close

Вход

Забыли?

вход по аккаунту

?

Lack of activation of C1 despite circulating immune complexes detected by two C1q methods in patients with rheumatoid arthritis.

код для вставкиСкачать
40
LACK OF ACTIVATION OF C1,
DESPITE CIRCULATING IMMUNE COMPLEXES
DETECTED BY TWO Clq METHODS,
IN PATIENTS WITH RHEUMATOID ARTHRITIS
C. ERIK HACK, ANKE J. M. EEKENBERG-RELMEK, U W I G . LIM,
JOOST HAVERMAN. and ROB C. AALBERSI;
The activation of C1 by circulating immune
complexes in patients
- - with rheumatoid arthritis was
investigated. ClrCls(Ci-In)z complexes in EDTA-plasma, reflecting C1 activation in vivo, were slightly raised
in 35 of 57 patients with rheumatoid arthritis, though
most patients had elevated levels of circulating immune
complexes as measured with either the '251-Cl q binding
test or the C l q solid phase assay. The activation of C l
by circulating immune complexes in vitro
_ was
_ investigated by measuring the generation of C l r C l ~ ( C i - l n ) ~
complexes during 60 minutes at 37°C in diluted recalcified EDTA-plasma. In 16 of the 57 patients, a slightly
increased C1 activation in vitro was observed. These
patients tended to have high levels of circulating immune complexes. However, the majority of the patients
with high levels of circulating immune complexes
showed a normal C1 activation in vitro. Therefore, it
was concluded that measurement of circulating immune
complexes by either of the two C l q methods in patients
From the Central Laboratory of the Netherlands Red Cross
Blood Transfusion Service and the Laboratory for Experimental and
Clinical Immunology. University of Amsterdam, Amsterdam. The
Netherlands.
Suppoi-ted by The Netherlands League against Kheumatism.
C . Erik Hack. MD: Research Fellow, Department of Autoimmune Diseases; Anke J . M . Eerenberg-Belmer: Technician,
Department of Autoimmune Diseases; Uwi G. Lim. MD: Kheumatologist, Scheperziekenhuis, Emmen, The Netherlands: Joost Haverman, MD: Rheumatologist, Jan van Breemen Institute, Centre for
Rheumatology and Rehabilitation, Amsterdam (present addrcss
Groote Gasthuis. ' s Hcrtogenbosch); Rob C. Aalberse. PhD: Research Worker, Department of Immunochemistry.
Address correspondence to C.E. Hack. c/o Publication
Secretariat, Central Laboratory of the Netherlands Red Cross Hlood
Transfusion Service. P.O. Box 9406, loo0 AK Amsterdam. The
Netherlapds.
Sllbmitted for publication February 28, 1983; accepted in
revised form August 12, 1983.
-.
-.
Arthritis and Rheumatism, Vol. 27, No. 1 (January 1984)
with rheumatoid arthritis does not imply that these
circulating immune complexes are able to activate C1.
Material believed to be circulating immune
complexes (CIC) can be detected by many methods
(1,2). Among these assays, those based on the interaction between complement and CIC have often been
preferred because complement activation by CIC
might be important for these complexes' phlogistic
properties. Two such methods that use the binding of
Clq to CIC have been described: the Clq solid phase
assay (Clq-SP) (3) and the '''I-Clq binding test (ClqBT) (4). With both assays, CIC can be detected in
patients with rheumatoid arthritis (RA) (5-7). Those
studies show that high lev& of CIC are more frequently detected in patients with active disease. Moreover, high levels of CIC are in general correlated with
low levels of complement factors, especially C4, and
with increased levels of complement breakdown products (5,7). Therefore, it was concluded from those
studies that CIC correlate with disease activity, probably because they activate complement. However,
those studies do not exclude the bossibility that the
CIC detected do not activate complement, but that
they appear in the circulation together with other
immune complexes which activate complement and
then subsequently are removed from the circulation by
the mononuclear phagocytic system or are deposited
in the tissues.
Recently, we found that the binding of Clq to
CIC in the Clq-BT is largely caused by the presence of
polyethylene glycol used in this test (8). Therefore, it
is possible that CIC detected by this test do not
activate complement. Similarly, because CIC detected
in the the Clq-SP bear neither C4 nor C3 (Hack et al:
LACK OF C1 IN RA
manuscript in preparation), it is also questionable
whether CIC detected by this assay can activate
complement. This issue is further complicated by the
finding that C1 may be activated by immune complexes without a detectable binding (9).
When C1 is bound and activated by immune
complexes, C f i and C E are dissociated from these
immune complexes by the action of the CT-inhibitor
(CT-In),
- - and released as complexes consisting of
ClrCls(CT-In)2 (10). Recently, we developed a sensitive radioimmunoassay, a CT-inhibitor complex assay
(INCA), to detect these CfiCi-s(Ci-ln), complexes,
thus enabling us to measure the activation of CI ( 1 1).
In the study reported here, the INCA was used
t o investigate whether CIC detected by either the ClqBT or the Clq-SP in sera from patients with rheumatoid arthritis are able to activate CI.
PATIENTS AND METHODS
Blood samples. Serum and EDTA-plasma were obtained from 25 normal donors and 57 patients fulfilling the
criteria of the American Rheumatism Association ( I ? ) for
classic or definite RA. The mean age of the patients was 64
years; 42 patients were female. The duration of the disease
varied from 6 months to 38 years. Twenty-five were outpatients, the others were hospitalized. Forty-three patients had
seropositive disease, although at the time of venipuncture
only 32 were positive for rheumatoid factors, as measured by
a latex fixation test (Behring Werke AG. Marburg, West
Germany). Twenty-four patients had extraarticular manifestations of the disease. mostly nodules.
Serum was obtained by clotting blood for 1 hour at
37°C and subsequent centrifugation. and En'l'A-plasma was
obtained by centrifugation of EDI'A blood immediately after
venipuncture. Both EDTA-plasma and serum were immediately stored in aliquots at -70°C.
Reagents. Human IgG was purified from Cohn fraction 11 by ion-exchange chromatography and aggregated by
incubating a 10-mgiml solution for 20 minutes at 63°C. By
ACA-34 gel filtration, approximately SO%, of the IgG was
excluded in the void volume. In all experiments, the unfractionated preparation was used as a source of aggregated
human IgG (AHG). 'Ihe AHG concentrations indicated were
not corrected for residual monomeric IgG.
Human C l q was isolated according to the method of
Yonemasu and Stroud (13) and radiolabeled with IZ51 as
previously described (8).
Monospecific antisera were obtained from the Department of Immune Reagents (Central Laboratory of the
Netherlands Ked Cross Blood Transfusion Service) and used
for thc determination of complement levels by radial
immunodiffusion. C3d was measured as described by Perrin
el al (14).
Immune complex assays. The '*'I-CIq binding test
(Clq-B'T) was performed according to the method of Zubler
el al (4). The Clq solid phase immune complex assay (ClqSP) was performed as described by Hay et al (3) with the
41
following minor modifications: 1 volume serum was incubated with 2 volumes 0.2M EDTA, pH 7.5 (4), and 15 pl of this
diluted serum was tested in polypropylene tubes (Microtest
tubes, EET-23, Beckman Instruments Nederland B.V., Mijdrecht, The Netherlands) that had been coated for 3 days at
4°C with 250 pl phosphate buffered saline (PBS) containing
10 mg C l q per liter. The test was further performed a s
described by Hay et al (3), except that a final volume of 250
p1 was used. For both tests, AHG dilutions in normal human
serum (NHS) were used as standard. Findings in the sera
from patients were expressed as microgram equivalents of
AHG per ml of serum.
Ci-inhibitor complex assay (INCA). This assay was
performed as de6cribed previously ( 1 1). Briefly, 1 pl plasma
or serum was incubated with 0.5 ml Sepharose-anti-C I s for 5
hours at roSmLemperature. The Sepharose was then washed
and the C l r C l s(CT-In)? complexes bound to the Sepharose
were detected by an overnight incubation with '"I-anti-CTIn at room temperature. Results are expressed as units per
_ One
_ hundred unit~imilliliteris the maximum amount of
ml.
ClrCls(Ci-In)2 complexes generated by AHG in I ml of
pooled NHS.
C l activation in serum. Fifty microliters of serum was
incubated for I hour at 37°C with 50 pI veronal buffered
saline. pH 7.4, containing 1 .O mM MgCI2 and 0.15 mM CaCI?
(VBS' !). Thereafter, 100 pl 0.2M EDTA, pH 7.5, was
added, and 4 pI of this mixture (i.e., l j l serum) was tested
in the INCA. The amount of CECIs(Ci-In)2 complexes
detected in this mixture was compared with the amount in a
mixture prepared in the same way but without the incubation
at 37°C (i.e., the EDI'A was added immediately). 'Ihe
difference was taken as a measure for C I activation. For
assessing the sensitivity of the assay, 50 pl of VBS' .
containing dilutions of AHG (3 1-1.000 p g h l ) was incubated
with SO pI serum. The mixtures were further tested as
described above.
C1 activation in EDTA-plasma. Ten microliters of
EDTA-plasma was incubated for I hour at 37°C with 490 pl
VBS' ' to which additional CaClz ( 1 mM) and heparin
(Hepalean, Harris Laboratories, Montreal, Canada) (0. I
units/ml, final concentration) were added. Thereafter, 500 pI
0.02M EDTA, pH 7.5. was addcd, and 100 pl of this mixture
(i.e., 1 pI plasma) was tested in the INCA. The C1 activation
was calculated in the same way as described for the sera, and
again a mixture without an incubation at 37°C was used as a
"zero" level. As a positive control, AHG dilutions (10 pl)
were added to EDTA-plasma ( I 0 pl). After an incubation for
15 minutes at 37°C. 480 pi VBS' ' containing I mM CaCI?
and 0.1 units h c p a r i n h l were added, and these mixtures
were further tested as described above.
Isolation of IgM rheumatoid factors (IgM-RF). An
appropriate amount of serum was dialyzed against 10 mM
acetate, pH 5.5. The precipitate formed was washed once
with the same buffer and then dissolved in a buffer consisting
of 10 mM acetate and 0.5M NaCI, p H 4.0, and applied to a
Sephacryl S300 column equilibrated in the same buffer. 'Ihe
IgM-containing fractions were collected, dialyzed against
PBS, and absorbed into Sepharose 4B to which rabbit IgG
had been coupled (10 mg IgG/300 mg Sepharose). The
Sepharose was washed and the IgM-RF were eluted with
O.1M glvcin HCI O.5M NaCI, pH 2.5. The eluate was
42
HACK ET AL
RESULTS
CIC
E
3
Levels of CiiCii(Ci-In)2 and CIC. CrKi$CTIn)z complexes in EDTA-plasma and serum were measured with the INCA; CIC in serum were detected
with the Clq-BT and the Clq-SP (Figure 1). Slightly
elevated levels of C i i ? ~ ( C i - I n ) z complexes in
EDTA-plasma were found in 35 of the 57 patients
compared with normal donors; this indicated a low
extent of C1 activation in these paticnts. However,
when serum rather than plasma was tested in 6 patients, considerably higher levels were found, i.e.,
more than twice the plasma value, indicating C 1
activation during coagulation. This was never observed when normal donors were tested-22
in a
previous study (1 1) and 25 in this
study.
- Correlations between ClrCls(CT-In)z complexes, CIC, and complement components were analyzed
by linear regression (Table 1). A weak but significant
correlation between CTfCE(Ci-In), levels in EDTAplasma and CIC, as measured with the Clq-RT, was
found. The correlation with CIC measured by the C l q SP was not significant. Significant correlations between CfiCg(CT-In), complexes and C4 and C3d
levels were also found. Furthermore, data from others
(5,7) on correlations between CIC and complement
factors were confirmed.
Activation of C1 by AHG in serum and EDTAthere
plasma. At present, - is no information about the
clearance time of C l r C l s(Ct-In), complexes in vivo.
So the levels of these complexes, low when compared
with the CIC levels, might be the result of a rapid
clearance in*vivo. Therefore, we developed an assay
to study activation of C1 by CIC in vitro. This assay
30
:
20
"t
10
0
i
-
a
Clq-BT
Clq-SP
I
-
EDTA-plasma
- _
serum
Figure 1. ClrCls(Cf-In)z complexes and circulating immune
_ _complexes (CIC) in 57 patients with rheumatoid arthritis. ClrCls(CiIn)2 complexes were measured in BDTA-plasma and in simultaneously obtained serum with the C I-inhibitor complex assay
(INCA) and expressed as units per ml. CIC wcre measured in serum
with the C l q binding test and the solid phase assay ( C l q - B l and
Clq-SP) and expressed as g g eqtiivalent aggregated human IgG/ml.
Normal values (mean + 2 SD) are indicated by the dotted lines.
adjusted to a pH of 7.0 by the addition of 0.1M NaOH and
dialyzed against PBS. By double immunodiffusion, it was
shown that the resulting preparations only contained IgM.
To test the effect of IgM-RF on the Clq-B?' and the
activation of C 1, a volume of IgM-RF was added to an equal
volume of NHS. One hundred microliters of this mixture
was then incubated with 0.4M EDTA, pH 7.5, and tested in
the Clq-BT. To assess the C1 activation of IgM-RF in NHS,
20 pl of the IgM-RF NHS mixture was added to 480 pl
VBS ' - and further tested as described above for C1 activation in EDTA-plasma.
_ _
Correlation by linear regression analysis between ClrCls(Ci-ln)2 complexes, circulating immune complexes, complement
components, and spontaneous C1 activation in 57 patients with rheumatoid arthritis*
Table 1.
cr;Cls(Ci-In)2
CECii(Ci-In)2(EDTA-plasma)
Clq-BT
Clq-SP
1
0.28
(<0.025)
0.13
(W
c4
c3
C3d
Spontaneous CI activation
* P value in parentheses, NS
-0.30
(=0.0 I )
-0.10
(NS)
0.28
( <0.025)
0.36
(<0.005)
ciq-m
Clq-SP
c4
c3
C3d
Spontaneous
C1 activation
I
0.33
(<0. 0 I )
-0.54
(C0.005)
-0.32
(<0.01)
0.34
(=O.O05)
0.58
(<0.005)
I
--0.27
(<0.05)
-0.34
(=0.005)
0.15
(W
0.44
(<0.005)
I
0.56
(<0.005)
-0.24
(C0.05)
-0.49
(<0.005)
1
-0.10
I
(NS)
-0.33
(<0.01)
= not significant. Clq-BT = Clq binding test. Clq-SP = Clq solid phase assay.
0.22
(=0.05)
1
LACK OF C1 IN RA
43
was performed by measuring the generation of
ClrCE(CT-In)2 complexes during an incubation at
37°C. As a positive control, AHG was used as a model
NHS
for IC. The addition of AHG to - resulted in a
dose-dependent generation of ClrC 1s(Ci-In), complexes that was complete within 60 minutes of incubation at 37°C. For further experiments, an incubation
time of 60 minutes was therefore used unless indicated
otherwise.
As will be discussed below, there were several
reasons for measuring C1 activation in EDTA-plasma
instead of in serum. To allow activation of C1, the
chelating effect of the EDTA had to be overcome.
Therefore, 10 pl EDTA-plasma was diluted in 490 pI
VBS'+ containing additional CaClz (1 mM) as well as
heparin (0.1 unitdml) to prevent coagulation in the
tubes. Experiments (not shown here) indicated that
this heparin concentration did not influence the generation of CiiCi-s(Ci-In)2 complexes in NHS by AHG.
So, although an inhibiting effect of heparin on the
enzymatic activity of C1 has been described (15). it
apparently did not interfere in our system, probably
because of the low concentrations used. Normal human plasma, to which AHG was added, was then
tested for C1 activation and compared with the results
obtained with serum. The resulting C1 activation in
EDTA-plasma by AHG was Comparable with that in
normal serum (Figure 2). In both assays, 31 pg
AHG/ml of plasma or serum resulted in a detectable
C1 activation. This was the lowest amount of AHG
detected by the Clq-SP; however, this amount was not
1
serum
E 5 0
0
0
500
1000
Fig AHG/ml serum or EDTA-plasma
Figure 2. Activation of C l by aggregated human IgG (AHG) in
normal serum and in diluted recalcified EDIA-plasma. Various
concentrations of AHG were added to serum and EDTA-plasma. C I
activation was measured
_ _ as described in Patients and Methods and
expressed as A U ClrCls(Ci-In)z
complexes per ml.
detected by the Clq-BT. With the highcr AHG concentrations, a lower assay response was obtained in
EDTA-plasma compared with serum. A lower assay
response was also obtained when AHG was added to
NHS pretreated with 10 mM EDTA. Sucrose gradient
density analysis revealed that, in the EDTA-plasma
system, the
concentrations of AHG generated
_ higher
not only C I rC I s(Ci-In), complexes sedimenting in the
9s position but also complexes that sedimented slightly more slowly. Evidently, when C1 is activated during
its assembly from its subcomponents, some of the
complexes formed by the action of the Ci-In are
composed of CiiCi-In and CfiCi-In. Since CECT-In
complexes do not bind to anti-Cls Sepharose, this
results in a lower binding of "'I-anti-Ci-In in the
INCA.
Spontaneous C1 activation in EDTA-plasma. To
assess the activation of C1 in vitro, presumably caused
by_ CIC,
_ the C1 activation-i.e., the generation of
C 1rC 1s(Cf-In), complexes-was measured during 60
minutes at 37°C in diluted recalcified EDTA-plasma.
This activation was called spontaneous C 1 activation
when no activator (i.e., AHG) was added. In normal
donors, the uppcr limit of this spontaneous C1 activation was 5.5 units/ml/hour. In 16 of the 57 RA patients,
a slightly increased activation was found compared
with that in normal donors (Figure 3). The highest
spontaneous C1 activation encountered in these patients was comparable with that caused in EDTAplasma by 125 pg AHG/ml (Figure 2). Yet several
patients had CIC levels of 1,000 pg equivalent
AHG/ml when measured by the C lq-BT. The addition
of AHG (1 mg/ml) resulted in a considerable C1
activation in all EDTA-plasma samples, indicating that
a low degree of spontaneous C1 activation was not due
to either a dysfunctional C1 or a dysfunctional Ciinhibitor. Furthermore, an inhibiting effect of the
heparin is also excluded by this control.
The correlation between spontaneous C1 activation, CIC, and complement factors was analyzed by
linear regression (Table I ) . The correlation between
spontaneous C1 activation and results of the Clq-BT is
given in Figure 4. In general, patients with an increased spontaneous C I activation tended to have high
levels of CIC, as measured with the Clq-BT. However, several patients with high CIC levels showed a
normal spontaneous Cl activation. The same pattern
was found when the spontaneous Cl activation was
plotted against results of the Clq-SP (Figure 5). Again,
some patients with elevated CIC levels did not show
an increased spontaneous C I activation. The reverse
HACK ET AL
44
0
0
0.
::
........
..........
......
.'........
.......
0
0
8".
0.
0
0
0.
0
all
0 1
I
.
50
100
150
n .
Y
250 545
Clq-SP(pg eq AHG/rnl)
200
Figure 5. Correlation between the spontaneous C1 activation in
diluted recalcified EDTA-plasma (expressed as A U per ml) and
circulating immune complex levels in serum as measured with the
Clq solid phase assay (Clq-SP) (expressed as pg equivalent aggregated human IgG/ml). Dotted lines indicate normal values (mean t- 2
SD). Seronegative patients are indicated by open circles.
0I
normal
donors
RA
Figure 3. Spontaneous activation of C1 in diluted recalcified
EDIA-plasma from normal donors and from rheumatoid arthritis
(RA) patients. CI activation during an incubation for 60 minutes at
37°C was measured with the CI-inhibitor complex assay and expressed as AL' per rnl.
.
..
E 16
c
0
.c
.-9 12
c
u
m
G
: a
i
c
m
c
C
0
a
v)
1
0
0
0
I
1
I
1
250
500
1
I
750
1000
Clq-BT(pg eqAHG/rnl)
Figure 4. Correlation between the spontaneous CI activation in
diluted ELYTA-plasma (expressed as AU per ml) and circulating
immune complex levels in serum as measured with the Clq binding
test (C Iq-B'T) (expressed as pg equivalent aggregated human
IgG/ml). Dotted lines indicate normal values (mean t 2 SD).
Seronegative patients are indicated by open circles.
was also found: no CIC were detected by the Clq-SP
in several patients who showed an increased spontaneous C1 activation. In these patients, however, CIC
were detected by the Clq-BT. All patients with an
increased spontaneous C1 activation had CIC detectable by either the Clq-BT or the Clq-SP or both.
The kinetics of the spontaneous C1 activation
was investigated in plasma and serum from the 2
patients with the highest spontaneous Cl activation.
The results obtained in 1 patient are shown in Figure 6,
and the results in the other were similar. Compared
with the C1 activation by AHG in normal plasma and
serum, a marked difference was noted: the spontaneous C1 activation still continued after 60 minutes of
incubation. The assay response was higher in serum
As explained above, the formathan in EDTA-plasma.
_ _
tion of ClrC1-In complexes accounts for this difference. The results also show that the increased spontaneous C1 activation was not dependent on the
presence of heparin. Furthermore, the spontaneous C 1
activation in serum in a 1 :2 dilution was much less
than in a 1 :50 dilution. This was also found, though
not so strikingly, when AHG was added to NHS.
Relation to clinical features. The patients were
divided into several groups based on either clinical
features or presence of IgM-RF. Results from the Clq-
LACK OF C1 IN RA
45
-E
.
2
100
-.-6
.?
0
?
6
-85
lu
3
0
0
2
4
incubation-time (hrs)
Figure 6. ‘The course of C1 activation in a rheumatoid arthritis
patient. Recalcified EDIA-plasma ( 1 :SO diluted) and serum ( 1 :50
diluted in the presence or absence of heparin, 0.1 unit/ml,
_ - and 1 : 2
diluted) were incubated at 37°C. At various times, ClrCls(CT-In)z
complexes reflecting C1 activation were measured with the CTinhibitor complex assay and expressed as units per ml.
- BT, Clq-SP, ClrCls(CT-In)zcomplexes and spontaneous CI activation obtained in these groups were analyzed by unpaired Student’s f-test. Table 2 shows the
following observations:
1. Patients with seropositive disease had higher levels
of CIC as detected by the Clq-BT, higher CfiCGCCiIn)?complexes, and a higher spontaneous CI activation
compared with patients who had seronegative disease.
2 . Patients with an elevated erythrocyte sedimentation rate
- and/or
- low hemoglobin levels had higher CIC
and C I rC 1~ ( C i - l nlevels
)~
and a higher spontaneous
Cl activation compared with patients who had inactive
disease. The differences, however, were not marked.
3. Patients with extraarticular manifestations had
slightly higher CIC levels as detected by the Clq-BT
and also higher levels of C ~ C ~ ( C ~ - Icomplexes
n)2
and a higher spontaneous C1 activation, compared
with patients who had only joint involvement.
Three patients with clinical evidence of vasculitis (i.e., nailfold lesions and skin ulcers) were included
in this study. All 3 showed an increased spontaneous
C1 activation (2 of them being the highest encountered
in this study) together with high CIC levels.
Relation to the presence of rheumatoid factors.
Comparison of the results from the patients with
seropositive disease with those from the patients with
seronegative disease (Table 2 ) suggested that the different assays might be influenced by the presence of
IgM-RF. Sera with a high RF titer tended to have a
high Clq-BT; a high Clq-SP was much less common
(Figures 4 and 5 ) . A significant correlation in the
seropositive patients was found between IgM-RF titer
and either results of the Clq-BT (r = 0.91, P < 0.005)
or results of the spontaneous C1 activation (r = 0.58, P
< 0.005). Therefore, the influence of IgM-RF on these
assays was further investigated. Isolated IgM-RF were
added to NHS in titers comparable with those found in
the patients. The resulting IgM-RF-NHS mixtures
were then tested in the Clq-BT and the C1-activation
assay as described in Patients and Methods. All IgMRF preparations in NHS gave a strongly positive ClqBT, sometimes accompanied by an increased C1 activation (Table 3). When the preparations were tested in
the absence of NHS, the results of the Clq-BT were
not higher than 1596, indicating that the increased ClqBT of the IgM-RF-NHS mixtures was not due to a
direct interaction of IgM-RF with Iz5I-Clq.
Table 2. Circulating immune complex (CIC) levels and C1 activation in vivo and in vitro in patients with various forms of rheumatoid arthritis
Form of rheumatoid arthritis
Number of
patients
Seropositive
Seronegative
32
25
Active
Inactive
38
Extraarticular
Only synovitis
24
33
Normal values (mean
19
+ 2 SD)
CIC*
CI activation
Clq-RT
Clq-SP
680 i 329
166 5 179
( P < 0.0001)
531 t_ 370
302 2 340
( P = 0.014)
555 i 375
382 1 362
(P = 0.04)
I25
84 101
62 r 54
(NS)
YO 2 47
43 30
( P = 0.02)
67 + 62
80 rf- YO
*
+
(NS)
30
In v i v o t
In vitrot
6.1 ? 1.1
5.0 I 1.2
(P = 0.0004)
5.8 2 1 . 1
5.1 ? 1.4
( P 0.02)
6.2 5 0.8
5.1 2 1.3
( P = 0.0005)
5.0
6.6 ? 2.9
4.1 2 1.5
( P = 0.0001)
6.1 t 2.8
4.3 2 1.9
(P = 0.008)
6.5 ? 3.0
4.8 I- 2.2
(P = 0.006)
5.5
7
* Expressed as pg equivalent aggregated human IgG/ml (mean 5 SD). Clq-BT = C l q binding test; Clq-SP = Clq solid phase assay. P value by
Student’s r-test, NS = not significant.
_ _
t Expressed as units ~ ~ ~ ~ s ( ~ i - l n(mean
) , / mL l SD).
t Expressed as AU ClrCls(C1-ln)z/ml (mean ? SD).
HACK ET AL
46
Influence of IgM-RF in normal human serum (NHS) on
the Clq-BT (Clq binding test) and the spontaneous C1 activation*
Table 3.
IgM-RE'
no.
Titer
1
400
2
3
4
PBS
400
200
80
-
96
C I q-RT
CI activation,
unitsiml
97
88
13
3
6.5
3.5
3
64
56
-
* Isolated IgM-rheumatoid factor (IgM-RF) was added to NHS. The
latex agglutination titer, the Clq-BT, and the CI activation of the
mixtures were measured during 60 minutes at 37°C. PBS = phosphate buffered saline.
DISCUSSION
Circulating immune complexes have often been
detected by Clq methods in patients with RA and have
been held responsible, at least in part, for the complement abnormalities observed in these patients (57,16,17).
There is, however, surprisingly little information about whether these CIC indeed activate complement. This is probably due to technical reasons,
because the methods proposed to assess complement
activation by CIC need either heat-inactivation (18) or
polyethylene glycol (PEG) precipitation (19) of the
blood sample; both procedures possibly change the
structure of CIC and thereby their complement activating properties. Furthermore, heat-inactivation of
serum may cause aggregation of immunoglobulins and,
therefore, yield false-positive results when complement activation by CIC is studied (20). We applied the
INCA, recently developed in our laboratory (1 I), to
answer the question of whether CIC detected by Clq
assay can activate complement. This INCA enabled us
to measure C1 activation not only in vivo, but also in
vitro without maltreating the blood samples.
- To assess C1 activation in vivo, ClrCls(Ci-In)z
complexes in EDTA-plasma and paired serum from
RA patients were measured. Only slightly elevated
levels were found in EDTA-plasma in 61% of the
patients, although several patients had high levels of
CIC, especially when measured with the Clq-BT.
When 125 pg
_ AHG
_ was added to 1 ml NHS, about 18
units of C 1rC 1s(CT-In)z complexes were generated
(Figure 2). Such levels of CcCG(Ci-In), were not
observed in EDTA-plasma from any of the patients,
though many patients had CIC levels exceeding 125 pg
_ _ Surprisingly, considerably higher
equivalent AHG/ml.
levels of ClrCl s(Ci-In)z complexes were found in
corresponding sera from some patients. Experiments
(not shown here) indicated that this difference was not
due to a factor in plasma interfering in the INCA.
Therefore, the difference indicates C 1 activation in
vitro during the coagulation process. Because this was
not found in normal donors, it is unlikely that this
activation of C1 in vitro is caused by the coagulation
process itself. Another explanation is that C1 is activated by CIC during coagulation. This explanation fits
with kinetics of the Cl activation by AHG in NHS,
because most of this activation occurs within 30 minutes after the addition of AHG to NHS. However, it is
also possible that the C1 activation during coagulation
is caused by the formation of new immune complexes.
A source of these newly formed immune complexes
might be antinuclear antibodies (21) reacting with
nuclear antigens that are released from damaged leukocytes during coagulation. For this reason, one might
wonder whether measurement of CIC should be performed in EDTA-plasma instead of in serum. Further
study is needed to clarify this point.
A rapid clearance of Ci?CK(Ci-In), complexes
in vivo may account for the rather low values of these
complexes, at least when compared with the CIC
levels, in plasma from the RA patients. Therefore, we
also studied the activation of C l by CIC in vitro. For
reasons explained above, we measured the spontaneous activation of C1 in EDTA-plasma. To overcome
the chelating effect of EDTA without increasing the
ionic strength unduly, EDTA-plasma was tested in a
1 :50 dilution. However, this system has some disadvantages: C1 is more easily activated when tested in a
1 :50 dilution compared with a 1 :2 dilution (Figure 6).
Furthermore, when CI is activated
_ _during its assembly
_ _
from the subcomponents, ClrCl-In and
_ _ClsC1-In
complexes are formed together with ClrC ls(Ci-In)2.
Third, owing to the presence of calcium ions, the
coagulation system is also activated, resulting in the
formation of fibrin. This problem was circumvented by
the addition of heparin in a dose just high enough to
prevent fibrin formation without affecting the activation of C1. Despite these limitations, when AHG was
added to normal EDTA-plasma, the resulting C1 activation was comparable with that in normal serum
(Figure 2). Therefore, the system is suitable to test the
spontaneous C1 activation in EDTA-plasma.
Several noteworthy observations were made
when the spontaneous C1 activation in the RA patients
was measured. First, an increased spontaneous C1
activation was found in 26% of the patients, whereas
89% had elevated CIC levels. So, 63% of the patients
had a normal spontaneous C1 activation despite elevated CIC levels. Because the assay is sensitive
enough to detect C l activation by small amounts of
47
LACK OF C l IN RA
AHG (Figure 2), we conclude that in these patients
CIC do not activate C1, although these CIC are
detected by Clq. This can also be derived from the
work by Berglund et al
_ (22),
_ although in that study for
technical reasons C I rC 1 s(Ci-In)? complexes were
measured in serum.
The finding of a normal C1 activation despite
high levels of CIC can be explained by false-positive
reactions in both Clq assays. Concerning the Clq-SP,
we have some evidence that the IgG that is detected by
this assay in sera from patients is indeed not always
immune complex in nature. because it often sediments
as monomeric IgG and does not bear either C4 or C 3
(Hack et al: manuscript in preparation). In the Clq-BT
assay false-positive results may also occur. Substances that in the presence of polyethylene glycol
directly bind to C lq may cause an increased precipitation of the ''sI-CIq, as has been reported for heparin
and fibrinogen ( 2 3 ) . The very significant correlation
between the titer of RF and the results of the Clq-BT
suggests that IgM-RF are largely responsible for the
results in the Clq-BT in the RA patients. In fact, the
addition of isolated IgM-RF to NHS resulted in a
positive Clq-HT. This is possibly due to the formation
of IgM-RF-IgG complexes (24). An unsolved problem
is whether this complex formation already occurs in
vivo and may therefore have biologic relevance, or
whether it only occurs in vitro, especially in the
presence of PEG. A specific assay for the detection of
IgM-RF-IgG complexes would solve this problem.
However, indirect evidence that some IgM-RF can
complex with IgG in the absence of PEG is the fact
that some IgM-RF added to NHS can activate C1
(Table 3 ) . Conversely, not all IgM-RF in NHS activate
CI . Therefore, these IgM-RF may form complexes
with IgG only in the presence of PEG and thus
represent false-positive reactions in the C Iq-BT.
Second, an increased spontaneous CI activation was found in 28% of the patients. This activation
was moderate when compared with the high CIC
levels encountered in these patients. Patients with an
increased spontaneous C1 activation tended to have
high CIC levels. This might suggest that this C1
activation is caused by CIC. However, the kinetics of
the spontaneous C1 activation in the patients differed
from that caused by AHG in normal plasma: the
spontaneous activation of CI in the RA patients still
continued after 60 minutes of incubation (Figure 6), in
contrast with the CI activation by AHG. We believe
that this continued activation is caused by the interaction of IgM-RF with IgG (24), an interaction of low
affinity, thereby resulting in the continuous associa-
tion and dissociation of the IgM-KF-IgG complex. In
favor of this explanation is the finding that the spontaneous C1 activation correlated significantly with the
titer of rheumatoid factors. Furthermore, it has been
demonstrated that the complement abnormalities in
RA patients correlate with the complement-activating
properties of the IgM-RF (25). However, an alternative explanation could be that the activation of CI is
not caused by CIC but, for example, by proteolytic
enzymes (26,27). Clearly, this point needs further
investigation.
_ _
Increased ClrC ls(CT-In), complexes, reflecting CI activation in vivo, and an increased spontaneous Cl activation, possibly caused by CIC, were more
often encountered in patients with extraarticular manifestations. Scott et a1 (28) reported that CIC levels
detected by the Clq-RT in patients with joint involvement alone were not different from those detected in
patients with vasculitis, whereas anticomplementary
activity was only observed in sera from patients with
vasculitis. The 3 patients with vasculitis in the present
study showed an increased spontaneous C1 activation.
One might therefore speculate, as has also been suggested by Bourke et a1 ( 2 9 , that extraarticular complications, especially vasculitis, are mediated by complement-activating CIC and that nonactivating CIC are
quite harmless in this respect. Careful clinical studies
are needed to confirm this point, especially concerning
the temporal relationship between the presence of
complement-activating CIC and the appearance of
extraarticular manifestations.
In conclusion, we have demonstrated that measurement of CIC by means of Clq (CIq-BT and/or
Clq-SP) does not mean that these CIC activate C1.
Future studies on the role of circulating immune
complexes in diseases should take this into account.
ACKNOWLEDGMENTS
We thank our colleagues Drs. T. A. Out, L. A.
Aarden, A. J . Hannema, T. E. W. Feltkamp, M. R. Daha,
and A. J . G. Swaak for helpful discussions and for critically
reading the manuscript. The assistance of Ms Carry Bloklander and Mrs. Jetty Gerritsen in typing the manuscript is
highly appreciated.
REFERENCES
1. Zubler RH, Lambert PH: Detection of immune complexes in human diseases. Prog Allergy 24: 1-48, 1978
2. Theofilopoulos A N , Dixon FJ: Biology and detection of
immune complexes. Adv Immunol 28:89-220, 1979
3. Hay FC, Nineham LJ, Roitt IM: Routine assay for the
HACK ET AL
detection of immune complexes of known immunoglobd i n class using solid-phase Clq. Clin Exp Immunol
24:396-400, 1976
4. Zubler KH, Lange G , Lambert PH, Miescher PA: Detection of circulating immune complexes in unheated
sera by a modified "'I-CIq-binding test: effect of heating on the binding of C l q to immune complexes and
application of the test to systemic lupus erythematosus.
J Immunol I16:232-235, 1976
5. Zubler RH, Nydegger UE, Perrin LH, Fehr K, McCormick J. Lambert PH, Miescher PA: Circulating and
intra-articular immune complexes in patients with rheumatoid arthritis: correlation of "'I-C 1q-binding activity
with clinical and biological features of the disease. J Clin
Invest 57:1308-1319, 1976
6. Hay FC, Nineham LJ, Perumal K , Koitt 1M: Intraarticular and circulating immune complexes and antiglobulins (IgG and IgM) in rheumatoid arthritis: correlalation with clinical features. Ann Rheum Dis 38:l-7,
I979
7. Nydegger UE, Zubler RH, Gabay R , Joliat G , Karagcvrekis CH, Lambert PH, Miescher PA: Circulating complement breakdown products in patients with
rheumatoid arthritis: correlation between plasma C3d,
circulating immune complexes and clinical activity. J
Clin Invest 59:862-868, 1977
8. Hack CE, Eerenberg-Belmer AJM, Hannema AJ, Out
TA, Aalberse KC: Polyethylene glycol enhances the
binding of C Iq to circulating immune complexes.
J Immunol Methods 44:211-221, 1981
9. Fiist G. Medgyesi GA, Rajnavolgyi E, Csecsi-Nagy M ,
Csikora K , Gergely J: Possible mechanisms of the first
step of the classical complement activation pathway:
binding and activation of C I . Immunology 35:873-884,
1978
10. Ziccardi RJ, Cooper NR: Active disassembly of the first
complement component, C i , by CT inactivator. J Immunol 123:788-792, 1979
11. Hack CE, Hannema AJ, Eerenberg-Belmer AJM, Out
TA, Aalberse RC: A CI-inhibitor complex assay
(INCA): a method to detect CI activation in vitro and in
vivo. J Immunol 127:1450-1453, 1981
12. Kopes MW, Bennett GA, Cobb S, Jacox R, Jessar RA:
1958 revision of diagnostic criteria for rheumatoid arthritis. Bull Rheum Dis 9:175-176, 1958
13. Yonemasu K , Stroud RM: Clq: rapid purification method for preparation of monospecific antisera and for
biochemical studies. J lmmunol 106:304-313, 1971
14. Perrin LH, Lambert PH, Miescher PA: Complement
breakdown products in plasma from patients with membranoproliferative or other glomerulonephritis. J Clin
Invest 56:165-176, 1975
15. Strunk R, Cohen HR: Inhibition of the enzymatic activity of the first component of complement (C]) by heparin. Clin Imniunol Inimunopathol 6:248-256, 1976
16. Hurd ER, Chubick A, Jasin HE, Ziff M: Increased Clqbinding immune complexes in Felty's syndrome: comparison with uncomplicated rheumatoid arthritis. Arthritis Rheum 22:697-702, 1979
17. Halla JT, Volanakis JE, Schrohenloher KE: Immune
complexes in rheumatoid arthritis sera and synovial
fluids: a comparison of three methods. Arthritis Rheum
22~440-448, 1979
18. Mowbray JF. Hofbrand AV, Holborrow EJ, Seah PP,
Fry L: Circulating immune complexes in dermatitis
herpetiformis. Lancet i:400-402, 1973
19. Harkiss GD. Brown DL: Detection of immune complexes by a new assay, the polyethylene-glycol precipitationcomplement-consumption test (PEG-CC). Clin Exp Immunol 36:117-129, 1979
20. Soltis RD. Hasz D, Morris MJ, Wilson ID: The elrect of
heat-inactivation of serum on aggregation of immunoglobulhs. Immunology 36:37-45. 1979
21. Ward DJ, Johnson GD, Holborrow EJ: Antinuclear
factor in rheumatoid arthritis, its incidence and clinical
significance. Ann Rheum Dis 23:306-310, 1964
22. Berglund K, Laurel1 A-B, Nived 0, Sjoholm AG, Sturfelt G: Complement activation, circulating C Iq-binding
substances and inflammatory activity in rheumatoid
arthritis: relations and changes on suppression of inflammation. J Clin Lab Immunol 4:7-14, 1980
23. Faaber P, Capel PJA, Koene KAP: The interference of
fibrinogen and heparin with the determination of circulating immune complexes in the Clq-binding assay. Clin
Exp Immunol 45590-594, 1981
24. Franklin EC, Holman HK, Miiller-Eberhard HJ, Kunkel
HG: An unusual protein component of high molecular
weight in the serum of certain patients with rheumatoid
arthritis. J Exp Med 105:425-438, 1957
25. Kaplan RA, Curd JG, Deheer DH. Carson DA, Pangburn MK, Muller-Eberhard HJ, Vaughan JH: Metabolism of C4 and factor B in rheumatoid arthritis: relation
to rheumatoid factor. Arthritis Kheum 23:911-920, 1980
26. Ratnoff OD, Naff GB: The conversion of CIS to C'I
esterase by plasmin and trypsin. J Exp Med 125:337358, 1967
27. Laurel1 A-B, Johnson U , Martensson U , Sjoholm AG:
Formation of complexes composed of Cfi, C G and C i
inactivator in human serum on activation of CI. Acta
Pathol Microbiol Scand [C] 86:2!39-306, 1978
28. Scott DGI, Bacon PA, Allen C, Elson CJ, Wellington 'r:
IgG-rheumatoid factor, complement and immune complexes in rheumatoid synovitis and vasculitis: comparative and serial studies during cytotoxic therapy. Clin
Exp Immunol 43:54-63, 1981
29. Bourke BE, Moss IK, Mumford P, Horsfall A, Maini
K N : The complement-fixing ability of putative circulating immune complexes in rheumatoid arthritis and its
relationship to extra-articular disease. Clin Exp Immunol 48:726-732, 1982
Документ
Категория
Без категории
Просмотров
0
Размер файла
842 Кб
Теги
two, detected, patients, immune, method, activation, complexes, circulating, despite, lack, arthritis, rheumatoid, c1q
1/--страниц
Пожаловаться на содержимое документа