Shift toward T lymphocytes with a T helper 1 cytokine-secretion profile in the joints of patients with rheumatoid arthritis.

ARTHRITIS & RHEUMATISM
Vol. 39, No. 12, December 1996, pp 1961-1969
0 1996, American College of Rheumatology
1961
SHIFT TOWARD T LYMPHOCYTES WITH A
T HELPER 1 CYTOKINE-SECRETION PROFILE IN THE JOINTS OF
PATIENTS WITH RHEUMATOID ARTHRITIS
RADBOUD J. E. M. DOLHAIN, ANNETTE N. VAN DER HEIDEN, NATALJA T. TER HAAR,
FERDINAND C. BREEDVELD, and ANDRE M. M. MILTENBURG
Objective. To investigate whether T cells in the
inflamed joints of patients with rheumatoid arthritis
(RA) preferentially produce the T helper 1 (Thl) cytokines, interferon-y (IFNy) and interleukin-2 (IL-2), or
the Th2 cytokine, IL-4, when compared with corresponding peripheral blood-derived T cells.
Methods. Synovial fluid mononuclear cells
(SFMC) and corresponding peripheral blood mononuclear cells (PBMC) from 10 patients with RA were
analyzed, either directly or after in vitro stimulation, for
the intracellular presence of T h l and Th2 cytokines.
The amount of secreted cytokine in the cell culture
supernatants was measured by enzyme-linked immunosorbent assay (ELISA).
Results. IFNy-containing cells were detected in
the unstirnulated SFMC, but not in the PBMC, of 3
patients with RA. Cells positive for IL-2 or IL-4 were not
detected in the unstimulated samples. Following stimulation, the mean percentage of cells containing T h l
cytokines was significantly increased in the SFMC compared with the PBMC; no differences were found in the
mean percentage of IL-4-containing cells. A comparable shift toward T h l cytokines was observed when the
amount of secreted cytokine was determined by ELISA.
Conclusion. A shift toward T cells with a T h l
cytokine profile was observed in the joints of patients
with RA. Since an imbalance between T h l and Th2 cells
is thought to be of pathogenic significance, this finding
Dr. Dolhain’s work was supported by a grant from “Het
Nationaal Reumafonds.”
Radboud J. E. M. Dolhain, MD, Annette N. van der Heiden,
Natalja T. ter Haar, Ferdinand C. Breedveld, MD, PhD, Andre M. M.
Miltenburg, PhD: Leiden University Hospital, Leiden, The
Netherlands.
Address reprint requests to Radboud J. E. M. Dolhain. MD,
Leiden University Hospital, Department of Rheumatology, C4-R, PO
Box 9600, 2300 RC Leiden, The Netherlands.
Submitted for publication March 19, 1996; accepted in revised
form July 1, 1996.
might have implications for the development of new
therapies for RA.
T cell-mediated autoimmune responses are considered to play a role in the pathogenesis of rheumatoid
arthritis (RA) (1,2). At least 2 functional subsets of T
cells can be distinguished according to their cytokinesecretion profiles: T helper 1 (Thl) cells, which produce
interleukin-2 (IL-2), interferon-y (IFNy), and lymphotoxin, and Th2 cells, which secrete IL-4, IL-5, and IL-10
(3,4). At present, an accumulation of evidence supports
the concept that a T cell response with either a T h l or a
Th2 character is associated with different manifestations
of disease (5). In animal models of T cell-mediated
autoimmune disease, T h l cells were responsible for a
pathogenic response, whereas a protective immunosuppressive response resided within the Th2 population
(6-9). In experimental Lyme disease, C3H/HeJ mice
developed severe arthritis based upon antigen recognition by T h l cells, whereas BALB/c mice that developed
a Th2 response have only shown mild disease (7).
Further evidence indicates that therapies that
induce a shift of the immune response from T h l to Th2,
or that directly influence the balance between T h l and
Th2 cells (e.g., the infusion of autoreactive T cell clones
with a Th2 character into animals with established
disease), both result in disease amelioration (8,9). Moreover, in human T cell-mediated autoimmune diseases, a
T h l response is believed to be pathogenic. In multiple
sclerosis, for example, proteolipid protein reactive
clones isolated from peripheral blood during an acute
attack exhibit a T h l character, whereas during remission, such clones can exhibit either Tho, T h l , or Th2
characteristics (10). Although in most studies the T cell
cytokine profile is determined following antigen-specific
stimulation, differences in the T cell cytokine-secretion
DOLHAIN ET AL
1962
Table 1. Characteristics of the study patients;
Patient
1
2
3
4
5
6
7
8
9
10
Medication
Sulfasalazine
NSAID
NSAID
None
Cyclosporine,
hydroxychloroquine
NSAID,
acetaminophen
Prednisone, NSAID
Hydroxychloroquine,
NSAID
NSAID
NSAID
Disease
duration
(years)
Ritchie Articular
Index score
ESR
CRP
(mmihour) (rnglliter)
Leukocytes in
synovial fluid
( X 103/mrn3)
4
2
24
20
22
14
71
103
47
36
91
78
60
NA
81
23
NA
13.4
5.2
4.6
17.0
8
38
111
134
10.1
4
13
22
15
74
44
21
32
8.7
8.0
3
15
7
5
NA
15
78
9.1
5.5
15
5
43
21
* ESR = erythrocyte sedimentation rate; CRP
nonsteroidal antiinflammatory drug.
=
profile between patients and controls have also been
observed following mitogenic stimulation (11-13).
The available data concerning the presence of T
lymphocytes with a Thl or Th2 character in the rheumatic joint are limited. T cell clones isolated from RA
synovial tissue were proven to be primarily of the Thl
type (14,15). However, only limited numbers of clones
have been analyzed, and the possibility cannot be excluded that the procedure for cloning T cells favored the
outgrowth of T cells with a specific Th type. Another
study demonstrated the dominant presence of IFNy
messenger RNA (mRNA), relative to IL-4, in the rheumatoid synovial membrane (16). Although this approach
reflects the situation in vivo, no conclusion can be drawn
regarding the full potential of T cells derived from the
inflamed joint to produce cytokines following appropriate stimulation.
To further investigate the balance (relative presence) of Thl and Th2 cells at the site of rheumatoid
inflammation, paired synovial fluid mononuclear cells
(SFMC) and peripheral blood mononuclear cells
(PBMC) were obtained from RA patients in order to
determine the percentage of cells producing the Thl
cytokines, IFNy and IL-2, and the Th2 cytokine, IL-4.
Staining for the different cytokines was performed either
directly, after activation with phorbol myristate acetate
(PMA) and a calcium ionophore, or after activation with
immobilized anti-CD3 antibodies. Cytokine-producing
cells were identified using cytokine-specific monoclonal
antibodies (MAb) and immunohistochemical detection
procedures.
The amount of secreted cytokine in the cell
culture supernatant was measured by enzyme-linked
immunosorbent assay (ELISA). To determine whether
differences in the T cell subsets present in the PBMC
36
C-reactive protein; NA
=
not available; NSAID
=
and SFMC could be responsible for a shift toward a
cytokine-secretion pattern with either a Thl or a Th2
character, the percentage of T cells and their activation
and maturation state in these 2 compartments was
evaluated by fluorescence-activated cell sorter (FACS)
analysis. Finally, triple-staining techniques were used to
examine the phenotype of cytokine-producing cells.
PATIENTS AND METHODS
Patients. Paired PBMC and SFMC were obtained
from 10 patients (Table 1) with active RA (mean 2 SD
number of swollen joints 6.2 ? 3.6; mean -C SD Ritchie
Articular Index [17] 16.9 -C 10.7; mean -C SD erythrocyte
sedimentation rate 64.4 2 31.0 mm/hour). All but 1 patient
(patient 7) were positive for rheumatoid factor, and all patients
had periarticular joint erosions detected on radiographs of the
hands and feet. RA was defined according to the 1987 revised
criteria of the American College of Rheumatology (formerly,
the American Rheumatism Association) (18). Synovial fluid
was obtained from knee joints (mean ? SD number of
leukocytes per mm3 of synovial fluid 9.1 2 4.1 X 10’; mean 2
SD visual analog scale for knee pain 5.9 -C 1.7, range 0-10).
PBMC were also obtained from 5 healthy individuals who were
matched for age and sex in order to compare cytokineproduction profiles.
Lymphocytes were isolated from blood or synovial
fluid by Ficoll-amidotrizoate gradient centrifugation and
stored in tissue culture medium (TCM) supplemented with
10% dimethyl sulfoxide (Merck, Darmstadt, Germany) in
liquid nitrogen. TCM consisted of Iscove’s modified DulbecCO’S medium (Whittaker, Baltimore, MD) supplemented with
the antibiotics penicillin and streptomycin (100 IU/ml and 100
pg/ml, respectively; Boehringer Mannheim, Mannheim, Germany) and 10% fetal calf serum (Gibco BRL, Breda, The
Netherlands).
Induction of cytokines. Paired PBMC and SFMC
samples were thawed and washed with phosphate buffered
saline (PBS). A portion of the cells was directly analyzed for
the presence of cytokine-producing cells. These cells were
1963
Thl CYTOKINE-SECRETION PROFILE IN RA
Table 2.
Specificity
-
Characteristics of the monoclonal antibodies used in the present study"
Antibody
CD2
CD3
CD4
D9D10
MD1
17H12
25D2
MT910
Leu-4
Leu-3A
CD8
Leu-2A
CD16
CD20
CDE
CD45RA
CLB-CD16
LEU-16
Tac
CD45RO
Leu-45RO
CD69
Leu-23
B8.11.2
-
-
-
-
LEU-18
Reactivity
Source
IFNy
IFNy
IL-2
IL-4
Pan T cell
Pan T cell
Helper/inducer T cells, subset
of monocytes
Cytotoxic/suppressor T cells,
natural killer cells
Natural killer cells
B cells
I L 2 receptor
Naive T cells, B cells,
monocytes
Memory T cells, B cells,
monocytes
Activated T and B cells
HLA-DR
Laskay et a1 (see ref. 19)
Van der Meide et al (see ref. 21)
DNAX (Palo Alto, CA) (see ref. 22)
DNAX (Palo Alto, CA) (see ref. 22)
Dakopatts (Glostrup, Denmark)
Becton Dickinson (Mountain View, CA)
Becton Dickinson (Mountain View, CA)
-
Biotin
-
FITC
PFITC
FITC
Becton Dickinson (Mountain View, CA)
P
CLB (Amsterdam, The Netherlands)
Becton Dickinson (Mountain View, CA)
ATCC (Rockville, MD)
Becton Dickinson (Mountain View, CA)
-
-
FITC
Becton Dickinson (Mountain View, CA)
P
Becton Dickinson (Mountain View, CA)
Rebai et a1 (see ref. 23)
-
* IFNy = interferon-y; IL-2 = interleukin-2; FITC = fluorescein isothiocyanate; P = R-phycoerythrin; CLB
Netherlands Red Cross Blood Transfusion Service; ATCC = American Type Culture Collection.
stained as described below. The remaining cells were stimulated in 96-well tissue culture plates (Greiner, Solingen, Germany) with a combination of PMA (1 ng/ml) (Sigma, St. Louis,
MO) and the calcium ionophore A23187 (ionomycin, 500
ng/ml; Sigma) or with immobilized anti-CD3 (OKT3; Cilag,
Herenthals, Belgium) in TCM. For the latter purpose, 96-well
plates were coated overnight at room temperature with OKT3
(10 pg/ml, 75 pl/well) in PBS. Wells were aspirated and
washed twice with PBS before use. The cells were cultured for
4 hours at 37°C and 5% CO, at a concentration of 2 X lo5
cells/well, after which cells and supernatants were harvested.
Supernatants were stored at -20°C until further analysis.
Staining procedure. Light microscopy. Both stimulated
and unstimulated cells were stained for the presence of the
Thl cytokines, IFNy and IL-2, and the Th2 cytokine, IL-4,
according to the paraformaldehyde saponin procedure (19,20).
The sources and specificities for all of the antibodies used are
listed in Table 2 (21-23). Cells were allowed to adhere t o
adhesion slides (Biorad, Munich, Germany), fixed with 4%
paraformaldehyde (P6148; Sigma) for 1 hour, and endogenous
peroxidase activity was blocked by incubating the cells with
0.3% H,O, (30%) and 0.1% NaN, in PBS for 10 minutes. Cells
were permeabilized with 0.1% saponin (Riedel de Haen,
Seelze, Germany) in PBS.
Cytobine-producing cells were detected using
cytokine-specific MAb (Table 2) followed by 2 steps of horseradish peroxidase-conjugated reagents (rabbit anti-mouse Ig
(P161), rabbit anti-rat Ig (P450), and swine anti-rabbit Ig
(P399) polyclonal antibodies; all from Dakopatts, Glostrup,
Denmark). All first-step reagents were diluted in PBS containing 0.1% saponin and 1% bovine serum albumin (BSA
Organon Teknika, Boxtel, The Netherlands). Second- and
third-step reagents were diluted in PBS containing 0.1%
saponin and 10% normal human serum. Between all incubation periods, slides were washed with PBS containing 0.1%
saponin. Histochemical revelation of horseradish peroxidase
was performed with diaminobenzidine tetrahydrochloride.
Label
=
Central Laboratory of The
Fluorescence microscopy (triple-stainingprocedures). In
order to identify cellular subsets involved in the production of
Thl cytokines, triple-staining techniques were developed. Both
stimulated and unstimulated cells were allowed to adhere to
reaction fields of adhesion slides. Cells were first incubated for
30 minutes at 4°C with unconjugated mouse anti-human MAb
against the maturation markers CD45RA and CD45RO or the
activation marker HLA-DR; these antibodies were diluted in
PBS with 1% BSA. Visualization of these antibodies was
performed by a second incubation period (30 minutes at 4°C)
of the reaction fields with a sheep anti-mouse Ig polyclonal
antibody conjugated to aminomethylcoumarin acetic acid
(1533983; Boehringer Mannheim) diluted in PBS with 1%
BSA. Subsequently, free binding sites of the sheep anti-mouse
Ig polyclonal antibody were blocked by incubating the reaction
fields with 20% normal mouse serum (NMS) in PBS for 20
minutes at 4°C.
To detect T cells, reaction fields were then incubated
with a mouse anti-human CD2 antibody conjugated to fluorescein isothiocyanate in PBS with 5% NMS for 30 minutes at
4°C. Between all incubation periods, slides were washed with
cold (4°C) PBS. When staining for surface markers was
completed, cells were fixed using 4% paraformaldehyde for 30
minutes at 4°C and permeabilized with saponin 0.1% in PBS.
To detect cytokine-producing cells, reaction fields were
incubated for 30 minutes at 37°C with either a biotinylated
mouse anti-human IFNy antibody or a rat anti-human IL-2
antibody, both diluted in PBS with 5% NMS and 0.1% saponin.
Incubation with the biotinylated anti-IFNy antibody was directly followed by a final incubation period of 30 minutes at
room temperature with streptavidin conjugated to tetramethyl
rhodamine isothiocyanate (A7169; Sigma) diluted in PBS with
0.1% saponin and 2% BSA.
For the detection of IL-2-producing cells, the final
incubation period (30 minutes at room temperature) was
preceded by an additional incubation with a biotinylated goat
anti-rat Ig polyclonal antibody (BA9400; Vector Laboratories,
DOLHAIN ET AL
1964
A
B
Figure 1. A, Morphologic aspects of the intracellular presence of a cytokine product, using interleukin-2 (IL-2) as an example. Positive cells
(in this case, activated synovial fluid mononuclear cells that stain positive for IL-2) are characterized by a distinct perinuclear dot, indicative
of Golgi localization (by immunoperoxidase staining). B, Negative control. (Original magnification X 250.)
Burlingame, CA) diluted in PBS with 0.1% saponin and 5%
NMS. During the procedure of intracellular staining for cytokines, slides were washed with PBS with 0.1% saponin between
all incubation periods. Primary, secondary, and tertiary reagents were titrated to obtain optimal results.
Specificity controls. The specificity of all antibodies
used has been described in detail (Table 2). As controls for
nonspecific interactions, isotype-matched irrelevant antibodies
were included and cytokine-specific antibodies were omitted
from the procedure. As a positive control for the accuracy of
the staining procedure, activated cells from a CD4+ Tho clone
were included in all experiments.
Examination of slides. Cytokine-producing cells were
identified by a clear perinuclear dot indicative of Golgi localization. To determine the total number of cytokine-producing
cells, at least 2,000 cells were scored in case the cells were not
stimulated in vitro or following stimulation with immobilized
OKT3. Following activation with PMA and ionomycin, at least
1,000 cells were analyzed. For double- and triple-staining
techniques, -200 cells expressing a specific marker were
scored on a Leitz Diaplane fluorescence microscope (Leica
Mikroskopie und Systeme, Wetzlar, Germany).
FACS analysis. Phenotyping of PBMC and SFMC was
performed by flow cytometry on nonstimulated cells using a
FACS scan (Becton Dickinson, Mountain View, CA). The
percentage of cells expressing CD14 (monocytes), CD16 (natural killer cells), and CD20 (B cells) was assessed using
monolabel fluorescence techniques. To determine the percentage of T cells and to analyze subpopulations of cells within the
T cell subset, double-label immunofluorescence techniques
were used. All antibodies used for FACS analysis are listed in
Table 2.
ELISA. ELISA was used to detect the amount of
cytokine secreted in the cell culture supernatant. The ELISA
for IFNy was performed as described previously (21). To
detect IL-2, the mouse anti-human IL-2 MAb 419 A7A3
(58.1 19.08; Medgenix Diagnostics SA, Fleurus, Belgium) was
used as a capture reagent. The biotinylated mouse anti-human
IL-2 MAb 297 C 16GZ (58.124.03; Medgenix Diagnostics SA)
was used as a secondary reagent. The IL-2 ELISA and IL-4
ELISA (M1914; CLB, Amsterdam, The Netherlands) were
performed according to the specifications of the manufacturer.
Statistical analysis. The Wilcoxon signed rank test was
used to analyze matched pairs. The Mann-Whitney U test was
used to compare group means. Correlation coefficients were
determined by Spearman’s rank correlation test.
RESULTS
Enhanced presence of Thl cytokine-producing
cells in SFMC of RA patients. PBMC and S F M C from
10 RA patients were stained either directly or after in
vitro activation for t h e presence of intracellular cytokines. Cells that stained positive for cytokines could b e
identified by a distinct perinuclear dot, indicative of
Golgi localization (Figure 1A). These dots were absent
when the first antibody was omitted from the procedure
or when it was replaced by an irrelevant isotype-matched
control antibody (Figure 1B). The occurrence of cells
that were found to spontaneously contain a cytokine
product in peripheral blood was below 1 in 1,000 cells.
T h l CYTOKINE-SECRETION PROFILE IN RA
3.01
N.S,
2.0
0.30
0.20
N.S
F
T
1.o
0.0
SFMC PBMC
SFMC PBMC
SFMC PBMC
Figure 2. Enhanced presence of T helper I cytokines in synovial fluid
mononuclear cells (SFMC) of patients with rheumatoid arthritis (RA).
SFMC and peripheral blood mononuclear cells (PBMC) of 10 patients
with RA were stimulated for 4 hours with phorbol myristate acetate
and ionomycin. The percentages of cells that stained positive for A,
interferon-y (IFNy) (W), B, interleukin-2 (IL-2) (
were determined using immunohistochemical detection methods. The
concentrations of these cytokines in the supernatants (D,IFNy. E,
1L-2, and F, IL-4) were detected by enzyme-linked immunosorbent
assay. Bars show the mean and SD. P values represent levels of
significance between paired SFMC and PBMC. N.S. = not significant.
However, the T h l cytokine, IFNy, could be detected in
3 (2%, 2%, and lo%, respectively) of the unstimulated
SFMC samples. No IL-2- or IL-4-containing cells could
be detected in unstimulated samples.
SFMC and corresponding PBMC samples were
stimulated with a combination of PMA and ionomycin to
investigate the full potential of cells from these 2 compartments to produce cytokines following stimulation. In
SFMC samples, more cells were found to produce IFNy
(P < 0.02) and IL-2 ( P < 0.05), compared with paired
PBMC (Figures 2A and B). IL-4-producing cells could
be detected in only 5 of 10 SFMC samples and 6 of 10
PBMC samples. The mean number of cells expressing
this cytokine in all samples tested was low and similar in
both populations (Figure 2C). PBMC of 5 age- and
sex-matched healthy volunteers were included in this
experiment to determine whether the difference in
frequencies of IFNy- and IL-2-producing cells could be
1965
attributed to the PBMC or SFMC fraction. Similar
frequencies of IFNy-, IL-2-, and IL-4-producing cells
were detected in the PBMC of healthy volunteers when
compared with PBMC of RA patients.
To investigate the potential for cytokine production after treatment with a more physiologic stimulus,
cells were activated using immobilized OKT3 antibodies.
With this stimulus, IFNy- and IL-Zexpressing cells were
detected in all PBMC and SFMC samples, although
lower numbers of cells that stained positive were present
when compared with PMA- and ionomycin-activated
cellular preparations. The mean ? SD percentages of
cells staining positive for IFNy were 2.9 ? 2.1 in SFMC
and 1.8 2 1.7 in PBMC samples (P = 0.06). The mean ?
SD percentages of IL-2-containing cells were 0.5 2 0.4
in SFMC and 0.3 ? 0.3 in PBMC samples ( P = 0.21).
Following stimulation with OKT3, only 1 SFMC sample
was found to contain IL-4-positive cells.
Enhanced presence of Thl cytokines in culture
supernatants of activated SFMC. In all supernatants of
cell cultures stimulated with PMA and ionomycin, IFNy
and IL-4 could be detected. The concentrations of IFNy
in the supernatants of stimulated SFMC were significantly higher when compared with those in supernatants
of activated paired PBMC (P C 0.01) (Figure 2D).
Similar concentrations of IL-4 could be detected in the
supernatants of PBMC and SFMC cultures (Figure 2F).
In 8 of 10 SFMC and 5 of 10 PBMC cell culture
supernatants, IL-2 concentrations above the detection
limit of the ELSA (625 pg/ml) were measured. Although the Supernatants of stimulated SFMC contained
more IL-2 when compared with those of activated paired
PBMC, the difference did not reach statistical significance (P = 0.07) (Figure 2E).
T o obtain more insight into the balance between
the production of T h l and Th2 cytokines in SFMC and
PBMC, ratios of the levels of IFNy to IL-4 were
calculated by dividing the concentration of IFNy in the
cell culture supernatants by the concentration of IL-4. In
9 of 10 patients, this ratio was higher in the SFMC
culture supernatants compared with paired PBMC culture supernatants (P < 0.01) (Figure 3). No differences
were observed in the concentrations of IFNy, IL-2, and
IL-4 in supernatants of stimulated PBMC of RA patients
compared with PBMC of 5 healthy controls.
To investigate whether the amount of a certain
cytokine in the cell culture supernatant was the result of
the number of cells producing that specific cytokine and
to evaluate the accuracy of the technique for intracellular staining, the percentage of cells that stained positive
for a certain cytokine was correlated with the concentration of that specific cytokine in the cell culture
DOLHAIN ET AL
1966
P 500
'
1
2
3
4
5
6
7
PATIENT
8
9
1
0
Figure 3. Ratio of the production of interferon-y (IFNy) to
interleukin-4 (IL-4) in synovial fluid mononuclear cells (SFMC) (H)
compared with peripheral blood mononuclear cells (PBMC) (H). To
obtain insight into the balance between T helper 1 and T helper 2
cytokines, the ratio of IFNy to IL-4 was calculated by dividing the
concentration of IFNy by that of IL-4 in the cell culture supernatant of
phorbol myristate acetate- and ionomycin-stimulated SFMC and
PBMC. P < 0.01. SFMC vs. PBMC.
supernatant. For all cytokines evaluated, statistically
significant correlations were found (IFNy p = 0.71, P <
0.0001; IL-2 p = 0.83, P < 0.0001; IL-4 p = 0.76, P <
0.000 1).
Phenotypic characterization of the different
PBMC and SFMC samples. Phenotyping of the different PBMC and SFMC samples revealed a significant
increase in the number of CD3-positive cells in the
SFMC samples compared with the PBMC samples,
accompanied by a lower percentage of cells expressing
the monocyte marker CD14, the marker for natural
killer cells CD16, and the B cell marker CD20 (Table 3).
The increase in the percentage of CD3-positive cells in
the SFMC samples when compared with the PBMC
samples cannot solely account for the difference in the
number of cytokine-producing cells following activation,
since, in patients who showed comparable frequencies of
T cells in their SFMC and PBMC, a higher number of
cytokine-producing cells could be demonstrated in the
SFMC (patient 9 SFMC CD3 38%, IFNy 10.6%, PBMC
CD3 48%, IFNy 6.2%; patient 10 SFMC CD3 52%,
IFNy 14.9%, PBMC CD3 42%, IFNy 3.5%).
To obtain more insight into the differences between T cell populations in SFMC and PBMC samples,
double-staining procedures were performed. No differences were observed regarding the percentage of T cells
expressing CD4 or CD8. However, the SFMC-derived
fraction contained significantly more memory T cells
(CD45RO) and fewer naive T cells (CD45RA) when
compared with the PBMC-derived fraction. In the
SFMC fraction, more T cells were found to express the
activation markers HLA-DR and CD69; the expression
of the activation marker CD25 on T cells was low, and T
cells expressing this marker could not be detected in all
samples (PBMC CD25 4 of 9 samples positive, SFMC
CD25 7 of 10 samples positive). Nevertheless, the expression of CD25 was significantly increased on SFMCderived T cells compared with PBMC-derived T cells
(Table 3).
Phenotype of the Thl cytokine-producing cells.
Triple stainings were performed to identify the phenotype of the T h l cytokine-producing cells (Figure 4). In
all 3 patients tested, the vast majority of the cytokineproducing cells coexpressed the T cell marker CD2
(IFNy PBMC mean 94% CD2+, SFMC mean 99%
CD2-t; IL-2 PBMC mean 80% CD2+, SFMC mean
98% CD2-t). I F N y and IL-2-producing cells could be
detected in both the CD2+, HLA-DR- and the CD2+,
HLA-DR+ subsets. No clear differences in frequencies
of cytokine-producing cells could be demonstrated in the
2 subsets tested. IFNy-producing cells were detected in
both the CD2+, CD45RA and CD2+, CD45RO subsets. In all patients the expression of IL-2 seemed to be
confined to the CD2+, CD45RO subset only. Triple
staining with CD69 as a marker was not performed,
since, upon activation, a very rapid increase in CD69+ T
cells was noted (data not shown). Unfortunately, using
immunofluorescence-based techniques, no IL-4producing cells could be demonstrated. This is probably
due to the fact that, as reported previously (20),
immunofluorescence-based techniques are somewhat
less sensitive in demonstrating cytokine-producing cells
when compared with techniques used with the light
microscope.
Table 3. Differences in phenotype of mononuclear cells isolated
from synovial fluid mononuclear cells (SFMC) and peripheral blood
mononuclear cells (PBMC)*
Mononuclear cells,
T cells, '36
%
Marker
PBMC
SFMC
Marker
PBMC
SFMC
CD3
CD14
CD16
CD20
45 t 9 t
23 t 11$
18 2 10
7 % 75
7 0 % 16
9 % I1
4 +5
2%2
CD4
CD8
CD45RA
CD45RO
HLA-DR
CD69
CD25
55 i 15
4 2 ? 15
43 ? lo$
6 2 t 121
11 t 5 t
4 -t 4 t
2t64
54 t 9
41 2 9
11 ? 9
93t4
53 t 12
57 z 11
6'10
* Values are the mean t SD.
f P < 0.01 versus SFMC.
$. P < 0.025 versus SFMC.
5 P < 0.05 versus SFMC.
Thl CYTOKINE-SECRETION PROFILE IN RA
1967
DISCUSSION
A
B
C
Figure 4. Phenotypic analysis of cytokine-producing cells using triplestaining immunofluorescence technology. The arrow in each figure
indicates a cell coexpressing A, interferon-y (using tetramethylrhodamine isothiocyanate), B, CD2 (using fluorescein isothiocyanate), or
C, CD45RO (using aminomethylcoumarin acetic acid). (Original
magnification x 400.)
The present data have provided evidence to
support a role for Thl cytokine-producing T cells in the
pathogenesis of RA. Cells that spontaneously produced
[FNy within the synovial fluid mononuclear cell population of RA patients were detected. Furthermore, a
shift toward lymphocytes with a Thl cytokine profile was
Dbserved in activated SFMC compared with paired
PBMC samples. This shift was observed not only when
the percentage of cells producing Thl or Th2 cytokines
was taken into account, but also when the amount of Thl
3r Th2 cytokmes in the cell culture supernatants was
determined. The increase in T cells expressing a Thl
cytokine profile was specific for the SFMC, since the
balance between the Thl and Th2 cells from the PBMC
of RA patients was not different from that of healthy
controls.
It was also shown that the SFMC-derived T cells
were more often of the memory type, and a higher
percentage of these cells expressed activation markers
when compared with PBMC-derived T cells. Triple
staining revealed that differences in maturation state
might have contributed to the increased production of
IL-2, but, for F N y , this was not the case. In addition, it
was concluded that differences in the expression of
activation markers were not associated with the increased production of cytokines.
In this study, highly significant correlation coefficients were found between the percentage of cytokineproducing cells and the amount of that specific cytokine
n the cell culture supernatant. These findings are conistent with those of a previous study (24) that suggested
I model of T cell activation in which, at a single-cell
evel, all-or-none rather than graded responses of cytoLine genes were dominant. Our data extrapolate from
he results obtained in murine models to human disease.
n addition, our results extend beyond the mRNA to the
)rotein level. The finding that the concentrations of
,ytokines, as detected by ELISA, correlated well with
he percentage of cells producing a specific cytokine, as
Jetected by intracellular cytokine staining, underlines
the usefulness of the latter method in quantifying cytokine production, even when, similar to IL-4, only a few
producer cells may be present.
The percentages of cells producing IFNy, IL-2,
and IL-4 were in the same range in the PBMC of RA
patients when compared with the PBMC of healthy
controls, and similar concentrations of cytokines were
measured in the supernatants of these cell popu~ations.
This implies that the shift toward a Thl response is
restricted to the actual site of inflammation. Such com-
1968
partmentalization of functional subsets of T cells in the
direction of a T h l or a Th2 response has been reported
previously. In experimental Lyme disease, a Thl-based T
cell response against the inciting antigen was associated
with severe disease manifestations. This polarization in
T cell function was present only at the site of inflammation, and not in distant lymphoid organs (7). Comparable observations have been made in patients with conjunctivitis (13).
Triple-staining techniques were used to determine whether the differences in activation and maturation states of the SFMC-derived T cells could be responsible for the increased production of T h l cytokines in
the SFMC fraction. Consistent with findings in the
literature, FACS analysis demonstrated that synovial
fluid-derived T cells were mainly of the CD45RO
memory type and showed an increased expression of the
activation markers HLA-DR and CD69 (25). Most
(26-29), but not all (30),studies have demonstrated that
CD45RO T cells produce more IFNy than CD45RA
cells. In this study, the percentage of CD45RO T cells
producing IFNy was not significantly higher than that of
CD45RA cells.
Furthermore, consistent with previous findings,
IL-2 production was mostly confined to the CD45RO
subset (29,31). No differences in production of IFNy
and IL-2 could be observed with regard to T cells with or
without HLA-DR expression. From these data it can be
concluded that an increase in CD45RO T cells, but not
differences in activation state, might have contributed to
the increased expression of T h l cytokines in SFMC.
The finding that the cytokine profile of SFMCderived T cells was significantly different from that of
PBMC-derived T cells underscores the importance of T
cells in maintaining RA. Although RA has been considered to be a T cell-mediated autoimmune disease (1,2),
the dominant role of T cells in the process of rheumatoid
synovitis has been questioned recently (32). Moreover,
mechanisms that lead to joint destruction but are not
under the regulation of T cells have been proposed (33).
The observations made in this study, however, favor the
idea of either a selective influx or a specific outgrowth of
T cells of the T h l type occurring in the synovial membrane. This suggests a role for T h l cytokine-producing T
cells in the pathogenic mechanisms underlying RA. As a
result, an increase in IFNy over I L 4 production may be
of pathogenic significance and may account for several
of the histomorphologic changes in RA, such as activation of monocytes/macrophages and up-regulation of
HLA class I1 antigens, both of which are prominent
hallmarks of the rheumatoid joint (1,34,35).
Since T h l and Th2 cytokines are believed to have
DOLHAIN ET AL
mutual regulatory interactions, the balance between T
cells with a T h l or Th2 phenotype is considered to be an
important regulatory mechanism of T cell function. This
balance is most profoundly influenced by those factors,
including cytokines, that influence the initial priming of
T cells in the direction of T h l or Th2. During priming,
the monokine IL-12 causes a shift toward a T h l response, whereas the T cell-derived cytokine IL-4 does so
for a Th2 response. The Th2 and monocyte-derived
cytokine IL-10 is able to down-regulate T h l function,
but does not influence the initial priming toward a T h l
response (36). This could be an explanation for the
predominance of T h l over Th2 cytokine-producing cells,
despite the presence of 1L-10 in the rheumatoid joint (37).
Therefore, similar to findings in animal models of
autoimmune disease, treatment strategies that result in
reversal of the observed shift in T cell function could be
of therapeutic benefit (8,9). In experimental arthritis,
infusion of IL-4 alone or in combination with IL-10 has
been shown to have beneficial effects on the severity of
disease (38,39). Based on the concept that such a
restoration of the Thl/Th2 balance may be critical in the
control of chronic inflammation, the Th2 cytokines IL-4
and IL-10 will be tested in clinical trials for their
therapeutic efficacy in RA.
The present study thus demonstrates an imbalance between T lymphocytes that produce T h l and Th2
cytokines at sites of inflammation in patients with RA.
Future research will indicate whether treatment modalities aimed at restoring this imbalance between T h l and
Th2 cells will have therapeutic applicability in the management of RA.
REFERENCES
1. Klareskog L, Forsum U, Scheynius A, Kabelitz D, Wigzell H:
Evidence in support of a self-perpetuating HLA-DR-dependent
delayed-type cell reaction in rheumatoid arthritis. Proc Natl Acad
Sci U S A 79:3632-3636, 1982
2. Panayi GS, Lanchbury JS, Kingsley GH: The importance of the T
cell in initiating and maintaining the chronic synovitis of rheumatoid arthritis. Arthritis Rheum 35:729-735, 1992
3. Mosmann TR, Coffman RL: Heterogeneity of cytokine secretion
patterns and functions of helper T cells. Adv Immunol46:111-147,
1989
4. Romagnani S: Human Thl and Th2 subsets: doubt no more.
Immunol Today 12:256-257, 1991
5. Romagnani S: Lymphokine production by human T cells in disease
states. Annu Rev Immunol 12:256-257, 1994
6. Shehadeh NN, Larosa F, Lafferty KJ: Altered cytokine activity in
adjuvant inhibition of autoimmune diabetes. J Autoimmun 6:291300, 1993
7. Matyniak JE, Reiner SL: T helper phenotype and genetic susceptibility in experimental Lyme disease. J Exp Med 181:1251-1254,
1995
8. Racke MK, Bonomo A, Scott DE, Cannella B, Levine A, Raine
CS, Shevach EM, Rocken M: Cytokine-induced immune deviation
T h l CYTOKINE-SECRETION PROFILE IN RA
as a therapy for inflammatory autoimmune disease. J Exp Med
180:1961-1966, 1994
9. Kuchroo VK, Das MP, Brown JA, Ranger AM, Zamvil SS, Sobel
RA, Weiner HL, Nabavi N, Glimcher LH: B7-1 and B7-2 costimulatory molecules activate differentially the Thl/Th2 development
pathways: application to autoimmune disease therapy. Cell 80:
707-718, 1995
10. Correale J, Gilmore W, McMillan M, Li S, McCarthy K, Le T,
Weiner LP: Patterns of cytokine secretion by autoreactive proteolipid protein-specific T cell clones during the course of multiple
sclerosis. J Immunol 154:2959-2968, 1995
11. Mahanty S, Abrams JS, King CL, Limaye AP, Nutman TB:
Parallel regulation of IL-4 and IL-5 in human helminth infections.
J Immunol 148:3567-3571, 1992
12. Del Prete G, Tiri A, Maggi E, De Cadi M, Macchia D, Parronchi
P, Rossi ME, Pietrogrande M, Ricci M, Romagnani S: Defective in
vitro production of y-interferon and tumor necrosis factor-cy by
circulating T cells from patients with the hyper-immunoglobulin E
syndrome. J Clin Invest 84:1830-1835, 1989
13. Maggi E, Biswas P, Del Prete G. Parronchi P, Macchia D,
Simonelli C, Emmi L, De Cadi M, Tiri A, Ricci M, Romagnani S:
Accumulation of Th-2-like helper T cells in the conjunctiva of
patients with vernal conjunctivitis. J Immunol 146:1169-1174, 1991
14. Miltenburg AMM, van Laar JM, de Kuiper R, Daha MR, Breedveld FC: T cells cloned from human rheumatoid synovial membrane functionally represent the Thl subset. Scand J Immunol
35~603-610, 1992
15. Quayle &, Chomarat P, Miossec P, Kjeldsen-Kragh J, Forre 0,
Natvig JB: Rheumatoid inflammatory T-cell clones express mostly
Thl but also Th2 and mixed (Tho-like) cytokine patterns. Scand
J Immunol38:75-82, 1993
16. Simon A&, Seipelt E, Sieper J: Divergent T-cell cytokine patterns
in inflammatory arthritis. Proc Natl Acad Sci U S A 91:8562-8566,
1994
17. Ritchie DM, Boyle JA, McInnes JM, Jasani MK, Dalakos TG,
Grieveson P, Buchanan WW: Clinical studies with an articular
index for the assessment of joint tenderness in patients with
rheumatoid arthritis. QJM 147:393-406, 1968
18. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF,
Cooper NS, Healey LA, Kaplan SR, Liang MH, Luthra HS,
Medsger TA Jr, Mitchell DM, Neustadt DH, Pinals RS, Schaller
JG, Sharp JT, Wilder RL, Hunder GG: American Rheumatism
Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 31:315-324, 1988
19. Laskay T, Andersson U, Andersson J, Kiessling R, Deley M: An
immunofluorescent method for identifying individual IFNyproducing lymphocytes. J Immunol Methods 951-7, 1986
20. Dolhain RJEM, Andersson U, ter Haar NT, Brinkman BMN,
Verweij CL, Daha MR, Breedveld FC, Miltenburg AMM: Detection of intracellular interferon-y by light microscopy using an
immunoperoxidase technique: correlation with the corresponding
mRNA and protein product. J Leukocyte Biol 54545-551, 1993
21. Van der Meide PH, Dubbeld M, Schellekens H: Monoclonal
antibodies to human immune interferon and their use in a sensitive
solid-phase ELISA. J Immunol Methods 79:293-305, 1985
22. Abrams JS, Roncarolo MG, Yssel H, Andersson U, Gilich GJ,
Silver J E Strategies of'anticytokine monoclonal antibody development: immunoassay of IL-10 and IL-5 in clinical samples.
ImmUnQl.Rev 1275-24, 1992
23. Rebai N, Malissen B, Pierres M, Accolla RS, Corte G, Mawas C
Distinct HLA-DR. epitopes . and distinct families of HLA-DR
molecules defined by 15 nionoclonal antibodies (mAb) either
anti-DR or allo-anti Iak cross-reacting with human DR molecule.
I. Cross-inhibition studies of mAb cell surface fixation and differential binding of. k 4 b :to 'detergent-solubilized HLA molecules
1969
immobilized to a solid phase by a first mAb. Eur J Immunol
13:106-111, 1983
24. Bucy RP, Panoskaltsis-Mortari A, Huang G, Li J, Karr L, Ross M,
Russell JH, Murphy KM, Weaver CT:Heterogeneity of single cell
cytokine gene expression in clonal T cell populations. J Exp Med
18011251-1262, 1994
25. Iannone F, Corrigall VM, Kingsley GH, Panayi GS: Evidence for
the continuous recruitment and activation of T cells into the joints
of patients with rheumatoid arthritis. Eur J Immunol 24:27062713, 1994
26. Salmon M, Kitas GD, Bacon PA: Production of lymphokine
mRNA by CD45R+ and CD45R- helper T cells from human
peripheral blood and human CD4+ T cell clones. J Immunol
1431907-912, 1989
27. Sanders ME, Makgoba MW, Sharrow SO, Stephany D, Springer
TA. Young HA, Shaw S: Human memory T lymphocytes express
increased levels of three cell adhesion molecules (LFA-3, CD2,
and LFA-1) and three other molecules (UCHLl, CDw29, and
Pgp-1) and have enhanced IFNy production. J Immunol 140:14011407, 1988
28. Sanders ME, Makgoba MW, June CH, Young HA, Shaw S:
Enhanced responsiveness of human memory T cells to CD2 and
CD3 receptor-mediated activation. Eur J Immunol 19:803-808,
1989
29. Conlon K, Osborne J, Morimoto C, Ortaldo JR, Young HA:
Comparison of lymphokine secretion and mRNA expression in the
CD45RA+ and CD45RO+ subsets of human peripheral blood
CD4+ and CD8+ lymphocytes. Eur J Immunol25:644-648, 1995
30. Bettens F, Walker C, Cauchat JF, Cauchat D, Wyss T, Pichler WJ:
Lymphokine gene expression related to CD4 T cell subset
(CD45R/CDw29) phenotype conversion. Eur J Immunol19:15691574, 1989
31. Andersson U, Andersson J, Lindfors A, Wagner K, Moller G,
Heusser CH: Simultaneous production of interleukin 2, interleukin 4 and interferon-y by activated human blood lymphocytes. Eur
J Immunol 20:1591-1596, 1990
32. Firestein GS, Zvaifler NJ: How important are T cells in chronic
rheumatoid synovitis? Arthritis Rheum 33:768-773, 1990
33. Gay S, Gay RE, Koopman WJ: Molecular and cellular mechanisms of joint destruction i n rheumatoid arthritis: two cellular
mechanisms explain joint destruction? Ann Rheum Dis 52:S39S47, 1993
34. Halloran PF: Interferon-y, prototype of the proinflammatory
cytokines: importance in activation, suppression, and maintenance
of the immune response. Transplant Proc 25 (suppl 1):10-15, 1993
35. Firestein GS, Zvaifler NJ: Peripheral blood and synovial fluid
monocyte activation in inflammatory arthritis. I. A cytofluorographic study of monocjte differentiation antigens and class I1
antigens and their regulation by y-interferon. Arthritis Rheum
301857-863, 1987
36. Seder RA, Paul WE, Davis MM, Fazekas de St. Groth B: The
presence of interleukin 4 during in vitro priming determines the
lymphokin'e-producing potential of CD4+ T cells from T cell
receptor transgenic mice. J Exp Med 176:1091-1098, 1992
37. Katsikis PD, Chu CQ, Brennan FM, Maini RN, Feldmann M:
Immunoregulatory role of interleukin 10 in rheumatoid arthritis. J
Exp Med 179:1517-1527,' 1994
38. Allen JB, Wong HL, Costa GL, Bienkowski MJ, Wahl SM:
Suppression of monocyte function and differential regulation of
IL-1 and IL-lra by IL-4 contribute to resolution of experimental
arthritis. J Immunol 151:4344-4351, 1993
39. Van den Berg WB, Joosten LAB, Jacobs M, Helsen MA, Durez P,
Delvaux A, Velu T, Goldman M: Suppression of established
collagen-induced arthritis with IL-4 and IL-10 (abstract). Arthritis
Rheum 37' fsuppl 9):S280, 1994