| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,*
*
Department of Clinical Immunology and Transfusion Medicine, University Hospital, Uppsala, Sweden; and
Department of Natural Sciences, University of Kalmar, Kalmar, Sweden
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Complement component C3 has been shown to be phosphorylated in vitro by at least five different protein kinases, in different domains of the molecule and with different biologic effects. The C3a moiety of C3 is known to be phosphorylated by PKA and PKC (11) as well as by an ecto-protein kinase from the human parasite Leishmania major (12). The effect of this phosphorylation is to protect the molecule against cleavage at the Arg77-Ser78 bond either by trypsin (all three protein kinases) or by both the classical and the alternative pathway convertases (PKA and PKC). In contrast, CK2 phosphorylates both polypeptide chains of C3 and increases its susceptibility to cleavage by elastase (5).
We have recently demonstrated that C3 is phosphorylated in vitro by a CK1 that is released from activated human platelets (13). This CK phosphorylates a Thr residue in the C3d moiety, which increases the resistance of C3b to cleavage by factor I, acting in concert with factor H (13). In addition to C3, the platelet CK also phosphorylated fibrinogen, vitronectin, and serum albumin (14).
We have also shown that plasma proteins in healthy blood donors contain
covalently bound phosphate at a level of approximately 0.09 mol
phosphate/mol protein; among these phosphorylated proteins is C3 (14).
Activation of isolated platelets by heat-aggregated 
-globulin (HAGG)
under the conditions used to release platelet CK was sufficient to
support the phosphorylation of exogenously added proteins (e.g., C3)
without further addition of ATP or divalent cations (14). In the same
study we also demonstrated that the phosphate content of plasma
proteins as a whole and of C3 and fibrinogen in particular are
substantially increased in systemic lupus erythematosus patients during
exacerbation, as a result of platelet activation. Taken together, these
observations strongly suggest that platelet CK-mediated phosphorylation
of plasma proteins, including C3, is likely to occur in the blood as a
result of platelet activation. However, direct evidence linking
platelet activation in vitro to extracellular phosphorylation of
proteins in whole blood has not yet been established.
The present study was undertaken to characterize this platelet protein kinase and to further elucidate the effect(s) of phosphorylation on C3 functions.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Polyclonal rabbit Abs against human C3a and C3c and
HRP-conjugated polyclonal Abs against human C3c and mouse Igs were
purchased from Dako (Glostrup, Denmark). Chicken Abs against rat
recombinant CK1 were purchased from Immunsystem (Uppsala, Sweden), a
mouse monoclonal IgG1 Ab (mAb) clone 1AD9 against the 
subunit of
human protein kinase CK2 was obtained from Boehringer Mannheim
(Mannheim Germany). Polyclonal rabbit Abs against phosphoserine and
phosphothreonine were bought from Zymed (San Francisco, CA).
HRP-conjugated streptavidin was purchased from Amersham (Slough, U.K.),
Falcon tubes from (Becton Dickinson, Meylan, France) were furnished with Corline heparin surface (Corline, Uppsala, Sweden) according to the manufacturer’s recommendation (15). The surface concentration of heparin was 0.5 µg/cm2, corresponding to approximately 0.1 IU/cm2, with an antithrombin III binding capacity of 2–4 pmol/cm2.
Preparation of reagents
C3 and factor B were purified from human plasma according to the procedures described by Hammer et al. (16), and Lambris and Müller-Eberhard (17), respectively. Factor D was purified from peritoneal fluid from patients with renal failure, as described by Catana and Schifferli (18). C3 was digested with trypsin (Sigma, St. Louis, MO) to C3a and C3b (1%, w/w, for 5 min at room temperature), and the fragments were separated on Sephadex G-100 (Pharmacia Upjohn, Stockholm, Sweden) equilibrated in PBS. Nascent C3b, which exposes a free sulfhydryl group, was covalently bound to activated thiol Sepharose (ATS; Pharmacia Upjohn) (19): 80 µg trypsin was added to 10 mg of C3 and 2 ml of ATS and incubated at 37°C for 15 min. The digestion was interrupted by incubation with 0.8 mg of soybean trypsin inhibitor (Sigma) for 15 min, followed by extensive washing of the Sepharose.
U937-derived microparticles were prepared according to the procedure of Paas and Fishelson (20). We have previously demonstrated the presence of CK2 on these membrane vesicles (5).
Human -globulin (Pharmacia Upjohn; 1 mg/ml in PBS) was heat
aggregated to HAGG by incubation for 30 min at 63°C. Chicken
polyclonal Abs against rat recombinant CK1, rabbit Abs against human
C3a, and phosphoserine and phosphothreonine were biotinylated using
biotin-amidocaproate N-hydroxysuccimide ester (Sigma) as
previously described (21). Polyclonal rabbit Abs to human C3c were
coupled to cyanogen bromide-activated Sepharose (Pharmacia Upjohn)
according to the manufacturer’s instructions.
For substitution experiments, serum from a patient deficient in C3 was used. The patient’s parents were cousins and were both of Turkish origin. High voltage electrophoretic analysis indicated that both parents had C3S only, and no C3 could be detected in samples from the patient. Serum from healthy blood donors was used as a control. Serum samples were collected within 1 h after blood was drawn and were stored at -70°C until used for analysis.
Phosphorylation of C3 and C3b
Platelets were precipitated from citrated blood and washed as described by Mustard et al. (22), with the modifications described previously (13, 23). Platelet protein kinase was prepared by resuspending the final platelet pellet to one-third of is original plasma volume in albumin-free Tyrode’s medium (pH 7.45) containing 137 mM NaCl, 2.7 mM KCl, 1 mM MgCl2, 0.36 mM NaH2PO4, 12 mM NaHCO3, 2 mM CaCl2, and 5.5 mM glucose. The platelets were activated with 10 µg/ml of HAGG for 15 min at room temperature and were centrifuged at 1100 x g for 10 min. After centrifugation, the supernatant was frozen in aliquots at -70°C. The major protein kinase activity released from platelets under these conditions is CK (13).
C3 and C3b were phosphorylated with the protein kinase preparation in
the presence of 10 mM Mn2+ or Ca2+ and 0.1 mM
ATP (final concentrations) at 37°C for 30 min. The reaction was
terminated by passing the samples through PD10 columns (Pharmacia
Upjohn). Unphosphorylated controls were treated identically, except for
the addition of ATP. The degree of phosphorylation was estimated from
parallel samples with added [-32P]ATP of known
specific activity (Amersham). Levels of 0.7–0.9 mol of phosphate/mol
protein were measured in the preparations of C3 and C3b used in this
study. The phosphorylated and unphosphorylated preparations of C3 and
C3b were divided into aliquots and stored at -70°C. No appreciable
dephosphorylation of C3 was found when samples were stored under these
conditions.
Partial characterization of platelet CK
Platelet CK was analyzed by the dot-blot technique using Abs against CK1 and CK2. Recombinant rat CK1 served as a positive control when platelet CK was detected by biotinylated polyclonal chicken Abs against CK1, followed by HRP-conjugated streptavidin. Human CK2 on U937-derived microparticles served as a positive control when platelet CK was analyzed using mAb anti-CK2, followed by HRP-conjugated anti-mouse Igs.
In separate experiments, ATS-C3b was phosphorylated with platelet CK in the presence of 10 mM Mn2+ and 0.1 mM ATP (Sigma) and with three different protein kinase inhibitors, present in concentrations of up to 1 mM. The compounds tested were N-(2-aminoethyl)-5-chloro-isoquinoline-8-sulfonamide (CKI-7; Seikagaku, Tokyo, Japan), which is specific for CK1 (24); 5,6-dichloro-1-ß-D-ribofuranosylbenzimidazole (Calbiochem, La Jolla, CA), which is specific for CK1 (25); and N-(2-aminoethyl)-5-chloronaphtalene-1-sulfonamide HCl (A3; Calbiochem), which is reported to inhibit CK1 and CK2; protein kinases A, C, and G; and myosin light chain kinase (26). After phosphorylation, Sepharose was washed three times with PBS containing 0.1% Tween-20, and the amount of protein-bound radioactivity was measured in a beta counter.
The ion dependence of platelet CK was compared with that of rat
recombinant CK1 (Calbiochem) using casein (3 µg; Sigma) and C3 (20
µg) as substrates in the presence of 0.1 mM
[-32P]ATP and 5 mM of Mn2+,
Mg2+, or Ca2+. To facilitate comparison of the
two enzymes, the recombinant CK1 was diluted in albumin-free Tyrode’s
medium to yield a sp. act. similar to that of the platelet CK
preparation, giving final concentrations of 0.25 mM Mg2+
and 0.5 mM Ca2+, respectively. After phosphorylation at
37°C for 30 min, the samples were subjected to 10% SDS-PAGE under
reducing conditions (27). Exposure of the dried gels and scanning were
performed in a PhosphorImager using ImageQuant software (Molecular
Dynamics, Sunnyvale, CA) for calculations.
Activation of platelets in whole blood or in platelet-rich plasma (PRP)
Twenty milliliters of blood was collected in a 50-ml Falcon tube (Becton Dickinson), which had been furnished with a Corline heparin surface and contained 20 IU of soluble heparin (Bio Iberica, Barcelona, Spain). PRP was produced from blood by centrifugation for 15 min at 150 x g and was used immediately. Platelets in PRP or whole blood were stimulated with 100 µM ADP (Sigma) or HAGG (up to 100 µg/ml) for up to 15 min without agitation at room temperature. After incubation, 10 mM EDTA (final concentration) was added, and the cells were removed by centrifugation at 1100 x g for 10 min. Blood collected in Vacutainer tubes containing EDTA (Becton Dickinson, Rutherford, NJ) was used as a control.
C3 was precipitated removed from these samples by incubation with Sepharose-coupled anti-human C3c (13). The samples were then subjected to SDS-PAGE followed by Western blot analysis. A standard curve for the quantification of phosphate content in isolated proteins, ranging from 0.1–150 pmol of phosphate, was constructed with mixed histone type II-AS (Sigma). By using a colorimetric technique (28), this preparation of histone was shown to contain 0.2 mol of phosphate/mol protein. Nitrocellulose membranes with precipitated proteins and histone were incubated in parallel with biotinylated rabbit Abs against phosphoserine and phosphothreonine, followed by HRP-conjugated streptavidin as described previously (14). The membranes were scanned using Studio Scan II (Agfa-Gaevert, Antwerp, Belgium), the intensity of the bands was quantified using NIH Image 1.54 for Macintosh, and the phosphate content of C3 from the samples was evaluated from the standard curve.
Activation of C3 in serum on an IgG-coated surface
Wells of microtiter plates were coated with 10 µg of human IgG/ml. Normal human serum or C3-deficient serum was diluted 3-fold (final concentration) in veronal-buffered saline (VBS; 1.8 mM sodium barbiturate, 3.1 mM barbituric acid, and 0.15 M NaCl, pH 7.4) containing 0.75 mM Ca2+ and 2.5 mM Mg2+. Before dilution, the C3-deficient serum was supplemented with 1 mg/ml of either phosphorylated or unphosphorylated C3. After incubation at 37°C for 20 min the serum was transferred to test tubes containing EDTA (10 mM, final concentration) and stored at -70°C until used for the detection of generated C3a.
C3 binding to IgG. After removal of the serum, the amount of C3 bound to the IgG-coated surface was detected by HRP-conjugated anti C3c. A sandwich ELISA (29) was used to assess the relative amount of C3/C3 fragments present on the IgG surface. This ELISA, which used a polyclonal anti-C3c Ab for capture and a HRP-conjugated polyclonal anti-C3c Ab for detection, was performed in parallel, with identical Ab dilutions and staining times. This technique had a lower detection limit of 0.5 ng C3/C3 fragments/ml, which corresponds to 0.05 ng of C3/C3 fragments/well.
ELISA for detection of C3a. Frozen serum samples were diluted 1/2000 (final dilution) and analyzed as described previously (29). mAb 4SD17.3 was used for capture, and bound C3a was detected with biotinylated anti-C3a followed by HRP-conjugated streptavidin. Zymosan-activated serum, calibrated against a solution of purified C3a, served as the standard.
Cleavage of C3 by trypsin
Phosphorylated or unphosphorylated C3 (10 µg) in PBS was incubated with trypsin in serial dilutions up to 0.5 µg for 30 min at 37°C. The reaction was then stopped by boiling the samples in electrophoresis sample buffer containing 15% 2-ME, and the samples were analyzed by SDS-PAGE, followed by densitometric scanning of the gels.
Activation of C3 by the alternative pathway convertase
Activation by the alternative pathway was assessed by incubating phosphorylated or unphosphorylated C3 (16 µg) with factor B (2.1 µg) and factor D (10 ng) diluted in VBS containing 2.5 mM Mg2+ for up to 5 min at 37°C. In an alternative procedure, phosphorylated or unphosphorylated C3 was added to preformed alternative pathway convertase complex molecules obtained by incubating 2 µg of unphosphorylated C3b in VBS containing 2.5 mM Mg2+ with factors B and D. The resulting cleavage was analyzed by SDS-PAGE, as described above.
Activation of C3 by the classical pathway convertase
Activation by the classical pathway was measured by incubating phosphorylated or unphosphorylated C3 diluted in VBS containing 0.75 mM Ca2+ and 2.5 mM Mg2+ with EAC14oxy2 cells bound to microtiter plates as described previously (30); microtiter plates were coated with EAC14oxy2 or EA (control) cells. Phosphorylated or unphosphorylated C3 in 3-fold serial dilutions from 0.04–30 µg/ml was incubated at 37°C for 60 min, and the bound C3 fragments were visualized by incubation with HRP-conjugated Abs to human C3c followed by staining. EDTA-serum served as a positive control. The deposition of C3/C3 fragments was estimated using the sandwich ELISA described above, and nonspecific binding, i.e., binding to plates with control cells, was subtracted from that to EAC14oxy2 cells.
Binding of trypsin-generated C3 fragments to IgG
Wells of microtiter plates were coated with 10 µg of human IgG/ml. Purified phosphorylated or unphosphorylated C3 (0.5 mg/ml) in PBS was incubated with trypsin in serial dilutions from 3.1 µg/ml for 30 min at 37°C. The deposited C3 was detected with HRP-conjugated anti-C3c. Quantitation of the relative amount of bound C3 fragments was made using the sandwich ELISA as described above.
Binding of factor B-generated C3 fragments to IgG
Purified phosphorylated or unphosphorylated C3 (0.5 mg/ml) in VBS containing 2.5 mM Mg2+ was activated in IgG-coated microtiter plates for 30 min at 37°C by incubation with factors B and D in serial dilution from 50 and 0.4 µg/ml, respectively. Detection was performed as described above, and quantitation of the deposited C3 fragments was performed.
Covalent binding of C3 to glycerol and glycine
The binding of glycerol and glycine to C3 was measured by a
modification of the technique described by Dodds and Law (31).
Phosphorylated or unphosphorylated C3 (10 µg) in PBS containing 1 mM
EDTA was activated with trypsin (0.03 µg) in the presence of either
2.5 mM [2-3H]glycine (41.1 Ci/mmol) or 10 mM
[2-3H]glycerol (200 mCi/mmol; both from New England
Nuclear, Boston, MA) for 60 min at 37°C. Control samples of
phosphorylated and unphosphorylated C3 were incubated in parallel, but
without the addition of trypsin. After incubation, the reaction was
terminated by boiling the samples in electrophoresis sample buffer,
followed by separation by SDS-PAGE. Under these conditions, C3 was
quantitatively activated, while no fragments beyond C3b were visible.
Thereafter, the gels were cut, and the band in each lane corresponding
to the '-chain of C3b was incubated with Biolute-S tissue
solubilizer (Zissner Analytic, Frankfurt, Germany), and the
radioactivity in the samples was determined after the addition of
Aquasafe 300 Plus scintillation mixture (Zissner Analytic). The sp.
act. of the added glycine and glycerol was determined, and the degree
of binding to the C3 samples, expressed as moles per mole of protein,
was calculated.
Statistical analyses
The results are expressed as the mean ± SEM. Statistical significance was calculated with Student’s t test for unpaired samples, using StatView 4.01 (Abacus Concepts, Berkeley, CA) for Macintosh.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We first compared platelet CK to recombinant rat CK1 with regard
to its immunological reactivity, dependence on divalent cations, and
inhibition by protein kinase inhibitors (Table I
). The immunological identity of
platelet CK was checked by means of dot blots using Abs against CK1 and
CK2 (Table I). Chicken Abs against CK1 reacted weakly with platelet CK
and strongly with rat CK1. In contrast, a rat anti-CK2 mAb reacted
with CK2 on U937-derived microparticles (not shown), but failed to
react with both platelet CK and CK1.
|
). Platelet CK was inhibited by CKI-7
(specific for CK1) and A3 (which inhibits multiple protein kinases,
including CK1 and CK2), but not by
5,6-dichloro-1-ß-D-ribofuranosylbenzimidazole (which is
specific for CK2). The Ki50 values obtained were
110 µM for CKI-7 and 800 µM for A3, respectively. These values are
substantially higher than those cited by the manufacturers for these
inhibitors against CK1.
Mn2+ and Ca2+ were more potent stimulators of
platelet CK activity than Mg2+ when equimolar amounts of
casein and C3 were used as substrates (Table I). In contrast,
activation of CK1 was most pronounced in the presence of
Mg2+, whereas inhibition was seen in the presence of
Mn2+.
Phosphorylation of C3 after activation of platelets in whole blood or in PRP
Western blot analysis (using a mixture of Abs against
phosphoserine and phosphothreonine) of C3 that had been isolated from
EDTA plasma had a phosphate content of 0.14 ± 0.02 mol/mol
protein. The level was increased in C3 that had been isolated from
activated whole blood or PRP. In whole blood the phosphate content in
C3 increased after incubation with HAGG (from 0.14 ± 0.02 to
0.52 ± 0.08 mol/mol; n = 7) or incubation with
ADP (to 0.60 ± 0.14 mol/mol; n = 7). A similar
increase was seen in PRP from 0.15 ± 0.02 to 0.56 ± 0.09
mol/mol (n = 7) after incubation with HAGG or to
0.65 ± 0.14 (n = 7) after incubation with ADP
(Fig. 1A).
|
B). Activation of C3 in serum on an IgG-coated surface
When the activation of C3 and binding of C3 fragments to an
IgG-coated surface were assessed by ELISA, significantly higher amounts
of split products from phosphorylated than from unphosphorylated C3
were detected in all serum concentrations tested (Fig. 2A). At the highest serum
concentration tested (1/3), substitution with unphosphorylated C3
resulted in deposition of approximately 10 ng of C3 fragments/well as
assessed by sandwich ELISA, whereas substitution with phosphorylated C3
resulted in deposition of 4 times higher amounts of C3 fragments/well
(n = 3; p < 0.0001).
|
B). At a serum dilution of 1/3 the
levels of C3a reached 3000 ng/ml (for phosphorylated C3) and 2200 ng/ml
(for unphosphorylated), respectively (n = 3;
p = 0.0286). Cleavage of C3 by trypsin
There was no significant difference in the rate of trypsin
cleavage after phosphorylation of C3 (Fig. 3B). In addition, cleavage
generated very similar or identical polypeptide fragments from
phosphorylated and unphosphorylated C3 (Fig. 3A).
|
Alternative pathway activation of phosphorylated and
unphosphorylated C3 was studied by incubation of mixtures of purified
factor B, factor D, and C3. The cleavage of phosphorylated C3 occurred
at a slightly (but not significantly) lower rate that of
unphosphorylated C3 (Fig. 4, A
and B). After 3 min of incubation, 60% of the
unphosphorylated C3 remained intact compared with 75% of the
phosphorylated C3.
|
, C and D). Activation of C3 by the classical pathway convertase
Physiological activation of C3 by the classical pathway convertase
was measured after incubation of phosphorylated or unphosphorylated C3
with EAC14oxy2 cells fixed to microtiter plates. Activation
was monitored by measuring the binding of C3 fragments on the
EAC14oxy2 cell surface. Approximately 2-fold higher binding
occurred when fragments were generated from phosphorylated C3 than from
unphosphorylated C3 after addition to the cells at concentrations >3
µg/ml (Fig. 5; n = 4;
p = 0.0006 for 30 µg/ml). Lower concentrations of
added C3 resulted in the same degree of binding.
|
The capacity of C3 to bind to IgG was assessed by incubating C3
together with trypsin in wells of microtiter plates coated with IgG.
The level of binding of phosphorylated C3 was significantly higher at
all concentrations of trypsin tested (Fig. 6A). When trypsin
concentrations of 1.5 µg/ml or higher were used, the binding of C3
fragments reached a plateau at approximately 10 ng of C3 fragments/well
for unphosphorylated C3 (as estimated by sandwich ELISA) and 1.6 times
higher for phosphorylated C3, respectively (n = 4;
p < 0.0001).
|
The capacity of C3 to bind to IgG was further assessed by
incubating C3 together with factor B and factor D in wells of
microtiter plates coated with IgG. The level of binding of
phosphorylated C3 was significantly higher at all tested
concentrations of factor B (Fig. 6B). When factor B at
concentrations of 6 µg/ml or higher were used, the binding of C3
fragments reached a plateau at an approximately 1.6 times higher level
for phosphorylated C3 compared with unphosphorylated C3
(n = 4; p = 0.0001).
Covalent binding of C3 to glycerol and glycine
The capacity of C3 to form ester and amide bonds was assessed by
examining the binding of radiolabeled glycerol and glycine to
unphosphorylated and phosphorylated C3 when trypsin was used to
activate the C3. The ratio of trypsin to C3 (indicated by an arrow in
Fig. 6A) was the lowest amount that gave maximal binding of
C3 fragments to the IgG surface. Under these conditions
unphosphorylated C3 was found to bind 0.454 ± 0.047 mol of
glycerol/mol of protein, whereas phosphorylated C3 bound 0.726 ±
0.065 mol of glycerol/mol of protein (n = 10;
p = 0.0148).
The binding of glycine decreased after phosphorylation, from 0.224
± 0.025 to 0.149 ± 0.011 mol of glycine/mol of protein,
respectively (n = 10; p = 0.0036).
Consequently, the glycerol/glycine ratio increased from 2.0 for
unphosphorylated C3 to 4.9 for phosphorylated C3 (Table II). The background levels of glycine
binding was 0.010 ± 0.002 mol/mol for unphosphorylated C3 and
0.011 ± 0.002 mol/mol for phosphorylated C3 (n =
4). The corresponding values for glycerol binding to control and
phosphorylated C3 were 0.020 ± 0.002 and 0.021 ± 0.002
mol/mol (n = 4), respectively.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The protein kinase that is released from platelets and phosphorylates
human C3 is a Ser/Thr protein kinase that has CK1-like properties, in
that it phosphorylates casein and is not inhibited by heparin; however,
in contrast to CK1, its activity is dependent on Mn2+ and
Ca2+ (13, 23). Phosphorylation by platelet kinase made
bound C3b less susceptible to cleavage by factor I, with factor H as a
cofactor (13). To characterize the platelet Ser/Thr protein kinase, we
have now compared this protein kinase to commercially available rat
recombinant CK1 with regard to its response to specific protein kinase
inhibitors, immunological reactivity, and dependence on various
divalent cations. Platelet CK was partially inhibited by compounds
reported to be specific for CK1, but with a 10-fold higher
Ki than that reported for CK1, suggesting similarity, but
not identity, between the two enzymes. This interpretation was
supported by the observation that Abs raised against CK1 reacted much
more weakly with platelet CK than with CK1. CK1 is extremely well
conserved between different species, e.g., 94% identity between rat
and human CK12 (32). There is even a considerable degree of homology
between mammalian and yeast CK1 (33). It is therefore most likely that
polyclonal Abs raised against rat CK1 should react with human CK1. Both
protein kinases phosphorylated C3 and casein, but while
Mn2+ ions potentiated phosphorylation of both substrates by
platelet CK, it inhibited phosphorylation by CK1. Taken together, these
results suggest that the platelet protein kinase is a CK that is
distinct from CK1. We have earlier excluded the possibility that it is
a CK2, since it did not react with a mAb to human CK2, and its activity
was not inhibited by heparin or 2,3-DPG (13).
The observation that phosphorylation of C3b by platelet CK decreased the susceptibility for factor I cleavage led us to speculate whether phosphorylation of C3 might increase the amount of C3b that binds to a target surface, e.g., an immune complex. The fact that the half-life of bound C3 fragments in the form of C3b is prolonged should provide greater opportunity for formation of alternative pathway convertase complexes and should thus potentially give rise to increased opsonization. We found that this was indeed the case when serum from a patient deficient in C3 was reconstituted with phosphorylated or unphosphorylated C3 and then incubated with an IgG-coated surface. Phosphorylation increased by >4-fold the amount of generated C3 fragments that bound to IgG. This substantial increase in the binding of C3 fragments was accompanied by a more modest increase in C3a generation, which indicated that other mechanisms were also involved in increased binding.
Phosphorylation of C3 by PKA and PKC has previously been shown to
increase resistance to cleavage of C3 to C3a and C3b by the classical
and alternative pathway convertases as well as by trypsin (11). We have
now conducted similar experiments with purified C3 phosphorylated by
platelet CK; exposure of phosphorylated C3 to the classical pathway
convertase on EAC14oxy2 cells resulted in approximately
2-fold higher binding of C3 fragments than that for unphosphorylated
C3. The binding approached a plateau, indicating binding to a limited
number of sites. In this assay the level of binding is dependent upon
the efficiency with which the cell-bound convertase molecules cleave C3
and upon the ability of the nascent C3b molecules to form covalent
bonds to hydroxyl or amino groups on the cell surface (30).
Phosphorylation of C3 by platelet CK did not significantly affect
cleavage of fluid phase C3 by a fluid phase C3bBb convertase or by
trypsin. It is thus unlikely that the increased binding seen on the
EAC14oxy2 cells was due to increased cleavage, since both
of the convertases and trypsin cleave C3 at the same peptide bond. On
the other hand, it is feasible that the difference in binding was
related to an increase in the number of available acceptor sites for
phosphorylated C3. This hypothesis was confirmed by our observation of
increased binding to surface-bound IgG of fragments generated from
phosphorylated C3 by either C3bBb convertase- or trypsin-mediated
cleavage. In these experimental systems, as in EAC14oxy2
cells, the binding of fragments generated from phosphorylated C3 was
significantly (
1.6 times) higher than that for fragments from
unphosphorylated C3.
The putative phosphorylation site for platelet CK in C3 is located
within a tryptic fragment, between Lys979 and
Lys1014, which also comprises the thiol ester as we have
reported previously (13). After extensive trypsin digestion of
phosphorylated ATS-bound C3b, this fragment containing 90% of the
total radioactivity still bound to the ATS. After hydrolysis of the
ATS-bound C3 fragment, only Thr-P was detected. Based on these results
we concluded that the most likely phosphorylation site for platelet CK
is Thr1009, 19 aa residues toward the C terminus of the C3
-chain from the thiol ester. The sequence DETEQWE surrounding
Thr1009 (34) contains several acid amino acid residues,
which makes it a potential phosphorylation site for CKs (35). One
possible explanation for the difference in binding efficiency reported
in the present study is that phosphorylation of C3 in some way alters
the binding properties of the thiol ester. To test whether this was the
case, we used trypsin to cleave phosphorylated or unphosphorylated C3
in the presence of either [2-3H]glycerol or
[2-3H]-glycine. Phosphorylation of C3 increased the
glycerol binding capacity (by 1.6-fold), while it decreased glycine
binding, resulting in an increased ratio of glycerol/glycine binding,
from 2.0 for unphosphorylated C3 to 4.9 for phosphorylated C3. These
changes are sufficient to account for the increased binding of C3
fragments to IgG, since it has been demonstrated that the binding of
C3b to IgG mainly occurs via ester bonds that are formed with hydroxyl
group-containing residues in the heavy chain of IgG (36, 37, 38). It should
also be noted that in the present study phosphorylation of C3 brought
about an analogous increase in glycerol binding and in binding of
generated C3 fragments to IgG after activation by trypsin (Fig. 6),
emphasizing that the binding to IgG is indeed ester linked. A recently
published crystal structure of human C3d (39) confirms that
Thr38 of C3d (which corresponds to Thr1009 of
intact C3) is exposed on the exterior of the molecule, in close
proximity to a cluster of acidic amino acid residues. In C3d,
Thr38 and the thiol ester are located at opposite ends of
the 1 helix, with Thr38 at the C terminus. For the
formation of ester bonds the thiol ester of C3 requires the sequence
Cys1010, Gln1013, and His1126 (40).
In C3d, these amino acid residues correspond to Cys17,
Gln20, and His133, which are located at
opposite ends of the T3 segment. The mechanism by which phosphorylation
of Thr1009 alters the binding properties of C3 remains to
be established; one possibility is that the negatively charged
phosphate group alters the distance between the thiol ester and
His1126, thereby contributing to the formation of ester
bonds instead of amides.
The overall effect of the phosphorylation of C3 by platelet CK should be to enhance the opsonization of immune complexes in whole blood. Handling of immune complexes involves a sequence of events that includes complement activation and covalent binding of C3b to the immune complex (41). One can propose a pathway in vivo by which platelets that become activated by immune complexes (triggered by Fc receptors, C1q, and/or C5b-9 complexes) mediate the phosphorylation of fluid phase C3 and immune complex-bound C3b. The consequence is an amplification effect by which immune complexes indirectly potentiate their own opsonization.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Kristina Nilsson Ekdahl, Department of Clinical Immunology, University Hospital, S-751 85 Uppsala, Sweden. E-mail address:
3 Abbreviations used in this paper: PKA, cAMP-dependent protein kinase; PKC, calcium- and phospholipid-dependent protein kinase; CK, casein kinase; C3a, C3b, C3c, C3d, and iC3b, proteolytic fragments of C3; HAGG, heat-aggregated -globulin; ATS, activated thiol Sepharose; CKI-7, N-(2-aminoethyl)-5-chloro-isoquinoline-8-sulfonamide; A3, N-(2-aminoethyl)-5-chloronaphtalene-1-sulfonamide HCl; PRP, platelet-rich plasma; VBS, veronal-buffered saline.
Received for publication January 11, 1999. Accepted for publication March 31, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2 (CKSNK1G2). Genomics 46:133.[Medline]
This article has been cited by other articles:
|
|
 |
|
  K. N. Ekdahl, D. Norberg, A. A. Bengtsson, G. Sturfelt, U. R. Nilsson, and B. Nilsson Use of Serum or Buffer-Changed EDTA-Plasma in a Rapid, Inexpensive, and Easy-To-Perform Hemolytic Complement Assay for Differential Diagnosis of Systemic Lupus Erythematosus and Monitoring of Patients with the Disease Clin. Vaccine Immunol., May 1, 2007; 14(5): 549 - 555. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Kawakami, Y. Shimoyama, and K. Ohtsuki Characterization of Complement C3 as a Glycyrrhizin (GL)-Binding Protein and the Phosphorylation of C3{alpha} by CK-2, Which Is Potently Inhibited by GL and Glycyrrhetinic Acid In Vitro J. Biochem., February 1, 2003; 133(2): 231 - 237. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Walter, A. Schindzielorz, B. Hartung, and C. Haass Phosphorylation of the beta -Amyloid Precursor Protein at the Cell Surface by Ectocasein Kinases 1 and 2 J. Biol. Chem., July 28, 2000; 275(31): 23523 - 23529. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) or unphosphorylated (
) C3 and diluted from 1/3
was incubated in an IgG-coated ELISA plate C3 at 37°C for 20 min. The
relative amount of bound C3/C3 fragments was assessed by a sandwich
ELISA that was performed in parallel as described in Materials
and Methods. A shows binding of C3 fragments to
the IgG-coated plate. The serum was analyzed for C3a
(B). The presented data represent the mean ± SEM
of three experiments.
