP modules to the serine protease domain do not affect the catalytic power of the serine protease active site on the synthetic substrates. We suspect that the dramatic increase of the C1s cleaving ability of the CCP2-SP fragment is due to additional substrate binding sites present on the surface of the CCP2 domain, and not to the stabilizing effect. However, the corresponding kcat/Km value of the CCP1-CCP2-SP fragment for C1s cleavage is smaller than that of the CCP2-SP fragment. As we showed above, the CCP1 domain is responsible for the dimerization of C1r. The catalytic site of C1r or the substrate binding residues on the CCP2 domain can be less accessible for the C1s in the CCP1-CCP2-SP dimers than in the CCP2-SP and SP monomers. It is also very likely that the CCP1 domain does not contain additional substrate binding sites for C1s. We should keep in mind that we are dealing with fluid phase reactions in our present study. However, inside the C1 complex the catalytic domains of C1r and C1s are precisely positioned; therefore, the efficiency of C1s cleavage by the C1r dimer can be significantly higher (31).
Our work with recombinant fragments provided valuable information concerning the autoactivation of C1r. The S654A mutant retains its zymogen form during the renaturation and purification procedure, whereas the wild-type C1r fragments are fully activated after the same treatment. We can conclude that this activation is a true autoactivation and that extrinsic (i.e., E. coli) proteases do not contribute to it. Because all three wild-type fragments autoactivate, two important conclusions can be drawn: 1) dimerization is not a prerequisite for autoactivation, and 2) autoactivation is an inherent property of the serine protease domain. Previously, autoactivation was shown to be a property of the dimers both in the case of the entire molecule and its CCP1-CCP2-SP fragment (7). Under acidic conditions (pH <5.5) the C1r dimer dissociates and the resulting monomers lose their ability of autoactivation (32). Because at acidic pH the catalytic activity of the serine protease active center is expected to decrease it was not clear which phenomenon was responsible for the loss of the autoactivation ability. Our results indicate that under physiological conditions the monomeric CCP2-SP fragment is capable of autoactivation, i.e., dimerization is not required for autoactivation. Because the SP domain itself retains the ability to autoactivate, the presence of even one CCP domain is not a prerequisite for autoactivation.
Autoactivation of C1r is supposed to be a two-step process. In
the initial step zymogen molecules activate zymogens, whereas in the
second step the generated active enzymes cleave zymogen molecules. The
fact that our wild-type fragments can autoactivate shows the existence
of the initial step. The second step could be studied in detail using
our zymogen mutants. The Ser654Ala mutants, which
have an inactive catalytic center but have a cleavable Arg-Ile bond,
were used as substrates for the wild-type fragments. We determined the
kinetic constants for these type of reactions (Table III). Each
Ser654Ala mutant was cleaved by its own wild-type
counterpart. The SP fragment showed effective self-cleavage ability.
The higher catalytic efficiency of the CCP2-SP construct compared with
that of the SP domain can be interpreted assuming that the CCP2 domain,
like in the case of the C1s cleavage, orients the Arg-Ile bond of one
C1r in a favorable position to be cleaved by the active site of the
other C1r. The CCP1-CCP2-SP fragment possesses significantly lower
kcat/Km
values than the other two fragments. Because on
(CCP1-CCP2-SP)2 we follow intermolecular
(interdimeric) cleavage, we may conclude that dimer formation partially
blocks the accessibility of either the catalytic site of the protease,
or the activation site of the proenzyme and therefore decreases the
efficiency of the proteolysis. To clarify this question we conducted
experiments with combinations of the different fragments. The fact that
the wild-type SP fragment exerts similar proteolytic activity on the
dimeric CCP1-CCP2-SP fragment and on the smaller monomer fragments
indicates that the Arg-Ile bond to be cleaved is accessible for
extrinsic cleavage. The catalytic efficiency of CCP2-SP on CCP1-CCP2-SP
also supports this observation. In a complementary experiment the
wild-type CCP1-CCP2-SP fragment showed decreasing catalytic efficiency
with the increasing size of the substrate proenzyme SP, CCP2-SP, and
CCP1-CCP2-SP S654A fragments. It is very likely that the catalytic site
of one CCP1-CCP2-SP is pointed to the "inside" (i.e., toward the
other CCP1-CCP2-SP molecule) in the dimer, and this positioning
facilitates the intramolecular autolytic cleavage.
The major conclusion of this work is that the serine protease module itself is an autonomous folding unit with inherent serine protease activity similar to that of intact C1r. The SP module has the ability to cleave C1s, the natural substrate of C1r, and autoactivation property is also retained.
Comparative measurements highlighted the role of the CCP modules in C1r as modulators of the catalytic functions through allosteric effects occurring upon binding to natural substrates and dimerization. The intimate interaction of the SP domain with the CCP2 domain is reflected in the sizeable stabilizing effect observed if CCP2 is attached to the SP module.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 J.K. and P.G. contributed equally to this work.
3 Address correspondence and reprint requests to Dr. Péter Závodszky, Institute of Enzymology, Biological Research Center, Hungarian Acedemy of Sciences, H-1113 Karolina út 29, Budapest, Hungary. E-mail address: zxp{at}enzim.hu
4 Abbreviations used in this paper: C1, the first component of the complement system; C1q, C1r, C1s, subcomponents of C1; CCP, complement control protein; SP, serine protease module including the activation peptide; DSC, differential scanning calorimetry; CUB, module found in C1r/C1s, Uegf, and bone morphogenetic protein-1; EGF, epidermal growth factor; FPLC, fast protein liquid chromatography.
Received for publication June 22, 2001. Accepted for publication September 5, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-hydroxyasparagine, sialic acid, and one of its two carbohydrate chains still reassembles with C1q and C1r to form a functional C1 complex. Biochemistry 31:4254.[Medline]
-hydroxylase: potential modification of endogenous epidermal growth factor-like modules. Biochem. Biophys. Res. Commun. 189:233.[Medline]
This article has been cited by other articles:
|
|
 |
|
  B. A. Zabel, S. J. Allen, P. Kulig, J. A. Allen, J. Cichy, T. M. Handel, and E. C. Butcher Chemerin Activation by Serine Proteases of the Coagulation, Fibrinolytic, and Inflammatory Cascades J. Biol. Chem., October 14, 2005; 280(41): 34661 - 34666. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Gal, V. Harmat, A. Kocsis, T. Bian, L. Barna, G. Ambrus, B. Vegh, J. Balczer, R. B. Sim, G. Naray-Szabo, et al. A True Autoactivating Enzyme: STRUCTURAL INSIGHT INTO MANNOSE-BINDING LECTIN-ASSOCIATED SERINE PROTEASE-2 ACTIVATIONS J. Biol. Chem., September 30, 2005; 280(39): 33435 - 33444. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Ambrus, P. Gal, M. Kojima, K. Szilagyi, J. Balczer, J. Antal, L. Graf, A. Laich, B. E. Moffatt, W. Schwaeble, et al. Natural Substrates and Inhibitors of Mannan-Binding Lectin-Associated Serine Protease-1 and -2: A Study on Recombinant Catalytic Fragments J. Immunol., February 1, 2003; 170(3): 1374 - 1382. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), and CCP1-CCP2-SP
(
). CCP1-CCP2-SP shows a dimer-monomer transition with a midpoint at
pH 5.5. Molecular masses were determined on a Superose-12 column
using a protein standard series at every pH, at 25°C, as described in
Materials and Methods.
