British Journal of Rheumatology 1997;36:729–734
EFFECTS OF TRANSFORMING GROWTH FACTOR b ON THE PRODUCTION
OF PROSTAGLANDIN E AND CASEINASE ACTIVITY OF UNSTIMULATED
AND INTERLEUKIN 1-STIMULATED HUMAN ARTICULAR CHONDROCYTES
IN CULTURE
F. W. FAWTHROP,*† A. FRAZER,* R. G. G. RUSSELL* and R. A. D. BUNNING‡
*Department of Human Metabolism and Clinical Biochemistry, Sheffield University Medical School, Sheffield S10 2RX and
‡Division of Biomedical Sciences and Health Research Institute, Sheffield Hallam University, Sheffield S1 1WB
SUMMARY
Transforming growth factor b (TGFb) has previously been shown to have actions on chondrocytes and cartilage both in vitro
and in vivo which suggest a role in connective tissue repair. In particular, some of its actions have been shown to be antagonistic
to those of interleukin 1 (IL-1). In this study, the effects of TGFb on prostaglandin E (PGE) production and caseinase activity,
in the presence and absence of IL-1, in human articular chondrocytes were investigated. TGFb1 and TGFb2 were shown to
modulate IL-1b-stimulated PGE production and caseinase activity. Both TGFb1 and b2 inhibited IL-1b-stimulated PGE
production in the absence of serum and augmented it in the presence of serum. TGFb1 and TGFb2 inhibited IL-1-stimulated
caseinase activity with and without serum. In general, the TGFbs had little or no effect on basal PGE or caseinase levels. TGFbs
may be important modulators of chondrocyte metabolism, their effects on PGE production may depend on cytokine interactions;
furthermore, their effects on caseinase activity may help prevent cartilage breakdown.
K : Transforming growth factor b, Interleukin 1, Prostaglandin E, Caseinase activity, Chondrocytes.
T transforming growth factors b (TGFbs) are a
family of homodimeric polypeptides with mol. wts of
025 kDa. There are at least five isoforms of TGFb, of
which the best characterized are TGFb1 and TGFb2
[1, 2]. Although TGFbs 1 and 2 exhibit only 71%
homology, they seem to share similar activities [2].
However, it has been shown that TGFb2 is a more
active stimulator of chondrogenesis, proteoglycan
synthesis and cell proliferation in vitro [3]. Our aim was
to compare the actions of TGFbs 1 and 2 on
chondrocyte metabolism in terms of prostaglandin E
(PGE) production and caseinase activity.
The TGFbs are multifunctional molecules and are
thought to be involved in processes such as connective
tissue repair [4]. In this context, TGFb has been shown
to stimulate the synthesis of extracellular matrix
components such as type I collagen and fibronectin [5],
to inhibit the production by fibroblasts of proteinases
capable of extracellular matrix breakdown, such as
plasminogen activator and stromelysin, and to
stimulate the production of inhibitors of plasminogen
activator and metalloproteinases [1, 6–8]. TGFb has
been shown to have a number of actions on
chondrocytes and cartilage, some being antagonistic to
those of interleukin 1 (IL-1). These include the
inhibition of IL-1-stimulated protease activity, the
reversal of IL-1-mediated inhibition of proteoglycan
synthesis in rabbit chondrocytes and the partial
inhibition of IL-1-stimulated resorption of porcine
cartilage in vitro [9, 10]. In organ cultures of bovine
cartilage, TGFb stimulates proteoglycan synthesis and
decreases its catabolism [11]. Human osteoarthritic
cartilage has been shown to be more sensitive to TGFb
than normal cartilage in terms of its ability to stimulate
proteoglycan synthesis [12]. Depending on the culture
conditions, TGFb may inhibit or stimulate the
proliferation of rabbit chondrocytes and may stimulate
proliferation in human chondrocytes [13, 14]. TGFb
has also been shown to have a marked protective effect
on cartilage destruction in vivo. Injection of TGFb
together with IL-1 into murine knee joints prevented
the cartilage destruction observed in injecting IL-1
alone [15].
Both IL-1 and TGFb are present in synovial fluids
of patients with osteoarthritis and rheumatoid arthritis
[16, 17], and may interact to modulate chondrocyte
metabolism. There are a number of potential sources of
synovial fluid TGFb. Rheumatoid synovium produces
TGFb [18] and there is some evidence for the
production of TGFb by human chondrocytes. mRNAs
for TGFbs 1, 2 and 3 have been demonstrated in
cultured human chondrocytes in addition to the
production of TGFb by these cells in culture medium
[19, 20]. Explants of human cartilage in culture have
also been shown to produce TGFb [12].
In this study, we have investigated the effect of
TGFb1 and 2, alone and in combination with IL-1, on
prostaglandin production and caseinase activity in
human chondrocytes. Caseinase activity is used as a
measure of stromelysin activity [21].
PGE is a potent mediator of inflammation [22] which
may also have actions on connective tissue cell
Submitted 10 May 1996; revised version accepted 13 December
1996.
Correspondence to: R. A. D. Bunning, Division of Biomedical
Sciences and Health Research Institute, Sheffield Hallam University,
Pond Street, Sheffield S1 1WB.
Present address: †Level D, Rotherham District General Hospital,
Moorgate, Rotherham S60 3UD.
= 1997 British Society for Rheumatology
729
730
BRITISH JOURNAL OF RHEUMATOLOGY VOL. 36 NO. 7
metabolism, PGE2 stimulating both bone resorption
and formation under certain conditions [23]. IL-1
stimulates the production of PGE2 by human
chondrocytes [24]. In addition, IL-1 and TGFb in
combination produce a synergistic stimulation of PGE2
production in human rheumatoid synovial cells [25].
However, little is known of the effects of TGFb alone
or in combination with IL-1 on chondrocyte PGE
production.
Stromelysin is probably involved in both normal and
pathological connective tissue turnover as it can
degrade collagen types III, IV, V and IX, and
proteoglycans [26], and can activate procollagenase
[27]. IL-1 has been shown to stimulate the stromelysin
activity of cultured human cartilage explants and
chondrocytes [28]. Furthermore, stromelysin activity is
increased in osteoarthritic cartilage [29].
Production of TGFb by chondrocytes and its actions
on chondrocytes may be important in cartilage repair
mechanisms both by limiting cartilage breakdown, and
promoting new matrix synthesis and cellular proliferation. Also of importance are its interactions with other
cytokines, such as IL-1, which are known to mediate
cartilage destruction. Much of the previous work has
been performed using animal cells and tissues, and to
date there are few published studies of effects of TGFb
on human articular chondrocytes.
MATERIALS AND METHODS
Cytokines
Recombinant human (rh) TGFb1 was a gift from Dr
E. Amento, Genentech Inc., San Francisco, USA,
supplied at a concentration of 28.6 m and diluted in
medium containing 1% bovine serum albumin (BSA).
rhTGFb2, a gift from Sandoz AG, Basel, Switzerland,
was obtained at a concentration of 4 × 10−5  and
diluted as required in medium containing 1% BSA.
rhIL-1b (108 U/mg) was generously donated by Dr A.
Shaw, Glaxo Wellcome Research, Stevenage, Herts.
Preparation and culture of adult human articular
chondrocytes
Macroscopically normal human articular cartilage
was obtained from femoral heads, removed following
trauma, and from the femoral and tibial condyles of
above-knee amputation specimens. Samples were from
a predominantly elderly age group, mean age 76 yr
(range 54–93 yr), and are therefore relevant to mature
adult cartilage. Chondrocytes were dispersed by
sequential enzymatic digestion as previously described
[30] and were cultured in Eagle’s minimum essential
medium containing 10% fetal calf serum, 100 U/ml
penicillin, 100 mg/ml streptomycin, 2 m glutamine
(MEM) and kept in a humidified incubator at 37°C
(5% CO2/95% air). Cells were used at first passage after
3–6 weeks in culture. Chondrocytes have been shown
to maintain their phenotype, i.e. produce type II
collagen, under these conditions [31]. For experimental
purposes, chondrocytes were passaged in 24 well plates
at a density of 3–4 × 104 cells/well and grown to
confluence. After washing the cells three times with
phosphate-buffered saline, test substances were added
in either serum-free medium containing 0.1% BSA or
medium containing 1% fetal calf serum. For each
experiment, chondrocytes from a single patient were
used, control and test conditions being set up in
quadruplicate wells. Each set of experiments was
carried out on cells from at least three different donors.
PGE assay
PGE in cell supernatants was measured by
radioimmunoassay as described previously [32] using
an antiserum raised in rabbits against PGE2 (Sigma,
Poole). This antibody does not discriminate between
PGE1 and PGE2, hence results are expressed as PGE.
Caseinase assay
The non-specific caseinase assay using 14C-acetylated
casein as substrate was used as previously described
[21]. Latent enzyme was activated by the addition of
4-aminophenylmercuric acetate (0.8 m). One unit of
stromelysin activity degrades 1 mg of casein per minute
at 37°C to peptides soluble in 3% trichloroacetic acid.
It has previously been shown that the casein-degrading
activity produced by human chondrocytes is mainly
due to a latent metalloproteinase, and therefore
probably represents stromelysin [33].
Statistical methods
Results are expressed as the mean 2 ... for each
group. Statistical analysis was by analysis of variance
using Scheffe’s method for multiple comparisons [34].
RESULTS
TGFb1 or TGFb2, 10−13–10−7  in the presence or
absence of 1% serum, had no significant effect on the
basal PGE production of adult human chondrocytes as
compared to controls. In all experiments, IL-1
markedly stimulated PGE levels.
In the absence of serum, both TGFb1 and TGFb2, at
10−9–10−7 , significantly inhibited IL-1-stimulated
PGE production. Since we found that the effects of
TGFb1 and TGFb2 on PGE production were similar,
results for TGFb2 only are shown (Fig. 1a). The
inhibitory activity of TGFb1 or TGFb2 on IL-1stimulated PGE production was generally reduced or
lost at low concentrations: 10−13–10−10 . Indeed, in
some experiments, TGFb1 or TGFb2 at 10−13 
enhanced IL-1-stimulated PGE production (Fig. 1b).
In the presence of 1% serum, 10−9–10−7  TGFb1 or
TGFb2 significantly enhanced IL-1-stimulated PGE
production (Fig. 2a). This effect was lost at lower
concentrations of TGFb1 (10−13–10−11 ) or TGFb2
(10−13–10−10 ) (Fig. 2b).
In both the presence and absence of 1% serum, IL-1
significantly stimulated the caseinase activity of human
chondrocytes. TGFb1 alone had no significant effect on
basal caseinase levels, whereas TGFb2, in the absence
of serum, produced a slight increase in basal levels and
in the presence of serum a slight decrease which
reached significance in some experiments. IL-1-stimulated caseinase activity was decreased by TGFb1 or
FAWTHROP ET AL.: TGFb EFFECTS ON PGE AND CASEINASE
731
F. 1.—The effect of rhTGFb2, 10−9–10−7  (a) and 10−13––10−10  (b), on the PGE production of human chondrocytes and its interaction
with rhIL-1 (10 U/ml). Chondrocytes were incubated with MEM, containing 0.1% bovine serum albumin instead of fetal calf serum and test
substances for 24 h. PGE was measured in cell supernatants by radioimmunoassay. Results are representative of at least three experiments and
replicates, from the same donor, are expressed as mean 2 ... (n = 4), *P E 0.05, **P E 0.01 in comparison to the rhIL-1-treated group,
Scheffe’s multiple range analysis.
TGFb2 at 10−9–10−7  in the presence or absence of
serum (Fig. 3a and b). In the presence of serum,
the lowering of IL-1-stimulated activity with TGFbs 1
and 2 did not always reach significance, as seen in
Fig. 3b. In the absence of serum, 10−13  TGFb1 or
TGFb2 enhanced the action of IL-1 on caseinase
activity (Fig. 4). Since observations for TGFb1 or
TGFb2 were similar, only the results for TGFb1 are
shown.
DISCUSSION
TGFb1 has previously been identified in fetal and
adult non-human cartilage [11, 35, 36], and cultured
explants of human cartilage and human chondrocytes
in culture have been shown to secrete TGFb into the
medium [20], suggesting that TGFb may have an
important role in the metabolism of normal human
articular cartilage.
TGFb1 and b2 had no significant effect on basal PGE
levels in the presence or absence of serum. However, in
the absence of serum, both TGFb1 and b2 inhibited
IL-1-stimulated PGE production, possibly the result of
the downregulation of IL-1 receptors on chondrocytes
by TGFb [37, 38]. In contrast, in the presence of 1%
serum, both growth factors augmented IL-1-stimulated
PGE production. The reason for this is uncertain, but
is probably the result of a complex interaction(s)
between IL-1 and/or TGFb with serum components.
Platelet-derived growth factor (PDGF) and fibroblast
growth factor, which are present in serum, have been
shown to augment IL-1-stimulated PGE production in
rabbit articular chondrocytes [39, 40]. The effect of
PDGF is thought to be mediated by the induction of
IL-1 receptors by PDGF. However, we have observed
that PDGF inhibits IL-1-stimulated PGE production
in human articular chondrocytes [41]. Another
possibility may be that the active concentration of
TGFb is reduced in the presence of serum by binding
F. 2.—The effect of rhTGFb2, 10−9–10−7  (a) and 10−13–10−10  (b), on the PGE production of human chondrocytes and its interaction with
rhIL-1 (10 U/ml). Chondrocytes were incubated with MEM containing 1% fetal calf serum and test substances for 24 h. PGE was measured
in cell supernatants by radioimmunoassay. Results are representative of at least three experiments and replicates, from the same donor, are
expressed as mean 2 ... (n = 4), **P E 0.01 in comparison to the rhIL-1-treated group, Scheffe’s multiple range analysis.
732
BRITISH JOURNAL OF RHEUMATOLOGY VOL. 36 NO. 7
F. 3.—The effect of rhTGFb1 (10−9–10−7 ) on the caseinase activity of human chondrocytes and its interaction with rhIL-1 (10 U/ml).
Chondrocytes were incubated with MEM, containing (a) 0.1% bovine serum albumin or (b) 1% fetal calf serum and test substances for 24 h.
Caseinase activity was measured in cell supernatants. Results are representative of at least three experiments and replicates, from the same donor,
are expressed as mean 2 ... (n = 4), **P E 0.01, compared with the rhIL-1-treated group, Scheffe’s multiple range analysis.
to a2-macroglobulin [42] to levels where TGFb
augments IL-1-stimulated PGE production, as observed by us at 10−13  TGFb1 and b2 in the absence of
serum (see Fig. 1b). Additionally, transcription factors
present in serum may effect the synthesis of mediators
involved in the responsiveness of cells to TGFb or IL-1
[43]. A systematic study of the effect of serum
components on the action of IL-1 and TGFb would be
required to explain the results fully.
In general, TGFb reduced IL-1-stimulated caseinase
activity either in the presence or absence of 1% serum.
This observation is in accord with those of Chandrasekhar et al. [9] and Bandara et al. [44] who noted
a TGFb1-mediated suppression of neutral metalloproteinase activity in rabbit chondrocytes, activated by
synoviocyte-conditioned medium. Interestingly, they
also observed an increase in PGE synthesis in these
activated cells. Little or no significant effect of TGFb
was seen on basal caseinase levels, possibly because
activity levels were so low. Augmentation of IL-1-xstimulated caseinase activity at low concentrations of
TGFb (10−13 ) in the absence of serum mirrors its
effects on PGE production. TGFb appears to have
biphasic effects, depending on concentration. This may
result from the presence of TGFb receptors on
chondrocytes with differing affinities and opposing
effects [45].
Stromelysin is thought to be an important mediator
of extracellular matrix degradation. Its key role may be
in the activation of other latent metalloproteinases,
since although it is capable of breaking down the
cartilage proteoglycan, aggrecan, the major site at
which aggrecan is cleaved in vivo is not a stromelysin
cleavage site [46]. In osteoarthritic cartilage, an increase
in stromelysin has been observed, suggesting a role in
cartilage destruction [29]. The lowering of IL-1-stimulated stromelysin activity by TGFb may be important
in the prevention of cartilage breakdown.
The action of rhTGFb2 was markedly similar to that
of rhTGFb1, which is not surprising as TGFb1 and b2
share 70% homology and in other systems have a
common action [2]. It is not possible to comment on
their relative potencies as the growth factors were
obtained from different sources and their specific
activities were not available.
The production of TGFb by chondrocytes and its
actions on them suggests that it may be an important
autocrine and/or paracrine modulator of chondrocyte
metabolism. In the arthritic joint, any effects on PGE
production may be dependent on cytokine interactions
and levels of TGFb. Effects on IL-1-stimulated
stromelysin activity may help tip the balance of
cartilage metabolism from breakdown towards attempts at repair.
F. 4.—The effect of rhTGFb1 (10−13–10−10 ) on the caseinase
activity of human chondrocytes and its interaction with rhIL-1
(10 U/ml). Chondrocytes were incubated with MEM, containing
0.1% bovine serum albumin and test substances for 24 h. Caseinase
activity was measured in cell supernatants. Results are representative
of at least three experiments and replicates, from the same donor, are
expressed as mean 2 ... (n = 4), *P E 0.05, **P E 0.01, compared with the rhIL-1-treated group, Scheffe’s multiple range
analysis.
FAWTHROP ET AL.: TGFb EFFECTS ON PGE AND CASEINASE
A
This project was supported by the Arthritis and
Rheumatism Council (UK) and the Biotechnology and
Biological Sciences Research Council.
R
1. Sporn MB, Roberts AB, Wakefield LM, de Crombrugghe B. Some recent advances in the chemistry
and biology of transforming growth factor beta. J Cell
Biol 1987;105:1039–45.
2. Roberts AB, Sporn MB. The transforming growth factor
bs. In: Sporn MB, Roberts AB, eds. Handbook of
experimental pharmacology, peptide growth factors and
their receptors I, Vol. 95. Heidelberg: Springer Verlag,
1990:419–72.
3. Okazaki R, Sakai A, Nakamura T, Kunugita N,
Norimura T, Suzuki K. Effects of transforming growth
factor bs and basic fibroblast growth factor on articular
chondrocytes obtained from immobilised rabbit knees.
Ann Rheum Dis 1996;55:181–6.
4. Sporn MB, Roberts AB. Transforming growth factor-b:
multiple actions and potential clinical applications. J Am
Med Assoc 1989;262:938–41.
5. Ignotz RA, Massague J. Transforming growth factorbeta stimulates the expression of fibronectin and collagen
and their incorporation into the extracellular matrix. J
Biol Chem 1986;264:4337–45.
6. Matrisian LM, Leroy P, Ruhlmann C, Gesnel M-C,
Breathnach R. Isolation of the oncogene and epidermal
growth factor-induced transin gene: complex control in
rat fibroblasts. Mol Cell Biol 1986;6:1679–88.
7. Keski-Oja J, Blasi F, Leof EB, Moses HL. Regulation of
the synthesis and activity in A549 human lung carcinoma
cells by transforming growth factor-b. J Cell Biol
1988;106:451–9.
8. Laiho M, Saksela O, Andreasen PA, Keski-Oja J.
Enhanced production and extracellular deposition of the
endothelial type plasminogen activator inhibitor in
cultured human lung fibroblasts by transforming growth
factor-beta. J Cell Biol 1986;103:2403–10.
9. Chandrasekhar S, Harvey AK. Transforming growth
factor-b is a potent inhibitor of interleukin-1 induced
protease activity and cartilage proteoglycan degradation.
Biochem Biophys Res Commun 1988;157:1352–9.
10. Andrews HJ, Edwards TA, Cawston TE, Hazelman BL.
Transforming growth factor-beta causes partial inhibition of interleukin-1-stimulated cartilage degradation
in vitro. Biochem Biophys Res Commun 1989;162:144–
50.
11. Morales TI, Joyce ME, Sobel ME, Danielpour D,
Roberts AB. Transforming growth factor-beta in calf
articular cartilage organ cultures: synthesis and distribution. Arch Biochem Biophys 1991;288:397–405.
12. Lafeber FPJG, Van der Kraan PM, Huber-Bruning O,
Van den Berg WB, Bijlsma JWJ. Osteoarthritic human
cartilage is more sensitive to transforming growth factor
b than is normal cartilage. Br J Rheumatol 1993;32:281–
6.
13. Vivien D, Galéra P, Lebrun E, Loyau G, Pujol J-P.
Differential effects of transforming growth factor-b and
epidermal growth factor on the cell cycle of cultured
rabbit articular chondrocytes. J Cell Physiol 1990;
143:534–45.
14. Frazer A, Bunning RAD, Russell RGG. Effects of
transforming growth factor b and interleukin-1b on [3H]
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
733
thymidine incorporation by human articular chondrocytes in vitro. Biochim Biophys Acta 1994;1226:193–200.
Van Beuningen HM, Van der Kraan PM, Arntz OJ, Van
den Berg WB. Protection from interleukin 1 induced
destruction of articular cartilage by transforming growth
factor b: studies in anatomically intact cartilage in vitro
and in vivo. Ann Rheum Dis 1993;52:185–91.
Wood DD, Ihrie EJ, Dinarello CA, Cohen PL. Isolation
of an interleukin-1-like factor from human joint
effusions. Arthritis Rheum 1983;26:975–83.
Fava R, Olsen N, Keski-Oja J, Moses HL, Pincus T.
Active and latent forms of transforming growth factor b
activity in synovial effusions. J Exp Med 1989;169:291–6.
Lafyatis R, Thompson NL, Remmers EF et al.
Transforming growth factor-b production by synovial
tissues from rheumatoid patients and streptococcal cell
wall arthritic rats. J Immunol 1989;143:1142–8.
Frazer A, Seid JM, Hart KA, Bentley H, Bunning RAD,
Russell RGG. Detection of mRNA for the transforming
growth factor b family in human articular chondrocytes
by the polymerase chain reaction. Biochem Biophys Res
Commun 1991;180:602–8.
Villiger PM, Kusari AB, ten Dijke P, Lotz M. IL-1b and
IL-6 selectively induce transforming growth factor b
isoforms in human articular chondrocytes. J Immunol
1993;151:44–50.
Cawston TE, Galloway WA, Mercer E, Murphy G,
Reynolds JJ. Purification of rabbit bone inhibitor of
collagenase. Biochem J 1981;195:159–65.
Vane JR. The mode of action of aspirin and similar
compounds. J Allergy Clin Immunol 1976;58:690–712.
Raisz LG. The role of prostaglandins in the local
regulation of bone metabolism. Prog Clin Biol Res
1990;332:195–203.
Campbell IK, Piccoli DS, Hamilton JA. Stimulation of
human chondrocyte prostaglandin E2 production by
recombinant human interleukin-1 and tumour necrosis
factor. Biochim Biophys Acta 1990;1051:310–8.
Taylor DJ, Feldmann M, Evanson JM, Woolley DE.
Comparative and combined effects of transforming
growth factors a and b, interleukin 1 and interferongamma on rheumatoid synovial cell proliferation,
glycolysis and prostaglandin E production. Rheumatol
Int 1989;9:65–70.
Woessner JF. Matrix metalloproteinases and their
inhibitors in connective tissue remodeling. FASEB J
1991;5:2145–54.
Murphy G, Cockett MI, Stephens PE, Smith BJ,
Docherty AJP. Stromelysin is an activator of procollagenase. Biochem J 1987;248:265–8.
Martel-Pelletier J, Zafarullah M, Kodama S, Pelletier
J-P. In vitro effects of interleukin 1 on the synthesis of
metalloproteinases, TIMP, plasminogen activators and
inhibitors in human articular cartilage. J Rheumatol
1991;18(suppl. 27):80–4.
Woessner J, Gunja-Smith Z. Role of metalloproteinases
in human osteoarthritis. J Rheumatol 1991;18(suppl.
27):99–101.
Meats JE, Elford PR, Bunning RAD, Russell RGG.
Retinoids and synovial factor(s) stimulate the production
of plasminogen activator by cultured human chondrocytes: a possible role for plasminogen activator in the
resorption of cartilage in vitro. Biochim Biophys Acta
1985;838:161–9.
Frazer A, Bunning RAD, Russell RGG. Modulation of
human articular chondrocyte phenotype by transforming
734
32.
33.
34.
35.
36.
37.
38.
39.
BRITISH JOURNAL OF RHEUMATOLOGY VOL. 36 NO. 7
growth factor-b and interleukin-1b in vitro. Osteoarthritis Cartilage 1994;2:235–45.
Richardson JH, Elford PR, Sharrard RM, Meats JE,
Russell RGG. Modulation of connective tissue metabolism by partially purified human interleukin 1. Cell
Immunol 1985;90:41–51.
Bunning RAD, Russell RGG. The effect of tumour
necrosis factor a on the resorption of human articular
cartilage and on the production of prostaglandin E and
of caseinase activity by human articular chondrocytes.
Arthritis Rheum 1989;32:780–4.
Godfrey K. Statistics in practice: comparing the means
of several groups. N Engl J Med 1988;313:1450–6.
Flanders KC, Thomson NL, Cissel DS et al. Transforming growth factor-b1: histochemical localization with
antibodies to different epitopes. J Cell Biol 1989;108:653–
60.
Ellingsworth LR, Brennan JE, Fok K et al. Antibodies
to the N-terminal portion of cartilage-inducing factor A
and transforming growth factor b: immunohistochemical
localization and association with differentiating cells. J
Biol Chem 1986;261:12362–7.
Harvey AK, Hrubey PS, Chandrasekhar S. Transforming growth factor-beta inhibition of interleukin-1 activity
involves down-regulation of interleukin-1 receptors on
chondrocytes. Exp Cell Res 1991;195:376–85.
Rédini F, Mauviel A, Pronost S, Loyau G, Pujol J-P.
Transforming growth factor b exerts opposite effects
from interleukin-1b on cultured rabbit articular chondrocytes through reduction of interleukin-1 receptor
expression. Arthritis Rheum 1993;36:44–50.
Smith RJ, Justen JM, Sam LM, Rohloff NA, Ruppel PL,
40.
41.
42.
43.
44.
45.
46.
Brunden MN et al. Platelet derived growth factor
potentiates cellular responses of articular chondrocytes
to interleukin-1. Arthritis Rheum 1991;34:697–706.
Bandara G, Lin CW, Georgescu HI, Evans CH.
Chondrocyte activation by interleukin-1: synergism with
fibroblast growth factor and phorbol myristate acetate.
Agents Actions 1989;27:439–41.
Fawthrop FW, Russell RGG, Bunning RAD. The
modulation of interleukin-1b-stimulated prostaglandin E
production and caseinase activity by platelet derived
growth factors AA, AB and BB in human articular
chondrocytes. Calcif Tissue Int 1992;50(suppl.):A24.
Danielpour D, Sporn MB. Differential inhibition of
transforming growth factor beta 1 and transforming
growth factor beta 2 activity by alpha 2-macroglobulin.
J Biochem 1990;265:6973–7.
Goldring MB. Degradation of articular cartilage in
culture: regulatory factors. In: Woessner JF, Howell DS,
eds. Joint cartilage degradation, basic and clinical
aspects. New York: Marcel Dekker Inc., 1993:281–345.
Bandara G, Lin CW, Georgescu HI, Evans CH. The
synovial activation of chondrocytes: evidence for
complex cytokine interactions involving a possible novel
factor. Biochim Biophys Acta 1992;1134:309–18.
Glansbeek HL, van der Kraan PM, Vitters EL, van den
Berg WB. Correlation of the type II transforming growth
factor b (TGFb) receptor with TGF-b responses of
isolated bovine articular chondrocytes. Ann Rheum Dis
1993;52:812–6.
Lohmander LS, Neame PJ, Sandy JD. The structure of
aggrecan fragments in human synovial fluid. Arthritis
Rheum 1993;36:1214–22.