Clin. exp. Immunol. (1990) 82, 250-256
Occurrence of two germline-related rheumatoid factor idiotypes in rheumatoid arthritis and in non-rheumatoid seropositive individuals T. KOURI*, J. CROWLEY*, K. AHOt, T. PALOSUOt, H. ISOMAKI$, R. VON ESSENT, M. HELIOVAARA§, D. CARSON* & J. H. VAUGHAN* * Department of Molecular and Experimental Medicine, Research Institute of Scripps Clinic, La Jolla, CA, USA, t National Public Health Institute, Helsinki, I Rheumatism Foundation Hospital, Heinola, and § Social Insurance Institution, Helsinki, Finland
(Acceptedfor publication 1 June 1990)
SUMMARY Human rheumatoid factor (RF) paraproteins express two distinct light chain cross-reactive idiotypes defined by the monoclonal antibodies 17.109 and 6B6.6. These germline gene-related cross-reactive idiotypes are both carried on VK3 light chains and are each present on about one-third of IgM RF paraproteins. We assessed the degree to which these idiotypes are represented in polyclonal RFs. We used rheumatoid arthritis (RA) and non-RA RF-positive sera selected from a large cross-sectional population study (the Mini-Finland Health Survey), and sera from a community-based follow-up study of recent-onset RA patients from Heinola, Finland. In the Mini-Finland Health Survey, elevated levels of the 17.109 RF idiotype were seen in sera of 13% of the RA and 19% of the non-RA group; 6B6.6 RF was seen in 26% of the RA and 28% of the non-RA group. In sera of the Heinola follow-up study, 17.109 RF was seen in 12% initially, but in only 3% at 8 years. Similarly, 6B6.6 RF was detected in 25% initially, but in only 7% at 8 years. Ten sera positive for RF prior to the onset of clinical RA were identified from individuals of a second large population study from Finland (North Karelia project); two of these sera exhibited the 6B6.6 idiotype; none exhibited the 17.109 idiotype. The data are consistent with the concept that these germline gene-related cross-reactive RF idiotypes occur frequently in the polyclonal RF of non-RA as well as RA sera, and that in RA the idiotypes may sometimes be reduced or lost as a consequence of somatic diversification of the RF through somatic mutation, usage of new germline genes, or both.
Keywords rheumatoid factors rheumatoid arthritis idiotypes
In the polyclonal IgM RF of rheumatoid arthritis (RA), the 17.109 idiotype was reported to be present in 25% of patients (two out of eight) and to account for up to 14% of the total IgM RFs present in these sera (Fong et al., 1986). The 6B6.6 idiotype was reported by Schrohenloher & Koopman (1986) to be present in small amounts in the RFs of one-third of RA patients. Most of the RFs in RA sera, however, represent private idiotypes (Nelson et al., 1987). The appearance of RFs in the serum may precede the onset of clinical RA by years (Aho et al., 1985a, 1985b; Walker et al., 1986; Jones & Jacoby, 1989), and the RF titre in the sera of non-RA individuals correlates with the risk to develop RA (del Puente et al., 1988). In fact, about 5% of the RF-positive non-RA subjects eventually proceed to develop RA. The current study was carried out to examine whether the polyclonal RF associated with RA exhibits quantifiable idiotype differences from the polyclonal RF found in apparently healthy non-RA individuals obtained from a cross-sectional
INTRODUCTION IgM rheumatoid factor (RF) paraproteins obtained from patients with mixed cryoglobulinaemia commonly express two distinct light chain cross-reactive idiotypes, defined by the monoclonal antibodies 17.109 and 6B6.6, which are each present in one-third of IgM RF paraproteins (Crowley et al., 1988). The V genes that encode these idiotypes are Humkv325 for 17.109 (Radoux et al., 1986) and probably Humkv328 for 6B6.6 (Chen et al., 1987). These idiotypes are expressed by paraproteins that have no, or only few differences in their amino acid sequences as compared with the germline-deduced sequences. The light chain variable regions of RFs positive for 17.109 belong to VK3b subgroup, whereas those positive for 6B6.6 belong to VK3a (Crowley et al., 1988). Correspondence: Dr Timo Kouri, MD, Department of Clinical Chemistry, Tampere University Hospital, SF-33520 Tampere, Finland.
250
Two rheumatoidfactor idiotypes population study (the Mini-Finland Health Survey); and whether RFs from RA patients early and late in the course of their disease differ in their expression of the idiotypes, using sera from a community-based follow-up study of arthritis patients (from Heinola, Finland). Additionally, some seropositive RA and pre-RA samples obtained from another Finnish crosssectional population study (North Karelia project) were examined. SUBJECTS AND METHODS
Subjects Sera from two large Finnish population studies and a smaller sequential study of RA were available for this investigation. The Mini-Finland Health Survey was a cross-sectional population study comprising 7200 individuals from 40 places of residence, selected to represent the Finnish population, aged 30 years or more (Aho et al., 1989). Twenty-three RA patients seropositive by sensitized sheep cell agglutination, the Waaler-Rose assay (WaRo) at a titre of 250 or more, 36 non-RA healthy individuals seropositive at similar WaRo titres, and 54 seronegative controls (WaRo < 32; latex slide test negative) matched for age, sex and place of residence were obtained. Later, the intermediate group of 40 non-RA and four RA cases with WaRo titres of 64128 was included in our study. The second study was a follow-up study of RA, carried out as a community-based cohort from Heinola, Finland, comprising cases with seropositive and/or erosive arthritis (predominantly polyarthritis) patients. The first serum samples were
obtained within 6 months of the onset of symptoms of arthritis, and later sera were collected at 1, 3 and 8 years (Kaarela, 1985; Tuomi et al., 1988). Seventy-three RA sera of this follow-up study were available. Fifty-nine were positive for RF at least once (IgM RF concentration in ELISA above 98th percentile of the Finnish control population). Three of the seropositive cases had non-erosive disease. The third, smaller study, was the North Karelia cohort, comprised of 10199 individuals (Aho et al., 1985b). Out of these sera were available from 14 patients with already established RA and from 18 individuals who did not have RA, but developed the disease 4 months to 5 years later (on the basis of subsequent records of the Finnish Sickness Insurance Register). Ten of the established RA patients and 10 of the pre-RA individuals were seropositive for RF in the latex slide test at the time of survey. These seropositive RA and pre-RA samples were used in our current studies.
Assays ELISAs were done on 96-well plates (Costar, Cambridge, MA), using human polyclonal IgG (Cohn fraction II) as a precoat (100 pi, 60 ,g/ml concentration in borate-buffered saline (BBS) pH 8-0; incubated overnight at 4-C). The plates were quenched with 10% bovine serum albumin (BSA) diluted in BBS (BSA-BBS solution) for 1 h at room temperature and the test sera were added at 1/100 dilution (in BSA-BBS) overnight at 4°C. Murine monoclonal anti-idiotype antibodies 17.109 (10 pg/ml; with a MOPC 195 isotypic control; Bionetics, Charleston, SC) and 6B6.6 (3 pg/ml; with a MOPC 21 control; Bionetics) were added for 3 h at room temperature. Alkaline phosphatase-conjugated goat anti-mouse IgG (Kierkegaard & Perry, Gaithersburg, MD; 1/1000 in BSA-BBS) was incubated for 1 h at room tempera-
251
ture. As a substrate we used I mg/ml p-nitrophenylphosphate (No. 104-0, Sigma, St Louis, MO), reading the colours after 2 h at room temperature with a Titertek Multiskan reader (Eflab, Helsinki, Finland). Quantification was made using IgM RF paraproteins of known idiotype (Glo for 17.109, and Cor for 6B6.6) as standards. The method of testing idiotypes on RFs described above relies on the fact that pentameric IgM RFs bound to human IgG also leave free binding sites capable of binding to monoclonal antibodies against idiotypic markers. The validity of this measurement was verified using IgM RF paraproteins (Glo for 17.109 and Cor for 6B6.6), which produced reproducible standard surves. The expression of V,3b-bearing RFs was measured by an ELISA using a monoclonal antibody specific for this subgroup (JG-B1) kindly provided by Dr G. Abraham (University of Rochester, Rochester, NY), as reported earlier (Crowley et al., 1988). Data on RF isotypes in these sera have been published (Aho et al., 1988; Tuomi, Palosuo & Aho, 1986; Tuomi et al., 1988; Tuomi, 1989) and were available for this study. For comparison, total IgM RF concentrations were also measured from Heinola RA sera using the Cohn fraction II precoat as above. Test sera were diluted from 1/500 to 1/5000 in BSA-BBS for overnight incubation. Alkaline phosphatase-conjugated goat anti-human IgM (Kierkegaard & Perry; 1/1500 dilution in BSA-BBS) was used as detecting reagent. Quantification was made with the paraproteins indicated above. For clinical correlations, age, sex, number of swollen joints at 8 years, and an X-ray index quantifying the overall radiographic progression of RA at 8 years was available for patients participating in the Heinola follow-up study (Kaarela, 1985).
Statistical analysis Non-parametric statistics were used for comparisons of groups, because of the small number of idiotype-positive sera available: the Mann-Whitney U-test was used to compare unpaired observations (Mini-Finland sera), Wilcoxon's signed ranks test to analyse paired observations (Heinola follow-up study). To estimate correlations, Spedrman's rank coefficient was used. Fisher's exact test was used to evaluate significance of coincidences.
RESULTS 17.109 idiotype-positive rheumatoidfactors The frequencies of occurrence of the idiotype-positive RFs in the studied samples are summarized in Tables 1 and 2. Eleven per cent (six out of 54) of the seronegative (WaRo titre < 32; latex slide test negative) matched controls of the Mini-Finland Health Survey had detectable amounts of 17.109 RF by ELISA, although at significantly lower concentrations (P < 0-01) than the non-RA seropositive individuals with WaRo titre . 250 (Fig. 1). Thus, some control sera, although negative in the standard RF agglutination tests, contained low concentrations of 17.109 RF. We therefore used the concentration at the 95th percentile of these 'seronegative' controls to define positivity as elevated levels of 17.109 RF. This concentration was 7-7 pg/ml. It left out two seronegative control cases (40%) with a median of 8-0 ,g/ml (Table 1). In the Mini-Finland nonRA sera (with a WaRo titre . 250), one case expressing 17.109
252
T. Kouri et al. Table 1. Occurrence of 17.109 rheumatoid factor (RF) and 6B6.6 in the Mini-Finland Health Survey samples Waaler-Rose (WaRo) titre 64-128
Occurrence of two germline-related rheumatoid factor idiotypes in rheumatoid arthritis and in non-rheumatoid seropositive individuals T. KOURI*, J. CROWLEY*, K. AHOt, T. PALOSUOt, H. ISOMAKI$, R. VON ESSENT, M. HELIOVAARA§, D. CARSON* & J. H. VAUGHAN* * Department of Molecular and Experimental Medicine, Research Institute of Scripps Clinic, La Jolla, CA, USA, t National Public Health Institute, Helsinki, I Rheumatism Foundation Hospital, Heinola, and § Social Insurance Institution, Helsinki, Finland
(Acceptedfor publication 1 June 1990)
SUMMARY Human rheumatoid factor (RF) paraproteins express two distinct light chain cross-reactive idiotypes defined by the monoclonal antibodies 17.109 and 6B6.6. These germline gene-related cross-reactive idiotypes are both carried on VK3 light chains and are each present on about one-third of IgM RF paraproteins. We assessed the degree to which these idiotypes are represented in polyclonal RFs. We used rheumatoid arthritis (RA) and non-RA RF-positive sera selected from a large cross-sectional population study (the Mini-Finland Health Survey), and sera from a community-based follow-up study of recent-onset RA patients from Heinola, Finland. In the Mini-Finland Health Survey, elevated levels of the 17.109 RF idiotype were seen in sera of 13% of the RA and 19% of the non-RA group; 6B6.6 RF was seen in 26% of the RA and 28% of the non-RA group. In sera of the Heinola follow-up study, 17.109 RF was seen in 12% initially, but in only 3% at 8 years. Similarly, 6B6.6 RF was detected in 25% initially, but in only 7% at 8 years. Ten sera positive for RF prior to the onset of clinical RA were identified from individuals of a second large population study from Finland (North Karelia project); two of these sera exhibited the 6B6.6 idiotype; none exhibited the 17.109 idiotype. The data are consistent with the concept that these germline gene-related cross-reactive RF idiotypes occur frequently in the polyclonal RF of non-RA as well as RA sera, and that in RA the idiotypes may sometimes be reduced or lost as a consequence of somatic diversification of the RF through somatic mutation, usage of new germline genes, or both.
Keywords rheumatoid factors rheumatoid arthritis idiotypes
In the polyclonal IgM RF of rheumatoid arthritis (RA), the 17.109 idiotype was reported to be present in 25% of patients (two out of eight) and to account for up to 14% of the total IgM RFs present in these sera (Fong et al., 1986). The 6B6.6 idiotype was reported by Schrohenloher & Koopman (1986) to be present in small amounts in the RFs of one-third of RA patients. Most of the RFs in RA sera, however, represent private idiotypes (Nelson et al., 1987). The appearance of RFs in the serum may precede the onset of clinical RA by years (Aho et al., 1985a, 1985b; Walker et al., 1986; Jones & Jacoby, 1989), and the RF titre in the sera of non-RA individuals correlates with the risk to develop RA (del Puente et al., 1988). In fact, about 5% of the RF-positive non-RA subjects eventually proceed to develop RA. The current study was carried out to examine whether the polyclonal RF associated with RA exhibits quantifiable idiotype differences from the polyclonal RF found in apparently healthy non-RA individuals obtained from a cross-sectional
INTRODUCTION IgM rheumatoid factor (RF) paraproteins obtained from patients with mixed cryoglobulinaemia commonly express two distinct light chain cross-reactive idiotypes, defined by the monoclonal antibodies 17.109 and 6B6.6, which are each present in one-third of IgM RF paraproteins (Crowley et al., 1988). The V genes that encode these idiotypes are Humkv325 for 17.109 (Radoux et al., 1986) and probably Humkv328 for 6B6.6 (Chen et al., 1987). These idiotypes are expressed by paraproteins that have no, or only few differences in their amino acid sequences as compared with the germline-deduced sequences. The light chain variable regions of RFs positive for 17.109 belong to VK3b subgroup, whereas those positive for 6B6.6 belong to VK3a (Crowley et al., 1988). Correspondence: Dr Timo Kouri, MD, Department of Clinical Chemistry, Tampere University Hospital, SF-33520 Tampere, Finland.
250
Two rheumatoidfactor idiotypes population study (the Mini-Finland Health Survey); and whether RFs from RA patients early and late in the course of their disease differ in their expression of the idiotypes, using sera from a community-based follow-up study of arthritis patients (from Heinola, Finland). Additionally, some seropositive RA and pre-RA samples obtained from another Finnish crosssectional population study (North Karelia project) were examined. SUBJECTS AND METHODS
Subjects Sera from two large Finnish population studies and a smaller sequential study of RA were available for this investigation. The Mini-Finland Health Survey was a cross-sectional population study comprising 7200 individuals from 40 places of residence, selected to represent the Finnish population, aged 30 years or more (Aho et al., 1989). Twenty-three RA patients seropositive by sensitized sheep cell agglutination, the Waaler-Rose assay (WaRo) at a titre of 250 or more, 36 non-RA healthy individuals seropositive at similar WaRo titres, and 54 seronegative controls (WaRo < 32; latex slide test negative) matched for age, sex and place of residence were obtained. Later, the intermediate group of 40 non-RA and four RA cases with WaRo titres of 64128 was included in our study. The second study was a follow-up study of RA, carried out as a community-based cohort from Heinola, Finland, comprising cases with seropositive and/or erosive arthritis (predominantly polyarthritis) patients. The first serum samples were
obtained within 6 months of the onset of symptoms of arthritis, and later sera were collected at 1, 3 and 8 years (Kaarela, 1985; Tuomi et al., 1988). Seventy-three RA sera of this follow-up study were available. Fifty-nine were positive for RF at least once (IgM RF concentration in ELISA above 98th percentile of the Finnish control population). Three of the seropositive cases had non-erosive disease. The third, smaller study, was the North Karelia cohort, comprised of 10199 individuals (Aho et al., 1985b). Out of these sera were available from 14 patients with already established RA and from 18 individuals who did not have RA, but developed the disease 4 months to 5 years later (on the basis of subsequent records of the Finnish Sickness Insurance Register). Ten of the established RA patients and 10 of the pre-RA individuals were seropositive for RF in the latex slide test at the time of survey. These seropositive RA and pre-RA samples were used in our current studies.
Assays ELISAs were done on 96-well plates (Costar, Cambridge, MA), using human polyclonal IgG (Cohn fraction II) as a precoat (100 pi, 60 ,g/ml concentration in borate-buffered saline (BBS) pH 8-0; incubated overnight at 4-C). The plates were quenched with 10% bovine serum albumin (BSA) diluted in BBS (BSA-BBS solution) for 1 h at room temperature and the test sera were added at 1/100 dilution (in BSA-BBS) overnight at 4°C. Murine monoclonal anti-idiotype antibodies 17.109 (10 pg/ml; with a MOPC 195 isotypic control; Bionetics, Charleston, SC) and 6B6.6 (3 pg/ml; with a MOPC 21 control; Bionetics) were added for 3 h at room temperature. Alkaline phosphatase-conjugated goat anti-mouse IgG (Kierkegaard & Perry, Gaithersburg, MD; 1/1000 in BSA-BBS) was incubated for 1 h at room tempera-
251
ture. As a substrate we used I mg/ml p-nitrophenylphosphate (No. 104-0, Sigma, St Louis, MO), reading the colours after 2 h at room temperature with a Titertek Multiskan reader (Eflab, Helsinki, Finland). Quantification was made using IgM RF paraproteins of known idiotype (Glo for 17.109, and Cor for 6B6.6) as standards. The method of testing idiotypes on RFs described above relies on the fact that pentameric IgM RFs bound to human IgG also leave free binding sites capable of binding to monoclonal antibodies against idiotypic markers. The validity of this measurement was verified using IgM RF paraproteins (Glo for 17.109 and Cor for 6B6.6), which produced reproducible standard surves. The expression of V,3b-bearing RFs was measured by an ELISA using a monoclonal antibody specific for this subgroup (JG-B1) kindly provided by Dr G. Abraham (University of Rochester, Rochester, NY), as reported earlier (Crowley et al., 1988). Data on RF isotypes in these sera have been published (Aho et al., 1988; Tuomi, Palosuo & Aho, 1986; Tuomi et al., 1988; Tuomi, 1989) and were available for this study. For comparison, total IgM RF concentrations were also measured from Heinola RA sera using the Cohn fraction II precoat as above. Test sera were diluted from 1/500 to 1/5000 in BSA-BBS for overnight incubation. Alkaline phosphatase-conjugated goat anti-human IgM (Kierkegaard & Perry; 1/1500 dilution in BSA-BBS) was used as detecting reagent. Quantification was made with the paraproteins indicated above. For clinical correlations, age, sex, number of swollen joints at 8 years, and an X-ray index quantifying the overall radiographic progression of RA at 8 years was available for patients participating in the Heinola follow-up study (Kaarela, 1985).
Statistical analysis Non-parametric statistics were used for comparisons of groups, because of the small number of idiotype-positive sera available: the Mann-Whitney U-test was used to compare unpaired observations (Mini-Finland sera), Wilcoxon's signed ranks test to analyse paired observations (Heinola follow-up study). To estimate correlations, Spedrman's rank coefficient was used. Fisher's exact test was used to evaluate significance of coincidences.
RESULTS 17.109 idiotype-positive rheumatoidfactors The frequencies of occurrence of the idiotype-positive RFs in the studied samples are summarized in Tables 1 and 2. Eleven per cent (six out of 54) of the seronegative (WaRo titre < 32; latex slide test negative) matched controls of the Mini-Finland Health Survey had detectable amounts of 17.109 RF by ELISA, although at significantly lower concentrations (P < 0-01) than the non-RA seropositive individuals with WaRo titre . 250 (Fig. 1). Thus, some control sera, although negative in the standard RF agglutination tests, contained low concentrations of 17.109 RF. We therefore used the concentration at the 95th percentile of these 'seronegative' controls to define positivity as elevated levels of 17.109 RF. This concentration was 7-7 pg/ml. It left out two seronegative control cases (40%) with a median of 8-0 ,g/ml (Table 1). In the Mini-Finland nonRA sera (with a WaRo titre . 250), one case expressing 17.109
252
T. Kouri et al. Table 1. Occurrence of 17.109 rheumatoid factor (RF) and 6B6.6 in the Mini-Finland Health Survey samples Waaler-Rose (WaRo) titre 64-128
