| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
*
Inflammation and
Cytokine Research Units, School of Pathology, University of New South Wales, Sydney, Australia; and
Department of Rheumatology, Prince of Wales Hospital, Sydney, Australia
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
when
stimulated with ß2GPI in vitro. In this study, we show that
stimulation of PBMCs from 20 APS patients with ß2GPI induced
substantial monocyte tissue factor (TF) (80 ± 11 TF stimulation
index (TF-SI)), whereas no induction was observed using PBMCs from 13
patients with aPL Abs without APS (6 ± 1 TF-SI) or 7 normal and 7
autoimmune controls (5 ± 1 and 3 ± 1 TF-SI, respectively)
(p < 0.0001). TF induction on monocytes by
ß2GPI was dose dependent and required CD4+ T lymphocytes
and class II MHC molecules. Because monocyte TF induction by ß2GPI
was observed in all patients with APS, but not in any patient with aPL
Abs without APS, this response is a potentially useful predictor for
APS in patients with aPL Abs, as well as providing mechanistic insight
into thrombosis and fetal loss in these
patients. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The APS is the commonest acquired cause of predisposition to recurrent thromboembolism, and management often requires lifelong high-dose anticoagulation (3). Although there is strong epidemiological evidence for an association between aPL Abs and APS (1), two issues remain problematic. The first relates to the difficulty clinicians face in defining risk of APS in an individual with aPL Abs, particularly when there is no history of a thrombotic event or pregnancy failure, because the presence of aPL Abs provides a relatively low predictive value for determining the risk of future events (4). Population studies provide some evidence that the IgG isotype and levels of aPL Abs correlate with increased risk for clinical complications of APS (5, 6), but this is controversial. Detection of Abs with defined specificity for ß2GPI may be more specific (7, 8, 9), but anti-ß2GPI assays lack standardization and are not currently used in routine patient care.
The second problem is a lack of understanding about mechanisms involved
in a prothrombotic diathesis. The traditional paradigm is that aPL Abs
themselves mediate thrombosis, but despite investigations of various
aspects of both procoagulant and anticoagulant pathways, no consistent
abnormalities have been found (10). Although ß2GPI is
clearly a strong autoantigen that stimulates a vigorous B cell-humoral
response, little attention has been focused on cellular immunity to
this Ag. Recently, we reported that some 44% of patients with APS
possess circulating, autoreactive, ß2GPI-specific
CD4+ T cells that proliferate and produce IFN-
when stimulated with ß2GPI in vitro (11). Importantly,
these responses to ß2GPI were only observed in patients with
histories of thromboses or fetal loss, suggesting it may be a reliable
marker for increased risk of these events in individual patients. One
consequence of Ag stimulation of T cells is induction of monocyte
procoagulant activity (PCA), measurement of which is a more sensitive
marker of cellular immunity than proliferation (12).
Elevated levels of plasma and monocyte-associated tissue factor (TF)
have been reported in patients with aPL Abs (13, 14), and
elevated monocyte TF correlates with histories of thrombosis
(14). Thus, we hypothesized that a procoagulant diathesis
in APS patients may be due to up-regulation of monocyte TF as a result
of activation of ß2GPI-specific autoreactive
CD4+ T lymphocytes.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Thirty-three patients (31 female and 2 male) with aPL Abs were
studied (Table I
). Twenty-nine had serum
aCL Abs of either IgG or IgM isotype at levels >5 GPL or MPL units,
respectively, at the time of this study. Three additional patients
(patients 7, 8, and 20) did not have elevated aCL Abs, but had
previously been documented to have serum IgG-aCL Abs of 26, 15, and 33
GPL units, respectively, 
1 year previous to this study. One of these
(patient 20), and another (patient 9), had LA. These four patients all
had elevated serum anti-ß2GPI Abs (Table I) and had been
diagnosed with the primary APS (PAPS) by their treating physicians. The
clinical diagnoses in the 33 patients studied were systemic lupus
erythematosus (SLE) in 14, PAPS in another 14, and 1 each with primary
Sjogren’s Syndrome, rheumatoid arthritis, giant cell arteritis, livedo
reticularis, and peripheral vascular disease.
|
). Group I
comprised 13 patients who had histories of thrombosis; five of these
(patients 2, 8, 10, 11, and 12) also had histories of recurrent fetal
loss. Nine of the 13 were being treated with warfarin. Group II
comprised seven patients who had suffered recurrent fetal loss but not
thrombosis. None of these patients were being treated with warfarin
concurrently. Thus 20 of the 33 (groups I and II) had the APS. The
remaining 13 patients (group III) had aPL Abs without clinical features
of the APS. Two control groups were also studied. Seven patients with various autoimmune diseases (SLE, rheumatoid arthritis, or giant cell arteritis) and seven healthy controls (both groups aCL Ab negative) were tested. The median ages for the patients with aPL Abs and the two control groups were 43, 53, and 33, respectively. Blood was collected in citrate dextrose, diluted 1:3 with PBS, layered over Lymphoprep (Nycomed Pharma, Oslo, Norway), and centrifuged at 1800 rpm. PBMCs were collected, washed twice in PBS, and resuspended at 1 x 107/ml in RPMI 1640 plus 20% FCS (Life Technologies, Gaithersburg, MD), then frozen in 7.5% DMSO and stored in liquid nitrogen until further use.
aPL Ab ELISAs
IgG and IgM-aCL Abs were measured in serum using a standardized kit (Medical Innovations, Sydney, Australia). To detect anti-ß2GPI Abs, purified ß2GPI was coated at 5 µg/ml on high-binding microtiter plates (Costar 3590, Corning Costar, MA) in Tris buffer, pH 8.4, at 4°C overnight. After washing four times with PBS, plates were blocked with 1% skim milk powder in PBS for 1 h at 37°C. Standards from the IgG-aCL Ab ELISA diluted 1:20 in 0.3% gelatin/PBS were used to generate a standard curve expressed in arbitrary units. Serum was assayed in duplicate at 1:100 dilution in 0.3% gelatin/PBS, 100 µl/well. Sera from four healthy individuals were used as negative controls. Plates were incubated at room temperature for 2 h then washed with PBS. Bound Abs were detected by addition of a 1:500 dilution of HRP-conjugated goat anti-human IgG Ab (Dako, Glostrup, Denmark) and development with tetramethylbenzidine substrate. After 15 min, 100 µl 1 M H2SO4/well was added, and the plate was read at A405 nm using a Titertek Multiskan Plus MKII plate reader (Labsystems, Lierbyen, Norway).
Cells and incubations
PBMCs from patients and normal and autoimmune controls were isolated and resuspended at 1.5 x 105/ml in AIM-V serum-free medium (Life Technologies). ß2GPI was purified by standard methods (11) under LPS-minimizing conditions and filtered sequentially through four Zetapore syringe filters (Cuno, Meriden, CT). LPS levels were measured using the Limulus amebocyte lysate assay (Pyrotell, Woods Hole, MA) to ensure that levels were <0.125 endotoxin units. PBMCs (1.5 x 106 in 100 µl AIM-V) were cultured with media alone, 100 µl ß2GPI (25 µg/ml), LPS (1 µg/ml) (Sigma, St. Louis, MO), or tetanus toxoid (CSL, Melbourne, Australia) (2.5 LF U/ml) in 96-well round-bottom plates (Nunc, Roskilde, Denmark) and incubated at 37°C in 5% CO2 in air for 24 h. Plates were centrifuged at 1400 rpm, supernatants were removed, PBMCs were resuspended in fresh AIM-V (200 µl/well), and plates were stored at -20°C. In some experiments, PBMCs were cultured with concentrations of ß2GPI ranging from 1 to 25 µg/ml.
To determine whether monocyte PCA was lymphocyte dependent, PBMCs from two patients were incubated at 2 x 106/ml in a flat-bottom 24-well plate in RPMI 1640 plus 10% FCS for 2 h at 37°C. Nonadherent cells were removed, and adherent monocytes were removed with the addition of warm (37°C) AIM-V medium. Monocyte and lymphocyte populations were centrifuged, resuspended in a small volume of AIM-V, and counted. For induction of PCA, lymphocytes (1 x 106/ml) or monocytes (2 x 105/ml) or combinations of both were incubated in the presence or absence of ß2GPI (25 µg/ml) at 37°C in 5% CO2 in air for 24 h, then centrifuged and washed as described above before assessment of PCA. In some experiments, PBMCs were treated with 84 µg/ml anti-MHC class II Ab (anti-DP, DQ, DR) (Serotec, Raleigh, CA) or an irrelevant mouse IgG isotype control (Dako) during culture with ß2GPI as previously described (11).
Measurement of monocyte PCA
Total PCA of cells subjected to two cycles of freeze-thawing was determined by a one-stage plasma recalcification time with triplicate 100 µl samples added to 100 µl prewarmed (0.03 M) CaCl2 and 100 µl cold citrated human platelet-poor plasma using an automatic coagulometer (Schnitger and Gross, Amelung, Germany), and the coagulation time was measured. Activity was calculated from a standard curve (log-log plot) using dilutions of human brain extract as the standard; one thousand arbitrary units corresponded to a recalcification time of 226 s. The results were expressed as a ratio of TF units of stimulated cells/TF units of unstimulated cells (TF-stimulation index (TF-SI)).
TF activity on lysed or intact cells was determined using a continuous fluorogenic assay (15). One hundred microliters of PBMCs from responders (n = 2) were stimulated with ß2GPI (25 µg/ml) or LPS (1 µg/ml) as above. Plates were centrifuged, PBMCs washed in HBSS, centrifuged again, and supernatants discarded. The final reaction concentration of reagents was 2 ng factor VII, 1 µg factor X, and 1.6 mM factor X fluorogenic substrate/well (all from American Diagnostica, Greenwich, CT). The plate was read at 1-min intervals for 1 h at 360 nm (reference, 460 nm) using a Cytofluor Series 4000 Multiwell plate reader (Perseptive Biosystems, Framingham, MA). To confirm TF involvement, PCA activity was neutralized with 4 µg anti-TF Ab (American Diagnostica) or isotype control added to PBMCs 60 min before factor Xa generation.
Statistical analysis
Data were expressed as the mean ± SD or SEM. Differences between patient and control groups were determined using one-way ANOVA. Correlations between TF-SI values and aCL or anti-ß2GPI Ab levels were examined using the two-tailed Pearson r test.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
PBMCs from normal (N) (n = 7), or autoimmune (A)
(n = 7) controls generated low levels of monocyte PCA
in response to ß2GPI (mean ± SEM TF-SI: 5 ± 1 or 3
± 1, respectively). In contrast, ß2GPI incubated with PBMCs from
patients with aPL Abs up-regulated monocyte PCA in 20 of 33 patients,
as defined by a TF-SI of >15, which is >4 SD above the mean response
of normal controls to ß2GPI. These 20 patients were designated
TF-responders (TF-R), and the remaining 13 were designated
TF-nonresponders (TF-NR) (Fig. 1A). A highly significant
difference in mean monocyte PCA levels generated in response to ß2GPI
was observed between these two groups (80 ± 11 vs 6 ± 1
TF-SI, p < 0.0001). Moreover, the TF-SI values for all
TF-R were >25, whereas those for all TF-NR, and all controls, fell
below 15, distinguishing individual TF-R from all other groups.
|
).
Thus, for the first time, we show an absolute correlation between APS
and ß2GPI-induced monocyte PCA in patients with aPL Abs. Both major
clinical features of the APS, thrombosis and recurrent fetal loss, were
associated with ß2GPI-induced monocyte TF expression, with no
significant difference in TF-SI values between groups I and II
(p = 0.98).
Although warfarin therapy is reported to suppress mitogen-induced
monocyte TF activity (16), there was no significant
difference in TF-SI values between patients who were concurrently
treated with warfarin and those who were not, either within group I
alone (p = 0.61) or when groups I and II were
considered together (p = 0.68). The mean
IgG-aCL Ab levels in patients with APS who were TF-R (groups I and II)
did not differ significantly from those without the APS who were TF-NR
(group III) (30 ± 9 vs 26 ± 10 GPL units, Table I)
(p = 0.75). However, patients in group II with
recurrent fetal loss only had a lower mean IgG-aCL level than group I
(14 ± 8 vs 39 ± 13 GPL units, respectively), but
differences were not statistically significant
(p = 0.18). The level of IgM-aCL Abs was not
significantly different between groups I and II, nor between TF-R
(groups I and II) and TF-NR (group III) (Table I)
(p = 0.45 and 0.99, respectively).
The mean anti-ß2GPI Ab levels were not significantly different
between group I and group II patients (p =
0.51). However, 8 of 13 patients in group III (APS-negative TF-NR), in
whom anti-ß2GPI Ab levels were available, had a lower mean
anti-ß2GPI Ab level than APS-positive TF-R patients (30 ± 6
vs 60 ± 3, respectively, p < 0.0001). There was
no significant correlation between TF-SI values and either IgM-aCL,
IgG-aCL, or IgG-anti-ß2GPI Ab levels in the 33 patients with aPL
Abs (r = 0.07, 0.29, and 0.30, respectively, all
p > 0.10) (Fig. 2).
|
B). There were no
significant differences in TF-SI values to LPS between TF-R (37 ±
12), TF-NR (44 ± 12), normal controls (60 ± 38), or
autoimmune controls (57 ± 11). In addition, basal PCA of PBMCs
cultured with media alone did not differ significantly between normal
controls (171 ± 79 TF units), autoimmune controls (213 ± 73
TF units), TF-R (130 ± 26 TF units), or TF-NR (165 ± 72 TF
units) (p > 0.18 for all comparisons).
Monocyte TF induction was not observed when PBMCs from a subset
of the study population were stimulated with an irrelevant Ag, tetanus
toxoid, with TF-SI values <11 observed in the 14 individuals studied.
There were no significant differences in TF-SI values to tetanus toxoid
between PBMCs from six patients from groups I and II (patients 3, 5, 9,
11, 15, and 18), six from group III (patients 21, 22, 29, 31, 32,
and 33), or two autoimmune controls (p > 0.26)
(Table II).
|
Induction of PCA by ß2GPI was dose dependent when PBMC from five
TF-R were tested (Fig. 3). Activity was
induced with as little as 6.25 µg/ml ß2GPI and was maximal at
12.5–25 µg/ml. The PCA induced was primarily TF. Activity was
dependent on factors VII and X, and ß2GPI-stimulated viable PBMCs
from two TF-R patients generated a mean of 12 ng factor Xa, compared
with 4 ng and 29 ng following culture with media alone or LPS,
respectively. A neutralizing anti-TF Ab completely abrogated factor
Xa generation, whereas the isotype control IgG had little effect (Table III).
|
|
Induction of monocyte TF by ß2GPI required stimulation of
ß2GPI-specific lymphocytes. Table IV
shows that TF on monocytes did not increase when stimulated with
ß2GPI, whereas lymphocytes cultured with monocytes at a ratio of 5:1
induced substantial activity, and this increased when lymphocyte
numbers were doubled. Moreover, ß2GPI-induced monocyte TF was
abrogated by a neutralizing anti-class II MHC Ab, but not isotype
control, confirming a requirement for Ag presentation to
CD4+ T lymphocytes (Table IV).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The results suggest that some patients with the APS may have chronic
low-grade stimulation of ß2GPI-specific T lymphocytes by continuous
exposure to this plasma protein, leading to persistently high monocyte
TF expression resulting in a prothrombotic diathesis. This hypothesis
is consistent with reports of elevated levels of plasma TF
(13) and monocyte TF (14) in patients with
aPL Abs, particularly those patients with histories of thrombosis
(14). The mechanism by which autoreactive
CD4+ T cells induce TF in the APS is unclear, but
may involve IFN-. Type 1 (Th1) but not type 2 (Th2) Th lymphocytes
induce monocyte TF, which is mediated, in part, by release of IFN-
(17). Moreover, production of IFN- by activated
ß2GPI-specific T lymphocytes may also contribute to the high
frequency of pregnancy failure in APS patients. Normal pregnancy has
been considered a type 2 (Th2) phenomenon immunologically, with
prominent production of IL-4, IL-5, and IL-10 in fetoplacental tissues
(18). Type 1 (Th1) cytokines such as IFN- have
detrimental effects on fetal development and the ability to sustain
pregnancy (19), and production of IFN- and
related cytokines by ß2GPI-specific T lymphocytes could compromise
pregnancy in patients with the APS.
In addition to providing mechanistic insight into the pathogenesis of clinical events in the APS, the results of this study also raise the possibility that positive cellular immunity to ß2GPI, as measured by induction of monocyte TF in vitro, could be a useful indicator of risk of thrombosis or fetal loss in patients who have aPL Abs. Current literature suggests that the level of aPL Ab and the presence of the IgG isotype provide some predictive value for clinical events of the APS (5, 6), but these issues are not reliable predictors in individual patients. The detection of Abs with defined specificity for ß2GPI appears to be a more specific marker for clinical features of the APS than aCL Ab (4). Our finding that patients with features of the APS had higher mean IgG-anti-ß2GPI Ab levels than those without the APS but similar IgG or IgM aCL Ab levels supports this literature. If the absolute correlation between ß2GPI-specific induction of monocyte TF and clinical events of the APS reported in this study is confirmed in prospective studies, the use of this assay represents a significant advance in the clinical management of individual patients with this syndrome.
|
Footnotes |
|---|
2 Address correspondence and reprint request to Dr. H. Patrick McNeil, Inflammation Research Unit, School of Pathology, University of New South Wales, Sydney, NSW 2052, Australia.
3 Abbreviations used in this paper: aPL, antiphospholipid; APS, aPL syndrome; ß2GPI, ß2-glycoprotein-I; PCA, procoagulant activity; TF, tissue factor; aCL, anticardiolipin; LA, lupus anticoagulant; PAPS, primary APS; SLE, systemic lupus erythematosus; TF-SI, TF stimulation index; TF-R, TF responders; TF-NR, TF nonresponders.
Received for publication November 10, 1999. Accepted for publication May 23, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-thrombin, factor Xa, kallikreins, and urokinase. J Biochem. 82:1495.This article has been cited by other articles:
|
|
 |
|
  M. van Lummel, M. T. T. Pennings, R. H. W. M. Derksen, R. T. Urbanus, B. C. H. Lutters, N. Kaldenhoven, and P. G. de Groot The Binding Site in {beta}2-Glycoprotein I for ApoER2' on Platelets Is Located in Domain V J. Biol. Chem., November 4, 2005; 280(44): 36729 - 36736. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. George, N. Yacov, E. Breitbart, L. Bangio, A. Shaish, B. Gilburd, Y. Shoenfeld, and D. Harats Suppression of early atherosclerosis in LDL-receptor deficient mice by oral tolerance with {beta}2-glycoprotein I Cardiovasc Res, June 1, 2004; 62(3): 603 - 609. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. E. Warkentin, W. C. Aird, and J. H. Rand Platelet-Endothelial Interactions: Sepsis, HIT, and Antiphospholipid Syndrome Hematology, January 1, 2003; 2003(1): 497 - 519. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. H. Rand Molecular Pathogenesis of the Antiphospholipid Syndrome Circ. Res., January 11, 2002; 90(1): 29 - 37. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Arai, K. Yoshida, J. Kaburaki, H. Inoko, Y. Ikeda, Y. Kawakami, and M. Kuwana Autoreactive CD4+ T-cell clones to {beta}2-glycoprotein I in patients with antiphospholipid syndrome: preferential recognition of the major phospholipid-binding site Blood, September 15, 2001; 98(6): 1889 - 1896. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
