British Journal of Rheumatology 1997;37:27–31
SAMe RESTORES THE CHANGES IN THE PROLIFERATION AND IN
THE SYNTHESIS OF FIBRONECTIN AND PROTEOGLYCANS INDUCED
BY TUMOUR NECROSIS FACTOR ALPHA ON CULTURED RABBIT
SYNOVIAL CELLS
S. GUTIERREZ, I. PALACIOS, O. SANCHEZ-PERNAUTE, P. HERNANDEZ, J. MORENO,*
J. EGIDO and G. HERRERO-BEAUMONT
Research Laboratory and Inflammation Unit, Fundación Jiménez Dı́az, Universidad Autónoma and *CSIC, Madrid,
Spain
SUMMARY
S-Adenosyl--methionine (SAMe) is a naturally occurring compound involved in transmethylation and trans-sulphuration
reactions. The administration of SAMe to patients with osteoarthritis (OA) seems to have a protective effect, although the
mechanisms of its action are largely unknown. We have studied the effect of SAMe as a protective agent against the modifications
induced by tumour necrosis factor alpha (TNFa) on synovial cell proliferation and extracellular matrix protein synthesis, two
important hallmarks of progressive articular diseases. The stimulation of cells with 100 U/ml TNFa for 24 h decreased the
proliferative rate (58 2 14% with TNFa vs basal 100%, P Q 0.05), fibronectin (FN) mRNA expression (36 2 14% vs basal,
P Q 0.05) and FN synthesis (79 2 20% vs basal, P q 0.05). By contrast, TNFa raised total protein and proteoglycan synthesis
(127 2 12% vs basal and 239 2 40% vs basal, respectively, P Q 0.05). The addition of increasing concentrations of SAMe
(10−10–10−6 ) to synoviocytes incubated with TNFa reversed the effects induced by the cytokine, while SAMe alone did not
modify significantly the metabolic processes studied. These results indicate that, in cultured synovial cells, SAMe restores basal
conditions after cell damage elicited by TNFa stimulation.
K : SAMe, TNFa, Synovial cells, Proliferation, Fibronectin, Proteoglycans.
S---methionine (SAMe) is a naturally
occurring molecule distributed almost universally in
body tissues and fluids. SAMe acts as a methyl group
donor in several biochemical reactions involving
enzymatic transmethylation which leads to the
synthesis, activation or metabolism of different
biological components such as hormones, nucleic acids,
proteins and phospholipids [1]. Through the transsulphuration pathway, SAMe is a precursor of
endogenous sulphur compounds and essential molecules such as cysteine and glutathione, the main
cellular antioxidant [1]. Polyamine synthesis, involved
in cell proliferation and protein synthesis, is another
important metabolic pathway in which SAMe takes
part [2, 3].
The beneficial effects of SAMe in various chronic
liver diseases, drug hepatotoxicity and affective
disorders have been reported previously [4–6]. Experimental and clinical findings have shown the efficacy of
SAMe treatment in the development of osteoarthritis
(OA) [7, 8], suggesting a potential role of SAMe as a
reparative/protective agent with a certain anti-inflammatory action in articular damage. In this sense, SAMe
has been proposed as an alternative (or an addition) to
NSAID therapy [7]. Nevertheless, the mechanisms
involved in SAMe effects remain unknown. Most
studies of SAMe action in the joint have been focused
on osteoarthritic cartilage, where this compound
increases chondrocyte proteoglycan synthesis [9] and
proliferation rate [10]. The effect of SAMe on synovial
cells, the other main cell population in the joint,
remains largely unknown. The aim of this work was to
study the effects of SAMe as a protective agent against
the cell damage induced by a proinflammatory
cytokine such as tumour necrosis factor alpha (TNFa)
on cultured rabbit synovial cells.
MATERIALS AND METHODS
Synovial cell culture, cell characterization, DNA
synthesis and proliferation assays
Synovial tissue was obtained aseptically from
healthy New Zealand rabbits as previously described
[11]. Synovial cells were cultured and characterized as
recently reported [12]. A DNA-dependent fluorescenceenhancement assay was employed to determine the
concentration of DNA [13], and results were expressed
as micrograms of DNA per millilitre. Otherwise,
synovial cells seeded in 96 well microtitre plates (Nunc,
Roskilde, Denmark) were employed in methylene blue
proliferation assays to assess the possible variations
in cell number [14], and results were expressed as
percentage of proliferation vs basal considered as
100%. Prior to all experiments, the cells were
maintained for 48 h in RPMI 1640 medium (Gibco
BRL, Paisley) containing 0.5% fetal calf serum (FCS)
(BioWhittaker Inc., Walkersville, MD, USA) to reach
a quiescent state.
Submitted 21 February 1996; revised version accepted 11 July
1996.
Correspondence to: J. Egido, Research Laboratory, Fundación
Jiménez Dı́az, Avda. Reyes Católicos 2, 28040 Madrid, Spain.
RNA extraction and Northern blot analysis
Subconfluent synoviocytes seeded on Petri culture
dishes (Nunc) were incubated with 100 U/ml rh-TNFa
= 1997 British Society for Rheumatology
27
28
BRITISH JOURNAL OF RHEUMATOLOGY VOL. 36 NO. 1
(Immugenex, Los Angeles, CA, USA), SAMe (Sigma
Chemical Co., St Louis, MO, USA), or both, for 24 h.
Total RNA extraction, electrophoresis (15 mg/lane),
hybridization and autoradiography were performed as
described by Chomczynski and Sacchi [15]. Plasmid
pFH154, encoding the gene for FN (American Type
Culture Collection, Rockville, MD), was prepared as
reported previously [16] and labelled by nick translation (Boehringer Mannheim, Indianapolis, IN, USA)
with [32P]dCTP (DuPont, New England Nuclear Corp.,
Boston, MA, USA). The relative signal of the bands
obtained by autoradiography was measured by
densitometric scanning and normalized by 28S
expression. Results were expressed as the percentage of
variation of RNA expression vs basal values considered
as 100%.
Total protein and fibronectin synthesis
Subconfluent quiescent synovial cells seeded in 24
well plates (Nunc) were incubated for 24 h with the
stimuli in RPMI 1640 medium without methionine
(Gibco) supplemented with 0.5% FCS and 100 mCi/ml
of [35S]methionine (DuPont). After incubation, supernatants and cells were collected separately and stored
at −80°C until use.
Total protein synthesis was measured in aliquots of
both supernatants and cell fractions by protein
precipitation with 5% trichloroacetic acid (TCA) on ice
for 20 min. The precipitates were washed twice with
ice-cold 5% TCA, solubilized in 0.5  NaOH and
neutralized with 0.5  HCl. Radioactivity was determined by a liquid scintillation counter and d.p.m.
values were normalized by the concentration of DNA
in each well. Results were expressed as the percentage
of [35S]methionine incorporation vs basal values
considered as 100%.
Fibronectin (FN) synthesis was measured by
immunoprecipitation, electrophoresis and autoradiography procedures as previously described [17],
employing a polyclonal goat anti-human FN antibody
(Seralab, Sussex). Laser densitometry was used to
quantify the relative intensity of the bands, and the
densitometric arbitrary units obtained were related to
DNA concentration in each well. Results were
expressed as the percentage of [35S]methionine content
vs basal values considered as 100%.
Proteoglycan synthesis
Total proteoglycan synthesis was measured based
upon the rate of sodium [35S]sulphate (1 mCi/ml,
DuPont) incorporation into glycosaminoglycan chains,
as previously reported [18]. Confluent quiescent
synovial cells seeded in 96 well microtitre plates were
incubated for 24 h with the stimuli in medium
containing 0.5% FCS and sodium [35S]sulphate
(1 mCi/ml). Supernatants were collected, cold 0.1 
NaOH was added and neutralized with equal volumes
of 0.1  HCl. Polysaccharides were precipitated by the
addition of 1% cetylpyridinium chloride, collected on
a Millipore filter disc (f25 mm, 0.45 mm pore size) and
washed with 0.3  NaCl to remove unincorporated
radioactivity. The retained radioactivity was measured
in a b-counter. Results were expressed as the
percentage of [35S]sulphate incorporation vs basal
values considered as 100%.
Statistical analysis
Data are expressed as the mean 2 ... Student’s
t-test was employed to compare groups. A value of
P Q 0.05 was considered as significant.
RESULTS
Effects of TNFa and SAMe on synovial cell proliferation
The incubation of subconfluent quiescent synovial
cells with 100 U/ml TNFa for 24 h reduced cell
proliferation (58 2 14% vs basal, P Q 0.05, n = 4)
(Fig. 1). This effect was blocked by the addition of
increasing concentrations of SAMe (10−10–10−5 ) after
3 h of TNFa stimulation, while SAMe alone had no
significant effect (ranging from 87 2 16% to
114 2 12% vs basal with 10−10–10−5  SAMe,
P q 0.05, n = 5). Only 10−4  SAMe, alone or with
TNFa, occasionally elicited a proliferative response in
synovial cells (142 2 8% vs basal, P Q 0.05, n = 4).
Since at this concentration a certain cytotoxic effect
was noted, it was not used in further studies.
Effects of TNFa and SAMe on FN mRNA expression
The amount of FN mRNA expression in untreated
and SAMe-treated synovial cells was determined by
hybridization of extracted total RNA with the
appropriate cDNA probe. Densitometric scanning
revealed that cell incubation with 100 U/ml TNFa for
24 h decreased FN mRNA expression (36 2 14% vs
basal, P Q 0.05, n = 3), and the addition of 10−6 
SAMe restored basal levels (131 2 31% with SAMe vs
F. 1.—Effects of TNFa and SAMe on synovial cell proliferation.
Subconfluent quiescent rabbit synoviocytes were incubated for 24 h
with 100 U/ml TNFa, in the absence or presence of increasing doses
of SAMe. Proliferation was measured based upon methylene blue
assays. Results were expressed as the mean 2 ... percentage of
proliferation vs basal values, considered as 100%, of 4–5 experiments
performed in triplicate (*P Q 0.05 vs basal).
29
GUTIERREZ ET AL.: SAMe RESTORES TNFa EFFECTS
TABLE I
Effects of SAMe on total protein synthesis in cultured rabbit synovial
cells
Percentage of protein synthesis
Stimuli
Basal
SAMe
SAMe
SAMe
SAMe
10−8
10−7
10−6
10−5




No TNFa
TNFa (100 U/ml)
preincubation
100
111 2 6
119 2 7
122 2 13
72 2 31
127 2 12*
135 2 33
100 2 17
129 2 25
77 2 8*
Total protein synthesis by synovial cells stimulated with increasing
doses of SAMe for 24 h was measured based upon 35S-methionine
content in TCA precipitable material. Results are expressed as
mean 2 ... percentage of radioactivity corrected by mg DNA vs
basal values taken as 100% in four experiments performed in
duplicate. The absolute basal value in protein synthesis studies was
280 × 103 2 11 × 103 dpm (*P Q 0.05 vs basal).
case, the incubation with the cytokine induced a small,
but not statistically significant, decrease in FN
synthesis (79 2 20% vs basal, P q 0.05, n = 4),
probably due to the translation of the remaining FN
mRNA. Again, the addition of 10−8–10−6  SAMe
raised FN synthesis levels (Fig. 3), although the
concentrations of SAMe employed did not induce
significant variations in this process (ranging from
120 2 15 to 81 2 20% of [35S]methionine content vs
basal).
F. 2.—Effects of TNFa and SAMe on FN expression by
synoviocytes. Subconfluent quiescent cells were stimulated with
100 U/ml TNFa, 10−6  SAMe, or both, for 24 h. Total RNA was
extracted and hybridized with the FN probe. The relative signal of
the bands was related to 28S expression. (A) Representative
Northern blot analysis of FN expression by synoviocytes. 1: Basal;
2: 10−6  SAMe; 3: 100 U/ml TNFa; 4: TNF + SAMe. (B)
Representation of the relative variations of FN expression. Results
were expressed as the mean 2 ... percentage of FN expression vs
basal 100% in 3–4 separate experiments (*P Q 0.05).
Effects of TNFa and SAMe on proteoglycan synthesis
Total proteoglycan synthesis was measured based
upon the rate of sodium [35S]sulphate incorporation
into glycosaminoglycan chains. In agreement with
previous studies [19], radioactivity levels appeared in
basal, P Q 0.05, n = 3) (Fig. 2). Again, the absence of
significant effects on FN mRNA expression with 10−6 
SAMe at 24 h was observed.
Effects of TNFa and SAMe on total protein and FN
synthesis
The incubation of synovial cells with TNFa for 24 h
induced an increase in total protein synthesis on
synovial cells (Table I). These protein levels returned to
basal values when 10−8–10−6  SAMe was added after
3 h of TNFa stimulation, and were significantly
reduced by 10−5  SAMe. Nevertheless, none of the
SAMe concentrations employed altered total protein
synthesis.
FN synthesis was determined by immunoprecipitation of supernatants and cell fractions from synoviocytes incubated with 100 U/ml TNFa for 24 h. In this
F. 3.—Effect of TNFa and SAMe on FN synthesis in cultured
rabbit synovial cells. Subconfluent synoviocytes were incubated in
[35S]methionine-labelled medium for 24 h with 100 U/ml TNFa,
increasing doses of SAMe, or both. After immunoprecipitation,
electrophoresis and autoradiography, the relative signal of the bands
was evaluated by laser densitometry and related to total DNA
content in each point. Results were expressed as the mean 2 ...
percentage of FN synthesis vs basal values, considered as 100%, of
4–5 separate experiments (*P Q 0.05).
30
BRITISH JOURNAL OF RHEUMATOLOGY VOL. 36 NO. 1
F. 4.—Effect of TNFa and SAMe on total proteoglycan synthesis
by synovial cells. Confluent quiescent synovial cells were incubated
in [35S]sulphate-labelled medium for 24 h. Results were expressed as
the mean 2 ... percentage of [35S]sulphate incorporation vs the
basal value (100%) after incubation of the cells with 100 U/ml
TNFa, increasing doses of SAMe, or both, in 3–4 experiments
performed in quadruplicate. The absolute value of [35S]sulphate
uptake in basal at 24 h was 9568 2 1694 d.p.m. (*P Q 0.05).
supernatant fractions, indicating that synovial proteoglycans were released into the medium (85%).
The stimulation of synovial cells with 100 U/ml
TNFa for 24 h induced an increase in total proteoglycan synthesis (239 2 40% of [35S]sulphate incorporation vs basal, P Q 0.05, n = 3) (Fig. 4). The addition of
SAMe, mainly at the higher doses (10−6–10−5 ),
diminished the TNFa effect (050–60% of reduction in
[35S]sulphate uptake vs basal), even when only the
concentration of 10−6  SAMe modified proteoglycan
synthesis (133 2 12% of [35S]sulphate uptake vs basal,
P Q 0.05, n = 4).
DISCUSSION
This study was focused on the protective effect of
SAMe against TNFa actions on proliferation, total
protein and extracellular matrix components synthesis
in cultured synovial cells. Although SAMe alone did
not significantly alter these processes in general, an
almost total reversion of TNFa action on synoviocytes
could be observed with SAMe concentrations ranging
from 10−10 to 10−6 .
The beneficial effect of SAMe administration has
been well demonstrated in various diseases [4–6]. The
studies on the possible participation of SAMe as a
therapeutic agent in articular diseases have been
focused on OA, where this compound increases
chondrocyte proliferation, and protein and proteoglycan synthesis [9, 10]. Moreover, the potential action of
SAMe on synovial cells, the other main cellular
component in the joint, has also been proposed [20].
Large amounts of TNFa have been detected in the
synovial fluid of patients with different chronic arthritis
[21]. Since TNFa is a proinflammatory cytokine
implicated in the pathogenesis of joint diseases, we
have examined the potential effects of SAMe on the
modifications induced by TNFa on cultured synovial
cells.
The incubation of the cells with TNFa caused a
decrease in synovial cell proliferation. Subsequent
studies showed that TNFa also decreased FN mRNA
expression and induced a small diminution of FN
synthesis in cultured synoviocytes, although the
cytokine raised total protein and proteoglycan
synthesis. The treatment with SAMe restored almost
basal values even though SAMe alone did not have any
significant effect on synoviocytes in most of the studies
considered. Similarly to cultured chondrocytes [9] and
skin fibroblasts [22], incubation of synovial cells with
SAMe did not affect cell proliferation. In agreement
with those data, non-significant variations in total
protein synthesis and production of FN were observed
in this study.
The almost universal distribution of SAMe and its
complex role in metabolic processes complicate the
characterization of its effects on cellular biochemistry.
The mechanisms involved in SAMe actions remain to
be established. Nevertheless, the results obtained in our
study suggest that SAMe provides a protective effect
against the metabolic changes induced by TNFa on
synovial cells.
As described previously by Vara et al. [23], the
protective effect of SAMe could be based on the
blockade of cytokine-induced glutathione depletion. In
this sense, the incubation of cultured hepatocytes with
a cytokine mixture caused a reduction in SAMe
synthetase and methyltransferase [24]. On the other
hand, SAMe induces the synthesis of polyamines,
which might stabilize the polyanionic macromolecules
of proteoglycans and potentially protect them from the
attack of proteolytic and glycolytic enzymes [25]. As
SAMe stimulated the incorporation of sulphate into
proteoglycans of cartilage chondrocytes [9], it could
certainly have a reparative role in osteoarthritic
cartilage.
Whether these modifications are related to the
improvement observed in rheumatic patients remains
to be determined. Further studies focused on the
potential role of SAMe in matrix proteins in
experimental and human joint diseases are necessary.
In conclusion, these results indicate that SAMe
restores basal conditions after cell damage elicited by
TNFa stimulation. Moreover, some of the beneficial
effects of SAMe observed in osteoarthritic patients
could be based on its capacity to protect the cells
against the activity of proinflammatory mediators.
A
We would like to thank L. Gulliksen for secretarial
assistance. This work was supported by a grant from
Fondo de Investigación de la Seguridad Social
(91/158), Ministerio de Eduación y Ciencia (PM
92/0042, PB 94/0211) and Fundación Renal Iñigo
Alvarez de Toledo. SG, IP and PH are fellows of the
Fundación Conchita Rábago.
GUTIERREZ ET AL.: SAMe RESTORES TNFa EFFECTS
R
1. Stramentinoli G. Pharmacological aspects of S-adenosylmethionine: pharmacokinetics and pharmacodynamics.
Am J Med 1987;83:35–42.
2. Tabor H, Tabor CW. Polyamines. In: Methods in
enzymology. New York: Academic Press, 1983.
3. Williams-Ashaman HG, Pegg AE. Aminopropyl group
transfer in polyamine biosynthesis. In: Morris DR,
Marton LJ, eds. Polyamines in biology and medicine.
New York: Marcel Dekker, 1981:43–68.
4. Friedel HA, Goa KL, Benfield P. S-adenosyl-L-methionine: a review of its pharmacological properties and
therapeutic potential in liver dysfunction and affective
disorders in relation to its physiological role in cell
metabolism. Drugs 1989;38:389–416.
5. Lieber C, DeCarlo IM, Lowe N, Sasaki R, Leo MA.
S-adenosyl-L-methionine attenuates alcohol-induced
liver injury in the baboon. Hepatology 1990;11:165–
72.
6. Agnoli A, Andreoli V, Casacchia M, Cerbo R. Effect of
S-adenosyl-L-methionine (SAMe) upon depressive symptoms. J Psychiatric Res 1986;13:43–54.
7. Di Padova C. S-adenosylmethionine in the treatment of
osteoarthritis. Review of the clinical studies. Am J Med
1987;83:60–5.
8. Bradley JD, Flusser D, Katz BP, Schumacher HR,
Brandt KD, Chambers MA et al. A randomized, double
blind, placebo controlled trial of intravenous loading
with S-adenosylmethionine (SAM) followed by SAM
oral therapy in patients with knee osteoarthritis. J
Rheumatol 1994;21:905–11.
9. Harmand MF, Vilamitjana J, Maloche E, Duphil R,
Ducassou D. Effects of S-adenosylmethionine on human
articular chondrocyte differentiation. An in vitro study.
Am J Med 1987;83:48–54.
10. Barceló HA, Wiemeyer JCM, Sagasta CL, Macias M,
Barreira JC. Effect of S-adenosylmethionine on experimental osteoarthritis in rabbits. Am J Med 1987;83:55–
9.
11. Fraser JRE, Catt KJ. Human synovial cell culture: use
of a new method in a study of rheumatoid arthritis.
Lancet 1967;ii:1437–9.
12. Gutierrez S, Palacios I, Egido J, Zarco P, Miguélez R,
González E et al. IL-1b and IL-6 stimulate the
production of platelet-activating factor (PAF) by
cultured rabbit synovial cells. Clin Exp Immunol
1995;99:364–8.
13. Adams CJ, Storrie B. A simple DNA-dependent
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
31
fluorescence enhancement assay for cell number. J
Histochem Cytochem 1981;29:326–8.
Oliver MH, Harrison NK, Bishop JE, Cole PJ, Laurent
GJ. A rapid and convenient assay for counting cells
cultured in microwell plates: Application for assessment
of growth factors. J Cell Sci 1989;92:513–8.
Chomczynski P, Sacchi N. Single-step method of RNA
isolation by acid guanidium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987;162:156–9.
Maniatis F, Fritsch EF, Sambrook J. Molecular cloning.
A laboratory manual. New York: Cold Spring Harbor
Laboratory Press, 1989.
Takasaki I, Chobanian AV, Brecher P. Biosynthesis of
fibronectins by rabbit aorta. J Biol Chem
1991;266:17686–94.
Kato Y, Nomura Y, Daikuhara Y, Nasu N, Tsuji M,
Asada A et al. Cartilage-derived factor (CDF). I.
Stimulation of proteoglycan synthesis in rat and rabbit
costal chondrocytes in culture. Exp Cell Res
1980;130:73–81.
Bustos C, González-Cuadrado S, Ruiz-Ortega M,
Gómez-Guerrero C, González E et al. Cyclosporin A
(CsA) modulates the glomerular production of inflammatory mediators and proteoglycans in experimental
nephrosis. Clin Exp Immunol 1995;102:608–13.
Schumacher HR. Osteoarthritis: The clinical picture,
pathogenesis, and management with studies on a new
therapeutic agent, S-adenosylmethionine. Am J Med
1987;83:1–4.
Arend WP, Dayer JM. Cytokines and cytokine inhibitors
or antagonists in rheumatoid arthritis. Arthritis Rheum
1990;33:305–15.
Casini A, Banchetti E, Milani S, Maggloni Moratti E,
Surrenti C. S-adenosylmethionine inhibits collagen
synthesis by human fibroblasts in vitro. Exp Clin
Pharmacol 1989;11:331–4.
Vara E, Arias-Dı́az J, Garcı́a C, Villa N, Simón C, Ortiz
P et al. S-adenosyl-L-methionine might protect transplanted hepatocytes against the toxic effects of cytokines.
Transplant Proc 1994;26:3363–5.
Arias-Dı́az J, Villa N, Garcı́a C, Garcı́a-Carreras C,
Ortiz P, Vara E et al. S-adenosyl-methionine and
N-acetyl-cysteine protect human hepatocytes differently
against the toxic effect of cytokines. Curr Opin Surg
Infect 1995;3:540–1.
Conroy PD, Simms DM, Pointon JJ. Occurrence of
ornithine decarboxylase and polyamines in cartilage.
Biochem J 1977;162:347–50.