- Email: [email protected]
Humoral immune function in severe, active rheumatoid arthritis
CLINICAL
IMMUNOLOGY
AND
Humoral
IMMUNOPATHOLOGY
43, 185-194 (1987)
Immune Function in Severe, Active Rheumatoid Arthritis
E. P. BOLING,* T. OHISHI,$ S. M. WAHL,$ J. MISITI,? R. WISTAR, JR.,? AND R. L. WILDERII,~ *Department of Medicine, Malcolm Grow USAF Medical Center, Washington, DC; l-The Naval Medical Research Institute, Bethesda, Maryland, and $Metabolism Branch, National Cancer Institute, #Cellular Immunology Section, National Institute of Dental Research, and “Arthritis and Rheumatism Branch, National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health. Bethesda, Maryland 20892
Peripheral blood mononuclear cells (PBMC) from 30 patients with definite or classic active rheumatoid arthritis who were on no remittive drugs were studied for spontaneous and pokeweed mitogen (PWM)-stimulated immunoglobulin plaque-forming cell frequency (IgPFC), spontaneous IgM-rheumatoid factor (IgM-RF) secretion, and in vitro proliferative responses to soluble recall antigens. Rheumatoid spontaneous total (IgG + IgM + IgA) IgPFCs were higher than those of normal controls when assayed after 7 days in culture. Spontaneous and PWM-stimulated IgM-PFCs, in contrast, were significantly less than normal regardless of when assayed. Spontaneous synthesis of IgM-RF was observed in 56% of the RA patients, but absolute amounts produced were widely heterogeneous. Spontaneous IgM-RF production by RA PBMC was associated with low or absent spontaneous IgM-PFC production. Moreover, a strong association was found between the median amount of IgM-RF secreted and depressed proliferative responses to soluble recall antigens. Our results define several abnormalities of immunoglobulin production in a clinically homogeneous and highly active rheumatoid population and delineate methodologic variations that can complicate the interpretation of similar data in the literature. In addition, our findings suggest that subgroups of rheumatoid patients that show distinct cellular and humoral immune abnormalities can be identified. o 1987 Academic
Press. Inc.
INTRODUCTION
Rheumatoid arthritis (RA) is often associated with serum rheumatoid factors and increased total serum immunoglobulin (l-3). The significance of these humoral abnormalities, their possible relationship to T-lymphocyte abnormalities in RA, and the extent of their involvement in disease pathogenesis all await further clarification. Several recent studies have evaluated the production of polyclonal and/or rheumatoid factor immunoglobulins by rheumatoid peripheral blood mononuclear cells (PBMCs) by enumeration of antibody plaque-forming cells (PFCs) (4-8) and quantitation of in vitro immunoglobulin production by immunoassay (4, 9-14). The total numbers of antibody PFCs and/or the total amount of immunoglobulin produced by rheumatoid PBMCs in these studies have been reported as low (4, 12), normal (8), or high (5-7). Similar discrepant results have been noted for each of the immunoglobulin isotypes when separately measured. 1 To whom requests for reprints should be addressed. 185 0090- 1229/87 $1.50 Copyright 0 1987 by Academic Press. Inc. All rights of reproduction in any form reserved.
186
BOLING
ET AL.
Moreover, although rheumatoid PBMCs as a group are known to spontaneously secrete IgM-rheumatoid factor in culture, considerable heterogeneity has been shown (2, 15) and the relationship of such synthesis to disease activity continues to be debated (13, 14). Interpretation of the above studies has been complicated by the inclusion of patients who were heterogeneous with respect to the amount of disease activity, the presence of rheumatoid factor positivity, and the concomitant use of “anti-rheumatic” drugs. We therefore quantitated numbers of spontaneous and pokeweed mitogen-stimulated immunoglobulin plaque-forming cells in rheumatoid PBMCs and measured IgM-rheumatoid factor production by an enzyme-linked immunosorbent assay in order to define what abnormalities of humoral immune function are present in a population of severely active seropositive RA patients who had been on no remittive therapy for at least 4 weeks at the time of study. In addition to the humoral defects, multiple abnormalities of cell-mediated immune function occur in RA (1, 3) and T-lymphocyte dependence of IgM-rheumatoid factor production has been shown in vitro and in vivo. Thus, an association between abnormal cell-mediated immune function and altered rheumatoid factor synthesis might be identified. Prior studies from our laboratories have defined a subgroup of rheumatoid patients whose PBMCs show low to absent (i.e., anergic) in vitro proliferative responses to soluble recall antigens (3, 16- 18). We therefore investigated whether any of the heterogeneity of spontaneous IgM-rheumatoid factor production by untreated, clinically active rheumatoid PBMCs could be associated with the previously described abnormalities in cell-mediated immune function. The present report demonstrates that clinically active RA patients as a group show (1) increased total spontaneous immunoglobulin plaque-forming cell frequency (IgPFC), (2) decreased spontaneous and PWM-stimulated IgM-PFC, and (3) overall increased spontaneous IgM-rheumatoid factor production. Moreover, the highest spontaneous rheumatoid factor production was found in the anergic rheumatoid patients. MATERIALS
AND METHODS
Patients and controls. The patient group consisted of 30 individuals including 19 females and 11 males ranging in age from 34 to 70 years (mean = 54.7 years). All patients had classic seropositive rheumatoid arthritis as detined by American Rheumatism Association criteria (19). None took drugs other than nonsteroidal antiinflammatory agents and/or prednisone (< 10 mg/day) at the time of study, and all cytotoxic and long-acting anti-rheumatic drugs had been discontinued for at least 4 weeks. All had active disease as defined by a Ritchie-Camp articular index (20) of ~40, with sustained disease activity during the 2 months before study. Prednisone was not administered 24 hr prior to obtaining blood. Controls were healthy laboratory and office personnel, consisting of 30 individuals and including 15 females and I5 males ranging in age from 24 to 65 years (mean = 42.9 years). Determination of total immunoglobulin and IgM-secreting cells by reverse hemolytic plaque assay. PBMCs were obtained from heparinized blood by centrifu-
gation on Ficoll-Hypaque
gradients.
Washed PBMC
were suspended in RPM1
HUMORAL
IMMUNE
FUNCTION
IN RHEUMATOID
ARTHRITIS
187
1640 with glutamine (4 m&f), streptomycin (50 pg/ml), penicillin (50 units/ml), and 10% heat-inactivated fetal bovine serum (Microbiological Associates, Rockville, MD). PBMCs (3 x 10s) in 1 ml of medium were plaqued either directly (Day 0) or after having been cultured for 7 days (Day 7) in 12 x 75-mm polypropylene tubes (Falcon, Oxnard, CA; No. 2063) with or without a 1:lOO dilution of pokeweed mitogen (PWM, GIBCO, Chagrin Falls, OH, Lot llK3413). A reverse hemolytic plaque-forming cell assay (RHPA) was used to detect immunoglobulin-secreting cells as previously described (21, 22). Briefly, 125 ~1 of the above cell suspension was added to 96-well flat-bottom plates and mixed with 25 ~1 of sheep red blood cell (SRBC)-absorbed guinea pig complement, 25 ~1 of 30% protein A-SRBC, and 25 t.~l of either rabbit anti-human Ig antiserum (Cappel Laboratories, used at 1:40 dilution) or rabbit anti-human IgM (Cappel Laboratories, 1:40 dilution). After a 2-hr incubation at 37”C, plaques were counted and expressed as the number of immunoglobulin-secreting cells/lo6 PBMCs cultured. Assay of ZgM-rheumatoid factor (ZgM-RF) produced in supernatants of cultured PBMCs. PBMCs were cultured in 35-mm dishes at 5 x lo6 PBMUculture
with 10% hypogammaglobulinemic human AB serum in a 2-ml volume of RPM1 1640 with penicillin (100 units/ml), streptomycin (100 kg/ml), L-glutamine (2 mM), and 2-mercaptoethanol(5 x 10es iV). After 10 days, supernatants were harvested and frozen at - 20°C until assayed for IgM-RF. Rheumatoid factor determination was done on multiple dilutions of supernatant by an enzyme-linked immunosorbent adaptation of a radioimmunoassay method previously described (15). Ninety-six-well Dynatech Immunolon II plates were coated with 100 ~1 per well of 10 p.g/ml polyclonal human IgG (Cohn fraction II, Red Cross, Bethesda, MD) in normal saline, blocked with 200 yl per well of borate-buffered saline (BBS) containing 5 g/liter bovine serum albumin (BSA), and rinsed three times with washing buffer (BBS with 0.1 g/liter BSA) before 100 ~1 of supernatant was applied to the plate. After incubation for 1 hr at 25”C, the plates were washed, and 100 ~1 of alkaline phosphatase-conjugated goat anti-human IgM (Tago, Burlingame, CA, used at a dilution of 1:2000) was applied for 1 hr. The plates were washed, p-nitrophenyl phosphate substrate was applied (1 mg/ml in diethanolamine buffer), and after 30 min optical density at 405 nm was assessed with a Dynatech Model 580 ELISA reader. With each assay a series of dilutions of purified Wallenberg idiotype monoclonal IgM-K rheumatoid factor (23, 24) (courtesy of Dr. Gerald Penn) was included so that the color developed by the supernatants could be related to a known concentration of rheumatoid factor. PBMC proliferative responses. Mononuclear cell proliferative responses were determined as previously described in detail (16). Briefly, 2 x lo* PBMCs were cultured in 96-well flat-bottom plates in the presence or absence of a variety of soluble recall antigens (PPD, streptokinase-streptodornase, tetanus toxoid, and Candida). Six days later, the cells were pulsed for 4 hr with tritiated thymidine ([3H]TdR, 6.7 Ci/mmol, Schwarz/Mann, Orangeburg, NY) and harvested, and the incorporated activity was measured. A stimulation index was determined by dividing [3H]TdR incorporated (counts per minute) in cultures with antigen by the [3H]TdR incorporation in unstimulated cultures. A stimulation index of <2 for all antigens tested was used to define the anergic subgroup, whereas an index of >2
188
BOLING
ET AL.
to at least one antigen was considered normal or nonanergic. PBMCs from virtually all normal control individuals respond to at least one antigen with an index of >2. Leukupheresis procedure. Three rheumatoid arthritis patients were selected for leukapheresis treatment based upon their anergic immunologic status as previously described (16). Leukapheresis (LP) was carried out as described by Karsh et al. (25). Statistics. Statistical probabilities were assigned by performing the Wilcoxon rank sum test for independent or for paired samples or by the Fisher exact test in comparisons involving small numbers of samples. RESULTS Total Immunoglobulin Plaque-Forming Cells
Total spontaneous IgPFCs from rheumatoid PBMC were enumerated immediately postcollection and compared to similarly assayed normal controls. Spontaneous total IgPFC assayed on Day 0 were significantly higher than those of normal controls (P < 0.01, Fig. la). However, we also enumerated total IgPFCs 7 days after culture initiation and found significant depression in the number of 5O.cQO - la,
b)
20,000
10,000
g a is
5ooo
K P
2000 -
<
low-
E
*
-
T
* .
. i
500 -
f I
T . :
. . .
.
.
AA
N
z
s
3PONTANEOU:
RA
N
+ PWM
FIG. 1. Circulating total immunoglobulin plaque-forming cells (IgPFC) in active rheumatoid (RA) and normal control (N) peripheral blood mononuclear cells (PBMC). (a) Total IgPFC enumerated immediately after PBMC collection on Day 0. (b) and (c) Total IgPFC after 7 days in culture in the absence or presence of pokeweed mitogen (PWM), respectively. Spontaneous IgPFC on Day 0 are increased in RA PBMC relative to N (P < 0.01). IgPFC in Day 7 RA PBMC are decreased relative to N whether assayed in the absence (P < 0.01) or in the presence (P < 0.01) of PWM. Statistical probabilities were calculated by the Wilcoxon rank sum test for independent samples.
HUMORAL
IMMUNE
FUNCTION
IN RHEUMATOID
ARTHRITIS
189
spontaneous (Fig. lb, P < 0.01) and PWM-stimulated (Fig. lc, P < 0.01) total IgPFCs in rheumatoid patients relative to normal controls. Moreover, the numbers of spontaneous total IgPFCs in individual RA patients decreased significantly over the I-/-day culture period (P < 0.05), whereas concurrently run normal PBMCs increased during the same period (P < 0.05). ZgM Plaque-Forming
Cells (ZgM-PFCs)
In contrast to the increase in spontaneous total IgPFCs found in immediately plaqued rheumatoid PBMCs (Fig. la), spontaneous Day 0 IgM-PFCs were significantly depressed (Fig. 2a, P < 0.01). Similar profound depression of IgM-PFCs was seen following 7 days in culture in the absence (Fig. 2b, P < 0.01) or presence (Fig. 2c, P < 0.01) of pokeweed mitogen. Furthermore, it was noted that, under all conditions examined, a subpopulation of RA patients existed whose PBMC showed little or no capacity to make detectable IgM-PFCs (Figs. 2a, b, c).
5000
- la)
,Ibi
Y
Cl :
: . *
. 2000
:
-
:.
i . .
: i
1000
7
t 500 -
.
Y P 8 -
200 -
.
Li .z E
loo-
* .
-
* .
ELI. I SPONTANEOU
.. . *
T
. .
:
!.
.
. -
P RA
s N
SPONTANEOU’
. .
.
-+----
N + PWN
-I I
FIG. 2. Circulating IgM-plaque-forming cells (IgM-PFC) in active RA and normal control (N) peripheral blood mononuclear cells (PBMC). (a) IgM-PFC enumerated immediately after PBMC collection on Day 0. (b) and (c) IgM-PFC after 7 days in culture in the absence or presence of PWM, respectively. Spontaneous IgM PFC on Day 0 are decreased in RA PBMC relative to N (P < 0.01). IgM-PFC on Day 7 are decreased relative to N whether assayed after spontaneous culture (P < 0.01) or in the presence (P < 0.01) of PWM. Statistical probabilities were calculated by the Wilcoxon rank sum test for independent samples.
190
BOLING
ET AL.
Zn Vitro ZgM-RF Production and Its Association with T-Lymphocyte Function
We found that 56% of unstimulated RA PBMC spontaneously produced measurable IgM-RF during in vitro culture in contrast with none of the controls (Fig. 3a). This prevalence of spontaneous IgM-RF secretion was consistent with that described in other reports (2, 15) but stood in sharp contrast to the overall decrease in polyclonal IgM-PFCs seen in these same cells (Figs. 2a, b). We therefore compared IgM-PFCs in rheumatoid PBMCs which either spontaneously secreted, or failed to secrete, IgM-RF in vitro. As shown in Table 1, we found that spontaneous secretors of IgM-RF were significantly more likely to show low (i.e., <20/106 PBMCs) numbers of IgM-PFCs than were nonsecretors (P = 0.028, Fisher exact test). In rheumatoid patients with low IgM-PFCs, 73% (8/l 1) spontaneously secreted IgM-RF, in contrast to 12% (l/8) of patients with normal IgMPFCs. Furthermore, we noted that extremely high production of IgM-RF was limited to PBMCs from a previously characterized subset of anergic patients (Fig. (bl .
. .
2 -P
1000 t .
. .
.
.
+ AN
NA RA
FIG. 3. Spontaneous IgM-rheumatoid factor (&M-RF) production in 1 l-day supernatants (SN) of cultured peripheral blood mononuclear cells (PBMC). IgM-RF is quantitatively expressed relative to binding of a purified monoclonal IgM-RF (Wallenberg). (a) The IgM-RF in SN of RA and normal (N) cultures; (b) anergic (AN) and nonanergic (NA) rheumatoid PBMC cultured. IgM-RF levels in RA PBMC-SN are higher than N-SN, but are heterogeneous. IgM-RF levels in anergic PBMC-SN are significantly higher than in nonanergic PBMC-SN (P < 0.01). Statistical probabilities were calculated by the Wilcoxon rank sum test for independent samples.
HUMORAL
IMMUNE
FUNCTION
IN RHEUMATOID
ARTHRITIS
191
3b). All four rheumatoid patients whose PBMCs produced ~1000 rig/ml of IgMRF in culture were anergic (median IgM-RF level, 2100 r&ml). In contrast, none of 18 nonanergic RA patients produced in vitro IgM-RF factor in these amounts (median IgM-RF level of all nonanergic samples ~15 &ml; median level of the IgM-RF-positive nonanergic samples, 210 rig/ml (Fig. 3b, P < 0.001)). Ten out of 18 nonanergic patients had no spontaneous elaboration of IgM-RF in vitro, in contrast to the high level production by anergic patients. The spontaneous IgMRF production did not appear to be correlated with clinical parameters such as corticosteroid usage or dose or serum RF titers (data not shown). Thus, these data support the notion that high level spontaneous IgM-RF production by PBMCs is associated with abnormal T-cell function. Because prior work has shown that anergic rheumatoid T-cell responses tend to normalize with serial leukapheresis therapy (16), we evaluated PBMC T-cell proliferative responses to soluble recall antigens in three anergic RA patients prior to and at the end of leukapheresis therapy and compared these values to concurrently measured in vitro IgM-rheumatoid factor production (Fig. 4). Supernatants of PBMCs from the three anergic patients examined showed a drop in spontaneously produced IgM-rheumatoid factor concurrent with a normalization of proliferative responses to one or more soluble antigens. Thus, these findings further support the view that hyporesponsive rheumatoid T-cell function and high level in vitro IgM-rheumatoid factor production are related. DISCUSSION
Because several recent studies have reported conflicting results regarding immunoglobulin production and RF synthesis in RA, we analyzed polyclonal Ig-PFC and spontaneous IgM-RF production in PBMCs of clinically homogeneous, severely active seropositive rheumatoid patients who were without disease-modifying drugs for a minimum of 4 weeks. We found that total (i.e., IgG + IgA + IgM) spontaneous immunoglobulin plaque-forming cells are elevated in freshly collected rheumatoid PBMCs. This finding is in agreement with most (5-7), but not all (4), previous reports and reinforces the concept that there are increased numbers of activated B cells in the rheumatoid patient’s peripheral blood compartment. In contrast, several recent studies using PFC (4) or quantitative immunoassay methods (4, 8, 12) have reported the apparently contradictory TABLE 1 NUMBERSOFRHEUMATOIDPATIENTSWITH Low(<20PFC1106PBMC) ORNORMAL (>20PFC/106 PBMC)IgM PFC THATSPONTANEOUSLYSECRETEORFAILTOSECRETE IgM-RF in Vitro No spontaneous IgM-RF production (< 15 rig/ml) Low IgM-PFC producers Normal IgM-PFC producers Total
3 7 10
Spontaneous IgM-RF production (> I.5 rig/ml) 8 1 9
Total 11 8 19
Note. RA patients with low IgM-PFC numbers (after ‘I-day spontaneous culture) were significantly more likely to spontaneously secrete IBM-RF in vitro (P = 0.028 by the Fisher exact test).
192
BOLING (al Patlent
ET AL.
/ Patient
s.s
R.B
1 Patient
D.K.
IgM-RF
Inglmll
2000 1000 500
Proliferation Index
1.02
t LPl
3.07
0.72
3.85
t
t
t
t
t
t
t
t
LP8
LPlO
LPI
LP3
LP6
LPl
LP3
LP6
FIG. 4. Spontaneous IgM-rheumatoid factor (IgM-RF) production versus maximum soluble recall antigen proliferative responses by peripheral blood mononuclear cells in three anergic RA patients (S.S., R.B., D.K.) undergoing serial leukapheresis treatments. Soluble recall antigen proliferative responses (Proliferation Index) from two separate leukaphereses are shown for each patient.
findings of decreased total IgPFC (4) and decreased (4, 12) or normal (8) quantitative immunoglobulin production in vitro. In this report, we have shown that the number of total immunoglobulin plaque-forming cells detected in rheumatoid PBMCs falls significantly with in vitro culture (Figs. la, b). This points to a probable methodologic explanation for some of the previously reported total IgPFC differences and suggests that one possible cause for the disparity between reported plaque-forming cell and immunoglobulin synthesis results may be a dropout of B cells that have been already activated in vivo. An important new finding in our study was the depression of IgM-PFCs in rheumatoid PBMC. This finding occurred under a variety of different culture conditions and was unexpected since some prior reports had shown normal (6) or increased (7) numbers of IgM-PFCs. However, our results are consistent with several previous studies which used immunoassay methods to show significantly diminished (4, 10) or normal (15) IgM production by rheumatoid PBMC in vitro. Furthermore, we noted a subgroup of consistently hyporesponsive patients (Figs. 2a, b, c) which could not be attributed to immunomodulating drug effects. The finding that spontaneous IgM-RF secretion in vitro was significantly more prevalent in this IgM-hyporesponsive subgroup is intriguing and suggests that these two phenomena may be causally related. It is possible that IgM-RF, complexing with either auto-IgG or antigen-IgG aggregates, induces suppressor T cells either directly or via suppressor inducer networks to down-regulate total numbers of IgM secreting cells. The absolute amounts of IgM-RF produced by RA PBMCs in our study were, as reported by others (2, 15>, widely heterogeneous and bore no discernible relationship to commonly used clinical criteria of the patients’ rheumatoid arthritis (as measured by disease duration, sedimentation rate, or Ritchie-Camp articular index). However, the highest levels of IgM-RF production were clearly found in
HUMORAL
IMMUNE
FUNCTION
IN RHEUMATOID
ARTHRITIS
193
the anergic patients. This subgroup of patients, defined by depressed PBMC proliferative responses to soluble antigens in vitro, has been shown to demonstrate clinical improvement after short-term leukapheresis (16), display distinctive PBMC subset phentotypes (17), and manifest predictable synovial immunohistopathologic findings including high-intensity T-helper/inducer lymphocyte and plasma cell infiltration (18). Because rheumatoid synovium is also rich in IgMRF-producing cells (6) and a relationship between peripheral blood RF-producing cells and similar cells in synovium has been shown, we postulated that an association between anergic status and IgPFC and/or IgM-RF production in rheumatoid PBMCs might exist. Because of the heterogeneity of the PFC responses in general and the prevalence of low or undetectable IgM-PFC responses in RA patients, we were unable to prove a relationship between anergy and IgM and/or total IgPFC numbers. However, our data do show that PBMC proliferative responses and very high spontaneous in vitro IgM-rheumatoid factor production are associated phenomena. The explanation for this association is unknown, although anergy, excessive IgM-RF production, and the increase in spontaneous total immunoglobulin production may be causally linked. It appears that rheumatoid T cells, activated and stimulated by a yet unidentified antigen, play a role in the increased IgM-rheumatoid factor secretion. In this regard, a population of E-rosette-positive, HLA-DR + activated T lymphocytes has been found in rheumatoid PBMCs, and these cells are increased in anergic RA patients (17). The enhanced immunoglobulin production stimulated by such activated T cells might subsequently lead to antigen-IgG-IgG complexes, which have recently been shown to be capable of prompting an increase in IgM-RF production in viva (26, 27). Alternatively, defective T-cell function may allow expansion of Epstein-Barr virus-infected B lymphocytes which secrete IgM-RF. T-cell defects regulating Epstein-Barr virus-infected B cells have been extensively documented in rheumatoid arthritis patients (3). However, since the function of rheumatoid factor and the mechanism underlying the defective T-cell function are unknown, many other explanations are also possible. Further work will be needed to further define the significance of our observations. In summary, in vitro humoral immune function was evaluated in PBMCs from a clinically homogeneous, highly active group of RA patients who were off antirheumatic drug therapy for a minimum of 4 weeks. Rheumatoid PBMCs as a group showed increased mean numbers of circulating, spontaneously secreting total immunoglobulin plaque-forming cells, profoundly decreased mean numbers of spontaneous and PWM-driven IgM-PFCs, and elevated median spontaneous in vitro IgM-rheumatoid factor production. Significant heterogeneity was noted when individual rheumatoid PBMC responses were compared. However, certain subgroups of RA patients were discernible, based on individual assay response patterns, Markedly elevated IgM-rheumatoid factor secretion by PBMCs was associated with hypoproliferation to soluble recall antigens, and RA PBMCs that spontaneously secreted IgM-RF were significantly more likely to show low total IgM-PFC. These findings suggest that certain subgroups of RA patients that show recognizable cellular (13- 15) and humoral immune response patterns can be
194
BOLING
ET AL.
identified. Further characterization of these subgroups may in turn lead to new insights regarding the cellular basis of rheumatoid humoral dysfunction and their relationship to altered cell-mediated immune function. ACKNOWLEDGMENTS We thank Janice Allen, Carole Cole, and Dr. David Yocum for the assistance in some experiments, and we also thank Dawn Smith for preparing the manuscript.
REFERENCES 1. Yu, D. T., and Peter, J. B., Semin. Arthritis Rheum. 4, 25, 1974. 2. Tsoukas, C. D., Carson, D. A., Fong, S., Pasquali, J-L., and Vaughan, J. H., J. Zmmunol. 1125,
125,
1980.
3. Decker, J. L., Malone, D. G., Haraoui, B., Wahl, S. M., Schrieber, L., Klippel, J. H., Steinberg, A. D., and Wilder, R. L., Ann. Intern. Med. 101, 810, 1984. 4. Panush, R. S., Katz, P., and Longley, S., Clin. Immunol. Immunopathol. 28, 252, 1983. 5. Pardo, I., and Levinson, A. I., Clin. Immun. Immunopathol. 29, 29, 1983. 6. Al-Balaghi, S., Strom, H., and Moller, E., Scand. J. Immunol. 16, 69, 1982. 7. Corke, C. F., GUI, V., Stierle, H. E., Huskisson, E. C., and Holborrow, E. J., Rheum. Znt. 4, 19, 1984. 8. Patel, V., Panayi, G. S., and Unger, A., J. Rheum. 10, 364, 1983. 9. Koopman, W. J., and Schrohenloher, R. E., Arthritis Rheum. 23, 706, 1980. 10. Olsen, N. J., and Jasin, H. E., Arthritis Rheum. 27, 985, 1984. 11. Olsen, N., Ziff, M., and Jasin, H. E., Rheum. Int. 2, 59, 1982. 12. Kotzin, B. L., Strober, S., Engleman, E. G., Calin, A., Hoppe, R. T., Kansan, G. S., Terrell, C. P., and Kaplan, H. S., N. Engl. J. Med. 305, 969, 1981. 13. Alarcon, G. S., Koopman, W. J., and Schrohenloher, R. E., Arthritis Rheum. 28, 356, 1985. 14. Olsen, N., Ziff, M., and Jasin, H. E., J. Rheumafol. 11, 17, 1984. 15. Koopman, W. J., and Schrohenloher, R. E., Arthritis Rheum. 23, 302, 1980. 16. Wahl, S. M., Wilder, R. L., Katona, I. M., Wahl, L. M., Allen, J. B., Scher, I.. and Decker, J. L., Arthriris Rheum. 26, 1076, 1983. 17. Haraoui, B., Wilder, R. L., Malone, D. G., Allen, J. B., Katona, I. M., and Wahl, S. M., J. Immunol. 133, 697, 1984. 18. Malone, D. G., Wahl, S. M., Tsokos, M., Cattell, H., Decker, J. L., and Wilder, R. L., J. Clin. Invest. 74, 1173, 1984. 19. Ropes, M. W., Bennett, G. A., Cobb, S., Jacox, R., and Jessar, R. A., Bull. Rheum. Dis. 9, 175, 1958. 20. Camp, A. V., Orthopedics 4, 39, 1971. 21. Ginsburg, W. W., Finkelman, E D., and Lipsky, P E., J. Zmmunol. 120, 33, 1978. 22. Lawrence, E. C., Blaese, R. M., Martin, R. R., and Stevens, P. M., J. Clin. Invest. 62,832, 1978. 23. Kunkel, H. G., Agnello, V., Joslin, F. G., and Winchester, R. J., J. Exp. Med. 137, 311, 1973. 24. Bongura, V. K., Kunkel, H. G., and Pernis, B., J. Clin. Invest. 69, 1356, 1982. 25. Karsh, J., Klippel, J. H., Plotz, P., Decker, J. L., Wright, D. C., and Flye, M. W., Arthritis Rheum. 24, 867, 1981. 26. Nemazee, D. A., J. Exp. Med. 161, 242, 1985. 27. Coulie, P. G., and Van &tick, J., J. Exp. Med. 161, 88, 1985. Received September 11, 1986; accepted with revision January 5, 1987
IMMUNOLOGY
AND
Humoral
IMMUNOPATHOLOGY
43, 185-194 (1987)
Immune Function in Severe, Active Rheumatoid Arthritis
E. P. BOLING,* T. OHISHI,$ S. M. WAHL,$ J. MISITI,? R. WISTAR, JR.,? AND R. L. WILDERII,~ *Department of Medicine, Malcolm Grow USAF Medical Center, Washington, DC; l-The Naval Medical Research Institute, Bethesda, Maryland, and $Metabolism Branch, National Cancer Institute, #Cellular Immunology Section, National Institute of Dental Research, and “Arthritis and Rheumatism Branch, National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health. Bethesda, Maryland 20892
Peripheral blood mononuclear cells (PBMC) from 30 patients with definite or classic active rheumatoid arthritis who were on no remittive drugs were studied for spontaneous and pokeweed mitogen (PWM)-stimulated immunoglobulin plaque-forming cell frequency (IgPFC), spontaneous IgM-rheumatoid factor (IgM-RF) secretion, and in vitro proliferative responses to soluble recall antigens. Rheumatoid spontaneous total (IgG + IgM + IgA) IgPFCs were higher than those of normal controls when assayed after 7 days in culture. Spontaneous and PWM-stimulated IgM-PFCs, in contrast, were significantly less than normal regardless of when assayed. Spontaneous synthesis of IgM-RF was observed in 56% of the RA patients, but absolute amounts produced were widely heterogeneous. Spontaneous IgM-RF production by RA PBMC was associated with low or absent spontaneous IgM-PFC production. Moreover, a strong association was found between the median amount of IgM-RF secreted and depressed proliferative responses to soluble recall antigens. Our results define several abnormalities of immunoglobulin production in a clinically homogeneous and highly active rheumatoid population and delineate methodologic variations that can complicate the interpretation of similar data in the literature. In addition, our findings suggest that subgroups of rheumatoid patients that show distinct cellular and humoral immune abnormalities can be identified. o 1987 Academic
Press. Inc.
INTRODUCTION
Rheumatoid arthritis (RA) is often associated with serum rheumatoid factors and increased total serum immunoglobulin (l-3). The significance of these humoral abnormalities, their possible relationship to T-lymphocyte abnormalities in RA, and the extent of their involvement in disease pathogenesis all await further clarification. Several recent studies have evaluated the production of polyclonal and/or rheumatoid factor immunoglobulins by rheumatoid peripheral blood mononuclear cells (PBMCs) by enumeration of antibody plaque-forming cells (PFCs) (4-8) and quantitation of in vitro immunoglobulin production by immunoassay (4, 9-14). The total numbers of antibody PFCs and/or the total amount of immunoglobulin produced by rheumatoid PBMCs in these studies have been reported as low (4, 12), normal (8), or high (5-7). Similar discrepant results have been noted for each of the immunoglobulin isotypes when separately measured. 1 To whom requests for reprints should be addressed. 185 0090- 1229/87 $1.50 Copyright 0 1987 by Academic Press. Inc. All rights of reproduction in any form reserved.
186
BOLING
ET AL.
Moreover, although rheumatoid PBMCs as a group are known to spontaneously secrete IgM-rheumatoid factor in culture, considerable heterogeneity has been shown (2, 15) and the relationship of such synthesis to disease activity continues to be debated (13, 14). Interpretation of the above studies has been complicated by the inclusion of patients who were heterogeneous with respect to the amount of disease activity, the presence of rheumatoid factor positivity, and the concomitant use of “anti-rheumatic” drugs. We therefore quantitated numbers of spontaneous and pokeweed mitogen-stimulated immunoglobulin plaque-forming cells in rheumatoid PBMCs and measured IgM-rheumatoid factor production by an enzyme-linked immunosorbent assay in order to define what abnormalities of humoral immune function are present in a population of severely active seropositive RA patients who had been on no remittive therapy for at least 4 weeks at the time of study. In addition to the humoral defects, multiple abnormalities of cell-mediated immune function occur in RA (1, 3) and T-lymphocyte dependence of IgM-rheumatoid factor production has been shown in vitro and in vivo. Thus, an association between abnormal cell-mediated immune function and altered rheumatoid factor synthesis might be identified. Prior studies from our laboratories have defined a subgroup of rheumatoid patients whose PBMCs show low to absent (i.e., anergic) in vitro proliferative responses to soluble recall antigens (3, 16- 18). We therefore investigated whether any of the heterogeneity of spontaneous IgM-rheumatoid factor production by untreated, clinically active rheumatoid PBMCs could be associated with the previously described abnormalities in cell-mediated immune function. The present report demonstrates that clinically active RA patients as a group show (1) increased total spontaneous immunoglobulin plaque-forming cell frequency (IgPFC), (2) decreased spontaneous and PWM-stimulated IgM-PFC, and (3) overall increased spontaneous IgM-rheumatoid factor production. Moreover, the highest spontaneous rheumatoid factor production was found in the anergic rheumatoid patients. MATERIALS
AND METHODS
Patients and controls. The patient group consisted of 30 individuals including 19 females and 11 males ranging in age from 34 to 70 years (mean = 54.7 years). All patients had classic seropositive rheumatoid arthritis as detined by American Rheumatism Association criteria (19). None took drugs other than nonsteroidal antiinflammatory agents and/or prednisone (< 10 mg/day) at the time of study, and all cytotoxic and long-acting anti-rheumatic drugs had been discontinued for at least 4 weeks. All had active disease as defined by a Ritchie-Camp articular index (20) of ~40, with sustained disease activity during the 2 months before study. Prednisone was not administered 24 hr prior to obtaining blood. Controls were healthy laboratory and office personnel, consisting of 30 individuals and including 15 females and I5 males ranging in age from 24 to 65 years (mean = 42.9 years). Determination of total immunoglobulin and IgM-secreting cells by reverse hemolytic plaque assay. PBMCs were obtained from heparinized blood by centrifu-
gation on Ficoll-Hypaque
gradients.
Washed PBMC
were suspended in RPM1
HUMORAL
IMMUNE
FUNCTION
IN RHEUMATOID
ARTHRITIS
187
1640 with glutamine (4 m&f), streptomycin (50 pg/ml), penicillin (50 units/ml), and 10% heat-inactivated fetal bovine serum (Microbiological Associates, Rockville, MD). PBMCs (3 x 10s) in 1 ml of medium were plaqued either directly (Day 0) or after having been cultured for 7 days (Day 7) in 12 x 75-mm polypropylene tubes (Falcon, Oxnard, CA; No. 2063) with or without a 1:lOO dilution of pokeweed mitogen (PWM, GIBCO, Chagrin Falls, OH, Lot llK3413). A reverse hemolytic plaque-forming cell assay (RHPA) was used to detect immunoglobulin-secreting cells as previously described (21, 22). Briefly, 125 ~1 of the above cell suspension was added to 96-well flat-bottom plates and mixed with 25 ~1 of sheep red blood cell (SRBC)-absorbed guinea pig complement, 25 ~1 of 30% protein A-SRBC, and 25 t.~l of either rabbit anti-human Ig antiserum (Cappel Laboratories, used at 1:40 dilution) or rabbit anti-human IgM (Cappel Laboratories, 1:40 dilution). After a 2-hr incubation at 37”C, plaques were counted and expressed as the number of immunoglobulin-secreting cells/lo6 PBMCs cultured. Assay of ZgM-rheumatoid factor (ZgM-RF) produced in supernatants of cultured PBMCs. PBMCs were cultured in 35-mm dishes at 5 x lo6 PBMUculture
with 10% hypogammaglobulinemic human AB serum in a 2-ml volume of RPM1 1640 with penicillin (100 units/ml), streptomycin (100 kg/ml), L-glutamine (2 mM), and 2-mercaptoethanol(5 x 10es iV). After 10 days, supernatants were harvested and frozen at - 20°C until assayed for IgM-RF. Rheumatoid factor determination was done on multiple dilutions of supernatant by an enzyme-linked immunosorbent adaptation of a radioimmunoassay method previously described (15). Ninety-six-well Dynatech Immunolon II plates were coated with 100 ~1 per well of 10 p.g/ml polyclonal human IgG (Cohn fraction II, Red Cross, Bethesda, MD) in normal saline, blocked with 200 yl per well of borate-buffered saline (BBS) containing 5 g/liter bovine serum albumin (BSA), and rinsed three times with washing buffer (BBS with 0.1 g/liter BSA) before 100 ~1 of supernatant was applied to the plate. After incubation for 1 hr at 25”C, the plates were washed, and 100 ~1 of alkaline phosphatase-conjugated goat anti-human IgM (Tago, Burlingame, CA, used at a dilution of 1:2000) was applied for 1 hr. The plates were washed, p-nitrophenyl phosphate substrate was applied (1 mg/ml in diethanolamine buffer), and after 30 min optical density at 405 nm was assessed with a Dynatech Model 580 ELISA reader. With each assay a series of dilutions of purified Wallenberg idiotype monoclonal IgM-K rheumatoid factor (23, 24) (courtesy of Dr. Gerald Penn) was included so that the color developed by the supernatants could be related to a known concentration of rheumatoid factor. PBMC proliferative responses. Mononuclear cell proliferative responses were determined as previously described in detail (16). Briefly, 2 x lo* PBMCs were cultured in 96-well flat-bottom plates in the presence or absence of a variety of soluble recall antigens (PPD, streptokinase-streptodornase, tetanus toxoid, and Candida). Six days later, the cells were pulsed for 4 hr with tritiated thymidine ([3H]TdR, 6.7 Ci/mmol, Schwarz/Mann, Orangeburg, NY) and harvested, and the incorporated activity was measured. A stimulation index was determined by dividing [3H]TdR incorporated (counts per minute) in cultures with antigen by the [3H]TdR incorporation in unstimulated cultures. A stimulation index of <2 for all antigens tested was used to define the anergic subgroup, whereas an index of >2
188
BOLING
ET AL.
to at least one antigen was considered normal or nonanergic. PBMCs from virtually all normal control individuals respond to at least one antigen with an index of >2. Leukupheresis procedure. Three rheumatoid arthritis patients were selected for leukapheresis treatment based upon their anergic immunologic status as previously described (16). Leukapheresis (LP) was carried out as described by Karsh et al. (25). Statistics. Statistical probabilities were assigned by performing the Wilcoxon rank sum test for independent or for paired samples or by the Fisher exact test in comparisons involving small numbers of samples. RESULTS Total Immunoglobulin Plaque-Forming Cells
Total spontaneous IgPFCs from rheumatoid PBMC were enumerated immediately postcollection and compared to similarly assayed normal controls. Spontaneous total IgPFC assayed on Day 0 were significantly higher than those of normal controls (P < 0.01, Fig. la). However, we also enumerated total IgPFCs 7 days after culture initiation and found significant depression in the number of 5O.cQO - la,
b)
20,000
10,000
g a is
5ooo
K P
2000 -
<
low-
E
*
-
T
* .
. i
500 -
f I
T . :
. . .
.
.
AA
N
z
s
3PONTANEOU:
RA
N
+ PWM
FIG. 1. Circulating total immunoglobulin plaque-forming cells (IgPFC) in active rheumatoid (RA) and normal control (N) peripheral blood mononuclear cells (PBMC). (a) Total IgPFC enumerated immediately after PBMC collection on Day 0. (b) and (c) Total IgPFC after 7 days in culture in the absence or presence of pokeweed mitogen (PWM), respectively. Spontaneous IgPFC on Day 0 are increased in RA PBMC relative to N (P < 0.01). IgPFC in Day 7 RA PBMC are decreased relative to N whether assayed in the absence (P < 0.01) or in the presence (P < 0.01) of PWM. Statistical probabilities were calculated by the Wilcoxon rank sum test for independent samples.
HUMORAL
IMMUNE
FUNCTION
IN RHEUMATOID
ARTHRITIS
189
spontaneous (Fig. lb, P < 0.01) and PWM-stimulated (Fig. lc, P < 0.01) total IgPFCs in rheumatoid patients relative to normal controls. Moreover, the numbers of spontaneous total IgPFCs in individual RA patients decreased significantly over the I-/-day culture period (P < 0.05), whereas concurrently run normal PBMCs increased during the same period (P < 0.05). ZgM Plaque-Forming
Cells (ZgM-PFCs)
In contrast to the increase in spontaneous total IgPFCs found in immediately plaqued rheumatoid PBMCs (Fig. la), spontaneous Day 0 IgM-PFCs were significantly depressed (Fig. 2a, P < 0.01). Similar profound depression of IgM-PFCs was seen following 7 days in culture in the absence (Fig. 2b, P < 0.01) or presence (Fig. 2c, P < 0.01) of pokeweed mitogen. Furthermore, it was noted that, under all conditions examined, a subpopulation of RA patients existed whose PBMC showed little or no capacity to make detectable IgM-PFCs (Figs. 2a, b, c).
5000
- la)
,Ibi
Y
Cl :
: . *
. 2000
:
-
:.
i . .
: i
1000
7
t 500 -
.
Y P 8 -
200 -
.
Li .z E
loo-
* .
-
* .
ELI. I SPONTANEOU
.. . *
T
. .
:
!.
.
. -
P RA
s N
SPONTANEOU’
. .
.
-+----
N + PWN
-I I
FIG. 2. Circulating IgM-plaque-forming cells (IgM-PFC) in active RA and normal control (N) peripheral blood mononuclear cells (PBMC). (a) IgM-PFC enumerated immediately after PBMC collection on Day 0. (b) and (c) IgM-PFC after 7 days in culture in the absence or presence of PWM, respectively. Spontaneous IgM PFC on Day 0 are decreased in RA PBMC relative to N (P < 0.01). IgM-PFC on Day 7 are decreased relative to N whether assayed after spontaneous culture (P < 0.01) or in the presence (P < 0.01) of PWM. Statistical probabilities were calculated by the Wilcoxon rank sum test for independent samples.
190
BOLING
ET AL.
Zn Vitro ZgM-RF Production and Its Association with T-Lymphocyte Function
We found that 56% of unstimulated RA PBMC spontaneously produced measurable IgM-RF during in vitro culture in contrast with none of the controls (Fig. 3a). This prevalence of spontaneous IgM-RF secretion was consistent with that described in other reports (2, 15) but stood in sharp contrast to the overall decrease in polyclonal IgM-PFCs seen in these same cells (Figs. 2a, b). We therefore compared IgM-PFCs in rheumatoid PBMCs which either spontaneously secreted, or failed to secrete, IgM-RF in vitro. As shown in Table 1, we found that spontaneous secretors of IgM-RF were significantly more likely to show low (i.e., <20/106 PBMCs) numbers of IgM-PFCs than were nonsecretors (P = 0.028, Fisher exact test). In rheumatoid patients with low IgM-PFCs, 73% (8/l 1) spontaneously secreted IgM-RF, in contrast to 12% (l/8) of patients with normal IgMPFCs. Furthermore, we noted that extremely high production of IgM-RF was limited to PBMCs from a previously characterized subset of anergic patients (Fig. (bl .
. .
2 -P
1000 t .
. .
.
.
+ AN
NA RA
FIG. 3. Spontaneous IgM-rheumatoid factor (&M-RF) production in 1 l-day supernatants (SN) of cultured peripheral blood mononuclear cells (PBMC). IgM-RF is quantitatively expressed relative to binding of a purified monoclonal IgM-RF (Wallenberg). (a) The IgM-RF in SN of RA and normal (N) cultures; (b) anergic (AN) and nonanergic (NA) rheumatoid PBMC cultured. IgM-RF levels in RA PBMC-SN are higher than N-SN, but are heterogeneous. IgM-RF levels in anergic PBMC-SN are significantly higher than in nonanergic PBMC-SN (P < 0.01). Statistical probabilities were calculated by the Wilcoxon rank sum test for independent samples.
HUMORAL
IMMUNE
FUNCTION
IN RHEUMATOID
ARTHRITIS
191
3b). All four rheumatoid patients whose PBMCs produced ~1000 rig/ml of IgMRF in culture were anergic (median IgM-RF level, 2100 r&ml). In contrast, none of 18 nonanergic RA patients produced in vitro IgM-RF factor in these amounts (median IgM-RF level of all nonanergic samples ~15 &ml; median level of the IgM-RF-positive nonanergic samples, 210 rig/ml (Fig. 3b, P < 0.001)). Ten out of 18 nonanergic patients had no spontaneous elaboration of IgM-RF in vitro, in contrast to the high level production by anergic patients. The spontaneous IgMRF production did not appear to be correlated with clinical parameters such as corticosteroid usage or dose or serum RF titers (data not shown). Thus, these data support the notion that high level spontaneous IgM-RF production by PBMCs is associated with abnormal T-cell function. Because prior work has shown that anergic rheumatoid T-cell responses tend to normalize with serial leukapheresis therapy (16), we evaluated PBMC T-cell proliferative responses to soluble recall antigens in three anergic RA patients prior to and at the end of leukapheresis therapy and compared these values to concurrently measured in vitro IgM-rheumatoid factor production (Fig. 4). Supernatants of PBMCs from the three anergic patients examined showed a drop in spontaneously produced IgM-rheumatoid factor concurrent with a normalization of proliferative responses to one or more soluble antigens. Thus, these findings further support the view that hyporesponsive rheumatoid T-cell function and high level in vitro IgM-rheumatoid factor production are related. DISCUSSION
Because several recent studies have reported conflicting results regarding immunoglobulin production and RF synthesis in RA, we analyzed polyclonal Ig-PFC and spontaneous IgM-RF production in PBMCs of clinically homogeneous, severely active seropositive rheumatoid patients who were without disease-modifying drugs for a minimum of 4 weeks. We found that total (i.e., IgG + IgA + IgM) spontaneous immunoglobulin plaque-forming cells are elevated in freshly collected rheumatoid PBMCs. This finding is in agreement with most (5-7), but not all (4), previous reports and reinforces the concept that there are increased numbers of activated B cells in the rheumatoid patient’s peripheral blood compartment. In contrast, several recent studies using PFC (4) or quantitative immunoassay methods (4, 8, 12) have reported the apparently contradictory TABLE 1 NUMBERSOFRHEUMATOIDPATIENTSWITH Low(<20PFC1106PBMC) ORNORMAL (>20PFC/106 PBMC)IgM PFC THATSPONTANEOUSLYSECRETEORFAILTOSECRETE IgM-RF in Vitro No spontaneous IgM-RF production (< 15 rig/ml) Low IgM-PFC producers Normal IgM-PFC producers Total
3 7 10
Spontaneous IgM-RF production (> I.5 rig/ml) 8 1 9
Total 11 8 19
Note. RA patients with low IgM-PFC numbers (after ‘I-day spontaneous culture) were significantly more likely to spontaneously secrete IBM-RF in vitro (P = 0.028 by the Fisher exact test).
192
BOLING (al Patlent
ET AL.
/ Patient
s.s
R.B
1 Patient
D.K.
IgM-RF
Inglmll
2000 1000 500
Proliferation Index
1.02
t LPl
3.07
0.72
3.85
t
t
t
t
t
t
t
t
LP8
LPlO
LPI
LP3
LP6
LPl
LP3
LP6
FIG. 4. Spontaneous IgM-rheumatoid factor (IgM-RF) production versus maximum soluble recall antigen proliferative responses by peripheral blood mononuclear cells in three anergic RA patients (S.S., R.B., D.K.) undergoing serial leukapheresis treatments. Soluble recall antigen proliferative responses (Proliferation Index) from two separate leukaphereses are shown for each patient.
findings of decreased total IgPFC (4) and decreased (4, 12) or normal (8) quantitative immunoglobulin production in vitro. In this report, we have shown that the number of total immunoglobulin plaque-forming cells detected in rheumatoid PBMCs falls significantly with in vitro culture (Figs. la, b). This points to a probable methodologic explanation for some of the previously reported total IgPFC differences and suggests that one possible cause for the disparity between reported plaque-forming cell and immunoglobulin synthesis results may be a dropout of B cells that have been already activated in vivo. An important new finding in our study was the depression of IgM-PFCs in rheumatoid PBMC. This finding occurred under a variety of different culture conditions and was unexpected since some prior reports had shown normal (6) or increased (7) numbers of IgM-PFCs. However, our results are consistent with several previous studies which used immunoassay methods to show significantly diminished (4, 10) or normal (15) IgM production by rheumatoid PBMC in vitro. Furthermore, we noted a subgroup of consistently hyporesponsive patients (Figs. 2a, b, c) which could not be attributed to immunomodulating drug effects. The finding that spontaneous IgM-RF secretion in vitro was significantly more prevalent in this IgM-hyporesponsive subgroup is intriguing and suggests that these two phenomena may be causally related. It is possible that IgM-RF, complexing with either auto-IgG or antigen-IgG aggregates, induces suppressor T cells either directly or via suppressor inducer networks to down-regulate total numbers of IgM secreting cells. The absolute amounts of IgM-RF produced by RA PBMCs in our study were, as reported by others (2, 15>, widely heterogeneous and bore no discernible relationship to commonly used clinical criteria of the patients’ rheumatoid arthritis (as measured by disease duration, sedimentation rate, or Ritchie-Camp articular index). However, the highest levels of IgM-RF production were clearly found in
HUMORAL
IMMUNE
FUNCTION
IN RHEUMATOID
ARTHRITIS
193
the anergic patients. This subgroup of patients, defined by depressed PBMC proliferative responses to soluble antigens in vitro, has been shown to demonstrate clinical improvement after short-term leukapheresis (16), display distinctive PBMC subset phentotypes (17), and manifest predictable synovial immunohistopathologic findings including high-intensity T-helper/inducer lymphocyte and plasma cell infiltration (18). Because rheumatoid synovium is also rich in IgMRF-producing cells (6) and a relationship between peripheral blood RF-producing cells and similar cells in synovium has been shown, we postulated that an association between anergic status and IgPFC and/or IgM-RF production in rheumatoid PBMCs might exist. Because of the heterogeneity of the PFC responses in general and the prevalence of low or undetectable IgM-PFC responses in RA patients, we were unable to prove a relationship between anergy and IgM and/or total IgPFC numbers. However, our data do show that PBMC proliferative responses and very high spontaneous in vitro IgM-rheumatoid factor production are associated phenomena. The explanation for this association is unknown, although anergy, excessive IgM-RF production, and the increase in spontaneous total immunoglobulin production may be causally linked. It appears that rheumatoid T cells, activated and stimulated by a yet unidentified antigen, play a role in the increased IgM-rheumatoid factor secretion. In this regard, a population of E-rosette-positive, HLA-DR + activated T lymphocytes has been found in rheumatoid PBMCs, and these cells are increased in anergic RA patients (17). The enhanced immunoglobulin production stimulated by such activated T cells might subsequently lead to antigen-IgG-IgG complexes, which have recently been shown to be capable of prompting an increase in IgM-RF production in viva (26, 27). Alternatively, defective T-cell function may allow expansion of Epstein-Barr virus-infected B lymphocytes which secrete IgM-RF. T-cell defects regulating Epstein-Barr virus-infected B cells have been extensively documented in rheumatoid arthritis patients (3). However, since the function of rheumatoid factor and the mechanism underlying the defective T-cell function are unknown, many other explanations are also possible. Further work will be needed to further define the significance of our observations. In summary, in vitro humoral immune function was evaluated in PBMCs from a clinically homogeneous, highly active group of RA patients who were off antirheumatic drug therapy for a minimum of 4 weeks. Rheumatoid PBMCs as a group showed increased mean numbers of circulating, spontaneously secreting total immunoglobulin plaque-forming cells, profoundly decreased mean numbers of spontaneous and PWM-driven IgM-PFCs, and elevated median spontaneous in vitro IgM-rheumatoid factor production. Significant heterogeneity was noted when individual rheumatoid PBMC responses were compared. However, certain subgroups of RA patients were discernible, based on individual assay response patterns, Markedly elevated IgM-rheumatoid factor secretion by PBMCs was associated with hypoproliferation to soluble recall antigens, and RA PBMCs that spontaneously secreted IgM-RF were significantly more likely to show low total IgM-PFC. These findings suggest that certain subgroups of RA patients that show recognizable cellular (13- 15) and humoral immune response patterns can be
194
BOLING
ET AL.
identified. Further characterization of these subgroups may in turn lead to new insights regarding the cellular basis of rheumatoid humoral dysfunction and their relationship to altered cell-mediated immune function. ACKNOWLEDGMENTS We thank Janice Allen, Carole Cole, and Dr. David Yocum for the assistance in some experiments, and we also thank Dawn Smith for preparing the manuscript.
REFERENCES 1. Yu, D. T., and Peter, J. B., Semin. Arthritis Rheum. 4, 25, 1974. 2. Tsoukas, C. D., Carson, D. A., Fong, S., Pasquali, J-L., and Vaughan, J. H., J. Zmmunol. 1125,
125,
1980.
3. Decker, J. L., Malone, D. G., Haraoui, B., Wahl, S. M., Schrieber, L., Klippel, J. H., Steinberg, A. D., and Wilder, R. L., Ann. Intern. Med. 101, 810, 1984. 4. Panush, R. S., Katz, P., and Longley, S., Clin. Immunol. Immunopathol. 28, 252, 1983. 5. Pardo, I., and Levinson, A. I., Clin. Immun. Immunopathol. 29, 29, 1983. 6. Al-Balaghi, S., Strom, H., and Moller, E., Scand. J. Immunol. 16, 69, 1982. 7. Corke, C. F., GUI, V., Stierle, H. E., Huskisson, E. C., and Holborrow, E. J., Rheum. Znt. 4, 19, 1984. 8. Patel, V., Panayi, G. S., and Unger, A., J. Rheum. 10, 364, 1983. 9. Koopman, W. J., and Schrohenloher, R. E., Arthritis Rheum. 23, 706, 1980. 10. Olsen, N. J., and Jasin, H. E., Arthritis Rheum. 27, 985, 1984. 11. Olsen, N., Ziff, M., and Jasin, H. E., Rheum. Int. 2, 59, 1982. 12. Kotzin, B. L., Strober, S., Engleman, E. G., Calin, A., Hoppe, R. T., Kansan, G. S., Terrell, C. P., and Kaplan, H. S., N. Engl. J. Med. 305, 969, 1981. 13. Alarcon, G. S., Koopman, W. J., and Schrohenloher, R. E., Arthritis Rheum. 28, 356, 1985. 14. Olsen, N., Ziff, M., and Jasin, H. E., J. Rheumafol. 11, 17, 1984. 15. Koopman, W. J., and Schrohenloher, R. E., Arthritis Rheum. 23, 302, 1980. 16. Wahl, S. M., Wilder, R. L., Katona, I. M., Wahl, L. M., Allen, J. B., Scher, I.. and Decker, J. L., Arthriris Rheum. 26, 1076, 1983. 17. Haraoui, B., Wilder, R. L., Malone, D. G., Allen, J. B., Katona, I. M., and Wahl, S. M., J. Immunol. 133, 697, 1984. 18. Malone, D. G., Wahl, S. M., Tsokos, M., Cattell, H., Decker, J. L., and Wilder, R. L., J. Clin. Invest. 74, 1173, 1984. 19. Ropes, M. W., Bennett, G. A., Cobb, S., Jacox, R., and Jessar, R. A., Bull. Rheum. Dis. 9, 175, 1958. 20. Camp, A. V., Orthopedics 4, 39, 1971. 21. Ginsburg, W. W., Finkelman, E D., and Lipsky, P E., J. Zmmunol. 120, 33, 1978. 22. Lawrence, E. C., Blaese, R. M., Martin, R. R., and Stevens, P. M., J. Clin. Invest. 62,832, 1978. 23. Kunkel, H. G., Agnello, V., Joslin, F. G., and Winchester, R. J., J. Exp. Med. 137, 311, 1973. 24. Bongura, V. K., Kunkel, H. G., and Pernis, B., J. Clin. Invest. 69, 1356, 1982. 25. Karsh, J., Klippel, J. H., Plotz, P., Decker, J. L., Wright, D. C., and Flye, M. W., Arthritis Rheum. 24, 867, 1981. 26. Nemazee, D. A., J. Exp. Med. 161, 242, 1985. 27. Coulie, P. G., and Van &tick, J., J. Exp. Med. 161, 88, 1985. Received September 11, 1986; accepted with revision January 5, 1987