- Email: [email protected]
Anti-vascular endothelial cell antibodies in patients with IgA nephropathy: Frequency and clinical significance
CLINICAL
IMMUNOLOGY
AND
IMMUNOPATHOLOGY
49, 450.462 f 1988)
Anti-vascular Endothelial Cell Antibodies in Patients with IgA Nephropathy: Frequency and Clinical Significance’ HUI KIM YAP,’ REBECCAS. SAKAI, LUCY BAHN, VALERIE RAPPAPORT, KENG THYE Woo, VATHSALA ANANTHURAMAN, CHENG HONG LIM, GILBERT S. C. CHIANG, STANLEY C. JORDAN Division of Pediatric Nephrology, Department of Pediatrich, Cedars-Sinai-UCLA Medical Center, Los Angeles, California; Department of Pediatrics, National University of Singapore: and Departments of Renal Medicine and Pathology, Singapore General Hospital, Singapore
This study examined the frequency of anti-vascular endothelial cell (VEC) antibodies (Ab) in 72 patients with IgA nephropathy (IgAN), and their possible relationship to clinical and histological parameters of the disease. An enzyme immunoassay was developed to measure the binding of sera to endothelial cells grown to a confluent monolayer. Thirty-two percent of IgAN patients had serum anti-VEC activity as compared to 4% of controls (P = 0.004) and 9% of patients with other primary glomerulonephritis (P = 0.017). This was shown to be due to anti-HLA class I Abs in 6 of the 23 IgAN patients. and in the I control positive for anti-VEC activity. Hence 17 IgAN patients had anti-VEC Abs, predominantly of the IgA subclass. Stimulation of the endothelial cells with interferon-y and interleukin 1 did not increase the binding of these Abs. There was no correlation with circulating immune complex (IC) levels, and removal of 1Cs in positive sera by ultracentrifugation did not decrease anti-VEC binding. Significant correlations were found between anti-VEC Abs and proteinuria greater than I g/day (P = 0.044). as we11as IgA anti-VEC Abs and C3 or IgA deposition in renal arterioles (P = 0.048). These IgA Abs may be an important marker of pathogenetic activity in IgAN. ‘I- 198x Academic Press. Inc.
INTRODUCTION
Antibodies directed against endothelial cell antigens have been implicated in the mediation of vasculitic disorders such as systemic lupus erythematosus (1). Similarly, in renal allograft rejection where vascular lesions play a prominent role, anti-vascular endothelial cell (VEC) antibodies (Ab) were demonstrated both in sera and eluates of the grafts, indicating a possible pathogenetic role (2, 3). Immunohistologic studies have also shown deposition of immunoglobulin along the endothelium of the peritubular vessels (4). IgA nephropathy is a glomerulopathy characterized by the deposition of IgA in the mesangium of the glomeruli. In addition, IgA and C3 deposits have been described in the walls of the renal arterioles (5). A relationship to anaphylactoid ’ This work was presented in part at the 20th annual meeting of the American Society of Nephrology, Washington, D.C., 1987. This work was supported by Grant DK-38831 from the NIADDK of the National Institutes of Health, The Rebecca Sakai Memorial Fund. and a Research Career Development Award (KO4-DK-01783) to Dr. S. C. Jordan. ’ To whom correspondence should be addressed at the Department of Pediatrics, National University of Singapore, Kent Ridge Road, Singapore 0511. 450 0090-1229/88 $1.50 Copyright 8 1988 by Academic Press. Inc. AU rights of reproduction in any form reserved
ANTI-VASCULAR
ENDOTHELIAL
CELL
451
ANTIBODIES
put-pm-a, a vasculitic disorder with IgA deposits in skin and glomeruli, has also been proposed (6). The aim of this study was to investigate the nature and possible role of antibodies directed against endothelial cell antigens in the genesis and progression of IgA nephropathy. MATERIALS
Patients
AND METHODS
and Controls
Seventy-two patients with IgA nephropathy (mean age 30 2 8 years, range 15 to 57 years), on follow-up at the Nephrology Clinic, Singapore General Hospital, were studied. The diagnosis was based on renal biopsy findings of predominant IgA mesangial deposits on immunofluorescence (7). In addition, the patients did not have any clinical evidence of liver disease or systemic disorders such as systemic lupus erythematosus, mixed cryoglobulinemia, or Henoch-Schonlein purpura. A control group consisting of 27 healthy adult volunteers was studied to determine the 95th percentile confidence limit for the anti-VEC Ab assays. Thirty-three patients with other forms of primary glomerulonephritis formed a third group as patient controls (Table 1). Sera were obtained from these patients and stored at - 70°C until tested. Serum creatinine concentrations and quantitative urinary protein excretion were measured in all patients. Isolation
of Human
Umbilical
Venous Endothelial
Cells
Endothelial cells were isolated from human umbilical cord veins as previously described (8, 9). Briefly, the endothelial cells were harvested by collagenase digestion of normal term human umbilical cord veins from single or pooled (two to six) donors. The cells were cultured in Medium 199 supplemented with 20% heat-inactivated fetal bovine serum, 125 U/ml penicillin, 125 pg/ml streptomycin, 2 mM L-glutamine, and 20 to 100 pg/ml of endothelial cell growth supplement (Sigma Chemical Co., St. Louis, MO). The primary cultures were grown from Day 0, on 96-well microtiter plates (Falcon, Becton-Dickinson, Oxnard, CA), until they formed a confluent monolayer with a “cobblestone” appearance. The cells were positive for Factor VIII antigen (Ag), using a mouse monoclonal Ab TABLE
1
HISTOLOGICALTYPES IN PATIENTCONTROLGROUP GN PATIENTS) Histological
type
Focal glomerulosclerosis Membranous nephropathy Mesangial proliferative GN Minimal lesion Membran+proliferative GN Crescenteric GN Diffuse sclerosing GN
Number of patients 18 6 4 2 1 1 1
452
YAP
ET
AL.
against anti-Factor VIIIR Ag (Cappel, Organon Teknika Co., West Chester, PA). The trays were ready for testing usually between Days 4 to 10. Treatment
of HUVE
Cells with Interferon-y
(IFN-y)
and Interleukin
I (IL-l)
The mediator, recombinant human IFN-7 (Amgen, Thousand Oaks, CA), was added at a concentration of 200 U/ml into the culture medium on Day I. The stimulated cells were tested after a minimum of 4 days culture, for HLA-DR expression, using a mouse monoclonal Ab to HLA-DR Ag (Cappel. Organon Teknika Co.) and for an increase in HLA class I expression using another monoclonal Ab against HLA-A,B,C (W6/32, Bioproducts for Science Inc., Indianapolis, IN). Another experiment was done with recombinant human IL-1 (Genzyme, Boston, MA) as mediator at a concentration of 5 U/ml. The HUVE cells were pretreated for 4.5 hr prior to performing the assay for anti-VEC activity. Enzyme Immunoassay
(EIA) for Anti-VEC
Activity
in Sera
The HUVE cells were fixed with 0.2% gluteraldehyde when a confluent monolayer was reached. The trays were then quenched with PBS-Tween 0.1% for an hour at room temperature. The test sera were diluted 1: 10 for the IgA specific assay, and 1:50 for the IgG specific anti-VEC assay. Fifty microliters of the diluted test sera were then incubated in quadruplicates with the endothelial cells for 2 hr at 37°C. The trays were washed three times with PBS-Tween 0.1% and were subsequently incubated at 37°C with 50 ~1 of peroxidase-conjugated affinitypurified sheep anti-human IgA Ab (Cappel. Organon Teknika Co.) or peroxidaseconjugated protein A (Sigma Chemical Co.) for the IgA specific and IgG specific assays, respectively. After washing four times with PBS-Tween O.l%, ophenylenediamine (OPD) (New England Nuclear, Boston, MA) in a 0.2 M phosphate buffer containing 0.1 M citrate, pH 6.3, was added. The reaction was stopped with 4.5 A4 sulfuric acid, and the color change was measured spectrophotometrically at 492 nm. inhibition Assays for the Detection of Anti-HLA AD To further characterize the nature of the anti-VEC activity, the presence of Abs directed against the HLA class I determinants on the endothelial cells was detected using an assay based on the ability of the test sera to inhibit the binding of mouse monoclonal anti-HLA-A,B,C Ab (W6132). Twenty-five microliters of the test sera, diluted 1:2, was added to 25 l~,l of the monoclonal Ab, W6132, diluted 1:50. The EIA was then performed with an initial incubation of 1 hr. Peroxidaselabeled affinity-puritied sheep anti-mouse IgG (Cappel, Organon Teknika Co.) was used as the second Ab. In addition, this assay was correlated with another method of detecting the anti-HLA class I Abs, where the sera was tested for its ability to competitively inhibit the binding of the monoclonal Ab, W6132, to class I determinants on platelets. Briefly, pooled platelets (Scantibodies Laboratory, Inc., Lakeside, CA) were coated onto 96-well microtiter plates (Titertek, Flow Laboratories, Inc.. McLean VA) by a previously described method (10). Sera was then added 1: I to
ANTI-VASCULAR
ENDOTHELIAL
CELL
ANTIBODIES
453
the monoclonal Ab, W6/32, and incubated for 45 min with the platelets. After three washes with PBS-Tween O.l%, peroxidase-labeled sheep anti-mouse IgG Ab was added, and the final color reaction was read at 492 nm on the spectrophotometer. Assays for Circulating
Immune
Complexes
(CZC)
The Raji cell and F(ab’), anti-C3 EIAs were used to measure the serum levels of IgA CICs (expressed as the ratio T/U where T = optical density (OD) of test sera, U = OD of upper 95th percentile confidence limit of normal) and IgG CICs, according to previously described methods (11, 12). In addition, the Clq solidphase EIA was also used to quantitate the IgG CICs, which were expressed as aggregated human IgG (AHG) equivalents. For the Clq solid-phase EIA, 96-well microtiter plates were coated with Clq (Calbiochem, Behring Diagnostics, La Jolla, CA) at 2 &ml in PBS, pH 7.2, stored at - 70°C until used. Test sera was diluted 1:3 in 0.2 M disodium EDTA and incubated for 30 min at 37°C. This was subsequently diluted 1:21 with PBS-Tween 0.1%. An AHG standard curve was constructed using serial dilutions of AHG in PBS-Tween 0.1% to which normal human sera inactivated with sodium EDTA was added. After quenching the tray for 15 min with PBS-Tween O.l%, 100 ~1 of the test sera and AHG standards were added in quadruplicates, and incubated for 1 hr at 37°C and 1 hr at 4°C. Following three washes with PBS-Tween O.l%, peroxidase-labeled protein A was added and incubated for another hour at 37°C. The subsequent color reaction was analyzed at 492 nm. Removal of Zmmune Complexes by Ultracentrifugation at 100,OOOg Sera positive for anti-VEC Ab was subjected to ultracentrifugation at 100,OOOg in an Airfuge (Beckman Instruments, Inc., Fullerton, CA) for 1 hr in order to remove immune complexes. The anti-VEC EIA was then repeated. Renal Histopathology Renal biopsy specimens were fixed in alcoholic Bouin’s. Sections were cut 2-pm thick, and were stained with hematoxylin and eosin, periodic acid-Schiff, and periodic acid-methenamine silver stains. Specimens for immunofluorescence were cut 4- to 5-pm thick and were tested with fluorescein-conjugated anti-sera specific for human IgG, IgA, IgM, IgD, IgE, C3, C4, Clq, and fibrinogen (Behringwerke AG, Marburg, GFR). The percentage of sclerosed glomeruli, the presence of epithelial crescents, and the deposition of immunoglobulins and complement components in the renal arteriolar walls were correlated with the presence or absence of anti-VEC Abs. Statistical Analysis To standardize the anti-VEC represent the distribution of the every EIA tray. A standardized four standard controls (M> on score = OD of test serum -
EIA, four controls were selected to statistically normal controls, and were subsequently tested on score was derived based on the mean OD of the each tray, and was expressed as: Anti-VEC Ab M. The upper 95th percentile confidence limit
454
YAP
ET AL.
(one-tailed test) of the control population was calculated, and scores of greater than 0.06 for IgA specific anti-VEC Ab, and greater than 0.05 for IgG specific anti-VEC Ab, were considered abnormal. The Student’s t test was used to examine the differences between the groups. The two-tailed Fisher’s exact test was used to analyze the correlations between the presence of anti-VEC Abs and clinical and histopathological parameters. RESULTS
Frequency
of Anti-VEC
Antibodies
Sera from 32% of IgAN patients were positive for anti-VEC activity as compared to 4% of controls (P = 0.004) and 9% of GN patients (P = 0.017). Of the IgAN patients, 21% had IgA specific anti-VEC activity as compared to 4% and 6% in the control group and GN group, respectively (P = ns). In addition, 15% of IgAN patients were positive for IgG specific anti-VEC activity as compared to none in the control group (P = 0.048) and 9% in GN patients. Three IgAN patients had both IgA and IgG specific anti-VEC Abs. The levels of anti-VEC activity in the three groups, expressed as the standardized score, are shown in Fig. I (IgA specific) and Fig. 2 (IgG specific). Characterization
of the Anti-VEC
Activity
An assay based on the ability of the patients’ sera to inhibit the binding of mouse monoclonal anti-HLA-A,B,C Ab to the HLA class I antigen on the endothelial cell was used to distinguish anti-HLA activity in the sera from “true” anti-VEC Ab activity. An inhibitory activity of greater than 75% was defined as positive for anti-HLA activity (95th percentile confidence limit for 18 normals studied). In addition, 3 patients on hemodialysis who were positive for anti-HLA .
..
-0.m
. . i
: . ........ . . .. ........ ... . . . . .. .. ... ... ... ... ... ... .. l *::::** iI
1 CONTROLS
IGAN
PATENTS
‘.‘.‘.‘.’ .. .. .. ... ... :. l :.
GN
.
.
PATIENTS
1. Distribution of IgA specific anti-VEC activity in IgAN patients, GN patients, and controls. Anti-VEC activity is expressed as a standard score. Dashed line represents the upper 95th percentile confidence limit of the normal population. FIG.
ANTI-VASCULAR
ENDOTHELIAL
CELL
455
ANTIBODIES
. 0 50
0.40
. s80: B 0.30 is; . g/p< 0.20 59 :. e“u) 5 0.10 . -----~-----~-0. . .IaJ:w:I I. .**y. i ----v 0.0 - .%z..3.~. :.:.:.J -0.10 - *-I” .*.*.* A .> ..B . :iY I -0.20 F
CONTROLS
IGAN PATIENTS
GN
PATIENTS
FIG. 2. Distribution of IgG specific anti-VEC activity in IgAN patients, GN patients, and controls. Dashed line represents the upper 95th percentile confidence limit of the normal population.
class I Abs on a lymphocyte microcytotoxicity assay had 78, 92, and 95% inhibitory activity. This assay was shown to correlate significantly with a platelet inhibition assay for the detection of anti-HLA class I Ab (r = 0.82, 0.000001) (Fig. 3). With these methods, 4 of 15 IgAN patients (27%) had IgA specific anti-VEC activity defined as Ab to HLA class I determinants (Table 2). In addition, 4 of 11 IgAN patients (36%) had IgG specific anti-HLA class I Abs (Table 3). The positive IgA anti-VEC activity in the normal control was also due to anti-HLA class I activity. None of the 3 GN patients positive for anti-VEC activity were positive on the anti-HLA class 1 assays. Figure 4 shows the anti-HLA class I inhibitory activity of sera from subjects positive on the anti-VEC EIA.
P<
loo,
”
I
0
i0
40
$0
% INHIBITION OF MO. AB W6/32 TO PLATELET MA - CLASS
ii0
lb0
BINDING 1 AG
FIG. 3. Correlation between the percentage of inhibition of mouse monoclonal antibody (MoAb) W6/32 binding to platelet (X axis) and VEC (r axis) HLA class I antigen. The subjects included in this correlation study were negative controls, positive controls, and IgA patients positive for anti-VEC activity.
456
YAP
TABLE OF NATURE
CHARACTERIZATION
(IgA
IgAN patient anti-VEC EIA
ET AL.. 2 OF IgA ANTI-VEC
ACTIVITY IgA CIC
+ )
Anti-HLA
Ab
Anti-C3
EIA
Raji
EIA
-t
I 2 3 4 5 6
-
pi
+
-.
HLA class 11 reactivity was excluded in these patients as the endothelial cells were unstimulated and therefore did not bind mouse monoclonal anti-HLA DR Ab. Stimulation
of Endothelial
Cells with IFN-7
and IL-1
Stimulation of the endothelial cells with IFN-)I resulted in HLA class 11 (DR) expression, as well as an increase in HLA class I expression (Fig. 5). However, there was no significant difference in the anti-VEC activity of the sera of IgAN patients when tested against unstimulated and stimulated cells. Similarly, IL-1 stimulation of endothelial cells did not result in an increase in anti-VEC binding to the cells. TABLE 3 OF NATURE OF IgG ANTI-VEC
CHARACTERIZATION IgAN patients (IgG Anti-VEC EIA
ACTIVITY IgG CIC
+ )
Anti-HLA
Ab
Clq
EIA
I
..-
--
2 3 4 5 6 7 8 9 IO
+ + + .+ -
-t t
-
11
-
-t
Raji EIA
Anti-C3
EIA
-
+
+
L
f
i
ANTI-VASCULAR
ENDOTHELIAL
CELL ANTIBODIES
457
.
I 01
. I
CONTROL
I
KiAN PATENTS
I
HD PATENTS POSITNE ON LYMPHOCYTE MlcRO~~m~OXtcrrY
FIG. 4. Anti-HLA class 1 inhibitory activity of IgAN patients positive for anti-VEC activity. Inhibition >75% was defined as positive (based on 18 normal controls shown).
Role of Immune Complexes in Endothelial Cell Binding
Forty percent of IgAN patients were positive for IgA CIC as measured by the F(ab’), anti-C3 EIA and Raji cell EIA. However, no significant correlation was seen between the presence of CIC and IgA anti-VEC activity (Table 3). Similarly, 060050a ew 040g z 0 ?b 0.30s g P 0.200.10-
o.oL
t3.N PATOICTS
FIG. 5. Anti-VEC activity of IgAN positive sera on endothelial cells stimulated with IFN-y. A denotes binding to unstimulated cells; 0 denotes binding to stimulated cells. (a) Increase in binding of mouse anti-HLA-A,B,C Ab on stimulated endothelial cells. (b) Increase in binding of mouse antiHLA-DR Ab on stimulated endothelial cells.
458
YAP
ET AL.
although 60% of IgAN patients were positive for IgG CIC, there was also no correlation with IgG anti-VEC activity (Table 4). To further exclude the possibility of immune complexes binding to the endothelial cells, immune complex positive sera were subjected to ultracentrifugation at 100,OOOg for 1 hr. All anti-VEC positive sera retained positivity after removal of immune complexes by centrifugation. Correlation
with Clinical
and Renal Biopsy Features
Quantitative data on the degree of proteinuria were available in 68 IgA patients. Thirteen of 16 IgA patients with anti-VEC Ab had significant proteinuria greater than 1 g/day (Table 5). Using the Fisher’s exact test, a significant correlation was demonstrated between the presence of anti-VEC Ab and proteinuria or greater than 1 g/day (P = 0.048) (Table 5). There was no correlation with serum creatinine greater than 1.5 mg/dl, the presence of more than 30% sclerosed glomeruli or the presence of crescents on renal biopsy. However, there was a positive correlation between the presence of IgA specific anti-VEC Ab in the sera with the presence of either C3 or IgA deposition in the walls of the renal arterioles, as shown on immunofluorescent examination of the renal biopsies (P = 0.044). DISCUSSION
In recent years, the activated endothelial cell has been increasingly recognized as an important participant in the genesis of immune-mediated vascular injury. Cytokines, such as interleukin 1 and tissue necrosis factor (TNF), which are produced by macrophages, stimulate cultured human umbilical venous endothe-
DEGREE
_.-
IN IgA
IgA anti-VEC Ab (OD score)
IgA patients 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
OF PROTEINURLA
~~~~~
0.08 0.15 0.34 0.07 0.15 0.08 0.08 0.08 0.08 0.09 0.03 -0.02 0.00 0.00 0.06 0.04
TABLE 4 PATIENTS
POSITIVE IgG
FOR ANTI-VEC
anti-VEC COD score) -- 0.07 - 0.09 0.02 -0.03 0.52 - 0.05 0.01 - 0.02 -- 0.03 0.03 0.22 0.08 0.10 0.32 0.06 0.08
ABS
Ab
Proteinuria (g/day)
_ _..._~~~
~~~c
3.04 0.24 1 .FO 1.08 2.25 1.56 2.00 1.30 I .20 0.93 1.19 0.32 I .20 1.20 2.25 1.84
ANTI-VASCULAR
CORRELATION
BETWEEN
IgA patients Anti-VEC Anti-VEC
Ab positive Ah negative
PRESENCE
ENDOTHELIAL
CELL
TABLE OF ANTI-VEC
Proteinuria
5 Ae AND SIGNIFICANT
> 1 g/day 13 26
459
ANTIBODIES
PROTEINURIA
>
1 g/day
Proteinuria < 1 g/day 3 26 ___~
lial cells to express tissue factor-like procoagulant activity (13, 14), leukocyte adhesion properties (15, 16), and unique activation antigens (17-19). The role of anti-VEC Abs in the pathogenesis of immune vascular disorders has been the subject of much debate. Such Abs have been described in the sera of patients with rheumatoid arthritis, scleroderma, systemic lupus erythematosus (SLE), dermatoyositis, polymyalgia rheumatica, and giant cell arteritis, as well as in about 14% of normal controls (1, 20, 21). Cines et al. (1) have shown that binding to endothelial cells of IgG and immune complexes found in the sera of patients with SLE resulted in the production of procoagulant tissue factor (22). In addition, the sera of patients with Kawasaki’s disease, a diffuse vasculitic disease common in children, contain Abs which cause lysis of activated endothelial cells in vitro (23, 24). Although endothehal activation antigens have been demonstrated in cultured glomerular cells (25), anti-VEC Abs have yet to be described in primary human glomerulonephritis. In our patients with IgAN, sera positive for anti-VEC activity was found in 32%, as compared to 4% in controls, and 9% in other forms of primary GN. The contribution of immune complexes to this activity was probably insignificant, as there was no correlation between anti-VEC activity and CIC levels in the patients. In addition, ultracentrifugation of the sera at 100,OOOg to remove immune complexes did not result in a decrease in anti-VEC binding. These Abs were predominantly of the IgA class, with three patients having both IgA and IgG Abs. Some of these Abs were directed against HLA class I antigens, as demonstrated by the significant inhibition of binding of the mouse monoclonal Ab (W6/32) to the framework region of the HLA class I antigen expressed on endothelial cells and platelets. The majority, however, were reactive against antigens present on unstimulated endothelial cells, as prior stimulation of the cells with IFN-?, or IL-l did not result in an increase in anti-VEC activity of the sera. It is conceivable that the endothelial cell in the human glomerulus is actively involved in the immune processes leading to glomerular injury (26). In a recently described model of experimental glomerulonephritis (27), an injection of heterologous anti-endothelial antibodies directed against angiotensin-converting enzyme in rabbits resulted in an increase in vascular permeability leading to “planting” of the heterologous IgG on the subepithelial part of the glomerular capillary wall, remniscent of classical Heymann nephritis (2&30). In our patients, a correlation was seen between the presence of IgA anti-VEC Abs with C3 and immunoglobulin deposition in the renal arteriolar walls. Additionally, a significant correlation was present between these Abs and the presence of proteinuria greater than 1 g per day in the IgAN patients. The presence of heavy
460
YAP ET AL.
proteinuria has been found to be an indicator of a more severe disease in our patients with IgAN (7, 31). Hence it is possible that binding of anti-VEC Abs to the glomerular endothelial cells could result in glomerular capillary wall damage via the production of mediators such as platelet-activating factor (32). and contribute to the progression of the renal lesion in these patients. On the other hand, 68% of the IgA patients did not have any anti-VEC activity, and only half of these patients had heavy proteinuria as compared to 81% in the group with anti-VEC Abs. Other mechanisms leading to progressive renal injury may be operative in these patients. In many patients with IgA nephropathy, the genesis of glomerular sclerotic lesions appears to be noninflammatory, as macrophages or T cells cannot be demonstrated in the glomeruli (33). Here, local hemodynamic factors have been postulated to be important in the pathogenesis of progressive glomerular sclerosis (34). In conclusion, this study demonstrated the presence of anti-VEC Abs in some patients with IgAN. The predominance of Abs of the IgA class may be a manifestation of the increase in IgA-secreting lymphocytes described in this disorder (35-37). The exact pathogenetic significance of these Abs is unclear. It would be exciting to postulate a role for these Abs in contributing to the progression of glomerular immune injury. On the other hand, it is also conceivable that damage to the vascular endothelium by a more primary process with exposure of endothelial cell antigens could have stimulated the production of these Abs. Further studies are required to characterize the nature of the endothelial antigen to which these Abs are reactive and their importance in mediating glomerular injury in &AN. ACKNOWLEDGMENTS The authors gratefully acknowledge the statistical assistance provided by Ms. F. Lennie Wong, biostatistician, Department of Medical Genetics. Cedars-Sinai Medical Center. We also thank Dr. C. Hobel, Department of Obstetrics/Gynecology. Cedars-Sinai Medical Center, for assistance in collection of umbilical cords; Mr. Y. K. Lau and staff of the Renal Laboratory, SGH, for their help in sera collection; and Dr. J. Prehn and Dr. D. Yamaguchi for their invaluable advice and encouragement.
REFERENCES 1. Cines, D. B., Lyss, A. P., Reeber, M.. Bina. M., and Dehoratius, R. J.. Presence of complementfixing anti-endothelial cell antibodies in systemic lupus erythematosus. J. Clin. Invest. 73, 61 I625, 1984. 2. Paul, L. C.. Claas, F. H. J., Van Es, L. A., Kalff, M. W.. and De Graeff. J.. Rejection of a renal allograft associated with pretransplantation antibodies directed against donor antigens on endothelium and monocytes. N. En@. J. Med. 300, 1258-1260, 1979. 3. Paul. L. C., Van Es, L. A., Kalff, M. W., and De Graeff, J., Intrarenal distribution of endothelial antigens recognized by antibodies from renal allograft recipients. Transplant. Proc. 11, 427-430, 1979. 4. Paul. L. C., Van Es, L. A., Van Rood, J. J., Van Leeuwen, A., De La Reviere, 6. B., and De Graeff, J., Antibodies directed against antigens on the endothelium of peritubular capillaries in patients with rejecting renal ahografts. Trun.~plantarion 27, 175-179, 1979. 5. Levy, M., Gonzalez-Burchard, G., Broyer, M., Dommergues. J. P.. Foulard, M., Sorez, J. P.. and Habib, R., Berger’s disease in children. Medicine 64, 157-180, 1985. 6. Nakamoto, Y., Asano, Y., Dohi. K., Fujioka, M., Iida, H., Kida. H.. Kibe. Y.. Hattori. N., and
ANTI-VASCULAR
7. 8. 9. 10. 11. 12. 13. 14.
15. 16. 17.
18. 19. 20. 21. 22. 23. 24.
25.
ENDOTHELIAL
CELL ANTIBODIES
461
Takeuchi, J., Primary IgA glomerulonephritis and Schonlein-Henoch purpura nephritis: Clinicopathological and immuno-histological characteristics. Q. J. Med. 47, 495-516, 1978. Woo, K. T., Edmondson, R. P. S., Wu, A. Y. T., Chiang, G. S. C., Pwee, H. S., and Lim, C. H., The natural history of IgA nephritis in Singapore. C&n. Nephrol. 25, 15-21, 1986. Jaffe, E. A., Nachman, R. L., Becker, C. G., and Minick, C. R., Culture of human endothelial cells derived from umbilical veins. J. Clin. Invest, 52, 2745-2756, 1973. Gimbrone, M. A., Jr., Culture of vascular endothelium. Prog. Hemostasis Throm. 3, l-28, 1976. Barkley, S. C., Sakai, R. S., Ettenger, R. B., Fine, R. N., and Jordan, S. C., Determination of antiidiotypic antibodies to anti-HLA IgG following blood transfusions. Transplantation 44, 30-34, 1987. Jordan, S. C., Gautier, E., Sakai, R., and Bahn, L., Quantitation of circulating immune complexes in human serum by the Raji cell and F(ab’), anti-C3 micro enzyme immunoassays. J. Immunol. Methods 83, 363-370, 1985. Yap, H. K., Sakai, R. S., Woo, K. T., Lim, C. H., and Jordan, S. C., Detection of bovine serum albumin in the circulating IgA immune complexes of patients with IgA nephropathy. Clin. Zmmunol. Immunopathol. 43, 395402, 1987. Bevilacqua. M. P., Pober, J. S., Wheeler, M. E., Cotran, R. S., and Gimbrone, M. A., Jr., Interleukin-1 activation of vascular endothelium: Effects on procoagulant activity and leukocyte adhesion. Amer. J. Pathol. 121, 393-403, 1985. Bevilacqua, M. P., Pober, J. S., Majeau, G. R., Fiers, W., Cotran, R. S., and Gimbrone, M. A., Jr., Recombinant tumor necrosis factor induces procoagulant activity in cultured human vascular endothelium: Characterization and comparison with the actions of interleukin 1. Proc. Nat/. Acad. Sci. USA 83, 45334537, 1986. Gamble, J. R., Harlan, J. M., Klebanoff, S. J., and Vadas, M. A., Stimulation of the adherence of neutrophils to umbilical vein endothelium by human recombinant tumor necrosis factor. Proc. Natl. Acad. Sci. USA 82, 8667-8671. 1985. Schleimer, R. P., and Rutledge, B. K., Cultured human vascular endothelial ceHs acquire adhesiveness for neutrophils after stimulation with interleukin I, endotoxin, and tumor-promoting phorbol diesters. J. lmmunol. 136, 649-654, 1986. Pober, J. S., Bevilacqua, M. P., Mendrick. D. L., Lapherre, L. A., Fiers, W., and Gimbrone, M. A., Jr., Two distinct monokines. interleukin 1 and tumor necrosis factor, each independently induce biosynthesis and transient expression of the same antigen on the surface of cultured human vascular endothelial cells. J. Immunol. 136, 1680-1687, 1986. Cotran, R. S., Gimbrone, M. A.. Jr., Bevilacqua. M. P.. Mendrick. D. L.. and Pober, J. S., Induction and detection of a human endothelial activation antigen in vivo. J. Exp. Med. 164, 661-666, 1986. Messadi, D. V., Pober, J. S., Fiers, W., Gimbrone, M. A., Jr., and Murphy, G. F., Induction of an activation antigen on postcapillary venular endothelium in human skin organ culture. J. Immunol. 139, 1557-1562, 1987. Lindqvist, K. I., and Osterland, C. K., Human antibodies to vascular endothelium. Clin. Exp. Immunol. 9, 753-760, 1971. Tan, E. M., and Pearson, C. M., Rheumatic disease sera reactive with capillaries in the mouse kidney. Arthritis Rheum. 15, 23-28, 1972. Tannenbaum, S. H., Finko, R., and Cines, D. B., Antibody and immune complexes induce tissue factor production by human endothelial cells. J. Immunol. 137, 1532-1537, 1986. Leung, D. Y. M., Collins, T., Lapierre, L. A., Geha, R. S., and Pober, J. S.. Immunoglobulin M antibodies present in the acute phase of Kawasaki syndrome lyse cultured vascular endothelial cells stimulated by gamma interferon. J. C/in. Invest. 77, 1428-1435, 1986. Leung, D. Y. M., Geha, R. S., Newburger, J. W., Bums, J. C., Fiers. W., Lapierre, L. A., and Pober. J. S., Two monokines, interleukin 1 and tumor necrosis factor, render cultured vascular endothelial cells susceptible to lysis by antibodies circulating during Kawasaki syndrome. J. Exp. Med. 164, 1958-1972, 1986. Hancock, W. W., and Cotran, R. S., Induction of activation antigens on human glomerular endothelium by interleukin 1 (IL-l), interferon gamma (IFN-g), and endotoxin (LPS) Kidney Int. 31, 322, 1987. [Abstract]
462
YAP ET AL.
26. Salant, D. J.. Nephrology forum: lmmunopathogenesis of crescentic glomerulonephritis and lung purpura. Kidney Znt. 32, 408-425, 1987. 27. Matsuo, S., Fukatsu, A., Taub, M. L., Caldwell, R. B., Brentjens. J. R., and Andres, G., Glomerulonephritis induced in the rabbit by antiendothelial antibodies. J. Clin. Invest. 79, 1798-1811. 1987. 28. Barabas. A. Z., and Lannigan, R., Induction of an autologous immune complex glomerulonephritis in the rat by intravenous injection of heterologous anti-rat kidney tubular antibody. I. Production of chronic and progressive immune complex glomerulonephritis. Brit. J. Exp. Pathol. 55, 47-55, 1974. 29. Couser, W. G., Steinmuller, D. R., Stilmant, M. M., Dalant, D., and Lowenstein, L. M.. Experimental glomerulonephritis in the isolated perfused rat kidney. J. Clin. Invest. 62, 1275-1287. 1978. 30. Jeraj, K.. Vernier, R. L., Sisson, S. P., and Michael, A. F., A new glomerular antigen in passive Heymann’s nephritis. Brit. 1. Exp. Pathol. 65, 485-198, 1984. 31. Woo, K. T., Edmondson, R. P. S., Yap. H. K.. Wu, A. Y. T., Chiang, G. C. S.. Lee, E. J. c‘. Pwee, H. S., and Lim, C. H., Effects of triple therapy on the progression of mesangial prohferative glomerulonephritis. C/in. Nephrol. 27, 56-64, 1987. 32. Camussi, G., Aghetta, M., Malavasi, F.. Tetta, C.. Piacibello, W., Sanavio, F.. and Bussolino. F.. The release of platelet-activating factor from human endothelial cells in culture. J. Zmmunol. 131, 2397-2403, 1983. 33. Woodroffe, A. J., Bannister, K. M., Bellon, M. S., Hale, G. M.. and Clarkson. A. R.. Do local hemodynamic alterations mediate injury in IgA nephropathy? Sem. Nephrol. 7, 341-343, 1987. 34. Brenner, B. M., Meyer, T. W.. and Hostetter. T. H., Dietary protein intake and the progressive nature of kidney disease: The role of hemodynamically mediated glomeruiar injury in the pathogenesis of progressive glomerular sclerosis in aging, renal ablation, and intrinsic renal disease, N. Engl. J. Med. 307, 652-659. 1982. 35. Nomoto. Y., Sakai, H., and Arimori, S., Increase of IgA-bearing lymphocytes in peripheral blood from patients with IgA nephropathy. Amer. J. Pathol. 71, 158-160. 1979. 36. Egido, J., Biasco, R., Sancho, J., Lozano, L., Sanchez-Crespo, M., and Hernando. L. 1Increased rate of polymeric IgA synthesis by circulating lymphoid cells in IgA mesangial glomerulonephritis. Clin. Exp. Zmmunol. 47, 309-316, 1982. 37. Schena. F. P., Mastrolitti, G., Fracasso, A. R.. Pastore. A., and Lad&. N., Increased immunoglobulin-secreting cells in the blood of patients with active idiopathic IgA nephropathy. C’lin. Nephrol. 26, 163-168, 1986. Received April 5, 1988: accepted with revision August 16, 1988
IMMUNOLOGY
AND
IMMUNOPATHOLOGY
49, 450.462 f 1988)
Anti-vascular Endothelial Cell Antibodies in Patients with IgA Nephropathy: Frequency and Clinical Significance’ HUI KIM YAP,’ REBECCAS. SAKAI, LUCY BAHN, VALERIE RAPPAPORT, KENG THYE Woo, VATHSALA ANANTHURAMAN, CHENG HONG LIM, GILBERT S. C. CHIANG, STANLEY C. JORDAN Division of Pediatric Nephrology, Department of Pediatrich, Cedars-Sinai-UCLA Medical Center, Los Angeles, California; Department of Pediatrics, National University of Singapore: and Departments of Renal Medicine and Pathology, Singapore General Hospital, Singapore
This study examined the frequency of anti-vascular endothelial cell (VEC) antibodies (Ab) in 72 patients with IgA nephropathy (IgAN), and their possible relationship to clinical and histological parameters of the disease. An enzyme immunoassay was developed to measure the binding of sera to endothelial cells grown to a confluent monolayer. Thirty-two percent of IgAN patients had serum anti-VEC activity as compared to 4% of controls (P = 0.004) and 9% of patients with other primary glomerulonephritis (P = 0.017). This was shown to be due to anti-HLA class I Abs in 6 of the 23 IgAN patients. and in the I control positive for anti-VEC activity. Hence 17 IgAN patients had anti-VEC Abs, predominantly of the IgA subclass. Stimulation of the endothelial cells with interferon-y and interleukin 1 did not increase the binding of these Abs. There was no correlation with circulating immune complex (IC) levels, and removal of 1Cs in positive sera by ultracentrifugation did not decrease anti-VEC binding. Significant correlations were found between anti-VEC Abs and proteinuria greater than I g/day (P = 0.044). as we11as IgA anti-VEC Abs and C3 or IgA deposition in renal arterioles (P = 0.048). These IgA Abs may be an important marker of pathogenetic activity in IgAN. ‘I- 198x Academic Press. Inc.
INTRODUCTION
Antibodies directed against endothelial cell antigens have been implicated in the mediation of vasculitic disorders such as systemic lupus erythematosus (1). Similarly, in renal allograft rejection where vascular lesions play a prominent role, anti-vascular endothelial cell (VEC) antibodies (Ab) were demonstrated both in sera and eluates of the grafts, indicating a possible pathogenetic role (2, 3). Immunohistologic studies have also shown deposition of immunoglobulin along the endothelium of the peritubular vessels (4). IgA nephropathy is a glomerulopathy characterized by the deposition of IgA in the mesangium of the glomeruli. In addition, IgA and C3 deposits have been described in the walls of the renal arterioles (5). A relationship to anaphylactoid ’ This work was presented in part at the 20th annual meeting of the American Society of Nephrology, Washington, D.C., 1987. This work was supported by Grant DK-38831 from the NIADDK of the National Institutes of Health, The Rebecca Sakai Memorial Fund. and a Research Career Development Award (KO4-DK-01783) to Dr. S. C. Jordan. ’ To whom correspondence should be addressed at the Department of Pediatrics, National University of Singapore, Kent Ridge Road, Singapore 0511. 450 0090-1229/88 $1.50 Copyright 8 1988 by Academic Press. Inc. AU rights of reproduction in any form reserved
ANTI-VASCULAR
ENDOTHELIAL
CELL
451
ANTIBODIES
put-pm-a, a vasculitic disorder with IgA deposits in skin and glomeruli, has also been proposed (6). The aim of this study was to investigate the nature and possible role of antibodies directed against endothelial cell antigens in the genesis and progression of IgA nephropathy. MATERIALS
Patients
AND METHODS
and Controls
Seventy-two patients with IgA nephropathy (mean age 30 2 8 years, range 15 to 57 years), on follow-up at the Nephrology Clinic, Singapore General Hospital, were studied. The diagnosis was based on renal biopsy findings of predominant IgA mesangial deposits on immunofluorescence (7). In addition, the patients did not have any clinical evidence of liver disease or systemic disorders such as systemic lupus erythematosus, mixed cryoglobulinemia, or Henoch-Schonlein purpura. A control group consisting of 27 healthy adult volunteers was studied to determine the 95th percentile confidence limit for the anti-VEC Ab assays. Thirty-three patients with other forms of primary glomerulonephritis formed a third group as patient controls (Table 1). Sera were obtained from these patients and stored at - 70°C until tested. Serum creatinine concentrations and quantitative urinary protein excretion were measured in all patients. Isolation
of Human
Umbilical
Venous Endothelial
Cells
Endothelial cells were isolated from human umbilical cord veins as previously described (8, 9). Briefly, the endothelial cells were harvested by collagenase digestion of normal term human umbilical cord veins from single or pooled (two to six) donors. The cells were cultured in Medium 199 supplemented with 20% heat-inactivated fetal bovine serum, 125 U/ml penicillin, 125 pg/ml streptomycin, 2 mM L-glutamine, and 20 to 100 pg/ml of endothelial cell growth supplement (Sigma Chemical Co., St. Louis, MO). The primary cultures were grown from Day 0, on 96-well microtiter plates (Falcon, Becton-Dickinson, Oxnard, CA), until they formed a confluent monolayer with a “cobblestone” appearance. The cells were positive for Factor VIII antigen (Ag), using a mouse monoclonal Ab TABLE
1
HISTOLOGICALTYPES IN PATIENTCONTROLGROUP GN PATIENTS) Histological
type
Focal glomerulosclerosis Membranous nephropathy Mesangial proliferative GN Minimal lesion Membran+proliferative GN Crescenteric GN Diffuse sclerosing GN
Number of patients 18 6 4 2 1 1 1
452
YAP
ET
AL.
against anti-Factor VIIIR Ag (Cappel, Organon Teknika Co., West Chester, PA). The trays were ready for testing usually between Days 4 to 10. Treatment
of HUVE
Cells with Interferon-y
(IFN-y)
and Interleukin
I (IL-l)
The mediator, recombinant human IFN-7 (Amgen, Thousand Oaks, CA), was added at a concentration of 200 U/ml into the culture medium on Day I. The stimulated cells were tested after a minimum of 4 days culture, for HLA-DR expression, using a mouse monoclonal Ab to HLA-DR Ag (Cappel. Organon Teknika Co.) and for an increase in HLA class I expression using another monoclonal Ab against HLA-A,B,C (W6/32, Bioproducts for Science Inc., Indianapolis, IN). Another experiment was done with recombinant human IL-1 (Genzyme, Boston, MA) as mediator at a concentration of 5 U/ml. The HUVE cells were pretreated for 4.5 hr prior to performing the assay for anti-VEC activity. Enzyme Immunoassay
(EIA) for Anti-VEC
Activity
in Sera
The HUVE cells were fixed with 0.2% gluteraldehyde when a confluent monolayer was reached. The trays were then quenched with PBS-Tween 0.1% for an hour at room temperature. The test sera were diluted 1: 10 for the IgA specific assay, and 1:50 for the IgG specific anti-VEC assay. Fifty microliters of the diluted test sera were then incubated in quadruplicates with the endothelial cells for 2 hr at 37°C. The trays were washed three times with PBS-Tween 0.1% and were subsequently incubated at 37°C with 50 ~1 of peroxidase-conjugated affinitypurified sheep anti-human IgA Ab (Cappel. Organon Teknika Co.) or peroxidaseconjugated protein A (Sigma Chemical Co.) for the IgA specific and IgG specific assays, respectively. After washing four times with PBS-Tween O.l%, ophenylenediamine (OPD) (New England Nuclear, Boston, MA) in a 0.2 M phosphate buffer containing 0.1 M citrate, pH 6.3, was added. The reaction was stopped with 4.5 A4 sulfuric acid, and the color change was measured spectrophotometrically at 492 nm. inhibition Assays for the Detection of Anti-HLA AD To further characterize the nature of the anti-VEC activity, the presence of Abs directed against the HLA class I determinants on the endothelial cells was detected using an assay based on the ability of the test sera to inhibit the binding of mouse monoclonal anti-HLA-A,B,C Ab (W6132). Twenty-five microliters of the test sera, diluted 1:2, was added to 25 l~,l of the monoclonal Ab, W6132, diluted 1:50. The EIA was then performed with an initial incubation of 1 hr. Peroxidaselabeled affinity-puritied sheep anti-mouse IgG (Cappel, Organon Teknika Co.) was used as the second Ab. In addition, this assay was correlated with another method of detecting the anti-HLA class I Abs, where the sera was tested for its ability to competitively inhibit the binding of the monoclonal Ab, W6132, to class I determinants on platelets. Briefly, pooled platelets (Scantibodies Laboratory, Inc., Lakeside, CA) were coated onto 96-well microtiter plates (Titertek, Flow Laboratories, Inc.. McLean VA) by a previously described method (10). Sera was then added 1: I to
ANTI-VASCULAR
ENDOTHELIAL
CELL
ANTIBODIES
453
the monoclonal Ab, W6/32, and incubated for 45 min with the platelets. After three washes with PBS-Tween O.l%, peroxidase-labeled sheep anti-mouse IgG Ab was added, and the final color reaction was read at 492 nm on the spectrophotometer. Assays for Circulating
Immune
Complexes
(CZC)
The Raji cell and F(ab’), anti-C3 EIAs were used to measure the serum levels of IgA CICs (expressed as the ratio T/U where T = optical density (OD) of test sera, U = OD of upper 95th percentile confidence limit of normal) and IgG CICs, according to previously described methods (11, 12). In addition, the Clq solidphase EIA was also used to quantitate the IgG CICs, which were expressed as aggregated human IgG (AHG) equivalents. For the Clq solid-phase EIA, 96-well microtiter plates were coated with Clq (Calbiochem, Behring Diagnostics, La Jolla, CA) at 2 &ml in PBS, pH 7.2, stored at - 70°C until used. Test sera was diluted 1:3 in 0.2 M disodium EDTA and incubated for 30 min at 37°C. This was subsequently diluted 1:21 with PBS-Tween 0.1%. An AHG standard curve was constructed using serial dilutions of AHG in PBS-Tween 0.1% to which normal human sera inactivated with sodium EDTA was added. After quenching the tray for 15 min with PBS-Tween O.l%, 100 ~1 of the test sera and AHG standards were added in quadruplicates, and incubated for 1 hr at 37°C and 1 hr at 4°C. Following three washes with PBS-Tween O.l%, peroxidase-labeled protein A was added and incubated for another hour at 37°C. The subsequent color reaction was analyzed at 492 nm. Removal of Zmmune Complexes by Ultracentrifugation at 100,OOOg Sera positive for anti-VEC Ab was subjected to ultracentrifugation at 100,OOOg in an Airfuge (Beckman Instruments, Inc., Fullerton, CA) for 1 hr in order to remove immune complexes. The anti-VEC EIA was then repeated. Renal Histopathology Renal biopsy specimens were fixed in alcoholic Bouin’s. Sections were cut 2-pm thick, and were stained with hematoxylin and eosin, periodic acid-Schiff, and periodic acid-methenamine silver stains. Specimens for immunofluorescence were cut 4- to 5-pm thick and were tested with fluorescein-conjugated anti-sera specific for human IgG, IgA, IgM, IgD, IgE, C3, C4, Clq, and fibrinogen (Behringwerke AG, Marburg, GFR). The percentage of sclerosed glomeruli, the presence of epithelial crescents, and the deposition of immunoglobulins and complement components in the renal arteriolar walls were correlated with the presence or absence of anti-VEC Abs. Statistical Analysis To standardize the anti-VEC represent the distribution of the every EIA tray. A standardized four standard controls (M> on score = OD of test serum -
EIA, four controls were selected to statistically normal controls, and were subsequently tested on score was derived based on the mean OD of the each tray, and was expressed as: Anti-VEC Ab M. The upper 95th percentile confidence limit
454
YAP
ET AL.
(one-tailed test) of the control population was calculated, and scores of greater than 0.06 for IgA specific anti-VEC Ab, and greater than 0.05 for IgG specific anti-VEC Ab, were considered abnormal. The Student’s t test was used to examine the differences between the groups. The two-tailed Fisher’s exact test was used to analyze the correlations between the presence of anti-VEC Abs and clinical and histopathological parameters. RESULTS
Frequency
of Anti-VEC
Antibodies
Sera from 32% of IgAN patients were positive for anti-VEC activity as compared to 4% of controls (P = 0.004) and 9% of GN patients (P = 0.017). Of the IgAN patients, 21% had IgA specific anti-VEC activity as compared to 4% and 6% in the control group and GN group, respectively (P = ns). In addition, 15% of IgAN patients were positive for IgG specific anti-VEC activity as compared to none in the control group (P = 0.048) and 9% in GN patients. Three IgAN patients had both IgA and IgG specific anti-VEC Abs. The levels of anti-VEC activity in the three groups, expressed as the standardized score, are shown in Fig. I (IgA specific) and Fig. 2 (IgG specific). Characterization
of the Anti-VEC
Activity
An assay based on the ability of the patients’ sera to inhibit the binding of mouse monoclonal anti-HLA-A,B,C Ab to the HLA class I antigen on the endothelial cell was used to distinguish anti-HLA activity in the sera from “true” anti-VEC Ab activity. An inhibitory activity of greater than 75% was defined as positive for anti-HLA activity (95th percentile confidence limit for 18 normals studied). In addition, 3 patients on hemodialysis who were positive for anti-HLA .
..
-0.m
. . i
: . ........ . . .. ........ ... . . . . .. .. ... ... ... ... ... ... .. l *::::** iI
1 CONTROLS
IGAN
PATENTS
‘.‘.‘.‘.’ .. .. .. ... ... :. l :.
GN
.
.
PATIENTS
1. Distribution of IgA specific anti-VEC activity in IgAN patients, GN patients, and controls. Anti-VEC activity is expressed as a standard score. Dashed line represents the upper 95th percentile confidence limit of the normal population. FIG.
ANTI-VASCULAR
ENDOTHELIAL
CELL
455
ANTIBODIES
. 0 50
0.40
. s80: B 0.30 is; . g/p< 0.20 59 :. e“u) 5 0.10 . -----~-----~-0. . .IaJ:w:I I. .**y. i ----v 0.0 - .%z..3.~. :.:.:.J -0.10 - *-I” .*.*.* A .> ..B . :iY I -0.20 F
CONTROLS
IGAN PATIENTS
GN
PATIENTS
FIG. 2. Distribution of IgG specific anti-VEC activity in IgAN patients, GN patients, and controls. Dashed line represents the upper 95th percentile confidence limit of the normal population.
class I Abs on a lymphocyte microcytotoxicity assay had 78, 92, and 95% inhibitory activity. This assay was shown to correlate significantly with a platelet inhibition assay for the detection of anti-HLA class I Ab (r = 0.82, 0.000001) (Fig. 3). With these methods, 4 of 15 IgAN patients (27%) had IgA specific anti-VEC activity defined as Ab to HLA class I determinants (Table 2). In addition, 4 of 11 IgAN patients (36%) had IgG specific anti-HLA class I Abs (Table 3). The positive IgA anti-VEC activity in the normal control was also due to anti-HLA class I activity. None of the 3 GN patients positive for anti-VEC activity were positive on the anti-HLA class 1 assays. Figure 4 shows the anti-HLA class I inhibitory activity of sera from subjects positive on the anti-VEC EIA.
P<
loo,
”
I
0
i0
40
$0
% INHIBITION OF MO. AB W6/32 TO PLATELET MA - CLASS
ii0
lb0
BINDING 1 AG
FIG. 3. Correlation between the percentage of inhibition of mouse monoclonal antibody (MoAb) W6/32 binding to platelet (X axis) and VEC (r axis) HLA class I antigen. The subjects included in this correlation study were negative controls, positive controls, and IgA patients positive for anti-VEC activity.
456
YAP
TABLE OF NATURE
CHARACTERIZATION
(IgA
IgAN patient anti-VEC EIA
ET AL.. 2 OF IgA ANTI-VEC
ACTIVITY IgA CIC
+ )
Anti-HLA
Ab
Anti-C3
EIA
Raji
EIA
-t
I 2 3 4 5 6
-
pi
+
-.
HLA class 11 reactivity was excluded in these patients as the endothelial cells were unstimulated and therefore did not bind mouse monoclonal anti-HLA DR Ab. Stimulation
of Endothelial
Cells with IFN-7
and IL-1
Stimulation of the endothelial cells with IFN-)I resulted in HLA class 11 (DR) expression, as well as an increase in HLA class I expression (Fig. 5). However, there was no significant difference in the anti-VEC activity of the sera of IgAN patients when tested against unstimulated and stimulated cells. Similarly, IL-1 stimulation of endothelial cells did not result in an increase in anti-VEC binding to the cells. TABLE 3 OF NATURE OF IgG ANTI-VEC
CHARACTERIZATION IgAN patients (IgG Anti-VEC EIA
ACTIVITY IgG CIC
+ )
Anti-HLA
Ab
Clq
EIA
I
..-
--
2 3 4 5 6 7 8 9 IO
+ + + .+ -
-t t
-
11
-
-t
Raji EIA
Anti-C3
EIA
-
+
+
L
f
i
ANTI-VASCULAR
ENDOTHELIAL
CELL ANTIBODIES
457
.
I 01
. I
CONTROL
I
KiAN PATENTS
I
HD PATENTS POSITNE ON LYMPHOCYTE MlcRO~~m~OXtcrrY
FIG. 4. Anti-HLA class 1 inhibitory activity of IgAN patients positive for anti-VEC activity. Inhibition >75% was defined as positive (based on 18 normal controls shown).
Role of Immune Complexes in Endothelial Cell Binding
Forty percent of IgAN patients were positive for IgA CIC as measured by the F(ab’), anti-C3 EIA and Raji cell EIA. However, no significant correlation was seen between the presence of CIC and IgA anti-VEC activity (Table 3). Similarly, 060050a ew 040g z 0 ?b 0.30s g P 0.200.10-
o.oL
t3.N PATOICTS
FIG. 5. Anti-VEC activity of IgAN positive sera on endothelial cells stimulated with IFN-y. A denotes binding to unstimulated cells; 0 denotes binding to stimulated cells. (a) Increase in binding of mouse anti-HLA-A,B,C Ab on stimulated endothelial cells. (b) Increase in binding of mouse antiHLA-DR Ab on stimulated endothelial cells.
458
YAP
ET AL.
although 60% of IgAN patients were positive for IgG CIC, there was also no correlation with IgG anti-VEC activity (Table 4). To further exclude the possibility of immune complexes binding to the endothelial cells, immune complex positive sera were subjected to ultracentrifugation at 100,OOOg for 1 hr. All anti-VEC positive sera retained positivity after removal of immune complexes by centrifugation. Correlation
with Clinical
and Renal Biopsy Features
Quantitative data on the degree of proteinuria were available in 68 IgA patients. Thirteen of 16 IgA patients with anti-VEC Ab had significant proteinuria greater than 1 g/day (Table 5). Using the Fisher’s exact test, a significant correlation was demonstrated between the presence of anti-VEC Ab and proteinuria or greater than 1 g/day (P = 0.048) (Table 5). There was no correlation with serum creatinine greater than 1.5 mg/dl, the presence of more than 30% sclerosed glomeruli or the presence of crescents on renal biopsy. However, there was a positive correlation between the presence of IgA specific anti-VEC Ab in the sera with the presence of either C3 or IgA deposition in the walls of the renal arterioles, as shown on immunofluorescent examination of the renal biopsies (P = 0.044). DISCUSSION
In recent years, the activated endothelial cell has been increasingly recognized as an important participant in the genesis of immune-mediated vascular injury. Cytokines, such as interleukin 1 and tissue necrosis factor (TNF), which are produced by macrophages, stimulate cultured human umbilical venous endothe-
DEGREE
_.-
IN IgA
IgA anti-VEC Ab (OD score)
IgA patients 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
OF PROTEINURLA
~~~~~
0.08 0.15 0.34 0.07 0.15 0.08 0.08 0.08 0.08 0.09 0.03 -0.02 0.00 0.00 0.06 0.04
TABLE 4 PATIENTS
POSITIVE IgG
FOR ANTI-VEC
anti-VEC COD score) -- 0.07 - 0.09 0.02 -0.03 0.52 - 0.05 0.01 - 0.02 -- 0.03 0.03 0.22 0.08 0.10 0.32 0.06 0.08
ABS
Ab
Proteinuria (g/day)
_ _..._~~~
~~~c
3.04 0.24 1 .FO 1.08 2.25 1.56 2.00 1.30 I .20 0.93 1.19 0.32 I .20 1.20 2.25 1.84
ANTI-VASCULAR
CORRELATION
BETWEEN
IgA patients Anti-VEC Anti-VEC
Ab positive Ah negative
PRESENCE
ENDOTHELIAL
CELL
TABLE OF ANTI-VEC
Proteinuria
5 Ae AND SIGNIFICANT
> 1 g/day 13 26
459
ANTIBODIES
PROTEINURIA
>
1 g/day
Proteinuria < 1 g/day 3 26 ___~
lial cells to express tissue factor-like procoagulant activity (13, 14), leukocyte adhesion properties (15, 16), and unique activation antigens (17-19). The role of anti-VEC Abs in the pathogenesis of immune vascular disorders has been the subject of much debate. Such Abs have been described in the sera of patients with rheumatoid arthritis, scleroderma, systemic lupus erythematosus (SLE), dermatoyositis, polymyalgia rheumatica, and giant cell arteritis, as well as in about 14% of normal controls (1, 20, 21). Cines et al. (1) have shown that binding to endothelial cells of IgG and immune complexes found in the sera of patients with SLE resulted in the production of procoagulant tissue factor (22). In addition, the sera of patients with Kawasaki’s disease, a diffuse vasculitic disease common in children, contain Abs which cause lysis of activated endothelial cells in vitro (23, 24). Although endothehal activation antigens have been demonstrated in cultured glomerular cells (25), anti-VEC Abs have yet to be described in primary human glomerulonephritis. In our patients with IgAN, sera positive for anti-VEC activity was found in 32%, as compared to 4% in controls, and 9% in other forms of primary GN. The contribution of immune complexes to this activity was probably insignificant, as there was no correlation between anti-VEC activity and CIC levels in the patients. In addition, ultracentrifugation of the sera at 100,OOOg to remove immune complexes did not result in a decrease in anti-VEC binding. These Abs were predominantly of the IgA class, with three patients having both IgA and IgG Abs. Some of these Abs were directed against HLA class I antigens, as demonstrated by the significant inhibition of binding of the mouse monoclonal Ab (W6/32) to the framework region of the HLA class I antigen expressed on endothelial cells and platelets. The majority, however, were reactive against antigens present on unstimulated endothelial cells, as prior stimulation of the cells with IFN-?, or IL-l did not result in an increase in anti-VEC activity of the sera. It is conceivable that the endothelial cell in the human glomerulus is actively involved in the immune processes leading to glomerular injury (26). In a recently described model of experimental glomerulonephritis (27), an injection of heterologous anti-endothelial antibodies directed against angiotensin-converting enzyme in rabbits resulted in an increase in vascular permeability leading to “planting” of the heterologous IgG on the subepithelial part of the glomerular capillary wall, remniscent of classical Heymann nephritis (2&30). In our patients, a correlation was seen between the presence of IgA anti-VEC Abs with C3 and immunoglobulin deposition in the renal arteriolar walls. Additionally, a significant correlation was present between these Abs and the presence of proteinuria greater than 1 g per day in the IgAN patients. The presence of heavy
460
YAP ET AL.
proteinuria has been found to be an indicator of a more severe disease in our patients with IgAN (7, 31). Hence it is possible that binding of anti-VEC Abs to the glomerular endothelial cells could result in glomerular capillary wall damage via the production of mediators such as platelet-activating factor (32). and contribute to the progression of the renal lesion in these patients. On the other hand, 68% of the IgA patients did not have any anti-VEC activity, and only half of these patients had heavy proteinuria as compared to 81% in the group with anti-VEC Abs. Other mechanisms leading to progressive renal injury may be operative in these patients. In many patients with IgA nephropathy, the genesis of glomerular sclerotic lesions appears to be noninflammatory, as macrophages or T cells cannot be demonstrated in the glomeruli (33). Here, local hemodynamic factors have been postulated to be important in the pathogenesis of progressive glomerular sclerosis (34). In conclusion, this study demonstrated the presence of anti-VEC Abs in some patients with IgAN. The predominance of Abs of the IgA class may be a manifestation of the increase in IgA-secreting lymphocytes described in this disorder (35-37). The exact pathogenetic significance of these Abs is unclear. It would be exciting to postulate a role for these Abs in contributing to the progression of glomerular immune injury. On the other hand, it is also conceivable that damage to the vascular endothelium by a more primary process with exposure of endothelial cell antigens could have stimulated the production of these Abs. Further studies are required to characterize the nature of the endothelial antigen to which these Abs are reactive and their importance in mediating glomerular injury in &AN. ACKNOWLEDGMENTS The authors gratefully acknowledge the statistical assistance provided by Ms. F. Lennie Wong, biostatistician, Department of Medical Genetics. Cedars-Sinai Medical Center. We also thank Dr. C. Hobel, Department of Obstetrics/Gynecology. Cedars-Sinai Medical Center, for assistance in collection of umbilical cords; Mr. Y. K. Lau and staff of the Renal Laboratory, SGH, for their help in sera collection; and Dr. J. Prehn and Dr. D. Yamaguchi for their invaluable advice and encouragement.
REFERENCES 1. Cines, D. B., Lyss, A. P., Reeber, M.. Bina. M., and Dehoratius, R. J.. Presence of complementfixing anti-endothelial cell antibodies in systemic lupus erythematosus. J. Clin. Invest. 73, 61 I625, 1984. 2. Paul, L. C.. Claas, F. H. J., Van Es, L. A., Kalff, M. W.. and De Graeff. J.. Rejection of a renal allograft associated with pretransplantation antibodies directed against donor antigens on endothelium and monocytes. N. En@. J. Med. 300, 1258-1260, 1979. 3. Paul. L. C., Van Es, L. A., Kalff, M. W., and De Graeff, J., Intrarenal distribution of endothelial antigens recognized by antibodies from renal allograft recipients. Transplant. Proc. 11, 427-430, 1979. 4. Paul. L. C., Van Es, L. A., Van Rood, J. J., Van Leeuwen, A., De La Reviere, 6. B., and De Graeff, J., Antibodies directed against antigens on the endothelium of peritubular capillaries in patients with rejecting renal ahografts. Trun.~plantarion 27, 175-179, 1979. 5. Levy, M., Gonzalez-Burchard, G., Broyer, M., Dommergues. J. P.. Foulard, M., Sorez, J. P.. and Habib, R., Berger’s disease in children. Medicine 64, 157-180, 1985. 6. Nakamoto, Y., Asano, Y., Dohi. K., Fujioka, M., Iida, H., Kida. H.. Kibe. Y.. Hattori. N., and
ANTI-VASCULAR
7. 8. 9. 10. 11. 12. 13. 14.
15. 16. 17.
18. 19. 20. 21. 22. 23. 24.
25.
ENDOTHELIAL
CELL ANTIBODIES
461
Takeuchi, J., Primary IgA glomerulonephritis and Schonlein-Henoch purpura nephritis: Clinicopathological and immuno-histological characteristics. Q. J. Med. 47, 495-516, 1978. Woo, K. T., Edmondson, R. P. S., Wu, A. Y. T., Chiang, G. S. C., Pwee, H. S., and Lim, C. H., The natural history of IgA nephritis in Singapore. C&n. Nephrol. 25, 15-21, 1986. Jaffe, E. A., Nachman, R. L., Becker, C. G., and Minick, C. R., Culture of human endothelial cells derived from umbilical veins. J. Clin. Invest, 52, 2745-2756, 1973. Gimbrone, M. A., Jr., Culture of vascular endothelium. Prog. Hemostasis Throm. 3, l-28, 1976. Barkley, S. C., Sakai, R. S., Ettenger, R. B., Fine, R. N., and Jordan, S. C., Determination of antiidiotypic antibodies to anti-HLA IgG following blood transfusions. Transplantation 44, 30-34, 1987. Jordan, S. C., Gautier, E., Sakai, R., and Bahn, L., Quantitation of circulating immune complexes in human serum by the Raji cell and F(ab’), anti-C3 micro enzyme immunoassays. J. Immunol. Methods 83, 363-370, 1985. Yap, H. K., Sakai, R. S., Woo, K. T., Lim, C. H., and Jordan, S. C., Detection of bovine serum albumin in the circulating IgA immune complexes of patients with IgA nephropathy. Clin. Zmmunol. Immunopathol. 43, 395402, 1987. Bevilacqua. M. P., Pober, J. S., Wheeler, M. E., Cotran, R. S., and Gimbrone, M. A., Jr., Interleukin-1 activation of vascular endothelium: Effects on procoagulant activity and leukocyte adhesion. Amer. J. Pathol. 121, 393-403, 1985. Bevilacqua, M. P., Pober, J. S., Majeau, G. R., Fiers, W., Cotran, R. S., and Gimbrone, M. A., Jr., Recombinant tumor necrosis factor induces procoagulant activity in cultured human vascular endothelium: Characterization and comparison with the actions of interleukin 1. Proc. Nat/. Acad. Sci. USA 83, 45334537, 1986. Gamble, J. R., Harlan, J. M., Klebanoff, S. J., and Vadas, M. A., Stimulation of the adherence of neutrophils to umbilical vein endothelium by human recombinant tumor necrosis factor. Proc. Natl. Acad. Sci. USA 82, 8667-8671. 1985. Schleimer, R. P., and Rutledge, B. K., Cultured human vascular endothelial ceHs acquire adhesiveness for neutrophils after stimulation with interleukin I, endotoxin, and tumor-promoting phorbol diesters. J. lmmunol. 136, 649-654, 1986. Pober, J. S., Bevilacqua, M. P., Mendrick. D. L., Lapherre, L. A., Fiers, W., and Gimbrone, M. A., Jr., Two distinct monokines. interleukin 1 and tumor necrosis factor, each independently induce biosynthesis and transient expression of the same antigen on the surface of cultured human vascular endothelial cells. J. Immunol. 136, 1680-1687, 1986. Cotran, R. S., Gimbrone, M. A.. Jr., Bevilacqua. M. P.. Mendrick. D. L.. and Pober, J. S., Induction and detection of a human endothelial activation antigen in vivo. J. Exp. Med. 164, 661-666, 1986. Messadi, D. V., Pober, J. S., Fiers, W., Gimbrone, M. A., Jr., and Murphy, G. F., Induction of an activation antigen on postcapillary venular endothelium in human skin organ culture. J. Immunol. 139, 1557-1562, 1987. Lindqvist, K. I., and Osterland, C. K., Human antibodies to vascular endothelium. Clin. Exp. Immunol. 9, 753-760, 1971. Tan, E. M., and Pearson, C. M., Rheumatic disease sera reactive with capillaries in the mouse kidney. Arthritis Rheum. 15, 23-28, 1972. Tannenbaum, S. H., Finko, R., and Cines, D. B., Antibody and immune complexes induce tissue factor production by human endothelial cells. J. Immunol. 137, 1532-1537, 1986. Leung, D. Y. M., Collins, T., Lapierre, L. A., Geha, R. S., and Pober, J. S.. Immunoglobulin M antibodies present in the acute phase of Kawasaki syndrome lyse cultured vascular endothelial cells stimulated by gamma interferon. J. C/in. Invest. 77, 1428-1435, 1986. Leung, D. Y. M., Geha, R. S., Newburger, J. W., Bums, J. C., Fiers. W., Lapierre, L. A., and Pober. J. S., Two monokines, interleukin 1 and tumor necrosis factor, render cultured vascular endothelial cells susceptible to lysis by antibodies circulating during Kawasaki syndrome. J. Exp. Med. 164, 1958-1972, 1986. Hancock, W. W., and Cotran, R. S., Induction of activation antigens on human glomerular endothelium by interleukin 1 (IL-l), interferon gamma (IFN-g), and endotoxin (LPS) Kidney Int. 31, 322, 1987. [Abstract]
462
YAP ET AL.
26. Salant, D. J.. Nephrology forum: lmmunopathogenesis of crescentic glomerulonephritis and lung purpura. Kidney Znt. 32, 408-425, 1987. 27. Matsuo, S., Fukatsu, A., Taub, M. L., Caldwell, R. B., Brentjens. J. R., and Andres, G., Glomerulonephritis induced in the rabbit by antiendothelial antibodies. J. Clin. Invest. 79, 1798-1811. 1987. 28. Barabas. A. Z., and Lannigan, R., Induction of an autologous immune complex glomerulonephritis in the rat by intravenous injection of heterologous anti-rat kidney tubular antibody. I. Production of chronic and progressive immune complex glomerulonephritis. Brit. J. Exp. Pathol. 55, 47-55, 1974. 29. Couser, W. G., Steinmuller, D. R., Stilmant, M. M., Dalant, D., and Lowenstein, L. M.. Experimental glomerulonephritis in the isolated perfused rat kidney. J. Clin. Invest. 62, 1275-1287. 1978. 30. Jeraj, K.. Vernier, R. L., Sisson, S. P., and Michael, A. F., A new glomerular antigen in passive Heymann’s nephritis. Brit. 1. Exp. Pathol. 65, 485-198, 1984. 31. Woo, K. T., Edmondson, R. P. S., Yap. H. K.. Wu, A. Y. T., Chiang, G. C. S.. Lee, E. J. c‘. Pwee, H. S., and Lim, C. H., Effects of triple therapy on the progression of mesangial prohferative glomerulonephritis. C/in. Nephrol. 27, 56-64, 1987. 32. Camussi, G., Aghetta, M., Malavasi, F.. Tetta, C.. Piacibello, W., Sanavio, F.. and Bussolino. F.. The release of platelet-activating factor from human endothelial cells in culture. J. Zmmunol. 131, 2397-2403, 1983. 33. Woodroffe, A. J., Bannister, K. M., Bellon, M. S., Hale, G. M.. and Clarkson. A. R.. Do local hemodynamic alterations mediate injury in IgA nephropathy? Sem. Nephrol. 7, 341-343, 1987. 34. Brenner, B. M., Meyer, T. W.. and Hostetter. T. H., Dietary protein intake and the progressive nature of kidney disease: The role of hemodynamically mediated glomeruiar injury in the pathogenesis of progressive glomerular sclerosis in aging, renal ablation, and intrinsic renal disease, N. Engl. J. Med. 307, 652-659. 1982. 35. Nomoto. Y., Sakai, H., and Arimori, S., Increase of IgA-bearing lymphocytes in peripheral blood from patients with IgA nephropathy. Amer. J. Pathol. 71, 158-160. 1979. 36. Egido, J., Biasco, R., Sancho, J., Lozano, L., Sanchez-Crespo, M., and Hernando. L. 1Increased rate of polymeric IgA synthesis by circulating lymphoid cells in IgA mesangial glomerulonephritis. Clin. Exp. Zmmunol. 47, 309-316, 1982. 37. Schena. F. P., Mastrolitti, G., Fracasso, A. R.. Pastore. A., and Lad&. N., Increased immunoglobulin-secreting cells in the blood of patients with active idiopathic IgA nephropathy. C’lin. Nephrol. 26, 163-168, 1986. Received April 5, 1988: accepted with revision August 16, 1988