Clin. exp. Immunol. (1978) 34, 213-218.
Circulating immune complexes in active Behcet's disease R. C. GUPTA,* J. D. O'DUFFY, F. C. MCDUFFIE, M. MEURER & R. E. JORDON Departments of Medicine, Dermatology and Immunology, Mayo Foundation, Rochester, Minnesota, USA (Received 12 April 1978)
SUMMARY
Levels of immune complexes (IC) were determined by the Raji-cell and Clq-binding radioimmunoassays in the sera of eighteen patients with Behqet's disease. Eight patients (44%4) were found to have significantly elevated levels of IC (range 56 to 1600 ,ug equivalent aliquots of heat aggregated IgG (AHG/ml) by the former test and nine by the latter (7 to 500 Pg equivalent AHG/ ml). The presence of IC showed a significant correlation with disease activity score (P = 0 003, Mann-Whitney Rank Sum Test). Abnormal values of IC tended to remain abnormal when sera were retested after 1 year. There was no correlation between IC and duration of disease or any specific organ involvement. The IC were found predominantly in fractions of about 19S and greater when fractionated by sucrose density gradient or Sephadex column techniques. The results suggest the possibility that IC may contribute to the pathophysiology of Behqet's disease.
INTRODUCTION Behqet's disease was initially described (Behset, 1937) as a triad of recurrent aphthous oral ulcerations, genital ulcerations and ocular inflammation. Other systemic features include arthritis, skin lesions, thrombophlebitis, colitis, pancreatitis, subungual infarctions, arteritis, and involvement of central and peripheral nervous system. The aetiology and pathogenesis of Behget's disease are unknown. It is characterized by the presence of vasculitis and of serum antibodies directed against oral and other mucosal epithelial cells (Oshima, Shimizu & Yokohari, 1963). The antibodies to oral mucosa have been demonstrated by indirect immunofluorescence (O'Duffy, Carney & Deodhar, 1971) and haemagglutination tests (Lehner, 1969). These antibodies belong predominantly to the IgM and to a lesser extent IgG class of antibodies (Lehner, 1969). Activation of complement preceding acute uveitis in Behget's disease has been noted (Shimada et al., 1974). The latter observations suggest that immune complexes may be contributing to the systemic inflammatory process. Accordingly we determined the presence of circulating immune complexes (IC) in sera of eighteen patients with Behget's disease using the Raji-cell and Clq-precipitin radioimmunoassays. We found a significant correlation between elevated levels of IC and the activity of the disease in these patients.
MATERIALS AND METHODS Patients and controls. The eighteen patients included in the present study fulfilled the criteria (O'Duffy & Goldstein, 1976) for the diagnosis of Behvet's disease. All patients were seen and followed at the Mayo Clinic. Eleven patients were females and seven were males. Their ages ranged between 24 years to 64 years with a mean of 40 4 years. One hundred normal healthy adults (volunteer blood donors) were used as controls. Collection ofserum samples. Peripheral blood was drawn from each patient at entry and about 1 year later. It was allowed to clot at room temperature for 2 to 3 hr, and the sera were separated and mailed to us by the patients' personal physicians.
Correspondence: Dr R. C. Gupta, Assistant Professor of Medicine, University of Colorado Medical Center, 4200 East Ninth Avenue, Box B-115, Denver, Colorado 80262, USA. 0099-9104/78/0110-0213$02.00 (D 1978 Blackwell Scientific Publications
213
214
R. C. Gupta et al.
All sera were stored in aliquots at - 20'C and thawed just before quantitation of IC. We have found that freezing and thawing does not affect the levels of IC in sera. Immune complex measurements: IC were determined by the techniques of Raji-cell and '3 I-Clq-precipitin radioimmunoassays. (a) Raji-cell radioimmunoassay. The method of quantitative detection of IC by Raji-cell radioimmunoassay was that of Theofilopoulos, Wilson & Dixon (1976) as modified by us (Gupta et al., 1978). Serum samples were incubated with fresh normal serum as a source of complement and then with Raji cells (lymphocytes belonging to a line established from a patient with Burkitt's lymphoma). IC present bound to Raji cells via cell surface receptors for Clq and C3d. The cells were washed and then incubated with 125I-labelled anti-human IgG. After further washing the cells were counted in a scintillation counter and the amount of IC determined by reference to a standard curve prepared from the data obtained with aliquots of heat aggregated IgG (AHG) added to normal serum. The amount of IC in each serum tested was expressed in terms of jug equivalent AHG/ml serum. 131I-Clq-Binding radioimmunoassay. The Clq-binding activity (ClqBA) in sera of patients and controls was determined using the technique of Nydegger et al. (1974) in a modification recently described by Tappeiner et al. (1977). Results were expressed as C1qBA equivalent to pug/ml AHG as determined from a standard curve. AHG in this assay was prepared by heating a solution of 25 mg/ml Cohn fraction 11 at 630 for 12 min, then cooling in an ice bath and passing over a Sepharose 6B column in 0-1M Tris HC1, 0-2M NaCl pH 8 5 to remove 7S IgG and small aggregates. The fractions in the exclusion peak were pooled and used. Determination ofsize of immune complexes. (a) Sucrose density-gradient studies: 0 5 ml of serum diluted 1:2 with saline was layered on a 5-30%/ sucrose gradient in 0 iM Tris, 0-2M NaCl, pH 8-0 buffer in a 12 x 75 mm tube. Centrifugation was carried out at 283,000g (at the bottom of the tube) for 12 hr at 4VC in a SW 283 rotor in an International B-60 ultracentrifuge. All samples were run in duplicate. The gradient tubes were then punctured and 18 fractions of 0-6 ml were collected. 0-25 ml of each fraction was mixed with equal volume of normal serum diluted 1: 2 with saline and the presence of IC in these fractions were determined by the Raji-cell radioimmunoassay. Markers were the IgG and IgM in serum as determined by radial immunodiffusion (Mancini, Carbonara & Heremans, 1965) with specific antisera made by us. (b) Sephadex G-200 column chromatography: Since high concentrations of sucrose were found to interfere with the Clqbinding radioimmunoassay, we employed Sephadex G-200 to fractionate sera and determine IC by this assay. 0 75 ml of serum, diluted 1:2 in pH 7 0 phosphate saline buffer was fractionated on a 1-5 x 30 cm of Sephadex G-200 equilibrated with the same buffer. Fractions of 1 0 ml were tested for content of IgM, and IgG by immunodiffusion and used as markers for size. Six sequential pools of the protein containing fractions were concentrated by ultrafiltration in a collodion tube (Schleicher & Schuell, Inc., Keene, NH), and the Clq binding activity of each determined. Disease activity score. Patients filled out a questionnaire which was designed to assess the persistence and severity of symptoms. Disease activity was assessed independently of IC results by an activity score which allotted 1 point for each system involved (aphthous ulceration, genital ulceration, uveitis, synovitis, skin vasculitis and meningoencephalitis) plus 1 point for moderate and 2 for severe activity as judged on clinical grounds. Serum immunoglobulin determinations. Serum IgG and IgM levels were determined by the radial immunodiffusion technique of Mancini et al. (1965) in plates containing agar mixed with monospecific antiserum prepared by us.
RESULTS Serum immune complexes by the Raji-cell radioimmunoassay The Raji-cell radioimmunoassay detects IC in serum via complement receptors on the surface of Raji cells (Theofilopoulos et al. 1976; Gupta et al., (1978). In our laboratory, sera of 100 normal healthy adults have been tested by this method and the upper limit of normal has been found to be 55-3 pug equivalent AHG/ml. The levels of serum IC determined on the initial examination of eighteen patients with Behset's disease are presented in Fig. 1. Eight patients (44%°) were found to have elevated levels of IC in their sera. Levels of IC were re-examined 1 year later in twelve patients including seven with elevated levels of IC initially. All seven patients with elevated levels of IC showed persistence of IC on the second examination, and only one of the five with normal levels found to have elevated levels of IC (Table 1). Serum immune complexes by 131IC1q-binding radioimmunoassay In the I3II-Clq-binding method radiolabelled Clq is mixed with test sample in the presence of EDTA followed by addition of polyethylene glycol which precipitates Clq bound to macromolecular complexes. The radioactivity of the precipitate is counted in a gamma counter. By this technique, sera of normal controls had values less than 5 pg equivalent AHG/ml. Only the initial serum samples of patients with Behqet's disease could be examined and elevated levels of Clq-binding activity were found in nine
Immune complexes in active Behfet's disease
215
2000 p 0
1000 F x
0
5001-
as
0
E
0
0
a)
c
E
_
0
*- I
E < '7
-
200
100 50
0
To
0 a)
0
-J
10 0
0 0
5
RAJI cell assay
Clq precipitin assay
FIG. 1. Distribution of levels of IC in sera of patients with Behset's disease measured by two radioimmunoassays. Patients with activity score < 3 (e) and < 4 (o). Broken line ---) represented the upper range of normal levels of IC by each assay. TABLE 1. Sequential measurements of the levels of IC by the Raji-cell radioimmunoassay and activity score in twelve patients with Behqet's disease
Initial measurements
Patients 1 2 3 4 5 6 7 8 9 10 11 12
One year follow-up measurements
Levels of IC
Activity
Levels of IC
Activity
score
1600 960 376 256 216 216 216 48 40 36 32 8
6 4 5 6 2 4 2 3 5 2 1 0
1248 176 896 1376 528 184 576 32 128 32 32 8
Improved* 3 6 4 4 4
score
2 5 6 1 3 0
* Patient had a peripheral neuropathy with cutaneous vasculitis and cryoglobulinaemia which improved on steroids.
of eighteen sera (Fig. 1), seven of which showed a marked increase in a range of 50-500 pg equivalent AHG/ml.
Serum immunoglobulin levels and circulating IC Since the IC measured in serum are presumably IgG or IgM aggregates capable of binding complement, we examined the relationship between levels of IC and the serum levels of these immunoglobulins. We found no significant correlation between levels of IC and levels of IgG and IgM.
R. C. Gupta et al.
216
Size ofIC To determine the approximate molecular size of the material measured as IC in this study, we performed sucrose density gradient ultracentrifugation and Sephadex G-200 column chromatography on the serum of one patient (patient no. 1, Table 1) with elevated levels of IC. The results of the Raji-cell assay performed on the sucrose density gradient fractions and the Clq-binding assay on the Sephadex G-200 fractions are presented in Figs 2 and 3, respectively. On density gradient ultracentrifugation, the bulk of the IC activity was in fractions approximately 19S. An additional small peak at 7S may represent 7S IgG bound to Fc receptors on the Raji cells. On G-200 chromatography, the Clq-binding material was present in the void volume indicating a molecular weight greater than 200,000. Additionally, the size of IC were determined by the Raji assay in three other patients. In two patients (patients 3 and 6, Table 1), the size of complexes were similar to patient no. 1 while in the third patient (no. 4, Table 1) complexes sedimented at 10-13S. Disease activity score and circulating IC Since there is no single measurement indicative of the severity and progression of the disease, we arbitrarily calculated a disease activity score based on the number and severity of systems involved. We 7S
19S
5.0 6.-
.c 4.5 E c3 -~
N.r-0 a
to
c
0
3L00 2
8 10 12 14 4 6 Sucrose gradient fractions
FIG. 2. Density gradient ultracentrifugation of serum in one patient with Behcet's disease and an elevated level of IC. IgM and IgG were used as molecular weight markers. Bottom tube to the left. The presence of IC in fractions was measured by the Raji-cell assay. ,
IgM (19S5 IgG (7S)
,
~~~~~~~2.0 OE
0
~ C
4
Circulating immune complexes in active Behcet's disease R. C. GUPTA,* J. D. O'DUFFY, F. C. MCDUFFIE, M. MEURER & R. E. JORDON Departments of Medicine, Dermatology and Immunology, Mayo Foundation, Rochester, Minnesota, USA (Received 12 April 1978)
SUMMARY
Levels of immune complexes (IC) were determined by the Raji-cell and Clq-binding radioimmunoassays in the sera of eighteen patients with Behqet's disease. Eight patients (44%4) were found to have significantly elevated levels of IC (range 56 to 1600 ,ug equivalent aliquots of heat aggregated IgG (AHG/ml) by the former test and nine by the latter (7 to 500 Pg equivalent AHG/ ml). The presence of IC showed a significant correlation with disease activity score (P = 0 003, Mann-Whitney Rank Sum Test). Abnormal values of IC tended to remain abnormal when sera were retested after 1 year. There was no correlation between IC and duration of disease or any specific organ involvement. The IC were found predominantly in fractions of about 19S and greater when fractionated by sucrose density gradient or Sephadex column techniques. The results suggest the possibility that IC may contribute to the pathophysiology of Behqet's disease.
INTRODUCTION Behqet's disease was initially described (Behset, 1937) as a triad of recurrent aphthous oral ulcerations, genital ulcerations and ocular inflammation. Other systemic features include arthritis, skin lesions, thrombophlebitis, colitis, pancreatitis, subungual infarctions, arteritis, and involvement of central and peripheral nervous system. The aetiology and pathogenesis of Behget's disease are unknown. It is characterized by the presence of vasculitis and of serum antibodies directed against oral and other mucosal epithelial cells (Oshima, Shimizu & Yokohari, 1963). The antibodies to oral mucosa have been demonstrated by indirect immunofluorescence (O'Duffy, Carney & Deodhar, 1971) and haemagglutination tests (Lehner, 1969). These antibodies belong predominantly to the IgM and to a lesser extent IgG class of antibodies (Lehner, 1969). Activation of complement preceding acute uveitis in Behget's disease has been noted (Shimada et al., 1974). The latter observations suggest that immune complexes may be contributing to the systemic inflammatory process. Accordingly we determined the presence of circulating immune complexes (IC) in sera of eighteen patients with Behget's disease using the Raji-cell and Clq-precipitin radioimmunoassays. We found a significant correlation between elevated levels of IC and the activity of the disease in these patients.
MATERIALS AND METHODS Patients and controls. The eighteen patients included in the present study fulfilled the criteria (O'Duffy & Goldstein, 1976) for the diagnosis of Behvet's disease. All patients were seen and followed at the Mayo Clinic. Eleven patients were females and seven were males. Their ages ranged between 24 years to 64 years with a mean of 40 4 years. One hundred normal healthy adults (volunteer blood donors) were used as controls. Collection ofserum samples. Peripheral blood was drawn from each patient at entry and about 1 year later. It was allowed to clot at room temperature for 2 to 3 hr, and the sera were separated and mailed to us by the patients' personal physicians.
Correspondence: Dr R. C. Gupta, Assistant Professor of Medicine, University of Colorado Medical Center, 4200 East Ninth Avenue, Box B-115, Denver, Colorado 80262, USA. 0099-9104/78/0110-0213$02.00 (D 1978 Blackwell Scientific Publications
213
214
R. C. Gupta et al.
All sera were stored in aliquots at - 20'C and thawed just before quantitation of IC. We have found that freezing and thawing does not affect the levels of IC in sera. Immune complex measurements: IC were determined by the techniques of Raji-cell and '3 I-Clq-precipitin radioimmunoassays. (a) Raji-cell radioimmunoassay. The method of quantitative detection of IC by Raji-cell radioimmunoassay was that of Theofilopoulos, Wilson & Dixon (1976) as modified by us (Gupta et al., 1978). Serum samples were incubated with fresh normal serum as a source of complement and then with Raji cells (lymphocytes belonging to a line established from a patient with Burkitt's lymphoma). IC present bound to Raji cells via cell surface receptors for Clq and C3d. The cells were washed and then incubated with 125I-labelled anti-human IgG. After further washing the cells were counted in a scintillation counter and the amount of IC determined by reference to a standard curve prepared from the data obtained with aliquots of heat aggregated IgG (AHG) added to normal serum. The amount of IC in each serum tested was expressed in terms of jug equivalent AHG/ml serum. 131I-Clq-Binding radioimmunoassay. The Clq-binding activity (ClqBA) in sera of patients and controls was determined using the technique of Nydegger et al. (1974) in a modification recently described by Tappeiner et al. (1977). Results were expressed as C1qBA equivalent to pug/ml AHG as determined from a standard curve. AHG in this assay was prepared by heating a solution of 25 mg/ml Cohn fraction 11 at 630 for 12 min, then cooling in an ice bath and passing over a Sepharose 6B column in 0-1M Tris HC1, 0-2M NaCl pH 8 5 to remove 7S IgG and small aggregates. The fractions in the exclusion peak were pooled and used. Determination ofsize of immune complexes. (a) Sucrose density-gradient studies: 0 5 ml of serum diluted 1:2 with saline was layered on a 5-30%/ sucrose gradient in 0 iM Tris, 0-2M NaCl, pH 8-0 buffer in a 12 x 75 mm tube. Centrifugation was carried out at 283,000g (at the bottom of the tube) for 12 hr at 4VC in a SW 283 rotor in an International B-60 ultracentrifuge. All samples were run in duplicate. The gradient tubes were then punctured and 18 fractions of 0-6 ml were collected. 0-25 ml of each fraction was mixed with equal volume of normal serum diluted 1: 2 with saline and the presence of IC in these fractions were determined by the Raji-cell radioimmunoassay. Markers were the IgG and IgM in serum as determined by radial immunodiffusion (Mancini, Carbonara & Heremans, 1965) with specific antisera made by us. (b) Sephadex G-200 column chromatography: Since high concentrations of sucrose were found to interfere with the Clqbinding radioimmunoassay, we employed Sephadex G-200 to fractionate sera and determine IC by this assay. 0 75 ml of serum, diluted 1:2 in pH 7 0 phosphate saline buffer was fractionated on a 1-5 x 30 cm of Sephadex G-200 equilibrated with the same buffer. Fractions of 1 0 ml were tested for content of IgM, and IgG by immunodiffusion and used as markers for size. Six sequential pools of the protein containing fractions were concentrated by ultrafiltration in a collodion tube (Schleicher & Schuell, Inc., Keene, NH), and the Clq binding activity of each determined. Disease activity score. Patients filled out a questionnaire which was designed to assess the persistence and severity of symptoms. Disease activity was assessed independently of IC results by an activity score which allotted 1 point for each system involved (aphthous ulceration, genital ulceration, uveitis, synovitis, skin vasculitis and meningoencephalitis) plus 1 point for moderate and 2 for severe activity as judged on clinical grounds. Serum immunoglobulin determinations. Serum IgG and IgM levels were determined by the radial immunodiffusion technique of Mancini et al. (1965) in plates containing agar mixed with monospecific antiserum prepared by us.
RESULTS Serum immune complexes by the Raji-cell radioimmunoassay The Raji-cell radioimmunoassay detects IC in serum via complement receptors on the surface of Raji cells (Theofilopoulos et al. 1976; Gupta et al., (1978). In our laboratory, sera of 100 normal healthy adults have been tested by this method and the upper limit of normal has been found to be 55-3 pug equivalent AHG/ml. The levels of serum IC determined on the initial examination of eighteen patients with Behset's disease are presented in Fig. 1. Eight patients (44%°) were found to have elevated levels of IC in their sera. Levels of IC were re-examined 1 year later in twelve patients including seven with elevated levels of IC initially. All seven patients with elevated levels of IC showed persistence of IC on the second examination, and only one of the five with normal levels found to have elevated levels of IC (Table 1). Serum immune complexes by 131IC1q-binding radioimmunoassay In the I3II-Clq-binding method radiolabelled Clq is mixed with test sample in the presence of EDTA followed by addition of polyethylene glycol which precipitates Clq bound to macromolecular complexes. The radioactivity of the precipitate is counted in a gamma counter. By this technique, sera of normal controls had values less than 5 pg equivalent AHG/ml. Only the initial serum samples of patients with Behqet's disease could be examined and elevated levels of Clq-binding activity were found in nine
Immune complexes in active Behfet's disease
215
2000 p 0
1000 F x
0
5001-
as
0
E
0
0
a)
c
E
_
0
*- I
E < '7
-
200
100 50
0
To
0 a)
0
-J
10 0
0 0
5
RAJI cell assay
Clq precipitin assay
FIG. 1. Distribution of levels of IC in sera of patients with Behset's disease measured by two radioimmunoassays. Patients with activity score < 3 (e) and < 4 (o). Broken line ---) represented the upper range of normal levels of IC by each assay. TABLE 1. Sequential measurements of the levels of IC by the Raji-cell radioimmunoassay and activity score in twelve patients with Behqet's disease
Initial measurements
Patients 1 2 3 4 5 6 7 8 9 10 11 12
One year follow-up measurements
Levels of IC
Activity
Levels of IC
Activity
score
1600 960 376 256 216 216 216 48 40 36 32 8
6 4 5 6 2 4 2 3 5 2 1 0
1248 176 896 1376 528 184 576 32 128 32 32 8
Improved* 3 6 4 4 4
score
2 5 6 1 3 0
* Patient had a peripheral neuropathy with cutaneous vasculitis and cryoglobulinaemia which improved on steroids.
of eighteen sera (Fig. 1), seven of which showed a marked increase in a range of 50-500 pg equivalent AHG/ml.
Serum immunoglobulin levels and circulating IC Since the IC measured in serum are presumably IgG or IgM aggregates capable of binding complement, we examined the relationship between levels of IC and the serum levels of these immunoglobulins. We found no significant correlation between levels of IC and levels of IgG and IgM.
R. C. Gupta et al.
216
Size ofIC To determine the approximate molecular size of the material measured as IC in this study, we performed sucrose density gradient ultracentrifugation and Sephadex G-200 column chromatography on the serum of one patient (patient no. 1, Table 1) with elevated levels of IC. The results of the Raji-cell assay performed on the sucrose density gradient fractions and the Clq-binding assay on the Sephadex G-200 fractions are presented in Figs 2 and 3, respectively. On density gradient ultracentrifugation, the bulk of the IC activity was in fractions approximately 19S. An additional small peak at 7S may represent 7S IgG bound to Fc receptors on the Raji cells. On G-200 chromatography, the Clq-binding material was present in the void volume indicating a molecular weight greater than 200,000. Additionally, the size of IC were determined by the Raji assay in three other patients. In two patients (patients 3 and 6, Table 1), the size of complexes were similar to patient no. 1 while in the third patient (no. 4, Table 1) complexes sedimented at 10-13S. Disease activity score and circulating IC Since there is no single measurement indicative of the severity and progression of the disease, we arbitrarily calculated a disease activity score based on the number and severity of systems involved. We 7S
19S
5.0 6.-
.c 4.5 E c3 -~
N.r-0 a
to
c
0
3L00 2
8 10 12 14 4 6 Sucrose gradient fractions
FIG. 2. Density gradient ultracentrifugation of serum in one patient with Behcet's disease and an elevated level of IC. IgM and IgG were used as molecular weight markers. Bottom tube to the left. The presence of IC in fractions was measured by the Raji-cell assay. ,
IgM (19S5 IgG (7S)
,
~~~~~~~2.0 OE
0
~ C
4
