Detection of serum soluble markers of immune activation in rheumatoid arthritis

The mutual correlation among soluble CD4 (sCD4), soluble CD8 (sCD8), and soluble CD23 (sCD23) has not yet been studied in patients with rheumatoid arthritis (RA), although previous studies have demonstrated that certain soluble markers of immune activation are elevated in RA. Thus, we examined this correlation based on the serum levels of sCD4, sCD8 and sCD23, and that of their levels with other serum markers such as immunoglobulin (Ig) subtypes (IgG, IgM and IgA), IgM-rheumatoid factor (IgM-RF) and C-reactive protein (CRP) in 25 RA patients, sCD4 was not elevated, whereas both sCD8 and sCD23 increased in RA patients compared with the healthy controls; a majority of RA patients, in particular, showed a high sCD23 level. The level of sCD23 showed a correlation with that of IgM-RF, but not with those of IgG, IgM, IgA and CRP. Importantly, a high level of sCD23 was not always accompartied with that of sCD8. The independent change between sCD23 and sCD8 levels was also observed in a one-year follow-up study of the two RA patients. These findings indicate that B cells might be generally activated in RA, whereas T-cell activation in variable in each patient with RA, suggesting that sCD23 is a more indicative marker for the immune status of RA patients than sCD8 from the clinical aspects.

THE mutual correlation among soluble CD4 (sCD4), soluble CD8 (sCD8), and soluble CD23 (sCD23) has not yet been studied in patients with rheumatoid arthritis (RA), although previous studies have demonstrated that certain soluble markers of immune activation are elevated in RA. Thus, we examined this correlation based on the serum levels of sCD4, sCD8 and sCD23, and that of their levels with other serum markers such as immunoglobulin (Ig) subtypes (IgG, IgM and IgA), IgM-rheumatoid factor (IgM-RF) and C-reactive protein (CRP) in 25 RA patients, sCD4 was not elevated, whereas both sCD8 and sCD23 increased in RA patients compared with the healthy controls; a majority of RA patients, in particular, showed a high sCD23 level. The level of sCD23 showed a correlation with that of IgM-RF, but not with those of IgG, IgM, IgA and CRP. Importantly, a high level of sCD23 was not always accompartied with that of sCD8. The independent change between sCD23 and sCD8 levels was also observed in a one-year follow-up study of the two RA patients. These findings indicate that B cells might be generally activated in RA, whereas T-cell activation in variable in each patient with RA, suggesting that sCD23 is a more indicative marker for the immune status of RA patients than sCD8 from the clinical aspects.

Introduction
Rheumatoid arthritis (RA) is a complex inflammatory disease of unknown cause and an autoimmune phenomenon is strongly suspected. 1'2 Previous studies have described the significant role of both T-cells and pro-inflammatory cytokines such as interleukin (IL)-I, IL-6 and tumour necrosis factor-z (TNF-00 in RA. -5 In addition, B cells in the rheumatoid synovium are believed to produce the rheumatoid factor (RF), of which controlled mechanisms are still unclear. 6 Similar to systemic lupus erythematosus, serum soluble CD8 (sCD8), soluble CD23 (sCD23) and soluble IL-2 receptor have been shown to become elevated in RA, [7][8][9] Because these molecules are recognized as being involved in Tand B-cell functions, 1'11 it is likely that the immune activation contributes in part to the pathogenesis of RA. However, a mutual correlation among the serum levels of these molecules has not yet been clarified in RA. Thus, we examined this issue based on the serum levels of sCD4, sCD8, and sCD23/(low-affinity receptor for the Fc portion of IgE), and their correlation with .other serum markers such as immunoglobulin (Ig) subtypes (IgG, IgM and IgA), IgM-rheumatoid factor (IgM-RF) and C-reactive protein (CRP) in addition to erythrocyte segmentation rate (ESR) in 25 RA patients. Furthermore, a longitudinal comparison in the changes in the serum levels of certain soluble markers was performed.

Materials and Methods
Twenty-five patients (19 females and six males) with flares of RA as defined by the revised criteria of the American College of Rheumatology 12 were included in this study. Characteristics of these patients were as follows: mean _ SD age 58.  Serum levels of the soluble markers were measured by antibody sandwich enzyme-linked immunosorbent assay (ELISA) using commercial kits (sCD8 and sCD23: T-cell Diagnostics, Inc., Cambridge, MA; sCD4: Boehringer Mannheim Biochemica, Germany). These levels were expressed as U/ml except for sCD4 which was expressed as ng/ml based on the standard curve prepared simultaneously using standard molecules provided in each kit. The detection limits B were 0.25 ng/ml, 50 U/ml and 15 U/ml for sCD4, sCD8 and sCD23, respectively. Serum samples were stored at-20C until use. All data were collected in a computer database and analysed using Spearman's rank correlation in the StatView-J 4.02 program. and sCD8 were measured simultaneously in the sera which were stored at -20C.

Results and Discussion
As shown in Fig. 1, the serum level of sCD4 showed no significant difference between RA patients and healthy controls, but that of sCD23 was significantly higher in all RA patients compared with controls. Concerning sCD8, these values were distributed in a wider range with a tendency that the majority of RA patients showed higher ones which were, as a whole, statistically no different from the controls. When the correlation of the level of sCD23 with that of each Ig subtype, CRP, or ESR was examined (Table 1), no correlation was obtained between sCD23 and each Ig subtype, or between sCD23 and CRP. However, an insignificant but considerable correlation was shown between sCD23 and ESR. Considering the fact that CRP is well known as a more acute phase reactant than ESR, this finding suggests that the serum level of sCD23 might be taken as a parameter reflecting the immune activation in the chronic phase rather than the acute phase. On the other hand, the correlation between serum levels of sCD23 and IgM-RF was statistically significant (Fig. 2A). These data confirm that sCD23 might play an important role on the specific production of IgM from RF-committed B cells.
A recent study has revealed that the sCD23 264 Mediators of Inflammation Vol 5 1996 molecule is not only a soluble activation marker, but also a proinflammatory cytokine achieving an important role in the control of the immune response including T-cell response. Thus, we further investigated the relationship between sCD23 and sCD8. As shown in Fig. 2, a correlation was not shown between these two molecules, indicating that B cells might be activated. generally in RA, whereas the degree of T-cell activation varies patient by patient in RA. Thus, it is likely that a higher level of sCD23 in RA patients does not always lead to the T-cell activation. The analysis of the relation between sCD23 and sCD4 was not performed, because the normal level of sCD4 was observed in RA (see Fig. 1). We further performed a longitudinal study on the changes in the serum sCD23, sCD8, and IgM-RF levels in addition to classical inflammation parameters (CRP and ESR) in two patients during one year. The results are shown in Fig. 3. In both cases, the serum levels of sCD23 were changing almost in parallel with that' of IgM-RF.
Comparing both cases, a changing pattern of sCD23 was correlated more closely to that of ESR than CRP. In contrast, time-related change of the serum level of sCD8 was also independent from that of sCD23 in this study as shown in the cross-sectional study described above. These observations might permit speculation that sCD23, but not sCD8, is taken as a more indicative marker reflecting the immune status of RA patients either in cross-section or longitudinal study.