close

Вход

Забыли?

вход по аккаунту

?

Plasma Protein Binding by Monosodium Urate Crystals.

код для вставкиСкачать
775
PLASMA PROTEIN BINDING BY
MONOSODIUM URATE CRYSTALS
Analysis by Two-Dimensional Gel Electrophoresis
ROBERT TERKELTAUB, ANDREA J. TENNER, FRANKLIN KOZIN, and MARK H. GINSBERG
Using 2-dimensional O’Farrell gel electrophore-
sis, we have mapped the proteins from undiluted plasma
and serum which bind to monosodium urate (MSU)
crystals. More than 30 crystal-associated polypeptides
were visualized, including anionic and cationic species.
Proteins increased on the crystals relative to plasma
included Clq, Clf, ClS, fibronectin, fibrinogen, and
kininogen. Crystal-bound polypeptides derived from
IgG, albumin, and transferrin were recovered in decreased amounts relative to plasma. Direct evidence for
activation of the complement and coagulation systems in
plasma was provided by the identification of crystalassociated activation fragments of C1 and kininogen.
Plasmas deficient in selected proteins (e.g., Clq and
IgG) were used to define the role of these proteins in
such activation events and confirmed activation of C1 in
immunoglobulin-deficient plasma by MSU crystals. In
summary, we have described a high resolution, semiPresented in part at the VIII Pan-American Congress of
Rheumatology, Washington, DC, June 1982.
From the Department of lmmunology and Division of
Rheumatology, Research Institute of Scripps Clinic and Research
Foundation, La Jolla, California. This is publication number 2823
from the Department of Immunology.
Supported by NIH grants no. AM-27214. AM-00720. and
AI-18042. Dr. Terkeltaub received support from the Arthritis Society of Canada.
Robert Terkeltaub, MD, FRCP(C): Postdoctoral Fellow;
Andrea J. Tenner, PhD: Assistant Member, Research Institute of
Scripps Clinic; Franklin Kozin, MD: Assistant Member, Research
Institute of Scripps Clinic; Mark H. Ginsberg, MD: Associate
Member, Research Institute of Scripps Clinic.
Address reprint requests to Mark Ginsberg, MD, Department of Immunology, Scripps Clinic and Research Foundation,
10666 North Torrey Pines Road, La Jolla, CA 92037.
Submitted for publication September 30, 1982; accepted in
revised form February 2. 1983.
Arthritis and Rheumatism, Vol. 26, No. 6 (June 1983)
quantitative approach to analyze protein binding to
crystals, have documented the complexity of crystalplasma protein interaction, and have provided direct
evidence for the binding of coagulation system proteins
and binding and activation of complement by MSU
crystals, in whole plasma and IgG-deficient plasma.
Crystals of monosodium urate (MSU) bind a
variety of proteins in vitro (1-4), with potential effects
on the inflammatory process in gout. For example, the
interaction of MSU crystals with proteins of the complement (5-7) and contact systems of plasma (8-10).
previously demonstrated in vitro, may play a role in
the triggering of gouty attacks. Furthermore, there is
evidence that the interaction of MSU crystals with
inflammatory cells, presently believed to be the central event in gout (1 1,12), may be modulated by certain
crystal-bound plasma proteins ( 1 3-15) and membrane
proteins (16).
Previous investigations of the binding of plasma
proteins to MSU crystals have utilized immunochemical, ultrastructural, and radiochemical techniques (13,17). These studies were directed at defining the
binding of selected proteins, e.g., IgG, rather than the
entire spectrum of protein binding by MSU crystals.
To obtain a comprehensive map of the plasma
proteins which coat MSU crystals, we analyzed proteins eluted from synthetic MSU crystals in vitro,
using 2-dimensional (2-D) gel electrophoresis (18-19).
The high resolution, semiquantitative nature, and lack
of inherent bias of this approach allowed the detection
of a variety of crystal-associated polypeptides. Several
of the identified crystal-bound proteins were unexpectedly enriched. As we will show, activation events
776
TERKELTAUB ET AL
requiring proteolytic cleavages on the crystal surface,
as well as simple binding events, could be identified by
this approach.
MATERIALS AND METHODS
MSU crystals were prepared under sterile conditions
and characterized as previously described (6). Crystals were
utilized in unheated form, and ranged from 15-45p in length.
Silica crystals (lop, unheated) were donated by the Pennsylvania Glass Sand Corporation, Pittsburgh, PA. Sodium
iodide (I2'I, I3'I) was purchased from Amersham, Arlington
Heights, IL. Ampholines were purchased from LKB,
Bromma, Sweden. Human Cohn Fraction I1 was purchased
from ER Squibb and Sons, New York, NY. All other
chemicals were reagent grade.
Buffers. Ten millimole phosphate buffered saline
(PBS) contained anhydrous NaH2P04 0.76 g m h t e r ,
Na2HP04 I .62 gmhter, NaCl 8.77 gmhiter.
Plasma and serum preparation. Plasmas from normal
donors were prepared from whole blood anticoagulated with
0.38% sodium citrate or 5 mM EDTA. Serum was collected
from whole blood allowed to clot at 37°C in glass tubes for 30
minutes. Sodim citrate (0.38%) or 5 mM EDTA was added to
serum thus obtained for certain experiments. Fibronectindepleted plasma was prepared by passage of plasma over a
gelatin-sepharose column as previously described (20). The
pooled effluent from the column contained less than 0.003
pg/ml fibronectin by radioimmunoassay (21). Plasma from
an asymptomatic patient with acquired agammaglobulinemia
was the generous gift of Dr. John Curd (Scripps Clinic). The
IgG concentration of this plasma was <38 pg/ml by the
Mancini method (22). IgM and IgA were undetectable.
C Iq-depleted serum was prepared by passage
through a Biorex 70 column (23), and had no Clq hemolytic
activity (24). C lq-depleted serum was reconstituted by addition of purified Clq, which restored Clq hemolytic activity
to greater than 80% of normal values.
Purified proteins. Clq was isolated and radiolabeled
with '*'I by the lactoperoxidase-glucose oxidase method as
previously described (23). C l r and CIP were the generous
gifts of Dr. Neil Cooper (Scripps Clinic) and were prepared
according to published methods (25). C l s and C1S (26) were
the generous gifts of Dr. Robert Ziccardi, Scripps Clinic.
Fibronectin was purified by gelatin-sepharose affinity chromatography as described previously (20). Fibrinogen
(27) was the generous gift of Dr. Edward Plow (Scripps
Clinic). Hageman factor and high molecular weight kininogen were isolated from fresh, citrated human plasma according to published methods (28,29).
Labeling of Cohn Fraction 11, fibronectin, fibrinogen, contact system proteins, and whole citrated human
plasma was done by a chloramine T method (30). For
example, to label plasma, 2 pl plasma was mixed with 98 pl
100 mM PBS, followed by the addition of 1 mCi Iz5I and 2 10pl additions of 5 mg/ml chloramine T, 5 minutes apart. The
reaction was terminated by the addition of 10 pl of 10 mg/ml
sodium metabisulfite, and the sample was dialyzed against 2
changes of PBS for 2 hours. One hundred percent of the 1251
counts in plasma were precipitable in 20% trichloracetic
acid.
Binding of proteins to crystals. MSU crystals were
washed once, resuspended in PBS pH 7.4 (25 mg/ml), and
then triturated through a 27 gauge needle. Crystals, 500 pI,
were mixed with 500 PI plasma in a 1.5 ml microfuge tube
(Beckman, Palo Alto, CA) and incubated with constant
agitation at 37°C for 30 minutes. The DH of the crvstalpiasma suspension was 7.5. Aliquots of 460 pl were removed
from the sample, layered over 1 ml of 20% sucrose in MSUsaturated PBS in a 1.5 ml microfuge tube, and centrifuged at
8,700g for 5 minutes in a Beckman Microfuge. The resulting
crystal pellet was resuspended and washed once in 1 ml
MSU-saturated PBS.
Elution of crystal-bound proteins. MSU crystalbound protein was eluted by heating at 100°C for 5 minutes in
a solution of 10% glycerol, 5% Nonidet-40 (NP40), 2% pmercaptoethanol, 1.5% sodium dodecyl sulfate (SDS) in
MSU-saturated PBS. The eluate was collected after sedimentation of the crystals in the Microfuge for 5 minutes.
Protein concentration of the eluates from MSU crystal
pellets was determined by the Lowry procedure (31), using a
bovine serum albumin standard (Sigma, St. Louis, MO).
Two-dimensional (2-D)gel electrophoresis. Plasma
and crystal-pellet samples containing 200 pg of protein and
samples of purified proteins were analyzed by isoelectric
focusing followed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) in the second dimension, as described by
O'Farrell(l8) with modifications by Anderson and Anderson
(19). This involved applying the samples in elution buffer as
above, in a total volume of 100 pl, to a tubular isoelectric
focusing gel containing 9M urea, 6.4% acrylamide, 0.38%
Bis-acrylamide, 3% NP40, and 3% ampholines (consisting of
2.8% pH 3.5-10, 0.1% pH 2.5-4, 0.1% pH 9-1 1). Gels were
prefocused at 200V for 1 hour, followed by sample application and focusing at 400V for 18 hours, followed by 1 hour at
700V (using a Buchler 3-1500 power supply, Fort Lee, NJ).
Isoelectric focusing gels were equilibrated for 30 minutes in
10% pmercaptoethanol, 10% glycerol, 2.2% SDS, and
0.25M Trizma base and stored at -70°C. SDS-PAGE in the
second dimension utilized an 8-20% exponential gradient
polyacrylamide slab gel with dimensions of 12 x 15 cm.
Proteins were detected by Coomassie blue staining
or autoradiography as previously described (16). To estimate
isoelectric points (PI), duplicate isoelectric focusing gels
were sliced into 5 mm segments and placed in 500 pI
degassed distilled water for 2 hours. A PHM 62 pH meter
(Radiometer, Copenhagen, Denmark) was used to measure
the pH of these gel segments. Values obtained for certain
consistently appearing polypeptides (e.g., albumin PI = 4.55.2) were used as internal standards as well. All experiments
involving incubation of normal and depleted plasmas and
sera with crystals were performed on 3 or more separate
occasions. A polypeptide observed on 3 or more 2-D gels of
crystal pellets from different incubations was accepted as
being crystal-associated.
RESULTS
Estimation of bound versus trapped plasma proteins. In order to evaluate the ability of the crystal
washing procedure to separate crystal-bound from
nonspecifically trapped polypeptides, '*'I labeled
plasma proteins and free 13'1 were added to citrated
plasma incubated with MSU crystals in triplicate,
followed by centrifugation through 20% sucrose as
777
PROTEIN BINDING BY MSU CRYSTALS
Table 1. Quantitation of crystal-bound versus trapped plasma
proteins*
'**I plasma
proteins
Free '"1
Starting sample
97,563 2 3,616
21,823
51 1
Monosodium urate pellet 5,041 2 143 (5.2%)
(through 20% sucrose)
39
Monosodium urate pellet 2,930 2 67 (3%)
(1 wash)
31
Eluted from pellet
1,994 5 82 (2%)
2
291
5
14 (0.2%)
5
10 (0.1%)
* 62 (2.3%)
* lZsI labeled plasma and free I3'l were added to 500 ~1 citrated
plasma and incubated with urate crystals in triplicate, as described
in the text. '*'I protein binding and trapping and I 3 l I trapping were
measured at each step of sample treatment. Results are expressed as
counts/30 seconds (mean 2 SD). Percent of counts applied is listed
in parentheses for each step.
described above. As seen in Table 1, the unwashed
crystal pellet contained 5.2% of the starting '"1 plasma protein counts and 2.3% of the applied free '-"I
counts. Thus it is estimated that 2.9% of the plasma
proteins were truly bound at this step. After one wash
of the urate pellet, only 0.2% of 13'1 was seen, whereas
3% of the '"I labeled plasma proteins remained. Nonspecific adsorption of 13'1 to the crystals cannot be
excluded; therefore the percent pellet-associated I3'l
represents an estimate of maximal trapping.
We thus estimated that urate crystals under
these conditions bound 2.8% of applied plasma proteins, and that trapped plasma proteins represented a
maximum of 7% of the total proteins present after this
wash step. Seventy percent of bound I2'I was recovered in the following elution in 1.5% SDS and 2% 0mercaptoethanol at 100°C for 5 minutes, indicating
that the bulk of crystal-bound polypeptides were recovered for analysis. An analysis of MSU crystalbound '*'I labeled IgG in isolation from plasma
showed 80% recovery of eluted IgG under the same
conditions (data not shown).
Characterization of crystal-bound polypeptides.
When the crystal eluate was subjected to 2-D gel
electrophoresis and the gels stained with Coomassie
blue, more than 30 crystal-bound polypeptides were
identified, including anionic and cationic moieties
(Figure 1). More than 15 polypeptides were seen to be
increased on the crystals relative to plasma. Examples
of these are indicated in Figure 1.
The 2-D gels of plasma proteins eluted from
MSU crystals after sucrose washing showed no differences from the pattern after simple sedimentation of
the crystal-plasma mixture followed by repeated washings in MSU-saturated buffer (not shown), ruling out a
significant effect of the sucrose on the final protein
pattern.
4 + 4 4 4 4 f
4.5 4.7
5
5.8 6.3 6.7 8
Apparent pl
Figure 1. Two-dimensional gel electrophoresis of urate crystalassociated polypeptides (top) and starting plasma (bottom). After
incubation of crystals with citrated plasma, the washed urate pellet
was prepared as described in Materials and Methods and run in an
isoelectric focusing gel under reducing conditions in the first dimension (right to left) and in an 8-20% exponential gradient sodium
dodecyl sulfate-polyacrylamide slab gel in the second dimension
(top to bottom). This and subsequent figures show the Coomassie
blue staining patterns of a 200 pg protein load of the indicated
samples. Arrows indicate plasma polypeptides corresponding to
albumin and IgG which are bound, but not increased, on the crystal
in comparison with starting plasma, as well as several of the
polypeptides which are increased on the crystal surface relative to
starting plasma (fibronectin. fibrinogen a,p. y chains, and Clq A,
B, C chains)
TERKELTAUB ET AL
778
Identified crystal-bound complement and coagulation system proteins. Figure 3 shows the 2-dimensiona1 gel of crystal-bound polypeptides in greater detail,
and indicates proteins of the complement and coagulation systems which were identified by the procedures
described above. The accompanying Table 2 lists the
arrow numbers used to designate these polypeptides
as well as the apparent Mr and pl, and other identification criteria. The following peptides were abundant on
the crystal pellet: Clq, ClP, and C1S as well as
fibronectin, fibrinogen, and kininogen (Figure 3 and
Table 2).
Regarding the identified components of C1, it
was noted that ClP and CIS were not visualized on 2dimensional gels of the crystal pellet of Clq depleted
serum or 5 mM EDTA plasma. Therefore the binding
and cleavage of C 1r and C 1s is C 1q and divalent cation
dependent, and indicates activation of C1 occurring on
the crystal surface (6). To examine the IgG dependence of C1 activation in plasma, we studied the
crystal pellet polypeptides after incubation with agammaglobulinemic plasma. Clq, Clr, and CIS were present on the crystal eluate from agammaglobulinemic
plasma, while IgG light and heavy chains were not
visualized (Figure 4). The crystal eluate after incuba-
Figure 2. Identification of umte-associated polypeptides, e.g., fibrinogen. Polypeptides that co-electrophoresed with '"1 fibrinogen in
the urate pellet of citrated plasma are encircled (dashed lines, top
panel). Bottom panel depicts the pattern from the urate pellet
obtained after incubation of crystals with serum. Dashed lines
indicate the regions of the gel where fibrinogen substituent chains
would run. Identification of specific polypeptides was corroborated
by the similarity to published values for molecular weight and
isoelectric paint (see text).
Several polypeptides were identihed by procedures analogous to those shown in Figure 2 for fibrinogen. Identification of fibrinogen polypeptides was
based on: 1) co-electrophoresis with the purified labeled or unlabeled protein; dotted lines in Figure 2
surround spots that co-electrophoresed with '"I fibrinogen; 2) lack of visualization of the polypeptide on
the crystal pellet of a sample run with starting plasma
or serum deficient in this protein. In the figure it is
seen that the crystal pellet of serum contains none of
the 3 series of fibrinogen-derived polypeptides indicated by dotted lines in plasma; and 3) similarity to
published values for molecular weight and PI for the
polypeptides in question, e.g., fibrinogen a chain Mr
67,000, pl6.1-7.1, p chain Mr 56,000, pl5.4-5.8, and y
chain Mr 47,000, PI 4.9-5.1 (32,33).
.. '
.
.. . .
.
..
(plasma) .
.:
Flgure 3. Urate-associated polypeptides of citrated plasma shown in
greater detail. Arrows point to identified polypeptides of the complement and coagulation systems (see accompanying Table 2 for
individual details). Arrow 1 = CIS heavy chain; arrow 2 = C1S light
chain; arrow 3 = C l i heavy chain; arrows4,5,6 = C l q A, B, and C
chains, respectively; arrow 7 = fibronectin; arrow 8 = kininogen
heavy chain.
779
PROTEIN BINDING BY MSU CRYSTALS
Table 2. Complement and coagulation proteins identified on urate crystals
Coagulation
system proteins
7
8
225
67
52
48
5-5.3
7.3-7.6
5.6-6.7
5
56
4.5
tion with Clq-depleted serum (not shown) demonstrated no obvious diminution in IgG, indicating that IgG
binding in normal serum was not C 1q-dependent . C 1q,
CIP, and C1S were also visualized in the crystal eluate
after coincubation with fibronectin-depleted plasma
(not shown), indicating that binding and activation of
C l was not fibronectin-dependent.
Unidentified polypeptides increased on the surface of MSU crystals. Table 3 lists the unidentified
MSU crystal polypeptides from plasma and from serum but not plasma. Of particular note are 2 polypep-
Figure 4. Urate-associated polypeptides of agammaglobulinemic
plasma. Polypeptides derived from Clq, CIP, and CIS are designated (arrows). The absence of staining for IgG heavy and light chains
is noted.
+
+
+
+
+
+
+
+
+
+
+
I.. 8
Fibronectin .
-
$
$
Fibrinogen
I.
I.
Kininogen .
+
a chain
p chain
y
chain
tides (Mr 42,000, pl 4.6 and Mr 39,000, PI 4.7) which
were visualized in the crystal eluate of serum, citrated
plasma, and citrated serum but not of 5 mM EDTA
plasma or 5 mM EDTA serum. These polypeptides
were observed in the crystal eluate of Clq-depleted
serum, and thus their presence on the urate crystals
was not Clq-dependent but was inhibited by 5 mM
EDTA. Five crystal-bound polypeptides were observed on incubation with serum but not with plasma
(Table 3). All these proteins were visualized on the
crystals after incubation with serum to which 0.38%
citrate or 5 mM EDTA was added, indicating that
absence from plasma-exposed urate crystals was not
due to the chelating agent.
Other MSU crystal-bound polypeptides. Table 4
lists a number of polypeptides which, though crystalassociated, were not increased relative to plasma.
These included albumin, transferrin, and IgG, which
are all abundant in starting plasma (Figure 1) and
present in reduced proportions in the crystal eluate.
Protein binding pattern of silica crystals. To
determine if protein binding was crystal-specific, we
incubated silica crystals with citrated plasma in the
proportions used for MSU and analyzed bound polypeptides as was done for MSU. Several differences
from MSU crystals were seen, including heavier binding of anionic polypeptides between Mr 30,000 and Mr
45,000 (Figure 5 ) and the absence of staining for Clq,
CIP, CIS, and fibronectin on silica crystals (compare
Figure 5 with Figure 3). As with MSU crystals,
fibrinogen and the heavy chain of kininogen were
abundant, and albumin, IgG, and transferrin were
present but not increased in eluates from silica crystals.
780
TERKELTAUB ET AL
Table 3.
Unidentified plasma and serum polypeptides increased on urate crystals relative to plasma
Unidentified crystal-bound
plasma polypeptides
(Figures I , 3)
Unidentified crystal-bound
serum polypeptides
(Figure 2)
Apparent
Mr (Kd)
148
72
65
45
42
39
34
Apparent
6-8
4.7
7.2-8
5
4.6
4.7
5.1
+
+
+
+
+
+
+
130
72
47
46
34
4.7-4.9
5
4.7
4.8
7.8-8
-
PI
Citrated
plasma
EDTA
plasma
Serum
Comments
+
+
+
+
+
+
+
+
+
+
+
*, t
*, t
f
f
f
f
f
t
* Observed in crystal pellet with Clq-deficient serum.
Observed in crystal pellet with citrated serum (0.38%) but not 5 mM EDTA serum.
f Observed in crystal pellet with citrated serum (0.38%) and 5 mM EDTA serum, indicating that
absence from plasma-exposed urate crystals was not due to chelating agent.
selective enrichment, the amounts of crystal-bound
Clq, as well as kininogen (both plasma concentrations
less than 100 pg/ml) were comparable with the amount
of IgG on the crystal (plasma concentration 7-18
mg/ml) (Figure 3).
One feature of this study was an estimation of
the relative proportions of bound and trapped proteins
associated with MSU crystals, by assay of the crystalassociated radioactivity after incubation with plasma
to which free 13'1 and trace '"I labeled plasma proteins had been added (Table 1). Maximal protein
trapping was less than 7% in the eluted material; thus
the majority of polypeptides visualized on 2-D gels
were truly crystal-bound. The fraction of recovery of
bound plasma proteins eluted from the crystal pellet
was 270%. While it is not possible to exclude the
failure to elute certain protein species possessing high
affinity binding to MSU crystals, this is unlikely given
the fact that SDS denaturation disrupts a majority of
noncovalent interactions of proteins, particularly in
DISCUSSION
When O'Farrell 2-D gel electrophoresis was
used to analyze plasma binding by MSU crystals,
more than 30 crystal-associated polypeptides were
observed. Anionic as well as cationic polypeptides
were bound, suggesting that the overall charge of a
polypeptide may not be the sole determinant of crystal
binding (1). Certain polypeptides abundant in starting
plasma, including albumin, transferrin, and the heavy
and light chains of IgG, were bound to MSU crystals
but not increased relative to their representation in
starting plasma (Figure 1, Table 4). Other polypeptides, which stained more intensely on 2-dimensional
gels of the crystal eluate than in starting plasma, were
said to be increased on the crystals relative to plasma
(Figure 1). These included fibronectin, fibrinogen a,p,
y chains, and Clq A, B, and C chains. Such polypeptides, listed in Tables 2 and 3, included species demonstrating highly selective enrichment. As examples of
Table 4.
Plasma polypeptides present but not increased on urate crystals (see Figure 1)
Apparent
Mr
Apparent
PI
Citrated
plasma
EDTA
plasma
169K
90K
80K
67 K
65K
48K
34K
28K
26K
22K
5
4.6-5
5.8-6.6
4.5-5.2
4.6-4.8
7.5-8
7.7
6.3-8
4.8
5.2
+
t
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
* Co-electrophoresed with purified I2'I labeled protein.
Serum
+
+
+
+
+
+
+
+
+
-
Comments
Probable ID
-
*
*
Transfemn
Albumin
IgG heavy chain
IgG light chain
-
PROTEIN BINDING BY MSU CRYSTALS
Figure 5. Silica crystal-associated polypeptides of citrated plasma.
The 2-dimensional gel electrophoresis pattern is distinct from that
seen with urate crystals. Note the absence of staining for fibronectin, Clf., CIS, and Clq (see Figure 3 to compare urate pellet) and the
greater abundance of several anionic polypeptides in the 30-45 Kd
range (small arrows). The polypeptide Mr 56 Kd, PI 4.5 (large arrow)
is kininogen heavy chain which is ennched on these crystals.
the presence of reducing agents (34). Furthermore, 2D gel analysis of the crystal absorbed supernatant (not
shown) showed no depletion of a protein species
relative to starting plasma, that could not be accounted
for in the crystal pellet.
IgG is known to be bound by MSU crystals
(2,3,17), and previous work using immunoelectrophoresis of eluted proteins from MSU crystals exposed to
progressive dilutions of serum showed the enrichment
of IgG relative to albumin, which increased with
increasing dilution of the starting material (1). The
present study, utilizing undiluted plasma (Figure 1)
and serum (Figure 2) confirmed the binding of IgG by
MSU crystals, but showed IgG to account for only a
small fraction of the bound proteins. Recovery of
eluted 1251labeled IgG bound to MSU crystals in
isolation was quantitatively similar to that of whole
plasma proteins. Thus, the relative dearth of IgG in the
crystal eluate is not likely to be due to selective failure
to elute crystal-bound IgG.
Although IgG-coated MSU crystals are effective stimuli for certain inflammatory cells (13,35,36),
previous studies with platelets (37), neutrophils ( 3 9 ,
and macrophages (36) have suggested complex inhibitory, as well as enhancing, effects of plasma and serum
proteins adsorbed to MSU crystals. Recently Abramson et a1 (38) documented the inhibitory effects of
78 1
precoating MSU crystals with plasma, on the production of superoxide by neutrophils. Our study supports
Abramson’s suggestion that the effects of plasma on
crystal-cell interaction cannot be accounted for simply
by IgG effects, since considerable quantities of other
proteins with known effects on cells were bound (see
below).
In this study we identified several polypeptides
present in abundance on the MSU crystal protein coat
in vitro, which may modify the crystals’ inflammatory
potential. For example, the collagen-like region of
crystal-bound C lq may interact with specific receptors
on polymorphonuclear leukocytes, monocytes, lymphocytes, or null cells (39,40) and mediate the efTects
of MSU crystals on these cells. Similarly fibronectin
receptors on neutrophils (41), platelets (42), and mononuclear cells (43) and fibrinogen receptors on platelets
and mononuclear cells (27,44) might be engaged by
these proteins on the crystal surface. The importance
of protein adsorption to MSU crystals is underscored
by these findings, but it is clear that the effects of
plasma on crystal-cell interaction will require further
analysis regarding each cell and crystal type. We have
here defined several of the major protein species
which must be considered.
The analysis by 2-D gel electrophoresis of polypeptides eluted from the crystal surface enabled us to
identify not only simple protein binding, but also
activation events involving cleavages of unmodified
complement and contact system proteins in whole
plasma. For example, MSU crystals have been shown
to induce C lq-dependent cleavage of radiolabeled C 1s
in a purified immunoglobulin-free system (6). We have
now demonstrated that cleavage of unmodified C1s
and C l r occurs in whole plasma in the presence of
MSU crystals, that the resulting C1S and CIS are
crystal-associated (Figure 3), and that appearance of
Clf and CIS on the crystal surface is divalent cation
and Clq-dependent. Moreover, using agammaglobulinemic plasma, Clq, Clf, and CIS were observed in
the crystal eluate in the absence of detectable immunoglobulin (Figure 4). Thus we have confirmed the
activation by MSU crystals of C l , in a plasma which
contained less than 0.2% of normal IgG (5).
Activation of the contact system of coagulation
by MSU crystals has been well recognized (8,9), and
was shown to be associated with prekallikrein and high
molecular weight kininogen (HMWK) dependent
cleavage of radiolabeled Hageman factor (10). In the
present study we detected an abundance of the
cleaved form of an unmodified contact system component, kininogen, on MSU crystals. Cleavage of
782
TERKELTAUB ET AL
HMWK is known t o yield bradykinin as well (45).
These data suggest the formation of a bradykiningenerating system on the surface of M SU crystals
exposed to plasma. This supports the previous observations of MSU crystal-induced kinin generation in
vitro (9) and in vivo (46) in synovial fluid.
That the interaction of plasma with MSU crystals is crystal-specific was shown by comparison with
the plasma protein binding pattern of silica crystals
(Figure 5). Of particular note was the lack of fibronectin and C1 associated with silica crystals as opposed to
MSU. The differences in protein binding between
these crystals in vitro may have some correlation with
their different biologic properties.
In summary, w e have mapped the plasma proteins which bind t o M SU crystals. Proteins which may
modulate inflammation, including Clq, CII-, CIS, fibronectin, fibrinogen, and kininogen, were in abundance on the crystal surface. Moreover, cleavage of
HMWK and IgG-independent cleavage of C 1 were
demonstrated in whole plasma. We have also defined
an approach whereby a biochemical map of the protein
coat of phlogistic crystals may be obtained. Such maps
should prove useful in investigation of the modification of the inflammatory potential of crystals by adsorbed proteins, and in t he study of the variable
clinical responses t o intraarticular MSU crystals.
ACKNOWLEDGMENTS
The authors gratefully acknowledge the technical
assistance of Bryon Jaques and Jane Forsyth and the cooperation of Dr. Edward Plow, Dr. Robert Zaccardi, and Dr. Neil
Cooper in providing purified proteins for study. We also
thank Dr. John Curd for providing agammaglobulinemic
plasma and Ms Betty Goddard for expert secretarial help.
REFERENCES
1. Kozin F, McCarty DJ: Protein binding to monosodium
urate monohydrate, calcium pyrophosphate dihydrate,
and silicon dioxide crystals. I. Physical characteristics. J
Lab Clin Med 89:1314-1325, 1977
2. Hasselbacher P: Binding of IgG and complement protein
by monosodium urate monohydrate and other crystals. J
Lab Clin Med 94532-541, 1979
3. Hasselbacher P, Schumacher HR: Immunoglobulin in
tophi and on the surface of monosodium urate crystals.
Arthritis Rheum 21:353-361, 1978
4. Hasselbacher P: Crystal-protein interactions in crystalinduced arthritis, Advances in Inflammation Research.
Vol. IV. Edited by G Weissmann. New York, Raven
Press, 1982,pp 25-44
5 . Naff GB, Byers PH: Complement as a mediator of
inflammation in acute gouty arthritis. I. Studies on the
reaction between human serum complement and sodium
urate crystals. J Lab Clin Med 81:747-760, 1969
6. Giclas PC, Ginsberg MH, Cooper NR: Immunoglobulin
G: independent activation of the classical complement
pathway by monosodium urate crystals. J Clin Invest
63:759-764, 1979
7. Hasselbacher P: C3 activation by monosodium urate
monohydrate and other crystalline material. Arthritis
Rheum 22571-578, 1979
8. Kellermeyer RW, Breckenridge RT: The inflammatory
process in acute gouty arthritis. I. Activation of Hageman factor by sodium urate crystals. J Lab Clin Med
65:307-315,1965
9. Kellermeyer RR, Breckenridge RT: The inflammatory
process in acute gouty arthritis. 111. Vascular permeability enhancing activity in normal human synovial fluid;
induction by Hageman factor activators and inhibition
by Hageman factor antiserum. J Lab Clin Med 70:372-
383, 1967
10. Ginsberg MH, Jaques B, Cochrane CG, Griffin JH:
Urate crystal-dependent cleavage of Hageman factor in
human plasma and synovial fluid. J Lab Clin Med
95:497-506, 1980
11. Phelps P,McCarty DJ: Crystal-induced inflammation in
canine joints. 11. Importance of polymorphonuclear leukocytes. J Exp Med 124:115-126, 1966
12. Chang Y-H, Gralla EJ: Suppression of urate crystalinduced canine joint inflammation by heterologous and
anti-polymorphonuclear leukocyte serum. Arthritis
Rheum 11:145-150, 1968
13. Kozin F,Ginsberg MH, Skosey JL: Polymorphonuclear
leukocyte responses to monosodium urate crystals:
modification by adsorbed serum proteins. J Rheumatol
61519-526, 1979
14. Ginsberg MH, Henson P, Henson J, Kozin F: Mechanisms of platelet response to monosodium urate crystals. Am J Pathol 94:549-568, 1979
15. Ginsberg MH, Kozin F: Mechanisms of cellular interaction with monosodium urate crystals: IgG-dependent
and IgG-independent platelet stimulation by urate crystals. Arthritis Rheum 21:896-903, 1978
16. Jaques BC, Ginsberg MH: The role of cell surface
proteins in platelet stimulation by monosodium urate
crystals. Arthritis Rheum 25508-521, 1982
17. Bardin T, Cherian PV, Clayburne G, Schumacher HR:
Transmission electron microscopic (TEM) demonstration of the binding of immunoglobulins (Ig) to monosodium urate (MSU) crystals (abstract). Arthritis Rheum
(suppl) 25:S76, 1982
18. O’Farrell PH: High resolution two-dimensional electrophoresis of proteins. J Biol Chem 250:4007-4021, 1975
19. Anderson L, Anderson NG: High resolution two-dimensional electrophoresis of human plasma proteins. Proc
Natl Acad Sci USA 745421-5425, 1977
20. Engvall E, Ruoslahti E: Binding of soluble form of
fibroblast surface protein, fibronectin, to collagen. Int J
Cancer 2O:l-5, 1977
PROTEIN BINDING BY MSU CRYSTALS
21. Ginsberg MH, Painter RG, Birdwell C, Plow E: The
detection, immunofluorescent localization, and thrombin induced release of human platelet-associated fibronectin antigen. J Supramol Struct Cell Biochem 11:167-
174, 1979
22. Mancini G, Carbonara 0, Heremans JF: Immunochemical quantitation of antigens by single radial immunodiffuse. Immunochemistry 2:235-242, 1965
23. Tenner AJ, Lesavre PH, Cooper NR: Purification and
radiolabeling of human Clq. J Immunol 127:648-653,
1981
24. Kolb WP, Kolb LM, Podack ER: Clq isolation from
human serum in high yield by affinity column chromatography and development of a highly sensitive hemolytic assay. J Immunol 122:2103-2111, 1979
25. Ziccardi RJ, Cooper NR: Physicochemical and functional characterization of the Clr subunit of the first complement component. J Immunol 116:496-509, 1976
26. Valet G, Cooper NR: Isolation and characterization of
the proenzyme form of the C l s subunit of the first
complement component. J Immunol 112:339-350, 1974
27. Marguerie GA, Plow EF, Edgington TS: Human platelets possess an inducible and saturable receptor specific
for fibrinogen. J Biol Chem 2545357-5363, 1979
28. Griffin JH, Cochrane CG: Human factor XII, Methods
in Enzymatology. Edited by L Lorand. New York,
Academic Press, 1976,pp 55-65
29. Kerbiriou DM, Griffin JH: Human high molecular
weight kininogen: studies of structure-function relationships and of proteolysis of the molecule occurring during
contact activation of plasma. J Biol Chem 254:12020-
12027, 1979
30. Hunter WM, Greenwood FC: Preparation of Iodine-131
labelled human growth hormone of high specific activity. Nature 194:495-496, 1962
31. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ:
Protein measurement with the fohn phenol reagent. J
Biol Chem 193:265-275, 1951
32. Weinstein MJ, Deykin D: Low solubility fibrinogen
examined by two-dimensional sodium dodecyl sulfate
gel electrophoresis and isoelectnc focusing. Thromb Res
13~36.5-377,1978
33. Doolittle RF: Fibrinogen and fibrin. Sci Am 245:126135, 1981
34. Weber K, Pringle JR, Osborn M: Measurement of
molecular weights by electrophoresis on SDS-acrylamide gels, Methods in Enzymology. Vol. XXVI, part C.
783
Edited by CHW Hirs, SN Timasheff. New York, Academic Press, 1972, pp 3-27
35. Kozin F, Ginsberg M, Karasek V, Skosey J, McCarty
D: Protein binding to monosodium urate crystals and its
effect on platelet degranulation. Adv Exp Med Biol
76B:201-207, 1976
36. McMillan RM, Hasselbacher P, Hahn JL, Harris ED:
Interactions of murine macrophages with monosodium
urate crystals: stimulation of lysosomal enzyme release
and prostaglandin synthesis. J Rheumatol 8555-562,
1981
37. Ginsberg MH, Kozin F, O’Malley M, McCarty DJ:
Release of platelet constituents by monosodium urate
crystals. J Clin Invest 6099-1007, 1977
38. Abramson S, Hoffstein ST, Weissmann G: Superoxide
anion generation by human neutrophils exposed to
monosodium urate: effect of protein adsorption and
complement activation. Arthritis Rheum 25:174-180,
1982
39. Tenner AJ, Cooper NR: Analysis of receptor-mediated
Clq binding to human peripheral blood mononuclear
cells. J Immunol 125:1658-1664, 1980
40. Tenner AJ, Cooper NR: Identification of types of cells in
human peripheral blood which bind Clq. J Immunol
126~1174-1179,1981
41. Mosher DF, Proctor RA, Grossman JE: Fibronectin:
role in inflammation, Advances in Inflammation Research. Vol. 11. Edited by G Weissmann. New York,
Raven Press, 1981,pp 187-206
42. Plow EF, Ginsberg MH: Specific and saturable binding
of plasma fibronectin to thrombin stimulated platelets. J
Biol Chem 256:9477-9482, 1981
43. Bevilacqua MP, Amrani D, Mosesson MW, Bianco C:
Receptors for cold-insoluble globulin (plasma fibronectin) on human monocytes. J Exp Med 153:42-60, 1981
44. Bianco C, Gotze 0, Cohn ZA: Complement, coagulation
and mononuclear phagocytes, Mononuclear Phagocytes. Edited by R van Furth. The Hague, Martinns
Nijhoff Publishers, 1980,pp 1443-1458
45. Silverberg M, Kaplan AP: Activation of Hageman factor, Advances in Inflammation Research. Vol. 11. Edited
by G Weissmann. New York, Raven Press, 1981, pp
165-185
46. Melmon KL, Webster ME, Goldfinger SE, Seegmiller
JE: The presence of a kinin in inflammatory synovial
effusion from arthritides of varying etiologies. Arthritis
Rheum 10:13-20, 1967
Документ
Категория
Без категории
Просмотров
2
Размер файла
841 Кб
Теги
crystals, urate, monosodium, protein, plasma, binding
1/--страниц
Пожаловаться на содержимое документа