Immunoregulatory Role of Interleukin 10 in
By Peter D. Katsikis, Cong-Qiu Chu, Fionula M. Brennan,
Ravinder N. Maini, and Marc Feldmann
From The Kennedy Institute of Rheumatology, Hammersmith, London W6 8LW,, UK
The presence and the role of interleukin 10 (IL-10), a potent cytokine synthesis inhibitory factor
and antiinflammatory cytokine, were investigated in rheumatoid arthritis (RA). The expression
of both mRNA and protein for IL-10 could be demonstrated in RA and osteoarthritis (OA)
joints. Human IL-10 mRNA could be demonstrated by polymerase chain reaction amplification
of cDNA made by reverse transcription of total RNA extracted directly from synovial tissue
in five out of five RA and four out of five OA patients. IL-10 protein was demonstrated by
specific immunoassay and immunohistology. IL-10 protein was spontaneously produced in all
11 RA and 17 OA synovial membrane cultures investigated, and this production was sustained
for up to 5 d in culture in the absence of any extrinsic stimulation. IL-10 protein could also
be detected by immunohistology in all five RA and four OA synovial membrane biopsies investigated,
but not three normal synovial membranes. Immunohistology revealed that the IL-10 was localized
to the synovial membrane lining layer and mononuclear cell aggregates. Immunofluorescence
double staining revealed that the sources of IL-10 were monocytes in the lining layer, and T
cells in the mononuclear cell aggregates. We found evidence that the IL-IO expression was
functionally relevant, as neutralization of endogenously produced IL-10 in the RA synovial
membrane cultures resulted in a two- to threefold increase in the protein levels of proinflammatory
cytokines tumor necrosis factor c~ (TNF-c 0 and IL-1B, although IL-6 and IL-8 levels were not
affected. The addition of exogenous recombinant IL-IO to the RA synovial membrane cultures
resulted in a two- to threefold decrease in the levels of TNF-ol and IL-1B. IL-8 levels were reduced
by day 5; however, IL-6 levels were not affected by exogenous IL-10. Neutralization of the
endogenous IL-10 in two out of seven RA synovial membrane cultures resulted in the expression
of detectable levels of interferon 7 (561-1,050 pg/ml). Taken together, the above findings suggest
that IL-10 is spontaneously produced in RA and OA and is an important immunoregulatory
component in the cytokine network of RA, regulating monocyte and in some cases T cell cytokine
leads to joint destruction (1). It is now established that
proinflammatory cytokines such as TNF-o~, IL-1, GM-CSF,
and IL-6 are all produced by the synovial membrane in RA,
and are considered to be important participants in the patho-
physiology of the disease (2-9). However, in addition to these
proinftammatory cytokines, a compensatory antiinflamma-
tory response is also observed in RA synovial membranes.
Thus there is expression of high levels of IL-1 receptor an-
tagonist (IL-1RA), soluble TNF receptors (of both the 55-
eumatoid arthritis (RA) t is an autoimmune disorder
that is characterized by a chronic synovitis which often
1 Abbreviations used in this paper: IL-1RA, IL-1 receptor antagonist; OA,
osteoarthritis, KA, rheumatoid arthritis.
and 75-kD receptor) and TGF-/5 in the RA synovium (8,
10-12, and Brennan, E M., manuscript in preparation), which
suggests that homeostatic mechanisms exist in the rheuma-
toid joint by which the immune system attempts to contain
the inflammation and limit joint destruction.
In contrast to the abundance of monocyte-derived pro-
inflammatory cytokines described above, T cell-derived
cytokine proteins have often proven difficult to detect in RA
synovium (13, 14). This is rather paradoxical, since the syno-
vial membrane-infiltrating T cells appear phenotypically ac-
tivated (15, 16). However, mRNA for T cell cytokines such
as IFN-3' and IL-2 has been demonstrated (17), and this sug-
gests that there may be T cell cytokine inhibitory factors oper-
ative in the rheumatoid synovium and/or cytokine consump-
tion. In addition, T cell proliferative responses to mitogens
are impaired in RA (18-20), again suggesting the presence
ofinhibitors. The report that most T cell clones derived from
J. Exp. Med. y The Rockefeller University Press i 0022-1007/94/05/1517/11 $2.00
Volume 179 May 1994 1517-1527
the RA synovium are of the Thl cell type (21, and Cohen,
S., unpublished observations), which produce predominantly
IFN-'y and IL-2 (22), raises the possibility that one of the
inhibitory factors present in RA could be IL-10, originally
defined as "a cytokine synthesis inhibitory factor" capable
of suppressing cytokine synthesis by murine Thl CD4 + cells
(23, 24). However the presence of IL-10 in RA synovium
remains to be documented.
IL-10 is a 35-kD homodimeric cytokine, which is produced
by human T cells, B cells, and monocytes (23, 25, 26). In
humans, IL-10 exhibits a cytokine synthesis inhibitory ac-
tivity (CSIF), inhibiting IFN-3,, GM-CSF, IL-4, and IL-5
production by T cells and can reduce Ag-specific T cell prolifer-
ation (23, 24). The latter chiefly involves inhibiting the ca-
pacity of monocytes to present Ag, a phenomenon mediated
at least in part through the downregulation of HLA class
II on monocytes (27). IL-10 has also been shown to have po-
tent antiinflammatory properties. IL-10 inhibits the synthesis
ofproinflammatory cytokines such as TNF-ot, IL-1, and GM-
CSF, as well as HLA class II expression by human mono-
cytes stimulated with LPS (25).
The reported antiinflammatory and CSIF properties of 1L-10
led us to investigate whether IL-10 is produced in RA and
osteoarthritis (OA), and whether it plays an important im-
munoregulatory role, regulating monokine and T cell-de-
rived cytokine production. Here we describe the presence of
IL-10 mRNA and protein in RA and OA synovial mem-
brane biopsies. We also report spontaneous IL-10 produc-
tion in RA and OA synovial membrane cultures, and evi-
dence that suggests that it has an important role in the
synovium as a regulator of production of the inflammatory
cytokines TNF-ot and IL-1. IL-10 may thus have a possible
role in new therapeutic approaches to inflammatory arthritis.
Materials and Methods
Patient Samples and Cell Cultures. 11 patients with RA fulfilling
the revised American Rheumatology Association (AIL/i) criteria
(28), and 17 patients with OA, all of whom underwent joint replace-
ment surgery, were included in this study. Synovial membrane cell
suspension cultures were prepared by collagenase and DNase diges-
tion of membranes as previously described (12). In brief, synovial
membrane tissue was digested in ILPMI 1640 containing 5% FCS
(Flow, High Wycombe, UK), 1 rag/m1 collagenase type A
(Boehringer, Mannheim, Germany), and 0.15 mg/ml DNase type
1 (Sigma Chemical Co., Poole, UK) for 2 h at 37~ After incuba-
tion, the tissue was pipetted through a 200-#m 2 nylon mesh into
a sterile beaker. Cells were then washed three times with ILPMI
1640 plus 5% FCS. Synovial membrane cells were cultured at
10+/ml/well in RPMI plus 5%/FCS in 24-well plates (Nunc, Ux-
bridge, UK) at 37 ~ in a 5% CO2 incubator, and supernatants were
harvested at 24, 72, and 120 h. Synovial membrane cell cultures
were treated with 2/zg/ml of neutralizing rat monoclonal anti-ID
10 antibody 9D7, rat isotype control antibody GLl13, or 10 ng/ml
rlL-10. Supernatants were frozen at -20~ until assayed. Synovial
fluid exudates from nine RA and three OA patients were aspirated
during needle biopsy, cells were spun down at 1,000 rpm, and su-
pernatants were assayed at a 1:2 dilution in PBS in the II+-10 and
isotype control ELISA. Synovial fluids were also assayed for IL-IO
after treatment with hyaluronidase. In brief, synovial fluids were
incubated with 4 U/m1 hyaluronidase (Sigma Chemical Co.) for
30 rain at 37~ (11), after centrifugation at 2,200 rpm for 12 min.
Supernatants were then assayed in the ID10 and isotype control
ELISA. All media and reagents were found to have <0.1 EU/ml
endotoxin contamination by the chromogenic Limulus Amoebo-
cyte Lysate assay (BioWhittaker, Walkersville, MD). The only ex-
ception was the rat monoclonal anti-IL-10 neutralizing antibody
9D7 which at the concentration of 2 #g/ml used had 0.13 EU/ml
endotoxin contamination. This led to the inculsion of 10/~g/ml
Polymyxin B (Sigma Chemical Co.) controls in three of the cul-
tures, to exclude the effect of endotoxin. In three other cultures,
the endotoxin-free rat monoclonal neutralizing anti-IL-10 antibody
12G8 at 2 #g/ml was also included as another control for the 9D7
ELISAs and Antibodies. Recombinant human II+-10, rat mono-
clonal anti-IIA0 neutralizing IgG1 antibody 9D7, rat monoclonal
isotype control IgG1 antibody GLl13, rat monoclonal neutralizing
IgG2 antibody 12G8, and specific ELISA for human/viral IL-10,
were generously provided by Drs. K. Moore andJ. Abrams (DNAX,
Palo Alto, CA). The sensitivity of the IL-IO ELISA was 100 pg/ml.
A control ELISA which differed from the IL-10 ELISA only in that
the coating antibody was the nonspecific isotype control mAb GLl13
was also used to exclude the effect of RF. A specific TNF-a ELISA
was kindly provided by Dr. W. Buurman (University of Limburg,
Maastricht, The Netherlands). The IL-1B ELISA was commercially
purchased (R & D Systems, Inc., Minneapolis, MN). The IL-6,
II+-8, and IFN-'y ELISAs were gifts from Drs. F. di Padova (Sandoz
Pharma AG, Basel, Switzerland), W. Lee Wong (Genentech Inc.,
S. San Francisco, CA), and D. Novick (Weismann Institute, Re-
hovot, Israel), respectively. Neutralizing anti-TNF~ and anti-ID
13 mouse mAbs A2 and IL-1RA were kind gifts from Dr. P. Dad-
dona (Centocor, Malvern, PA) and Dr. D. Tracey (Upjohn,
Kalamazoo, MI). Apart from the ILl0 ELISA, the neutralizing
anti-IL-10 antibodies did not interfere in any of the above ELISA
Immunohistology. Synovial membrane biopsies of patients with
RA (n = 5) and OA (n = 4) were taken at arthroscopy of knee
joints and snapffozen in iso-pentane cooled on liquid nitrogen.
Normal synovial membrane specimens (n = 3) were obtained from
the knee joint of patients undergoing amputation for osteogenic
sarcoma at a site distant for the knee joint. The synovial mem-
branes were sectioned at 6 #m and immunohistochemical staining
was performed as described previously (8). The sections were fixed
in acetone at -20~ and blocked with 20% normal goat serum,
then incubated with 10/~g/ml rat monodonal IgG1 anti-IL-10 an-
tibody 2A5 (DNAX) overnight at 4~ Binding was detected by
incubating the sections twice with goat anti-rat Ig alkaline phos-
phatase conjugate (Dako, High Wycombe, UK) for 30 rain at room
temperature, followed by alkaline phosphatase-antialkaline phos-
phatase (APAAP) complexes (Dako) for 30 min at room tempera-
ture and visualized with Fast red (4-chloro-2-methylbenzenediazo-
nium zinc chloride). Normal rat monoclonal IgG1 (Serotec,
Kidlington, UK) at the equivalent protein concentration served
as a negative control. The staining of the 2A5 anti-Ibl0 antibody
was abrogated by preincubating the antibody with 10 #g/ml human
rIL-10 for 2 h at 37~ before application to the sections. For double
immunofluorescence staining, the anti-IL-10 antibody was detected
with a biotinylated donkey anti-rat IgG with minimum cross-
reactivity to mouse serum proteins (Jackson Immunoresearch,
Luton, UK) and followed by streptavidin-fluorescein (Amersham,
Aylesbury, UK). Mouse anti-human CD3, CD20, and CD68 mAbs
(Dako) were detected with Texas red--conjugated donkey anti-mouse
IgG with minimum cross-reactivity with rat serum proteins (Jackson
Immunoregulatory Role of IL-10 in Rheumatoid Arthritis
PCR Amplification and Northern Blot. RNA was extracted
directly from synovial membrane tissues by the guanidinium
isothiocyanate method (29). To synthesize cDNA, 1/zg of total
RNA was incubated with 2/xg oligo (dt)02_ls) primers (Pharmacia,
Milton Keynes, UK), 5 ng BSA, 0.5 mM dNTPs, 10 mM DTT,
10/zl xl0 RT buffer (Y00146; GIBCO BILL, Uxbridge, UK), and
20 U murine Moloney leukemia virus reverse transcriptase (GIBCO
BILL) (30). PCIL reactions were performed using 5/A of this cDNA
in the presence of 50 mM KC1, 10 mM Tris, pH 8.3, 1.5 mM
MgC12, gelatin (0.001%), 0.2 mM dNTPs, 1 #M of each primer,
and 1.25 U Taq Polymerase (Perkin Elmer Cetus, Beaconsfield, UK)
in a total of 50/A. PCR analysis was performed using two human
IL-10 27-mer primers, spanning nucleotides 323-349 (sense strand)
and 648-674 (antisense strand) (23). These primers do not amplify
the viral homologue of I1r BCILF1. Samples were overlaid with
50 #1 of light mineral oil (Sigma Chemical Co.) and were incubated
at 94~ for 1 min, 55~ for 1 min, and 72~ for 2.5 min. This
cycle was repeated 35 times using a DNA thermal cycler (Perkin
Elmer Cetus) (31). 10 /xl PCIL products were electrophoresed
through a 6% polyacrylamide gel and visualized by ethidium bro-
mide staining. 10/A amplified DNA was digested with 1 U HaelII
restriction enzyme (New England Biolabs, Bishop's Stortford, UK)
for 1 h at 37~ and yielded the expected 123- and 229-bp size
For Northern blot analysis, total ILNA was extracted as above
from single cell suspension cultures of two RA synovial membranes
which were treated for 24 h with either nothing, 2/~g/ml anti-Ib
10 neutralizing rat mAb 9D7, 2/~g/ml isotype control rat mAb
GLll3, or 10 ng/ml rill0. Total RNA was size fractionated by
electrophoresis after denaturation with glyoxal and DMSO and then
transferred to a Hybond-N nylon membrane (Amersham). After
transfer, the membrane was baked for 2 h at 80~
ization, glyoxal was removed from RNA by washing the filter with
20 mM Tris-HC1, pH 8.0, at 65~
bridized in prehybridization buffer: 6 x SSC (1 x SSC = 150 mM
NaC1, 15 mM Na Citrate, pH 7.0), 5 x Denhardt's reagent (1 x
Denhardt's = 0.05% Ficoll 400, 0.05% polyvinylpyrrolidone,
0.05% BSA), 0.5% SDS, 100 #g/ml salmon sperm DNA (Sigma
Chemical Co.), and 50% formamide (BDH, Poole, UK) for 2 h
at 42~ Membranes were hybridized overnight at 42~ with a
cDNA probe to TNF-ot (kind gift H. M. Shepard, Genentech) la-
beled with ~_[32p] dCTP (Amersham). Membranes were washed
with 2x SSC and 0.5% SDS at room temperature, 2x SSC and
0.1% SDS at room temperature, 0.1 x SSC, and 0.5% SDS at 37~
and 0.1x SSC and 0.5% SDS at 65~
posed to Hyperfilm-MP (Amersham) at - 70~ The TNF-c~ probe
was stripped from membranes by immersing membranes in 1 mM
Tris-HC1, pH 8.0, 1 mM EDTA, pH 8.0, and 0.1x Denhardt's
reagent for 2 h at 75~ Membranes were prehybridized and then
hybridized with a probe to the 7B6 housekeeping gene (32). Au-
toradiographs were scanned on a dual wavelength TLC scanner C8-
930 (Shimadzu, Kyoto, Japan).
Statistics. Results were analyzed with Wilcoxon's signed rank
The membrane was prehy-
Membranes were then ex-
IL-IO Is Produced by RA and OA Synovial Membranes. IL-10
was produced by all 11 RA and 17 OA synovial membrane
cultures tested. The IL-10 levels were measured by specific
ELISA, and at 24 h ranged from 358 to 5,501 pg/ml for RA
Figure 1. ILl0 protein and mtLNA is expressed in RA (O) and OA
(A) synovial membranes. (A) Ibl0 is produced spontaneously by RA
(n = 11) and OA (n = 17) synovial membrane cultures, assayed by ELISA
at 24 h. (B) PCR of human IF10 and 3-actin amplified from cDNA made
by reverse transcription of total RNA extracted directly from synovial tissue.
Synovium from five RA patients (lanes I-5), synovium from five OA pa-
tients (lanes 6-10), unstimulated normal PBMs (lane 11), and anti-
CD3-stimulated normal PBMs (lane 12).
1519 Katsikis et al.
Figure 2. Immunohistological staining for Ibl0 in RA synovial membranes. (.-4) Ibl0 staining in the synovial membrane lining layer. (B) Preincuba-
tion of anti-IL-10 antibody with rlL-10 results in abrogation of staining. (C) Lack of staining with isotype control rat mAb. (D) Ibl0 staining in the
RA synovial membrane mononuclear aggregates. (E) No II:10 staining detected in normal synovial tissue. (F and G) IL-10 and monocyte immunofluores-
cence double staining in RA synovial membrane lining layer. (F) Ibl0 staining (9I)7 anti-IL-10 rat mAb, followed by biotinylated donkey anti-rat
IgG and streptavidin-fluorescein). (G) Monocyte/macrophage staining (anti-human CD68 mouse mAb, followed by Texas red-conjugated donkey anti-mouse
IgG antibody). (H and/) IL-10 and T cell immunofluorescence double staining in RA synovial membrane mononuclear aggregate. (/-/) IL-10 staining
(9D7 anti-Ibl0 rat mAb, followed by biotinylated donkey anti-rat IgG and streptavidin-fluorescein). (/) T cell staining (anti-human CD3 mouse mAb,
followed by Texas red-conjugated donkey anti-mouse IgG antibody). (Arrows) Individual cells that double stain for IL-10 and CD3.
and 310 to 3,390 pg/ml for OA cultures (Fig. 1 A). IL-10
levels remained detectable up to 5 d in culture. No difference
in IL-10 was observed between synovial cultures of RA and
OA patients. IL-10 was not detected (<100 pg/ml) in either
PBS-diluted or hyaluronidase-treated synovial fluids of nine
RA and three OA patients (data not shown). All samples
were tested in the control ELISA (isotype-matched, antibody-
coated plate) to exclude RFs and showed no binding. IL-IO
mRNA was demonstrated by PCR amplification of reverse
transcribed cDNA in five out of five RA patients, and four
out of five OA patients (Fig. 1 B). IL-10 was also detected
by immunohistological staining of synovial membrane tissue
in all five RA and four OA membranes examined. Cell staining
with IL-IO was found in the synovial membrane limiting layer
and the mononuclear aggregates. (Fig. 2). This staining was
shown to be specific for IL-10 since it could be abrogated
by preincubation of 2A5 staining antibody with recombinant
human IL-10, and the isotype control mAb yielded no staining.
(Fig. 2). There was no IL-10 staining in the three normal
synovial membranes examined (Fig. 2). By immunofluores-
cence double staining, IL-10 was shown to be produced by
monocytes in the lining layer and T cells in the lymphocytic
aggregates (Fig. 2).
Regulatory Role of lL-lO in Proinflammatory Cytokine Produc-
tion in Rheumatoid Synovial Membranes. Blocking IL-IO with
the neutralizing rat monoclonal anti-IL-10 antibody 9D7
resulted in an increase in the proinflammatory cytokines pro-
duced by the rheumatoid synovial membrane cultures. After
24 h in culture, the TNF-o~ and IL-1B levels measured by
specific ELISAs increased to 208 _+ 16% and 198 + 25%
(x -+ SEM) of the control levels respectively (p <0.01 and
p <0.01). The effect of anti-IL-10 was sustained up to day
5, when TNF-c~ was increased to 322 _+ 91% and IL-1B
to 235 + 65% of control levels. In contrast, blocking IL-10
had no significant effect on IL-6 and IL-8 production (Table
1, Figs. 3 and 4). Northern blot analysis was performed on
1520 Immunoregulatory Role of ID10 in Rheumatoid Arthritis
two 24-h RA synovial membrane cultures. Treatment of these
two cultures with anti-IL-10 resulted in a twofold increase
in TNF-ol protein produced, however the TNF-o~ mKNA
levels did not change (data not shown). Addition of 10 #g/ml
polymyxin B to three RA synovial membrane cultures treated
with the 9D7 antibody, and treatment of three RA mem-
branes with the endotoxin-ffee neutralizing anti-IL-10 anti-
body 12G8 gave identical results to the 9D7 treatment (data
not shown). The rat isotype control mAb GLl13 had no effect
on TNF-ol and IL-I~ produced by the synovial membranes
(Fig. 3). It was of interest, in view of the original function
of IL-10 described (cytokine synthesis inhibition) that anti-
IL-10 treatment resulted in detectable levels of IFN-y in two
RA membranes (561 and 1,050 pg/ml, respectively) out of
seven 24-h cultures that were assayed. In contrast, in the control
cultures, or isotype control antibody-treated cultures, IFN-'y
levels were undetectable (<100 pg/ml).
Exogenous Addition of lL-lO to Synovial Membrane Cultures
Inhibits Proinflammatory Cytokine Production. The levels of
IL-10 produced by the synovium membrane cultures were
all, with the exception of one (5 ng/ml), <2 ng/ml. We there-
fore investigated whether the IL-10/IL-10 receptor system
in these cultures was saturated, and whether addition of ex-
ogenous IL-10 would have any effect on monokine produc-
tion. Addition of 10 ng/ml of IL-10 resulted in marked de-
crease in cytokine production. At 24 h, TNF-o~ was reduced
to 49 _+ 6% and IL-13 to 64 _+ 11% of control (p <0.01
andp - 0.02, respectively), and this reduction was even more
marked by day 5 (TNF-ol to 42 + 9% and IL-13 to 51 -+
9% of control) (Table 2, Figs. 3 and 4). Exogenous IL-10
did not reduce TNF-ot mR.NA levels at 24 h in two RA
cultures in which the IL-10 protein levels were reduced by
twofold (data not shown). IL-6 levels were not affected by
1521 I~tsikis et al.
Table 1. Effect of Anti-IL-lO Treatment on Cytokine Production by Synovial Membrane Cultures
TNF-ot IL-1B IL-6 IL-8
O'S/toO O'S/toO (ng /mO (ng /mt)
1,161 + 370 (n = 10)*
2,236 • 649
1,168 + 380
1,549 _+ 659 (n = 9)
2,899 _+ 991
1,542 + 660
151 _+ 61 (n = 7)
160 + 61
142 _+ 56
242 _+ 75 (n = 7)
271 + 76
258 • 79
301 • 133 (n = 8)
1,025 + 505
295 + 134
1,630_+ 1,159(n = 6)
2,513 _+ 1,665
1,761 • 1,284
264 + 83 (n = 7)
280 _+ 80
276 + 83
473 _+ 148 (n = 7)
514 + 142
510 +_ 161
185 + 100 (n = 6)
1,024 • 781
173 • 92
472 _+ 182 (n = 5)
1,170 + 516
381 • 210
269 _+ 66 (n = 6)
272 • 59
282 + 73
438_+ 100 (n = 6)
535 _+ 121
419 + 95
* Mean _+ SE, n = number of cultures.
exogenous IL-IO addition to the RA cultures. IL-8 produc-
tion was only affected by day 5, reduced to 68 + 7% of con-
trol (p = 0.05). (Table 2, Figs. 3 and 4).
Blocking of TNF-cr and IL-1 Downregulates IL-IO Production
by the Synovial Membrane Cultures. IL-1RA (10 #g/ml) con-
sistently inhibited IL-IO production, IL-10 was reduced at
24 h to 68 _+ 8%, at day 3 to 57 + 7%, and at day 5 to
50 + 13% of control. (data not shown). Anti-TNF-c~ neu-
tralizing mAb antibody A2 (5 #g/ml) also resulted in a de-
crease, albeit lesser, in IL-IO levels produced by the synovial
membrane cultures at 24 h to 80 _+ 8% (p <0.05), at day
3 to 76 _+ 8%, and at day 5 to 75 + 11% of control (data
not shown). However, anti-TNF-ce and IL-1RA, when used
in combination, showed no additive or synergistic effect (data
not shown), suggesting that TNF-c~ and IL-1 regulate IL-10
levels by a common pathway,
TNF~ 2000 ]
1 3 5
1 3 5
i i i
1 3 5
1522 Immunoregulatory Role of IL-10 in Rheumatoid Arthritis
Figure 3. RA synovial mem-
brane cultures: effect of neutralizing
endogenously produced II.-10 and of
the addition of exogenous 1I.-10.
Representative experiment showing
TNF-ce, IL-1B, IL-6, and IL-8 levels
produced by RA synovial mem-
brane culture SM1376. Synovial
membrane culture was treated with
either neutralizing rat monoclonal
anti-ILl0 antibody 9D7 (2 #g/ml),
isotype-matched control rat mono-
clonal GLl13, or 10 ng/ml ofrlbl0
for 24, 72, and 120 h. (Similar
studies have been performed on 5-10
membranes; see Table 1). (-K]-)
Control; (-0-) anti-IL-10; (-O--)
isotype control; (-I-) IL-10.
Figure 4. Neutralization of spontaneously produced IL-10 and addition
of exogenous II.,10 in RA synovial membrane cultures. Pooled data from
5-10 RA synovial membrane cultures. (A) Neutralization of endogenous
Ibl0 with the 9D7 neutralizing rat monodonal anti-IL-10 antibody (2
#g/ml) results in increased levels of TNF-c~ and IL-1B, but not Ib6 and
Ib8. (B) Exogenous rlL-10 (10 ng/ml) decreases TNF-c~ and IL-1/~ levels
in RA cultures. (Dashed lines) 100% of control untreated cultures.
As judged by quantity, the major participants in the cytokine
network of rheumatoid synovitis are the monocyte and fibro-
blast-derived inflammatory cytokines, including TNF-ot,
IL-1/3, IL-6, GM-CSF, and IL-8. These are produced sport-
taneously in the diseased synovial membrane, and are consid-
ered to play an important role in the pathophysiology of the
chronic synovitis (33-35). Recently, TGF-B and cytokine in-
hibitors such as IL-1RA and soluble TNF receptors (both
p55 and p75 ffNFK) have also been found in synovial fluids
and shown to be produced by synovial membrane cultures
from RA patients (8, 10-12, and Brennan, F. M., manuscript
in preparation), and presumably represent a homeostatic at-
tempt to contain the inflammation. The recent description
of the potent antiinflammatory properties of IL-10 (23, 25)
prompted us to investigate whether IL-10 was produced by
the synovial membrane in RA, and if it played a regulatory
role in the cytokine network in synovitis.
In this study, IL-10 was detected at both the mRNA and
protein level in RA synovial membranes. PCR of reverse tran-
scribed total RNA extracted directly from synovial membrane
tissue showed that human IL-10 mRNA is present in these
membranes. Biologically significant quantities of IL-10 were
spontaneously produced in cell suspension cultures of RA
synovial membranes, and persisted for the 5-d duration of
the synovial cultures. IL-10 protein was also demonstrated
in RA and OA synovial tissues by immunohistology, but
not in the normal synovium. IL-10 immunostaining showed
that monocytes in the synovial membrane lining layer, and
T cells in the mononuclear aggregates stained for IL-10. IL-IO
however could not be detected in RA and OA synovial fluids
The IL-10 produced spontaneously by the RA cultures was
shown by neutralization experiments to play an important
regulatory role in KA in the synovial membrane cultures.
Blocking IL-10 in these cultures with a neutralizing mAb
resulted in a two- to threefold increase in the production of
TNF-cz and IL-1/3. This result probably has pathogenetic con-
sequences, as TNF-cx and IL-1/3 are implicated in the pathology
of KA, since they can both induce bone resorption and carti-
lage destruction, and can stimulate PGE2 release and col-
lagenase production (33), and blocking TNF-o~ in RA pa-
tients with a high affinity neutralizing chimeric anti-TNF-cz
antibody has been shown to result in clinical benefit (36).
The suppression of these cytokines by IL-10 thus appears to
Table 2. Effect of lL-10 Treatment on Cytokine Production by Synovial Membrane Cultures
TNF-o~ IL-1/~ IL-6 IL-8
Day 1 Control
1,145 + 418 (n = 9)*
563 • 210
1,708 -+ 850 (n = 7)
1,098 +_ 725
135 • 63 (n = 7)
125 + 56
227 + 77 (n = 7)
200 + 69
Day 3 Control
301 • 133 (n = 8)
106 • 36
1,563 + 1,168 (n = 6)
1,317 • 1,041
248 _+ 86 (n = 7)
242 • 83
479 • 147 (n = 7)
449 • 135
Day 5 Control
185 _+ 100 (n = 6)
55 • 24
407 _+ 147 (n = 5)
204 • 89
241 +_ 65 (n = 6)
231 -+ 60
438 • 100 (n = 6)
308 • 76
* Mean +_ SE, n = number of cultures.
1523 Katsikis et al.
be of potential importance in diminishing the synovial inflam-
mation. The observation that IL-10 upregulates the produc-
tion of IL-1RA and soluble TNF receptors by human mono-
cytes (Joyce, D., D. Gibbons, P. Green, M. Feldmann, and
F. M. Brennan, manuscript submitted for publication) also
supports this concept. There is thus increasing evidence that
in RA, IL-10, IL-1RA, and soluble TNF receptors are con-
stituents of an important antiinflammatory response. It is
noteworthy that blocking IL-IO in the RA cultures had no
effect on the IL-6 and IL-8 levels, which was somewhat un-
expected, since IL-IO inhibits both IL-6 and IL-8 production
by LPS-activated human monocytes (25). A possible expla-
nation may lie in the fact that in P,A synovial membrane
cultures another important source of these two cytokines are
fibroblasts and endothelial cells (6, 37, 38). In our labora-
tory, IL-10 has failed to have any significant effect on syno-
vial fibroblast IL-6, IL-8 or PGE2 production (Butler, D.,
F. M. Brennan, and M. Feldmann, unpublished observation).
Exogenous addition of IL-IO to RA synovial membrane
cultures resulted in a two- to threefold decrease of TNF-cz
and IL-13 production, noticeable by 24 h, the first assay point.
However, IL-8 was only affected by day 5 in culture, and this
may reflect an indirect effect oflL-10 on IL-8 expression sub-
sequent to downregulating IL-1 and TNF. As mentioned
above, IL-IO induces soluble TNF receptors and IL-1RA
production by human monocytes, and thus the net effect of
IL-IO on the biological activity of TNF-c~ and IL-1 in the
synovial membranes would be expected to be even greater
than that resulting exclusively from downmodulation of
TNF-(x and IL-1 protein levels. The above results also indi-
cate that the IL-IO/IL-IO receptor system is not saturated
in these RA cultures. This implies that there is a "relative
deficit" in IL-IO production in RA, which suggests the pos-
sibility of a novel therapeutic strategy for RA by augmenting
TGF-3 and IL-4 are both cytokines with potent antiinflam-
m atory properties (39, 40), and therefore could possibility
be antiinflammatory agents of use in RA. However in ex-
perimental animal models of arthritis, TGF-3 treatment has
yielded conflicting results (41, 42). In a report using RA sy-
novial membrane explants cultured for 10 d, IL-4 was shown
to inhibit IL-6, IL-13, and TNF-cz (43). However, the failure
to characterize the cellular sources of cytokine production
during the 10-d period of culture limits the interpretation
of these results. In contrast to IL-IO, TGF-3 and IL-4 have
both failed in our hands to inhibit inflammatory cytokine
production in RA synovial membrane cell suspension cul-
tures (12, 44). Suppression of LPS induced PBMC monokine
production by TGF-~/, and IL-4 depends on the pretreatment
of these cells with these cytokines (12, 45). Thus, the in-
ability of TGF-3 and IL-4 to have any effect in RA cultures
could be due to the chronic activation of monocytes. IL-IO
is therefore the first cytokine shown to have an inhibitory
effect on RA synovial membrane culture cytokine synthesis.
IL-1 and TNF-cz were found to participate in the regula-
tion of IL-10 production by the RA synovial membrane. This
is in concordance with our recent data which indicate that
IL-1, and to a lesser extent TNF-cz, are both involved in the
regulation of IL-10 production by LPS-activated human mono-
cytes (Katsikis, P., unpublished observation). In the RA cul-
tures it is possible that the anti-TNF-cz-mediated downregu-
lation of IL-IO may be the indirect result of its inhibition
of IL-1 (46). This possibility is also supported by the lack
of an additive or synergistic effect of anti-TNF-oe with IL-
1RA, when used in combination in the RA synovial mem-
An interesting paradox in the studies of the pathogenesis
of RA which has been highlighted by Firestein et al. (13,
14), has been the failure to detect abundant T cell cytokines,
although activated T cells as judged by cell surface markers
are present in the synovium (15, 16). However, we have pre-
viously reported the detection of T cell cytokine mRNA (IL-2
and IFN-'y) in the RA synovium (17), suggesting that an
inhibitor may be present and responsible for the low protein
detection. Cloning of these T cells has revealed a Thl cytokine
profile (21, and Cohen, S., unpublished observations), making
IL-10 a candidate for such an inhibitor, since IL-IO has been
reported to be a potent inhibitor ofThl cytokine production
(23, 24, 27). Preliminary data shows that in two out of seven
RA membranes studied, IFN-7 which was undetectable in
control cultures, could be measured at significant concentra-
tions after the neutralization of endogenous IL-IO. Clearly,
more rheumatoid membranes need to be studied, but the data
already obtained suggest that IL-IO may be the (or one of
the) factor responsible for the "elusiveness" of T cell-derived
cytokines. T cell proliferative responses are also impaired in
RA (18-20). TGF-3 has been implicated as a major but not
the sole immunosuppressive factor in RA synovial fluids (20).
The direct inhibitory effect of IL-IO on T cell proliferation
and IL-2 production (48, 49), and its indirect effect via APCs
(24, 27, 50) raises the possibility that IL-IO also contributes
to the low T cell proliferative responses in RA. This, how-
ever, remains to be established.
The data reported here indicate that in RA, IL-10 plays
an important role in the cytokine network, by inhibiting the
cytokine production of both monocytes and T cells. In the
RA synovium, IL-IO regulates the production of TNF-oJIL-1
and vice-versa. In rheumatoid synovium, IL-10 and IFN-y
may be reciprocally regulated also, IL-IO possibly inhibiting
IFN-y production, and IFN-y regulating IL-10 (51). Although
most of the properties of IL-IO suggest that it has a negative
regulatory role, recent observations indicate that IL-IO may
not be a general inhibitor of immune responses, since it can
stimulate monocyte expression of the Fc receptor for mono-
meric IgG, FcyRI (CD64), and enhance antibody-dependent
cellular cytotoxicity (52). Another recent report (53) indi-
cates that IL-IO might be a potent recruitment signal (or in-
ducer of) for leukocyte migration in vivo. Thus, animal studies
are needed to establish whether IL-10 treatment in vivo may
modulate the arthritis. Recent reports of IL-IO treatment of
experimental endotoxemia in mice have shown that in vivo
administration of IL-IO significantly reduces TNF-o~ levels
and mortality (54, 55), suggesting that IL-IO could be an
effective antiinflammatory in vivo.
1524 Immunoregulatory Role of IL-10 in Rheumatoid Arthritis
IL-10 was also detected in OA synovial membrane cultures
and biopsies. Monokine-derived proinflammatory cytokines
such as TNF-ot, IL-1, and GM-CSF (9, 46) and antiinflamma-
tory products such as soluble TNF receptors and IL-1RA
(7, 8, 10, 11) are all produced by the OA synovial membrane
macrophage. The role of proinflammatory cytokines in the
pathogenesis of osteoarthritis, which is considered a nonau-
toimmune but inflammatory disease, has not yet been studied
in detail. It seems that in OA as in R.A, IL-10 may be pro-
duced alongside soluble TNF receptors and IL-1RA to limit
the deleterious effect of proinflammatory cytokines.
In this study we have shown that both mRNA and pro-
tein for IL-10 are present in the RA and OA synovium, and
that IL-IO is an important participant in the cytokine net-
work of the rheumatoid synovial membrane, playing an im-
munoregulatory role in inflammatory and possibly T cell
cytokine production. Reciprocally, IL-10 itself is regulated
by IL-1 and TNF-ot, thus IL-10 appears to be an important
component of the complex cytokine network of rheumatoid
synovitis. Finally, exogenous IL-10 was shown to inhibit both
TNF-ot and IL-1/~ production by RA synovial membrane
cultures. These findings raise the possibility of new therapeutic
strategies for the treatment rheumatoid arthritis.
The authors thank Drs. E Taylor and A. Cope for providing synovial fluids and biopsies, Dr. M. Bayliss
for normal synovium, and Miss G. Harris for technical assistance.
Address correspondence to Professor Marc Feldmann, The Mathilda & Terence Kennedy Institute of Rheu-
matology, Sunley Building, 1 Lurgan Avenue, Hammersmith, London W6 8LW, UK.
Received for publication 29 September 1993 and in revised form 13 January 1994.
1. Harris, E.D., Jr. 1990. Rheumatoid arthritis: pathophysiology
and implications for therapy. N. Engl. J. Med. 322:1277.
2. Di Giovine, F.S., G. Nuki, and G.W. Duff. 1988. Tumor necrosis
factor in synovial exudates. Ann. Rheum. Dis. 47:768.
3. Saxne, T., M.A. PaUadino, D. Heinegard, N. Talal, and F.A.
Wollheim. 1988. Detection of turnout necrosis factor ct but
not tumour necrosis factor 3 in rheumatoid arthritis synovial
fluid and serum. Arthritis Rheum. 31:1041.
4. Miyasaka, N., K. Sato, and M. Goto. 1988. Augmented inter-
leukin 1 production and HLA-DR expression in the synovium
of rheumatoid arthritis patients: possible involvement in joint
destruction. Arthritis Rheum. 31:480.
5. Hirano, T., T. Matsuda, M. Turner, N. Miyasaka, G. Buchan,
B. Tang, K. Sato, M. Shimizu, R.N. Maini, M. Feldmann,
and T. Kishimoto. 1988. Excessive production of interleukin
6/B cell stimulatory factor-2 in rheumatoid arthritis. Eur. J.
6. Guerne, P.A., B.L. Zuraw, J.H. Vaughan, D.A. Carson, and
M. Lotz. 1989. Synovium as a source of interleukin 6 in vitro.
Contribution to local and systemic manifestations of arthritis.
J. Clin. Invest. 83:585.
7. Chu, C.Q., M. Field, M. Feldmann, and R.N. Maini. 1991.
Localization of tumor necrosis factor et in the synovial tissues
and at the cartilage-pannusjunction in patients with rheuma-
toid arthritis. Arthritis Rheum. 34:1125.
8. Deleuran, B., C.-Q. Chu, M. Field, F.M. Brennan, P. Kat-
sikis, M. Feldmann, and R.N. Maini. 1992. Localization of
intefleukin lct (IIAc 0, type 1 IL-1 receptor and interleukin
1 receptor antagonist protein in the synovial membrane and
cartilage/pannus junction in rheumatoid arthritis. Br.J. Rheum.
9. Haworth, C., F.M. Brennan, D. Chantry, M. Turner, R.N.
Maini, and M. Feldmann. 1991. Expression of granulocyte-
macrophage colony-stimulating factor in rheumatoid arthritis:
regulation by tumor necrosis factor-or. Eur.J. Immunol. 21:2575.
10. Firestein, G.S., A.E. Berger, D.E. Tracey, J.G. Chosay, D.L.
Chapman, M.M. Paine, C. Yu, and N.J. Zvaifler. 1992. Ibl
receptor antagonist protein production and gene expression
in rheumatoid arthritis and osteoarthritis synovium.J. Immunol.
11. Cope, A., D. Aderka, M. Doherty, H. Engelmann, D. Gibbons,
A.C. Jones, F.M. Brennan, R.N. Maini, D. Wallach, and M.
Feldmann. 1992. Increased levels of soluble tumor necrosis factor
receptors in the sera and synovial fluid of patients with rheu-
matic diseases. Arthritis Rheum. 35:1160.
12. Brennan, F.M., D. Chantry, M. Turner, B. Foxwell, Ik. Maini,
and M. Feldmann. 1990. Detection of transforming growth
factor-beta in rheumatoid arthritis synovial tissue: lack of effect
on spontaneous cytokine production in joint cell cultures. Clin.
Exp. Immunol. 81:278.
13. Firestein, G.S., and N.J. Zvaifler. 1987. Peripheral blood and
synovial fluid monocyte activation in inflammatory arthritis.
II. Low levels of synovial fluid and synovial tissue interferon
suggest that 3'-interferon is not the primary macrophage ac-
tivating factor. Arthritis Rheum. 30:864.
14. Firestein, G.S., W.-D. Xu, K. Townsend, D. Broide, J. Alvaro-
Garcia, A. Glasebrook, and N.J. Zvaifler. 1988. Cytokines in
chronic inflammatory arthritis. I. Failure to detect T cell lym-
phokines (interleukin 2 and interleukin 3) and presence of mac-
rophage colony-stimulating factor (CSF-1) and a novel mast
cell growth factor in rheumatoid synovitis, j. Exlx Med. 168:
15. Cush, J.J., and P.E. Lipsky. 1988. Phenotypic analysis of sy-
novial tissue and peripheral blood lymphocytes isolated from
patients with rheumatoid arthritis. Arthritis Rheum. 31:1230.
16. Hovdenes, J., G. Gaudernack, T.K. Kvien, and T. Egeland.
1989. Expression of activation markers on CD4 § and CD8 +
cells from synovial fluid, synovial tissue and peripheral blood
of patients with inflammatory arthritides. Stand. j. ImmunoL
17. Buchan, G.S., K. Barrett, T. Fujita, T. Taniguchi, R.N. Maini,
and M. Feldmann. 1988. Detection of activated T cell prod-
1525 Katsikis et al.
2, II~2 receptor and interferon % Clin. ExI~ Immunol. 71:295.
18. Silverman, H.A., J.S. Johnson, J.H. Vaughan, and J.C. Mc-
Glammory. 1976. Altered lymphocyte reactivity in rheuma-
toid arthritis. Arthritis Rheum. 19:509.
19. Lotz, M., C.D. Tsoukas, C.A. Robinson, C.A. Dinarello, D.A.
Carson, andJ.H. Vaughan. 1986. Basis for defective responses
of rheumatoid synovial fluid lymphocytes to anti-CD3 (T3)
antibodies. J. Clin. Invest. 78:713.
20. Lotz, M., J. Kekow, and D.A. Carson. 1990. Transforming
growth factor-fl and cellular immune responses in synovial
fluids. J. Immunol. 144:4189.
21. Miltenburg, A.M.M., J.M. Van Laar, R. De Kniper, M.R.
Daha, and F.C. Breedveld. 1992. T cell clones from human
rheumatoid synovial membrane functionally represent the
TH1 subset. Scand. J. Immunol. 35:603.
22. Street, N.E., and T.R. Mosmann. 1991. Functional diversity
of T lymphocytes due to secretion of different cytokine pat-
terns. FASEB (Fed. Am. SoL Exp. Biol.) J. 5:171.
23. Vieira, P., R.. de Waal-Malefyt, M.-N. Dang, K.E. Johnson,
R. Kastelein, D.F. Fiorentino, J.E. de Vries, M.-G, Roncarolo,
T.R. Mosmann, and K.W. Moore. 1991. Isolation and expres-
sion of human cytokine synthesis inhibitory factor cDNA
clones: homology to Epstein-Barr virus open reading frame
BCRF1. Proc Natl. Acad. Sci. USA. 88:1172.
24. Dd Prete, G., M. De Carl, F. Almerigogna, M.G. Gindizi,
R. Biagiotti, and S. Romagnani. 1993. Human I1"10 is pro-
duced by both type 1 helper (Thl) and type 2 helper (Th2)
T cell clones and inhibits their antigen-specific proliferation
and cytokine production, f Immunol. 150:353.
25. de Waal-Malefyt, R., J. Abrams, B. Bennett, C.G. Figdor, and
J.E. de'Cries. 1991. Interleukin 10 (IL-IO) inhibits cytokine syn-
thesis by human monocytes: an autoregulatory role of 1I-10
produced by monocytes. J. Exp. Med. 174:1209.
26. Yssel, H., R. de Waal-Malefyt, R.-M. R.oncarolo, J.S. Abrams,
R. Lahesmaa, H. Spits, and J.E. de Vries. 1992. I1"10 is pro-
duced by subsets of human CD4 + T cell clones and periph-
eral blood T cells. J. Immunol. 149:2378.
27. de Waal-Malefyt, R., J. Haanen, H. Spits, M.-G. Roncarolo,
A. te Velde, C. Figdor, K. Johnson, R. Kastelein, H. Yssel,
and J.E. de Vries. 1991. Interleukin 10 (11"10) and viral I1"10
strongly reduce antigen-specific human T cell proliferation by
diminishing the antigen-presenting capacity of monocytes via
downregulation of class II major histocompatibility complex
expression. J. Exp. Med. 174:915.
28. Arnett, F.C., S.M. Edworthy, D,A. Bloch, D.J. McShane, J.F.
Fries, N.S. Cooper, L.A. Healey, S.R. Kaplan, M.H. Liang,
H.S. Luthra et al. 1988. The American rheumatism associa-
tion 1987 revised criteria for the classification of rheumatoid
arthritis. Arthritis Rheum. 31:315.
29. Chomczynski, P., and N. Sacchi. 1987. Single step method
of RNA isolation by acid guanidinium thiocyanate-phenol-
chloroform extraction. Anal. Biochem. 165:156.
30. Sambrook, J., E.F. Fritsch, and T. Maniatis. 1989. Molecular
Cloning. A Laboratory Manual. 2nd ed. C. Nolan, editor. Cold
Spring Harbor Laboratory, Cold Spring Harbor, NY.
31. Saiki, ILK., S. Scharf, F. Faloona, K.B. Mullis, G.T. Horn,
H.A. Erlich, and N. Arnheim. 1985. Enzymatic amplification
of beta-globin genomic sequences and restriction site analysis
for diagnosis of sickle cell anemia. Science (Wash. DC). 230:1350.
32. Kaczmarek, L., B. Calabretta, and R. Baserga. 1985. Expres-
sion of cell cycle-dependent genes in phytohemagglutinin-
stimulated human lymphocytes. Pro~ Natl. Acad. Sci. USA.
33. Feldmann, M., F.M. Brennan, M. Field, and R.N. Maini. 1992.
Pathogenesis of rheumatoid arthritis: cellular and cytokine in-
teractions. In Rheumatoid Arthritis. J. Smolen, J. Kalden, and
K.N. Maini, editors. Springer-Verlag, Heidelberg, 41-54.
34. Eastgate, J.A., J.A. Symons, N.C. Wood, F.M. Grinlington,
F.S. Di-Giovine, and G.W. Duff. 1988. Correlation of plasma
interleukin 1 levels with disease activity in rheumatoid arthritis.
35. Alvaro-Garcia, J.M., N.J. Zvaifler, C.B. Brown, K. Kaushansky,
and G.S. Firestein. 1991. Cytokines in chronic inflammatory
arthritis. VI Analysis of the synovial cells involved in granulo-
cyte-macrophage colony-stimulating factor production and gene
expression of rheumatoid arthritis and its regulation by Ibl
and tumor necrosis factor-t~. J. Immunol. 146:3365.
36. Elliott, M.J., K.N. Maini, M. Feldmann, A. Fox-Long, P.
Charles, P. Katsikis, M. Brennan, J. Walker, H. Bijl, J. Ghrayeb,
and J. Woody, 1993. Treatment of rheumatoid arthritis with
chimeric monoclonal antibodies to TNFa, safety, clinical
efficacy and regulation of acute phase response. Arthritis Rheum.
37. Strieter, R.M., S.L. Kunkel, H.J. Showell, D.G. Remick, S.H.
Phan, P.A. Ward, and R.M. Marks. 1989. Endothelial cell gene
expression of a neutrophil chemotactic factor by TNFc~, LPS
and IblB. Science (Wash. DC). 243:1467.
38. Schroder, J.M., M. Sticherling, H.H. Henneicke, W.C. Preis-
sner, and E. Christophers. 1989. Iblct or tumor necrosis
factor-c~ stimulate release of three NAP-1/Ib8 related neutro-
phil chemotatic proteins in human dermal fibroblasts. J. Im-
39. Chantry, D., M. Turner, E.g. Abney, and M. Feldmann. 1989.
Modulation of cytokine production by transforming growth
factor ~1. J. lmmunol. 142:4295.
40. Hart, P.H., G.E Vitti, D.R, Burgess, G.A. Whitty, D.S. Picolli,
and J.H. Hamilton. 1989. 1989. Potential anti-inflammatory
effects of Ib4: suppression of human monocyte tumor necrosis
factor ct, interleukin-1 and prostaglandin Ez. Proc Natl. Acad.
Sci. USA. 86:3803.
41. Kuruvilla, A.P., R. Shah, G.M. Hochwald, H.D. Liggitt, M.A.
Palladino, and G.J. Thorbecke. 1991. Protective effect of trans-
forming growth factor B1 on experimental autoimmune dis-
eases in mice. Pro~ Natl. Acad. Sci. USA. 88:2918.
42. Fava, fLA., N.J. Olsen, A.E. Postlethwaite, K.N. Broadley,
J.M. Davidson, L.B. Nanney, C. Lucas, and A.S. Townes. 1991.
Transforming growth factor fll (TGF-/~I) induced neutrophil
recruitment to synovial tissues: implications for TGF-fl-driven
synovial inflammation and hyperplasia.J. Exi~ Med. 173:1121.
43. Miossec, P., J. Briolay, J. Dechanet, J. Wijdenes, H. Martinez-
Valdez, andJ. Banchereau. 1992. Inhibition of the production
of proinflammatory cytokines and immunoglobulins by inter-
leukin-4 in an ex vivo model of rheumatoid synovitis. Arthritis
44. Cope, A.P., D.L. Gibbons, D. Aderka, B.M. Foxwell, D. Wal-
lach, R.N. Maini, M. Feldmann, and F.M. Brennan. 1993.
Differential regulation of tumor necrosis factor receptors
(TNF-R) by Ib4; upregulation of p55 and p75 TNF-R on
synovial joint mononuclear cells. Cytokine. 3:205.
45. Essner, R., K. Rhoades, W.H. McBride, D.L. Morton, and
J.S. Economous. 1989. 11"4 down-regulates I1"1 and TNF gene
expression in human monocytes. J. Immunol. 142:3857.
1526 Immunoregulatory Role of II,10 in Rheumatoid Arthritis
46. Brennan, F.M., D. Chantry, A. Jackson, R.N. Maini, and M. Download full-text
Feldmann. 1989. Inhibitory effect of TNFex antibodies on sy-
novial cell interleukin-1 production in rheumatoid arthritis.
47. Taga, K., and G. Tosato. 1992. Ibl0 inhibits human T cell
proliferation and Ib2 production, f Immund. 148:1143.
48. Taga, K., H. Mostowski, and G. Tosato. 1993. Human
intedeukin-10 can directly inhibit T cell growth. Blood. 81:2964.
49. de Waal Malefyt, R., H. Yssel, andJ.E, de Vries. 1993. Direct
effect of Ibl0 on subsets of human CD4 + T cell clones and
resting T cells. Specific inhibition of Ib2 production and
proliferation. J. Immunol. 150:4754.
50. Macatonia, S.E., T.M. Doherty, S.C. Knight, and A. O'Garra.
1993. Differential effect of IL-10 on dendritic ceU-induced T
cell proliferation and IFNy production.J. Immunol. 150:3755.
51. Chomarat, P., M.-C. Rissoan, J. Banchereau, and P. Miossec.
1993. Interferon "y inhibit interleukin 10 production by mono-
cytes. J. Exp. Med. 177:523.
52. Te Velde, A.A., R. de Waal Malefijt, R.J.E Huijbens, J.E. de
Vries, and C.G. Figdor. 1992. Ibl0 stimulates monocyte Fc'yR
surface expression and cytotoxic activity. Distinct regulation
of antibody-dependent cellular cytotoxicity by IFN-% Ib4, and
Ibl0. j. ImmunoL 149:4048.
53. Wogensen, L., X. Huang, and N. Sarvetnick. 1993. Leuko-
cyte extravasation into the pancreatic tissue in transgenic mice
expressing interleukin 10 in the islets of Langerhans. J. Exp.
54. G~rard, C., C. Bruyns, A. Marchant, D. Abramowicz, P. Van-
denabeele, A. Delvaux, W. Fiers, M. Goldman, and T. Velu.
1993. Interleukin 10 reduces the release of tumor necrosis factor
and prevents lethality in experimental endotoxemia.J. Exl~ Med.
55. Howard, M., T. Muchamuel, S. Andrade, and S. Meaon. 1993.
Interleukin 10 protects mice from lethal endotoxemia.J. Ex F
1527 Katsikis et al.