Expression of the major rheumatoid factor cross-reactive idiotype in pediatric patients with systemic lupus erythematosus

CLINICAL

IMMUNOLOGY

AND

IMMUNOPATHOLOGY

60,

232-243 (lt)91)

Expression of the Major Rheumatoid Factor Cross-Reactive ldiotype in Pediatric Patients with Systemic Lupus Erythematosus VINCENT R. BONAGURA,“.’ NORMAN T. ILOWITE,* LYNDA HATAM,* DAVID J. VALACER,” AND JOSIAH F. WEDGWOOD?~ *Divisions of Allergyllmmunology and Rheumatology, tNeonatology Department of Pediatrics, Schneider Children’s Hospital, Long Island Jewish Medical Center, New Hyde Park, New York 11042. Long Island Campus for the Albert Einstein College of Medicine, Yeshiva University Bronx, New York 10461 Rheumatoid factor cross-reactive idiotype (RF-CRI) is expressed in high concentrations in the sera of some patients with rheumatoid arthritis (RA) and juvenile rheumatoid arthritis (JRA). To determine if RF-CR1 is specifically expressed in rheumatic disease or if it is secondary to polyclonal B-cell activation, we examined sera of 23 children with SLE, 16 adolescents with infectious mononucleosis (IM), and age-matched pediatric controls for RF-CR1 expression. Concentrations of RF-CR1 in serum, determined by an inhibition ELISA, were 24 ? 17 pg/ml (mean ? SD) in 25 normal children, 31 ? 17 in 16 young adults with IM, and were significantly increased, 70 2 80 &ml, in the 23 children with SLE @ < 0.036). Eleven of 23 SLE patients had serum RF-CRI greater than the mean r 2 SD for normal children. Ten of 23 SLE sera contained IgM rheumatoid factor (RF) activity. One patient with IM had a borderline elevated RF-CR1 level, and 5 IM patients had RF in their sera. The serum IgM concentrations in sera were: SLE (192 2 93 mg/dl) and IM (234 2 77 mg/dl) sera. These levels were significantly elevated compared to controls (132 + 44 mg/dl), p < 0.031 for SLE and p < 0.001 for IM, suggesting that polyclonal activation of B cells was present in SLE and IM patient groups. Increased expression of RF-CR1 in the SLE patients correlated directly with high titer anti-DNA antibody values (r = 0.3965, p < 0.05) and RF activity when human IgG (r = 0.5026, p < 0.05) was used as the RF binding substrate and inversely with serum C3 levels (r = 0.3925, p < 0.05). RF-CR1 expression did not correlate with RF that bound rabbit (r = 0.3123, p > 0.05). Increased serum RF-CRI expression is not a result of polyclonal B-cell activation. RF-CR1 may be selectively up-regulated in patients with SLE. o 1991Academic Press. Inc.

INTRODUCTION Within the network of idiotypes produced by polyclonal activated B cells from patients with systemic lupus erythematosus (SLE), anti-DNA antibodies have been extensively studied as putative pathogenic and immunoregulatory autoantibodies (l-5). In comparison, the relationship of IgM rheumatoid factors (RF) and their associated cross-reactive or “public” idiotypes in SLE patients have been less well characterized (6). The major RF cross-active idiotype (RF-CRI) originally described by Kunkel et ’ To whom correspondence and reprint requests should be addressed. 2 Present address: Division of Neonatology, Department of Pediatrics, Mount Sinai School of Medicine, One Gustave L. Levy Place, New York City, NY 10029. 232 0090-1229191 $1.50 Copy&b1 All rights

8 1991 by Academic Press, Inc. of reproduction in any form reserved.

RF CROSS-REACTIVE

IDIOTYPE

EXPRESSION

IN

SLE

233

al. is a dominant rheumatoid factor-related public idiotype defined by the Wa

group of monoclonal IgM RF paraproteins (7). This RF-CR1 is best identified by select polyclonai rabbit anti-idiotypic antibodies (g-l 1) that react with idiotypes encoded by a germ line VkIIIb light chain variable region gene product (12,13,14) and one of a select group of heavy chain V genes (14, 15). Conformationaldependent idiotopes encoded by a germ line VL and select VH genes are associated with the antigen-combining site or paratope (7, 14). To determine if the RF-CR1 is present within the repertoire of autoantibodies found in SLE and indicative of RF autoantibody production, we studied pediatric patients with SLE and age-matched controls for IgM RF and RF-CRI expression. Children with infectious mononucleosis (IM) were also studied because enhanced RF-CR1 expression might merely reflect polyclonal B-cell activation which is characteristic of both IM and SLE (16, 17). The relationship of RF-CR1 expression with serologic evidence of active disease in SLE was also studied to determine if this autoantibody marker parallels the course of active disease as does some public anti-DNA idiotypes expressed in SLE (1, 2). We and others have reported the expression of the major RF-CR1 in adults with rheumatoid arthritis (RA) @-lOa) and children with juvenile rheumatoid arthritis (JRA) (11). We described the coexpression of increased RF-CR1 expression and classic and/or hidden IgM RFs in some seropositive (RF + ) (8,9) and seronegative (RF-) (10) RA patients. However, in children with JRA, this association is not usually observed (11). We hypothesized that RF-CR1 + , RF- JRA patients express RF-CR1 + RFs that bear an isotype other than cr.or a parallel set of RF-CR1 + antibodies that do not bind IgG (18). Such a parallel set would be analogous to that reported within the anti-DNA network of idiotypes which are expressed by MRL mice, by SLE patients, and by the first-degree relatives of SLE patients (19-22). In SLE, most anti-DNA antibodies bearing a public idiotype bind DNA, while those expressed by first-degree relatives of SLE patients do not (21, 22). The parallel set of antibodies bearing public idiotypes related to anti-DNA antibodies in SLE and that associated with RFs in RA may respectively regulate anti-DNA and RF autoantibody production in these diseases. This possibility would be predicted by the model proposed by Oudin and Cazenave (18). Alternatively, the expression of antibodies bearing public idiotypes could simply be the result of polyclonal B-cell activation induced by the initiating events or antigens that cause SLE and RA. The associations of RFs, RF-CRI, and anti-DNA antibodies in children with SLE are the focus of this paper. MATERIALS

AND METHODS

Study Subjects

Twenty-three sera taken from pediatric patients with SLE, 16 girls and 7 boys (means age 15.3 -+ 3.2 years) who fulfilled American College of Rheumatology criteria (23) for the diagnosis of SLE, were studied. Clinical characteristics are summarized in Table 1. Sixteen patients with IM, 2 girls and 14 boys (means age 17.3 + 2.1 years),

234

BONAGURA

ET AL.

without evidence of arthritis were also studied along with 26 pediatric controls (13 girls, 12 boys; mean age 11.2 + 3.0 years). RF Assays A. Classic RF assays. In the Rose-Waaler hemagglutination assay, washed sheep red blood cells (SRBC) were coated with subagglutinating concentrations of purified rabbit IgG anti-SRBC antibodies (24). Aliquots of sera, 20 ~1 in each well, were tested in serial dilutions, neat through 212 (1:40%) in 0.01 M phosphatebuffered saline (PBS), pH 7.2, with 1% fetal bovine serum (FBS) that was preabsorbed with SRBC, before adding the rabbit IgG-coated SRBC indicator cells. In the Rose-Waaler assay, a titer greater than 1: 16 was considered positive. In the latex fixation method, latex particles coated with purified human IgG (Hyland laboratories, Costa Mesa, CA) were used as previously described (25). B. RF ELBA. IgM RF was also detected in sera by a modification of an ELISA method previously published (26). Flat-bottom %-well plastic plates (Immulon II, Dynatech, VA) were coated with 0.5 pg/well of pooled polyclonal human IgG (hIgG) or rabbit IgG (rIgG) overnight at 4°C. Residual binding sites were blocked TABLE CLINICAL

Patient

CHARACTERISTICS

Age/sex

1 2 3

19/F 14/F 14/F

4

20/F

5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

20/M 15/F 16/M 13/F 17/M 15/M 14/F 15/F 14/M 16/F 20/F 16/F 8/F 16/M 19/F 15/M IO/F 10/F 16/F

OF PATIENTS

1

WITH

SYSTEMIC

LUPUS

ERYTHEMATOSIS

Renal disease class”: other organ systems

Steroid Rxedb

IV: pericarditis, arthritis, fever V: Raynaud’s phenomenon, arthritis IV: cerebritis, thrombocytopenia, mucocutaneous, arthritis, fever None: pneumonitis, pericarditis, arthritis, fever V: hepatitis, mucocutaneous, arthritis II: cerebritis, mucocutaneous IV: mucocutaneous, arthritis IV: mucocutaneous, arthritis, fever IV: mucucutaneous, arthritis, fever None: pulmonary emboli, mucocutaneous, fever None: thrombocytopenia, mucocutaneous None: chorea, mucocutaneous III: pneumonitis, mucocutaneous, arthritis III: cerebritis, mucocutaneous, arthritis IV: cerebritis, pleuritis, arthritis II: thrombocytopenia, mucocutaneous, arthritis None: thrombocytopenia, mucocutaneous II: cerebritis, mucocutaneous, arthritis IV: mucocutaneous, arthritis, fever IV: pneumonitis, mucocutaneous IV: mucocutaneous, arthritis III: pneumonitis, arthritis IV: mucocutaneous

Yes No No

0 Disease class according to World Health Organization. ’ Steroid treated at the time the blood specimen was obtained.

Yes Yes No Yes No Yes Yes No Yes No Yes Yes No Yes No Yes Yes No No No

RF CROSS-REACTIVE

IDIOTYPE

EXPRESSION

IN SLE

235

with 2.5% bovine serumalbumin (BSA) for 1hr at 37°C.After washingthreetimes with a buffer containing PBS, 0.05% Tween 20, aliquots of sera were addedat various dilutions and incubated for 4 hr at 20°C. After further washing, bound IgM was detected with horseradish peroxidase-coupled goat anti-human IgM (p chain specific) (Tago, Burlingame, CA) and developed with O-phenylenediamine/H,0,. A standard curve was prepared using a purified monoclonal IgM RF on each plate. Results were expressed as RF binding units per milligram of IgM added. Values greater than 2 SD above the mean obtained with sera from 15 normal children were considered positive. C. Hidden RF assays. Hidden rheumatoid factors were determined by modiflcations of either the IgG-binding ELISA (27) or a hemolytic assay as described (28). Briefly 250 pl of sera was subjected to ion-exchange separation (QAE Sephadex A-50, Pharmacia, Piscataway, NJ) or System 3 columns (Isolab Inc., Akron, OH) under acid conditions to separate autologous IgG from IgM. The IgM fractions were tested either by ELISA using both hIgG and r&G as substrates (11) or by a hemolytic assay on SRBC coated with rabbit anti-SRBC IgG hemolysin as previously reported (28). Values greater than 2 SD above the mean for the normal childhood controls in the ELISA and titers greater than 1:16 in the hemolytic assay were considered positive. Production,

Screening,

and Preparation

of Anti-RF-CRI

Antibodies

Anti-RF-CR1 antibodies raised in Flanders giant rabbits were obtained as previously described in detail (8-11). In short, Ea monoclonal pentameric IgM (RF + , member of the major RF-CRI group) was injected subcutaneously in complete Freund’s adjuvant into rabbits. Sequential postimmunization bleedings were performed and the anti-RF-CR1 activity was assayed as previously described (8). Rabbit anti-Ea F(ab’), antibody fragments, isolated from high titer anti-RF-CR1 + sera, were subjected to exhaustive absorption on Sepharose-coupled immunoabsorbents including pooled polyclonal IgG and RF- monoclonal IgM K immunoglobulins (8). The affinity-purified rabbit F(ab’), anti-RF-CR1 antibodies were then absorbed on a Sepharose-linked column of RF + 19s monoclonal IgM from patient Wa (member of the major RF-CR1 group). These affinity-purified antiRF-CR&enriched antibodies were further absorbed with RF- control polyclonal IgM, 4 mg/ml final concentration, in fluid phase, before use in the inhibition ELISA as previously described (8-11). Affinity-purified anti-RF-CRI antibodies did not agglutinate SRBC coated with four different RF- IgM K monoclonal immunoglobulins or SRBC coated with control RF- polyclonal IgM or IgG antibodies in titers greater than 2l. These affinity-purified anti-RF-CR1 antibodies reacted with SRBC coated with 19s Ea and Wa in high titer (>2”) following fluid phase absorption. Inhibition

ELISA for Detection

of RF-CRI

in Serum

The presence of serum RF-CR1 was assayed by an inhibition ELISA technique which is described in detail (P-11). Briefly, polyvinylchloride wells were coated with purified monoclonal IgM RF from patient Wa. Nonspecific binding sites were saturated with 2.5% BSA. Anti-RF-CR1 antibodies described above were mixed

236

BONAGURA

ET AL.

with serum samples at various dilutions and were then added to the wells. The amount of anti-RF-CR1 bound to the plate was determined using horseradish peroxidase-conjugated goat anti-rabbit IgG. Results are expressed as micrograms of Wa IgM equivalents per milliliter of serum. Wa IgM equivalents were computed using the means of titratable values at two dilutions. Results do not vary by more than 20% using two dilutions within the Wa standard curve generated on each plate. All determinations were performed on at least three separate occasions (with the identical serum specimen) and the resulting means of the three values were taken as the RF-CR1 value for that patient. Variation among the three values was <30%. Quantitation

of Serum IgM,

C3, and C4 Concentrations

To determine if polyclonal B-cell activation in SLE patients accounted for the increase in RF-CR1 expression, we determined the serum concentrations of IgM in SLE, IM, and control sera by nephelometry (Beckman, Brea, CA) (29). Serum concentrations of C3 and C4 were also determined in these sera by nephelometry as an indicator of clinical disease status. Patients were considered to have a low complement level if either their C3 or C4 levels were decreased. Determination

of Anti-Double-Stranded

DNA Antibodies

To identify the presence of anti-double-stranded DNA antibodies in the sera of study subjects, sera were tested by ELISA (Diametix, Miami, FL), as previously described in detail (30). Results were reported in international units per milliliter of serum, traceable to the WHO anti-nuclear factor serum (homogenous), human 1st reference preparation 1970 (31). Concentrations greater than 250 IU were considered positive. Statistical

Methods

Results from normal and patient groups were compared by the Wilcoxon rank sum test and compared for multiple comparisons where applicable (32). Results of RF tests were compared using the one-tailed Spearman rank correlation test (32). Results in all assays were considered positive if they were greater than the mean value +2 SD in the normal control population. The Fisher exact test was used to test associations of normal and abnormal values in the RF-CRI, antidouble-stranded DNA, and complement assays among SLE patients. RESULTS

Serum ZgM Concentrations The patients with SLE and IM expressed significantly greater serum concentrations of IgM compared with controls (132 + 44 mg/ml in controls, 192 + % mg/ml in SLE patients, p < 0.031, and 234 ? 77 mg/ml in IM patients, p < 0.001) (see Table 2). The elevated IgM concentrations in both the SLE and IM patients are consistent with polyclonal B-cell activation in these diseases.

RF CROSS-REACTIVE

IDIOTYPE TABLE

EXPRESSION

237

IN SLE

2

SEROLOGIC CHARACTERIZATION OF PATIENTS WITH SYSTEMIC LUPUS ERYTHEMATOSUS SLE RF-CR1

dsDNA

__-

c3

c4

RF (human IiS) --__

RF-CRI+ 3 7 9 10 13 14 15 16 17 18 21 RF-CR1 1 2 4 5 6 8 11 12 19 20 22 23 SLE total IM Controls

70” 154 110 82 369 171 64 82 72 82 81 122 + 89d

2196 700 2576 280 626 1509 234 7213 299 322 1038 1365 ?z 2068

106’ 86 60 56 27 106 130 97 65 66 79 + 31

16 16
71 43 626 246 1749 2365 183 101 31 136 243 526 2 786 946 2 1586 -

123 108 118 74 101 38 166 197 136 % 119 116 f 43 91 + 41 -

8’ 19 9 11 5 21 34 8 9 11 926 33 24 17 17 9 10 31 37 29 16 29 23 + 10 16 k 10 -

74d 91 278 97 1140 156 197 641 I15 303 119 291 -c 325

76d 75 242 33 328 52 74 347 50 31 36 122 -t 121

62’ 197 172 78 500 252 247 279 158 67 290 209 + 127

178 74 52 91 81 97 208 184 41 199 231 80 126 _f 68 206 + 240 217 + 214 88 k 44

53 64 II 65 37 97 206 93 23 127 84 28 74 + 52 97 + 94 35 + 46 58 + 27

171 244 78 212 246 175 240 116 121 166 155 1% 177 + 52 192 k % 233 + 77 132 + 44

a Wa equivalents/ml of serum. b International units (IU) >250 = positive :. ’ mg/dl of serum. a RF binding units/mg of IgM added to wells coated with human or rabbit IgG generated using a monoclonal IgM RF. e Mean 2 SD.

from a standard curve

Expression of Classic IgM RFs

Sera from the SLE and IM patients and controls were screened for classic IgM RFs by ELISA using both human and rabbit IgG as binding substrates. Ten of the SLE sera, 11 of the IM sera, and 1 control sera expressed IgM RFs that bound human IgG. The mean value for the controls was 89 2 44 RF units, for the SLE patients 204 k 240 RF units (p < 0.019) and for the IM patients 217 + 214 RF units (p < 0.006) (see Table 2). Four SLE, one IM, and one control sera expressed high concentrations of IgM RFs that bound rabbit IgG. The mean value for the SLE patients was 97 + 94 RF units, 35 + 46 RF units for IM patients, and 58 k 27 RF units for controls. The expression of IgM RFs that bound rabbit IgG was not significantly different than controls in SLE patients @ = 0.33). However, the mean values of RF units that bound rabbit IgG were significantly lower for IM patients compared to controls (p < 0.0013). No SLE serum contained RF that selectively bound rabbit IgG.

238 Expression

BONAGURA

ET AL.

of RF-CR1

Serum RF-CR1 concentrations were increased in the SLE patients as compared with controls (24 + 17 Wa equivalents/ml in controls, 70 2 80 Wa equivalents/ml in SLE patients, p < 0.036) (Fig. I). Eleven of the 21 SLE patients had increased RF-CR1 concentrations in their sera compared with controls. Only 1 of the 16 patients with IM had a mildly elevated concentration of RF-CRI. The remaining 15 and none of the controls expressed high concentrations of RF-CRI. As a group, the IM patients did not express high concentrations of RF-CR1 compared with controls, (31 + 17 Wa equivalents/ml in IM patients, p > 0.25) or with the SLE patients 0, = 0.24). Association

of RF-CRI

and RF Expression

Of the 11 SLE patients who expressed increased concentrations of RF-CR1 in their sera, 5 had classic IgM RFs that bound human IgG and 3 of these also had RFs that bound rabbit IgG. The association of RF-CR1 expression by the SLE patients with the expression of RF that bound human IgG was significant (r = 0.5026, p < 0.05). The association of RF-CR1 expression and RFs that bound rabbit IgG was not significant (r = 0.3123, p > 0.05).

Six of the RF-CR1 + SLE sera did not express classic IgM RFs that bound human or rabbit IgG. None of these sera contained hidden RFs that bound human or rabbit IgG. 400

r 0

CTf?LS (~25) FIG. 1. Serum concentrations of the major rheumatoid factor cross-reactive idiotype. Open circles (0) represent individual study subjects. Closed circles (0) are the mean RF-CRI concentrations in Wa (monoclonal IgM RF, prototype of the major RF-CRI group) equivalents/ml of serum for each study group. Bars are 1 SD from the mean RF-CRI concentration for each group. Serum RF-CM concentrations are increased in the SLE patients compared with controls (p < 0.036).

RF CROSS-REACTIVE

IDIOTYPE

EXPRESSION

IN

SLE

239

Association RF-CRI and Anti-DNA Expression Eight of 10 RF-CR1 + SLE sera and O/10 RF-CR1 - SLE sera contained high titer (>250 IU/ml) anti-double-stranded DNA antibodies. The association of RFCRI expression in the SLE sera and the expression of anti-double-stranded DNA antibodies was significant (r = 0.3965, p < 0.05). Five of 6 RF-CRI + sera without IgM RFs contained high titer anti-double-stranded DNA antibodies.

of RF-CRI Expression and C3 and C4 Concentrations Seven of 10 RF-CR1 + SLE sera and none of 10 RF-CR1 - SLE sera had low concentrations of complement, C3K4, or both. RF-CRI expression was associated with low C3 concentrations in SLE sera, (r = -0.3925, p < 0.05), but not with low C4 levels (r = -0.374, p > 0.05).

Association

of RF-CRI Expression and IgM Concentration The association of RF-CR1 expression in SLE or IM sera with concentrations of serum IgM, which were increased in both patient groups, was not significant (r = 0.1679, p > 0.05). Association

DISCUSSION

Although the expression of RFs in SLE has been well documented (33), the role of these autoantibodies in the disease process is disputed (34, 35). We have detected the major RF-CRI, which is defined by prototypic human monoclonal antibodies and expressed by some RA and JRA patients (g-1 I), in the sera of some SLE patients. RF-CR1 expression in SLE correlates with serologic evidence of active disease, including the expression of anti-DNA antibodies. To determine whether RF-CR1 expression is specifically up-regulated in rheumatic disease or is merely the result of polyclonal B-cell activation, we also studied patients with IM. In IM, RFs are expressed in some instances, but RFCR1 expression is not associated with RF expression in these IM patients. However, in SLE, RF expression is associated with RF-CR1 expression. In all but one SLE patient with an increased RF-CR1 serum concentration, a RF or high titer anti-DNA antibody was also present in their set-a.Therefore, RF-CR1 expression appears to be specifically expressed in SLE and may indicate coexpression of RFs or other autoantibodies such as anti-DNA antibodies. RF-CR1 expression in SLE may be regulated by similar mechanisms shown to be active in anti-DNA antibody production (36-38). The expression of both RFCR1 and anti-DNA antibodies may occur through the loss of specific suppressor T cells that are operative in controls that do not express these autoantibodies or through the recruitment of specific T-helper cells that foster RF-CR1 expression in patients with rheumatic disease. Enhanced RF-CR1 expression in SLE may also be linked to a perturbation of the internal B-cell network of interacting idiotypes that regulate RF-CRI+ antibody production. Specifically, changes in the concentration, character, or expression of autologous anti-idiotypic antibodies may explain increased RF-CR1 expression in SLE patients, Such anti-idiotypes have been shown to function within

240

BONAGURA

ET AL.

the series of vectorial idiotypes that comprise the anti-DNA network that regulates anti-DNA antibody expression in SLE (39-41). High concentrations of RF-CR1 in sera of RF- RA or SLE patients may indicate the presence of a parallel set of regulatory RF-CRI+ antibodies related to RFs without RF activity, as we have proposed for RF-CR1 expression in some RFJRA patients (8-10, 42). This would be predicted by the observations of Oudan and Casenave in which antibodies sharing a defined idiotype have different ligandbinding specificities (18). Such antibodies have been described in the anti-DNA system (19). Members of the parallel set might include antibodies such as those specific for cytomegalovirus that contain conserved light chain primary amino acid sequences identical to those found within monoclonal RFs that bear the RF-CR1 (43). In this regard, RF-CR1 expression in sera containing high titer anti-DNA antibodies without RFs may indicate that immunoglobulin light chain and/or heavy chain V genes that encode antibodies bearing RF-CR1 may also share idiotopes within their variable regions with some antibodies that bind DNA. In the murine system, some rheumatoid factors and anti-DNA antibodies have been shown to share idiotypes (44). The immunologic mechanisms that generate RFs and anti-DNA antibodies share some similarities. Some RFs, like anti-DNA antibodies, have been shown to be derived from germ line V genes (12, 45) included in the set of “natural” antibodies found in the newborn (46) and expressed by the Lyl + /CD5 + B-cell subset (47). Molecular mimicry between specific antigens, such as those found on some bacteria, parasites, or drugs, has been implicated as direct or indirect stimuli for RF and anti-DNA antibody production (48-51). Two divergent hypotheses supported by strong evidence have been put forth to explain the production of anti-DNA antibodies in SLE; one which holds that anti-DNA antibodies are “natural” antibodies derived from germ line V genes from which antigen-specific antibodies are generated (52), and the other which holds that germ line-encoded, antigen-specific V genes that are mutated during the immunologic events, involved in the development of SLE, into autoantibodies that bind DNA (53). Both hypotheses may also hold true for RF expression. Given these similarities, the identification of RF-CR1 coexpression on antiDNA antibodies or other autoantibodies in SLE sera without RF activity should be pursued to determine the relationship of RF-CR1 with autoantibodies other than RFs. Isolation of B-cell clones that bear RF-CR1 that lack RF activity may definitively identify the parallel set of RF-CR1 + antibodies associated with RFs and provide reagents to study these antibodies within the network of idiotypes that control the production of RF and other RF-CRI+ antibodies in the sera of patients with rheumatic disease. ACKNOWLEDGMENTS This work was presented in part at the American College of Rheumatology meeting in Seattle, Washington, 1990, and supported by grants from the SLE Foundation Inc. and the Arthritis Foundation.

REFERENCES 1. Isenberg, D. A.. Dudeney, C., Wojnaruska,

F., Bhogal, B. S., Rauch, J., Schnatter, A., Na-

RF CROSS-REACTIVE

IDIOTYPE

EXPRESSION

IN SLE

241

parstek, Y., and Duggan, D., Detection of cross-reactive anti-DNA antibody idiotypes on tissuebound immunoglobulins from skin biopsies of lupus patients. J. Immunol. 135, 261-264, 1985. 2. Isenberg, D. A., and Collins, C., Detection of cross-reactive anti-DNA antibody idiotypes on renal tissue-bound immunoglobulins from lupus patients. J. Clin. Invest. 76, 287-294, 1985. 3. Hahn, B. H., and Ebling, F. M., Suppression of NZB/NZW mmine nephritis by administration of a syngeneic monoclonai antibody to DNA: Possible role of anti-idiotypic antibodies. J. Clin. Invest. 71, 1728-1736, 1983. 4. Nasu, H., Chia, D. S., Taniguchi, O., and Bamett, E. V., Characterization of the F(ab’), antibodies in SLE patients: Evidence for cross-reacting auto-anti-idiotypic antibodies. Clin. Zmmunol. Immunopathol. 25, 80-90, 1982. 5. Taniguchi, O., Chia, D. S., and Bamett, E. V., Auto-anti-anti-DNA antibodies from SLE patients and normals J. Rheumatol. 11, 291-297, 1984. 6. Agnello, V., Arbetter, A., Ibanez deKasep, G., Powell, R., Tan, E. M., and Joslin F., Evidence for a subset of rheumatoid factors that cross-react with DNA-histone and have a distinct crossidiotype. J. Exp. Med. 151, 1514-1527, 1980. 7. Kunkel, H. G., Angello, V., Joslin, F. G., Winchester, R. J., and Capra, J. D., Cross-idiotypic specificity among monoclonal IgM proteins with anti-gamma-globulin activity. J. Exp. Med. 137, 331-342, 1983. 8. Bonagura, V. R., Kunkel, H. G., and Per&, B., Cellular localization of rheumatoid factor idiotypes. J. Clin. Invest. 69, 1356-1365, 1982. 9. Bonagura, V. R., Pemis, B., Agostino, N., Ilowite, N. T., Hatam, L., and Wedgwood, J. F., The major rheumatoid factor cross-reactive idiotype in rheumatic disease. Int. Rev. Immunol. 5, 139-151, 1989. 10. Bonagura, V. R., Wedgwood, J. F., Agostino, N., Hatam, L., Mendez, L., Jaffe, I., and Pemis, B., Seronegative rheumatoid arthritis, rheumatoid factor cross-reactive idiotype expression, and hidden rheumatoid factors. Ann. Rheum. Dis. 48, 488-495, 1989. 10a. Fotre, O., Doblong, J. H., Michaelson, T. E., and Natvig, J. S., Evidence of similar idiotypic determinants on different rheumatoid factor populations. Stand. J. Immunol. 9, 281-289, 1979. Il. Ilowite, N. T.. Wedgwood, J. F., and Bonagura, V. R., Expression of the major rheumatoid factor cross-reactive idiotype in juvenile rheumatoid arthritis. Arthritis Rheum. 32,265-270, 1989. 12. Chen, P. P., Albrandt, K., Orida, N. K., Roudoux, V., Chen, E. Y., Schrantz, R., Liu, F. T., and Carson, D. A., Genetic basis for the cross-reactive idiotypes on the light chains of human IgM and anti-IgG antibodies. Proc. Natl. Acad. Sci. USA 83, 8318-8322, 1986. 13. Roudoux, V., Chen, P. P., Sorge, J. A., and Carson, D. A., A conserved human germhne V gene directly encodes rheumatoid factor light chains. J. Exp. Med. 164, 2119-2124, 1986. 14. Agnello, V., and Barnes, J. L., Human rheumatoid factor cross-idiotypes. I. WA and BLA are heat-labile conformational antigens requiring both heavy and light chains. J. fip. Med. 164, 1809-1814, 1986. 15. Silverman, G. J., Goldben, R. D., Chen, P., Mageed, R. A., Jefferis, R., Goni, F.. Frangione, B., Fong, S., and Carson, D. A., Idiotypic and subgroup analysis of human monoclonal rheumatoid factors: Implication for structural and genetic basis of autoantibodies in humans. J. Clin. Invest. 82, 469-475, 1988. 16. Tosato, G., and Blaese, R. M., Epstein-Barr virus infection and immunoregulation in man. Adv. Immunol. 37, 99-149, 1985. 17. Klinman, D. M., and Steinberg, A. D., Systemic autoimmune disease arises from polyclonal B-cell activation. J. Exp. Med. 165, 1755-1760, 1987. 18. Oudin, J., and Cazenave, P. A., Similar idiotypic specificities in immunoglobulin fractions with different antibody functions or even without detectable antibody function. Proc. Natl. Acad. Sci. USA 68, 261G2620., 1971. 19. Teitelbaum, D., Rauch, J., Stollar, B. D., and Schwartz, R. S.. In vivo effects of antibodies against a high frequency idiotype of anti-DNA antibodies in MRL mice. J. Immunol. 132, 12821285, 1984. 20. Isenberg, D. A., Schonfeld, Y., Madaio, M. P., Rauch, J., Reichlin, N.. Stollar, B. D., and Schwartz, R. S., Anti-DNA idiotypes in systemic lupus erythematosus. Lancet 2, 417-422, 1984.

242

BONAGURA

ET

AL.

21. Mackworth-Young, C.. and Schwartz, R. S., Autoantibodies to DNA. CRC Crit. Rev. Immunol. 8, 147-173, 1988. 22. Isenberg, D. A., Schoenfeld, Y., Walport, M., Mackworth-Young, C. G., Dudeney, C., ToddPolropek, A.. Brill. S., Weinberger, A., and Pinkas, J., Detection of cross-reactive anti-DNA antibody idiotypes in the serum of systemic lupus erythematosus patients and of their relatives. Arthriris Rheum, 28, 999-1007. 1985. 23. Tan, E. M., Cohen, A. S., Fries, J. F., Masi, A. T., McShane, D. J., Rothtield, N. F., Schaller, J. G., Talal, N., and Winchester, R. J., The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum. 25, 1272-1277, 1982. 24. Rose, H. M., Ragan, C., Pearce, E., and Lipman, M. 0.. Differential agglutination of normal and sensitized sheep erythrocytes by sera of patients with rheumatoid arthritis. Proc. Sot. Exp. Biol. Med. 68a, 1-6, 1948. 25. Singer, J. M., and Plotz, C. M., The latex fixation text, I. Application to the serologic diagnosis of rheumatoid arthritis. Am. J. Med. 21, 882-892, 1956. 26. Faith, A., Pontesilli, O., Unger, A., Panagyi, G. S., and Johns, P., ELISA assays for IgM rheumatoid factors. J. Immunol. Methods 55, 169-177, 1982. 27. Masaam, J., Ferjencik, P., Tempels, M., and Dippell, J., A new method for the detection of hidden IgM rheumatoid factor in patients with juvenile rheumatoid arthritis. J. Rheum. 14,~%7, 1987. 28. Moore, T. L., Domer, R. W.. Weiss, T. D., Baldassare, A. R., and Zucker, J., 19s IgM hidden rheumatoid factor in juvenile rheumatoid arthritis. Pediatr. Res. 14, 1135-l 138, 1980. 29. Stemberg, J. C.. A rate nephelemeter for measuring specific protein by immunoprecpitin reactions. C/in. Chem. 23, 14561464, 1977. 30. Karsh, J., Halbert, S. P.. Anken, M., Klima, E., and Steinberg, A. D., Anti-DNA, antideoxyribonucleoprotein and rheumatoid factor measured in ELISA in patients with systemic lupus erythematosus, Sjogren’s syndrome and rheumatoid arthritis. Int. Arch. Allergy Appl. Immunol. 68, 60-69, 1982. 31. Biologic Substances, International Standards, Reference Preparations and Reference Reagents, pp. ml, World Health Organization, Geneva, 1977. 32. Armitage, P., and Berry, G., “Statistical Methods in Medical Research,” 2nd ed., pp. 411-418, Blackwell, Oxford, 1987. 33. Quismorio, F. P., Other serologic abnormalities in systemic lupus erythematosus. In “Dubois’ Lupus Erythematosus” (D. J. Wallace and E. L. Dubois, Eds.), 3rd ed., Lea & Febiger, Philadelphia, 1987. 34. Hill, G. S., Hinglais, N., Tron, F., and Bach, J. F., Systemic lupus erythematosus: Morphologic correlations with immunologic and clinical data at the time of biopsy. Am. J. Med. 64,61-79, 1978. 35. Rosen, R. D.. Reisberg, M. A., Sharp, J. T., Sucki, W. S., Schroeder, F. X., Hill, L. I., and Eknoyan, G., Antiglobulins and glomerulonephritis: Classification of patients by the reactivity of their sera and renal tissue with aggregated and native IgG. J. C/in. Invest. 56, 427-437, 1975. 36. Alarcon-Segovia, D., Alcocer-Varela, J., and Diaz-Jouaneu, E., The connective tissue diseases as disorders of immune regulation. Clin. Rheum. Dis. 11, 451469, 1985. 37. Mach, P. S., Brouilhet, H., and Amor, B., The induction of human antinuclear antibodies by D-penicillamine: Activation of inducer helper T-cells in the absence of irradiation sensitive suppressor T-cells. C/in. Exp. Immunol. 63, 408, 1986. 38. Mach, P. S., Brouilhet, H., and Amor, B., D-Penicillamine: A modulator of anti-DNA antibody production. Clin. Exp. Immunol. 63, 41H18, 1986. 39. Halpem, R., Schiffenbauer, J., Solomon G., and Diamond, B.. Detection of masked anti-DNA antibodies in lupus by a monoclonal anti-idiotype. J. Zmmunol. 133, 1852-1856, 1984. 40. Tamate, E., Sasaki, T., Muryoi, T., Taki, O., Otani, K.. Tada, K., and Yoshinaga, K., Expression of idiotype on the surface of human B-cells producing anti-DNA antibody. J. tmmunol. 136, 1241-1246, 1986. 41. Sasaki, T., Muryoi, T., Takai, O., Tamate, E., Ono, Y., Koide, Y., Ishida, N., and Yoshinaga, K., Selective elimination of anti-DNA antibody-producing cells by anti-idiotypic antibody conjugated with neocarzinostatin. J. C/in. Invest. 77, 1382-1386, 1986. 42. Pemis, B., and Bonagura. V.. Idiotype regulation and rheumatoid factors. In “Idiotypes” (M. Reichlin and J. D. Capra, Eds.), pp. 31 l-314, Academic Press, Orlando, 1986.

RF CROSS-REACTIVE

IDIOTYPE

EXPRESSION

IN SLE

243

43. Newkirk, M. M., Gram, H., Heinrich, G. H., Ostberg, L., Capra, J. D., and Wasserman, R. L., The complete protein sequence of the variable regions of the cloned heavy and light chains of a human anti-cytomegalovirus antibody reveal a striking similarity to human monoclonal rheumatoid factors of the Wa idiotypic family. J. C/in. Invest. 81, 1511-1518, 1988. 44. Monestier M., Bonin B., Migliorini, P., Dang, H., Datta, S., Kupper, R., Rose, N., Maurer, P., Tahd, N., and Bona, C., Autoantibodies of various specificities encoded by genes from the VH J558 family bind to foreign antigens and shared idiotypes of antibodies specific for self and foreign antigens. J. Exp. Med. 166, 1109-1124, 1987. 45. Kofler, R., Noonan, D. J., Levy, D. E., Wilson, M. C., Moller, N. P., Dixon, F. J., and Theotilopoulos, A. N., Genetic elements used for a murine lupus anti-DNA antibody are closely related to those for antibodies to exogenous antigens. J. Exp. Med. 161, 805-815, 1985. 46. Guilbert, B., Dighiero, G., and Avrameas, S., Naturally occurring antibodies against nine common antigens in human sera. J. Immunol. 128, 2779-2787, 1982. 47. Casah, P., Burastero, S. E., Nakamura, M., Inghirami, G.. and No&ins, A. L., Human lymphocytes making rheumatoid factor and antibody to ssDNA belong to the Leu-l+ B-cell subset. Science 236,77-81, 1987. 48. Nardella, F. A., Teller, D. C., Barber, C. V., and Mannik, M., IgG rheumatoid factors and staphylococcal protein A bind to a common molecular site on IgG J. Exp. Med. 162, 1811-1824, 1985. 49. Hillyer, G. V., Deoxyribonuclear acid (DNA) and antibodies to DNA in the serum of hamsters and man infected with schistosomes. Proc. Sot. Exp. Viol. 136, 880.883, 1971. 50. Vilches, A. R., and Williams, D. G., Persistent anti-DNA antibodies and DNA-anti-DNA complexes in post-streptococcal glomerulonephritis. Clin. Nephrol. 22, 97401, 1984. 51. Yamauchi, Y.. Litwin, A., Adams, L., Zimmer, H., and Hess, E. V., Induction of antibodies to nuclear antigens in rabbits by immunization with hydralazine-human serum albumin conjugates. .I. Clin. Invest. 5, 958-%9, 1975. 52. Manser, T., Huang, S. Y., and Gefter, M. L., Intluence of clonal selection on the expression of immunoglobulin variable region genes. Science 226, 1283-1288, 1984. 53. Diamond, B., and Schartf, M. D., Somatic mutation of the T-15 heavy chain gives rise to an antibody with autoantibody specificity. Proc. Narl. Acad. Sci. USA 81, 5841-5844, 1984. Received October 20, 1990; Accepted with revision March 27, 1991.