CARTILAGE Articular carti lage proteoglycans in osteoarthrit ic STR/Ort mice
BY JOHN D. GAFFEN*, SUSAN J. GLEAVE*, MARGOT V. CROSSMAN*, MICHAEL T. BAYLISSt AND ROGER M. MASON *
• Department of Biochemistry, Charing Cross and Westminster Medical School, London, W68RF, U.K. and tDivision of Biochemistry, Kennedy Institute of Rheumatology, Bute Gardens,
London W6 7D W, U.K.
S u m m a r y
Compared to controls, the lateral and medial tibial articular cartilage chondroitin sulfate (CS) content in male STR/Ort mice was elevated between 8 and 19 weeks of age, fell at 24-26 weeks and increased again thereafter. The CS cartilage content of CBA mice remained relatively unchanged.
Cartilage CS content was measured in female CBA and STR/Ort mice and in male and female Balb C mice, all at 18 weeks of age. In every case the content was much lower than that found in male STR/Ort mice.
CS unsaturated disaccharides were analysed by capillary electrophoresis after digestion of the glycosaminoglycans with chondroitinase ABC. Chondroitin-4-sulfate (C4S) was the predominant isomer at all ages in both strains. The C4S: C6S isomer ratio in CBA mice was much higher in the lateral than the medial cartilage. This difference was much less marked in STR/Ort knee joints: these mice have relatively more C6S than CBA mice.
Proteoglycans, extracted from STR/Ort and CBA male mouse cartilage, were characterized by large pore gel electrophoresis and Western blotting. All cartilages contained two slow mobility bands identified as aggrecan by the reactivity with the mab 1C6. Both bands contained C4S and C6S chains. A third band of faster mobility contained only C4S and was 1C6 negative~ It was present in both strains. Thus the STR/Ort cartilage contained a normal spectrum of murine articular cartilage proteoglycans. Key words: Proteoglycans, Osteoarthritis, Animal model, Cartilage.
I n t r o d u c t i o n
ANIMAL MODELS of h u m a n disorders p rov ide infor- ma t ion a bou t b iochemica l changes in t issues pre- ced ing ove r t express ion of disease. The Pond-Nuk i model has been inves t iga ted ex tens ive ly in osteo- a r th r i t i s (OA). The model is acu te ly induced by sec t ion ing the can ine a n t e r i o r c ruc i a t e l i gamen t [1] and p roduces a rap id d e t e r i o r a t i o n of the art ic- u l a r ca r t i l age [2]. Ea r ly changes occu r in the b iochemis t ry of the a r t i c u l a r ca r t i l age proteogly- cans [3, 4]. In c o n t r a s t to induced disease, ea r ly b iochemica l changes in n a t u r a l l y - o c c u r r i n g ani- mal o s t e oa r th r o se s have been less well s tudied. STR/Or t mice prov ide a we l l -charac te r ized model of id iopa th ic OA [5]. Abou t 85% of male STR/Or t mice develop h i s to log ica l lesions in the media l t ibial p l a t e au which c lose ly resemble those of h u m a n OA. Les ions develop f rom abou t 20 weeks
Submitted 2 May 1994; accepted 27 September 1994. Supported by the Arthritis and Rheumatism Council (U.K.). Address correspondence to: Professor R.M. Mason, Depart-
ment of Biochemistry, Charing Cross and Westminster Medical School, Fulham Palace Road, London, W6 8RF, U.K.
of age, usua l ly in the ca r t i l age ad jacen t to the inse r t ion of the a n t e r i o r c r u c i a t e l igament , and m ay progress to invo lve all a reas of the media l p la teau . The l a te ra l condy le is a lmost a lways spared.
The STR/Or t l ine was deve loped f rom the STR/1N l ine [6]. P r o t e o g l y c a n s ex t r ac t ed f rom STR/1N mouse a r t i c u l a r ca r t i l age showed no d i f ferences in t he i r abi l i ty to aggrega te , no r in the su l fa t ion of t he i r g lycosaminog lycans , r a t i o of chondroi t in-4-su l fa te (C4S) to chondro i t in -6-su l fa te (C6S), b~ l ink prote ins , w h en compared to STR con t ro l mice, which h av e a m u c h lower i nc idence of OA [7]. However , 4 M guan id ine h y d r o c h l o r i d e ex t r ac t ed a g rea t e r p ro p o r t i o n of 85S-labelled pro- t e o g l y c a n f rom the a r t i c u l a r ca r t i l age of STR/1N mice compared to cont ro ls . No q u a n t i t a t i v e b iochemica l m e a s u r e m e n t s of the p r o t e o g l y c a n co n t en t of e i t he r STR/1N or STR/Or t c a r t i l age have been repor ted . However , severa l ea r ly meta- bolic d i s tu rbances h av e been found in the la t te r . H i s tochemica l s tudies by A l tm an [8] ind ica ted t h a t l a c t a t e d e h y d r o g e n a s e and succ ina t e dehydrogen- ase def ic iency is p r e sen t in ch o n d ro cy t e s p r i o r to
96 Gaffen et al: P r o t e o g l y c a n s in o s t e o a r t h r i t i c m i c e
major histological evidence of degeneration. Also, there are changes in the or ientat ion of the articu- lar carti lage proteoglycans and a dis turbance in the act ivi ty of chondrocyte glucose-6-phosphate dehydrogenase [9, 10]. Recently, Chambers et al. [11] reported abnormal activity and distr ibution of monoamine oxidase in male STR/Ort mice.
We have taken advar_tage of recent technical developments to investigate the s t ructure of art icular cart i lage proteoglycans during the development of idiopathic OA in male STR/Ort mice and to compare this with age-matched control CBA mice. The aim was to investigate whether changes in proteoglycan s t ructure and content precede overt expression of disease.
M a t e r i a l s a n d M e t h o d s
MAINTENANCE OF THE STR/ORT COLONY
A colony was established from breeding pairs of STR/Ort mice obtained from Har lan Olac Ltd (Shaw's Farm, Blackthorn, Bicester, Oxon., OX6 OTP, U.K.). The animals were inbred by strict s is ter-brother mating. Mice were maintained on a diurnal rhythm of 12 h light, 12 h dark (tempera- ture, 20-21°C; humidity, 40-50~/o) and all the animals were fed with irradiated CRM nuts (Special Diet Services, Witham, Essex, U.K.). Mice were caged in groups of up to 4-5 animals, since their behaviour was docile.
EXTRACTION OF MOUSE TIBIAL PLATEAU CARTILAGE
Full thickness medial and lateral t ibial plateau art icular carti lage was dissected from STR/Ort mice of about 9, 17, 25, 32 and 45 weeks and from age-matched CBA control mice. The carti lage was transferred to pre-weighed and pre-cooled Eppen- dorf tubes which were sealed rapidly to prevent water loss, and re-weighed to obtain the carti lage wet weight. Proteoglycans were extracted with a sequential procedure (4 M guanidine HC1 (GuHC1); ul t rapure collagenase, 4M guanidine HC1; NaOH) [12] or with 0.5 m NaOH alone. Car- tilage was extracted on a rotary shaker at 4°C for 48h with 100/~1 of 4M GuHC1 (El) containing protease inhib i tors (100mM aminohexanoic acid, 10 mM Na2 EDTA, 5 mM benzamide HC1, 1 mM N- ethylmaleimide, 1 mM phenyl methyl sulfonylfluo- ride; Sigma Chemical Co.). After centrifugation, the tubes were placed on ice and the supernatant removed using duck-billed pipette tips (Bioquote Ltd, The Raylor Centre, J a m e s Street, York, North Yorkshire, YO1 3DW, U.K.). The carti lage residues were washed twice with 100pl Tris
washing buffer (1 mM Tris-HC1, 2 mM CaC12, pH 7.2) to remove the GuHC1 and then incubated with 100#l buffer (25mM Tris-HC1, 10mM CaC1, 150mM NaC1) containing 132 units of type IH collagenase (Advanced Biofactures, New York) and protease inhibitors (0.1 mM pepstatin, 0.1 mM leupeptin, 5 mM N-ethylmaleimide and 1 mM phenyl methyl sulfonylfluoride; Sigma Chemical Co.) at 37°C for 48h. After centrifugation, the supernatant was removed and the residue ex- t racted again with 4 mM GuHC1. The supernatant was combined with that from the collagenase di- gest (E2). Final ly 100t£1 of 0.5M NaOH were added to the t issue residuel and the tubes incu- bated at 4°C for 24h on a rotary shaker (E3). Some samples were extracted with alkali alone, using the same protocol as described for E3. All the extract supernatants were stored at -20°C before drop dialysis and freeze drying.
DROP DIALYSIS AND FREEZE DRYING
To remove buffer salts, GuHC1, or NaOH, the supernatants were diluted twice with distilled water and subjected to 24 h drop-dialysis at 4°C. Due to the small sample volume, dialysis was performed in inverted Eppendorf tubes with the centre of their lids cut out. The dialysis mem- brane was stretched across the lid and held be- tween the lid and the body of the Eppendorf. After dialysis, the samples were snap frozen and freeze dried using an Edwards Super Modulyo freezer dryer.
CHONDROITINASE ABC DIGESTIONS
After drop dialysis and freeze drying, one-sixth of the total alkali extracts were saved for 1°3ruthenium red dot blot analysis to quant i ta te glycosaminoglycans. This assays low nanogram amounts of glycosaminoglycan, and its use to measure murine art icular carti lage glycosamino- glycans has been fully described elsewhere [13]. The remainder was digested with chondroit in ABC lyase (0.03 units/ incubation, ICN Biomedicals Ltd, High Wycombe, U.K.) in a buffer containing 50 mM Tris base, 60 mM sodium acetate, pH 8.0, at 37°C, for 3 h. This generated unsatura ted chondroit in sulfate (CS) disaccharides for subsequent analysis by capillary zone electrophoresis (CZE). Samples were reduced to dryness by centrifugal evapor- ation (30 min, 40°C, Savant Speed Vac SC100) and reconst i tuted in 20 ~1 ultra-pure H20 immediately before loading onto the CZE.
O s t e o a r t h r i t i s a n d C a r t i l a g e Vol. 3 No. 2 97
CAPILLARY ZONE ELECTROPHORESIS
Unsa tura ted CS disaccharide isomers were sep- ara ted by the method of Carney and Osborne [14] using a CZE appara tus (Model 270 A, Ap- plied Biosystems, Warr ington, Cheshire, U.K.). S tandard unsa tu ra ted C4S and C6S disaccharides (ADi-4S, ADi-6S, ICN Biomedical) were used to cons t ruc t cal ibrat ion curves against which the unknown samples were assayed. Samples and standards were loaded by negat ive pressure (2 s load) in a 200mM orthophosphor ic acid buffer (pH3). Electrophoresis was performed using reverse polar i ty in the same buffer at 15kV, 40°C on a 72 cm column (50 #m i.d.). The column was moni tored at 232nm. Before loading, the column was pre-washed with 0.1M NaOH for 2rain followed by a 5min rinse with running buffer.
LARGE PORE GEL ELECTROPHORESIS AND
WESTERN BLOTTING
Whole proteoglycans ext rac ted from mouse ar- t icular cart i lage by 4 M GuHCI(E1) and quantified using the l°~ruthenium dot blot assay, were separated dissociatively on large pore agarose- acrylamide slab gels [15]. Samples were dissociated overnight in 8 M urea, and 10-15 ng loaded per lane. The running buffer was 10 m~ Tris /aceta te (pH6.8), 0.25mM Na2SO4 and 4M urea. Samples were run into the gel at 60 V and then separated at 160-180 V unti l the bromophenol blue marker had migrated 3cm. Separated proteoglycans were t ransferred on to nylon membranes using a Hoeffer semi:dry b l o t t e r appara tus (1 h, 80 mA). The blot t ing buffer was the same as the running buffer but wi thout urea. After blocking and chondroiti- nase digestion, the membranes were probed with monoclonal antibodies (mabs) 3B3 and 2B6 to de- tect C6S and C4S l inkage region stubs, respect ively [16]. The membranes were reprobed with mab 1C6 after reduct ion and a lkyla t ion [17]. This anti- body recognizes epitopes in the G1 and G2 domains of aggrecan and requires prior reduct ion and a lkyla t ion for opt imum exposure of the amino acid sequence-QAAY [18]. Antibody binding was detected with an anti-mouse poly- valent immunoglobulin l inked to horseradish peroxidase and visualized by enhanced chemi- luminescence using Hyperfilm-MP (Amersham Internat ional) .
STATISTICAL ANALYSIS
Stat ist ical analysis was performed using the
Student ' s t-test for unpaired data. Data is presented in all figures as the mean + S.E.M.
R e s u l t s
CARTILAGE CS CONTENT
The total sulfated glycosaminoglycan con ten t of mur ine a r t icu la r car t i lage was measured using the l°aruthenium red assay and expressed as # g/mg wet weight of car t i lage (Fig. 1). The amount of gly- cosaminoglycan in the STR/Ort car t i lage was
20L (a) ** 1
10
0 h~
2o
10
tm : 1
(b) **
0 ~ 8-9 16-19 24-26 31-34 41-48
Age in weeks
FIG. 1. Quantitation of mouse cartilage glycosaminogly- cans~'with the l°%uthenium red assay. Proteoglycans were extracted from the lateral (a) and medial (b) tibial articular cartilage of STR/Ort ( . ) and CBA mice (f~) with either a sequential procedure (4 M GuHC1, collagen- ase, 4 M GuHC1, NaOH) or with 0.5 M NaOH alone and assayed for glycosaminoglycan content using the 1°3ruthenium red dot blot assay. Results are expressed as the mean + S.E.M. from 3-5 animals in each group. In this experiment and subsequently, cartilage extracts from the left and right legs were analysed separately and the data combined. Asterisks denote statistically significant differences between CBA and STR/Ort mice: *P < 0.05; **P ~< 0.01; ***P ~< 0.002. For the overall mean difference between the two strains P ~< 0.002 and <0.001 for lateral and medial tibial cartilage respectively.
98 Gaffen et al: P r o t e o g l y c a n s in os teoarthr i t ic mice
cons i s t en t ly h i g h e r in all age g roups t h a n in CBA car t i lage , by a f a c to r of be tween 1.35-3.8 t imes (overal l m e a n di f ference, P < 0.001 for bo th media l and l a te ra l t ib ia l car t i lage; Fig. 1). P r i o r d iges t ion wi th c h o n d r o i t i n lyase ABC e l imina ted all i n t e r a c t i o n wi th dye, i nd ica t ing t h a t the g l y c o s a m i n o g l y c a n is a t t r i b u t a b l e sole ly to c h o n d r o i t i n / d e r m a t a n sulfate.
The me a n CS c o n t e n t of CBA ca r t i l age in y o u n g animals (8-9 weeks) was abou t 5#g /mg . S imi la r va lues were seen in old an imals (41-48 weeks) wi th an a p p a r e n t dec rease in be tween these ages (24-26 weeks). In con t ras t , the CS c o n t e n t of bo th the medial and l a t e ra l t ibial ca r t i l age of STR/Or t mice was h ighe r in 8-9-week-old an imals ( ~ 7 . 5 pg/mg), inc reased g rea t ly by 16-19 weeks ( ~ 16/~g/mg) and t he n dec reased g rea t ly by 24-26 weeks ( ~ 3 #g/mg). T h e r e a f t e r i t i nc reased aga in as age progressed, r e a c h ing a level of a b o u t 8 #g /mg by 41-48 weeks (Fig. 1).
The CS c o n t e n t of m u r i n e a r t i c u l a r ca r t i l age was also m e a s u r e d by assay ing C4S and C6S unsa tu- r a t ed d i sacchar ides by CZE, fo l lowing d iges t ion wi th c hond r o i t i na se ABC (Fig. 2). Of the 24 mice used for these measu remen t s , 13 were f rom the series used for the dot b lo t assays above, whi ls t 11 were f rom a new series. The to ta l d i saccha r ide co n t e n t (ADi - 4S + ADi - 6S) was aga in m u c h g rea t e r in STR/Or t ca r t i l age t h a n in the age- ma t c he d con t ro l CBA ca r t i l age (P < 0.001 for the overa l l m e a n d i f fe rence for bo th l a t e ra l and media l p la teaux; Fig. 2). The two methods of m e a s u r e m e n t gave resu l t s for each age group and s t r a in which were gene ra l ly in close a g r e e m e n t wi th one a n o t h e r (Figs 1 & 2). However , the level of proteo- g lycan in the 17-19 week STR/Or t mice was no t so h igh in the e l e c t r o p h o r e t i c ana lys i s (Fig. 2) as in the dot b lo t ana lys i s (Fig. 1). Never the less , t he re was a m a r k e d fall in the media l ca r t i l age CS con t e n t a t 25-26 weeks fol lowed by an inc rease at l a t e r ages [Fig. 2(b)]. The fall o c c u r r e d l a t e r in the la te ra l car t i lage , bu t aga in the re was an inc rease in CS levels in o lder mice [Fig. 2(a)]. In con t ras t , the CS c o n t e n t in CBA mouse a r t i c u l a r ca r t i l age r ema ined a lmos t c o n s t a n t across the age r an g e studied. No di- or t r i su l fa ted d i sacchar ides were found in CS f rom e i the r s t ra in .
The C4S i somer is p r e d o m i n a n t a t all ages in bo th s t ra ins (Fig. 2). In the CBA mice the chon- d ro i t in 4:6 su l fa te i somer ra t io was g r e a t e r in the la te ra l t h a n the media l t ibial p l a t e au [Fig. 3(a), P <0.001 for the overa l l m e a n difference]. However , this d i f fe rence was far less p r o n o u n c e d in the STR/Or t mice [Fig. 3(b)]. In addi t ion, the STR/Or t mice have re l a t ive ly more C6S t h a n the CBA mice s ince the i r 4 :6 r a t i o s were gene ra l ly
15
10
5
0
15
10
i h
S C S C S C S C S C 9 17-19 25-26 31-34 41-44
Age in weeks FIG. 2. Analysis of mouse cartilage CS by CZE. Total alkali extracts were subject to chondroitinase ABC digestion. The resultant mixture of CS disaccharide isomers were quantified by capillary zone electrophor- esis. The ADi-4S ( I ) and ADi-6S ([]) isomer content of the mouse lateral (a) and medial (b) tibial articular cartilage is expressed as mg per mg wet cartilage. Each point is the mean _+ S.E.M. from 2-3 animals. Asterisks denote statistically significant differences between CBA (C) and STR/Ort (S) mice: *P<0.1 ; **P<0.05; ***P <0.02; ****P < 0.002 (P ~< 0.001 for the overall mean difference between the two mouse strains for both the lateral and medial cartilage).
lower (P < 0.001 f o r the overa l l m e a n d i f fe rence for bo th media l and l a t e ra l p la teaux) .
To examine w h e t h e r the e leva ted CS c o n t e n t in STR/Or t ca r t i l age is due on ly to a d i f fe rence in gender or s t ra in we compared 18-week-old STR/Or t mice of b o th sexes wi th male and female CBA and Balb C mice of s imi la r age. M e a s u r e m e n t s were made wi th bo th an a ly t i c a l methods . The resu l t s (Fig. 4) d e m o n s t r a t e t h a t the h igh CS c o n t e n t is a ch a r ac t e r i s t i c of on ly male STR/Or t mice. 18-week female STR/Or t mice as well as male and female CBA and Balb C mice all had a m u c h lower t ib ia l a r t i cu l a r ca r t i l age CS.
Osteoarthrit is and Carti lage Vol. 3 No. 2 99
EXTRACTABILITY OF PROTEOGLYCAN FROM M O U S E
ARTICULAR CARTILAGE
Table I shows the proportion of the total proteo- glycan solubilized in the first GuHC1 extract (El) for both medial and lateral tibial cart i lage in STR/Ort and CBA mice at 8, 24 and 48 weeks of age. Comparing CBA and STR/Ort mice at each age there were no P-values smaller than 0.23. Thus there are no obvious differences in extractabil i ty between the two strains. Nor do proteoglycans of the STR/Ort medial plateau, where lesions develop, appear to be more readily extracted than those of the lateral plateau.
12 t (a)
10
4
2 o
0 00 ~O ~9 "" 12
c.) 10
0 9 17 25 31-33 41-43
Age in weeks
FIG. 3. Chondroitin 4:6 sulfate isomer ratios in the cartilage of CBA and STR/Ort mice. The CS disaccharide isomers were quantified by CZE as described in Fig. 2. The results, presented as the mean_ S.~..M. show the ratio of ADi-4S:ADi-6S isomers in the lateral (m) and medial ([]) tibial articular cartilage for both CBA (a) and STR/Ort mice (b). Asterisks denote statistically significant differences between the lateral and medial tibial articular cartridge: *P < 0.1; **P < 0.05; ***P <0.02 and ****P <0.01. For the overall mean difference between lateral and medial cartilage P < 0.0001 and P = 0.003 for CBA and STR/Ort mice respectively.
IDENTIFICATION OF M U R I N E ARTICULAR CARTILAGE
PROTEOGLYCANS
The extracted proteoglycans were analysed by large pore gel electrophoresis and Western blot- ting. After t ransferr ing the proteoglycans to a nylon membrane and digesting with chondroi t in lyase ABC they were probed with mabs 2B6 a n d 3B3. These detect the 4- and 6-sulfated unsa tura ted disaccharides left on the nonreducing end of the l inkage region of the CS chains removed by the lyase. Neither antibody revealed any notable differences in the proteoglycan species present in CBA and STR/Ort ar t icular cart i lage at 8, 24 or 48 weeks of age (Fig. 5). Three proteoglycan bands of differing mobility were substi tuted with ADi-4S and presumably carried chains of the same compo- sition. The two slower mobility bands also contained ADi-6S and similarly are presumed to carry C6S chains.
The mab 1C6, which detects epitopes in the core protein of aggrecan, was used to re-probe the membranes: Only the two bands of slower mobili ty were detected in both species at each age exam- ined. (Representative photographs obtained for 48 week male STR/Ort mice are shown in Fig. 6.) Thus these two out of the three proteoglycan species are identified as aggrecan. The third C4S-rich proteo- glycan is as yet unidentified.
D i s c u s s i o n
Osteoarthri t ic lesions appear in the medial tibial carti lage of STR/Ort mice in the Charing Cross colony from about 20 weeks of age and by 48 weeks histopathological changes are found in 85~/o of the male animals. Initially, roughening of the cart i lage surface occurs, followed by the development of erosions, fibrillations, and in some animals loss of carti lage fragments down to the level of t h e t idemark [19]. In contrast, the incidence of OA in CBA mice is extremely low. We have not detected histological evidence of the disease in serial sec- tions of knee joints from over one hundred CBA mice (Chambers and Mason, unpublished results). Thus CBA male mice were used as controls in the present experiments.
We have shown tha t from an early age, and prior to the development of histological OA lesions, the CS content in the tibial ar t icular cart i lage of STR/Ort mice is elevated compared to tha t of CBA controls. Both C4S and C6S are increased. Proteo- glycan biosynthesis involves the co-ordination of protein synthesis and post-translational glycosyl transferase act ivi ty to form the glycosaminoglycan chains [20]. Thus increased
100 Gaffen et al: P r o t e o g l y c a n s in o s t e o a r t h r i t i c m i c e
20
10
0
h~ i5
10
(a)
F % *****
20
i0
I i
(b)
15
10
0 0 MST FST MCB FCB MBC FBC MST FST MCB FCB MBC FBC
FIG. 4. Measurement of glycosaminoglycans in 18 week mice of different strain and sex. Glycosaminoglycans were alkali extracted from the medial (a, c) and lateral (b, d) tibial articular cartilage of 18 week male (M) and female (F) STR/Ort (ST), CBA (CB) and Balb C (BC) mice. The extracts were quantified by l°3ruthenium red blot assay (a, b). In addition, each extract was digested with chondroitinase ABC and the resultant ADi-4S ( i ) and ADi-6S ([]) isomers quantified by CZE analysis (c, d). Data is expressed as mean ± S.E.M. with 2-4 mice per group. A statistical comparison was made between the male STR/Ort mice and all other categories: *P<0.1 ; **P<0.02; ***P <0.01; ****P < 0.005; *****P < 0.001. (For the overall mean difference between male STR/Ort mice vs the rest P < 0.0001 for Fig. 4a and 4b, P < 0.003 for Fig. 4c and P < 0.013 for Fig, 4d).
amoun t s of CS could represen t an inc rease in normal ly g lycos lya ted p ro teog lycans , or be due to subs t i tu t ion of the core p ro te in wi th e i ther addi t iona l or longer CS chains. Al te rna t ive ly , addi t iona l p r o t e o g l y c a n s of a different species m a y
be synthesized. The resul ts of the Wes te rn ana lyses rule ou t the lat ter . Two of the th ree p ro t eog lycans in STR/Orr ca r t i l age were identif ied as a g g r e c a n by thei r r eac t iv i ty wi th the mab 1C6. This an t i body does no t have any cross r eac t iv i t y wi th la rge
Table I Extractability of proteoglycans from STR /Ort and CBA mouse cartilage. Tibial articular cartilages from CBA and STR/Ort mice aged 8, 24 and
48 weeks were dissected and extracted sequentially with 4 M GuHC1 (El), ultrapure collagenase, 4 M GuHCl (E2) and 0.5 M NaOH (E3). The amount of proteoglycan present in the El , E2 and E3 extracts was quantified by
l°3ruthenium red dot blot analysis. The data, expressed as the mean +_ S.E.M. from two mice of each strain and age, shows the proportion of proteoglycan extracted by the first GuHCl extract (El) expressed as a
Fro. 5. Western blot analysis of mouse cartilage proteoglycans. Proteoglycans were extracted with 4 M GuHC1 (El) from the lateral (L) and medial (M) tibial articular cartilage of 24 week STR/Ort (S) and CBA (C) mice. After quantification by ~°~ruthenium red dot blot analysis, extracted proteoglycans were separated by dissociative large pore agarose-acryl- amide gel electrophoresis. Separated proteoglycans were blotted onto nylon membranes and probed with mabs 2B6 (C4S) and 3B3 (C6S). Epitopes were visualized using an HRP-conjugated secondary antibody coupled with ECL detection.
aggregating CS proteoglycans of the versican family [21]. Both STR/Ort aggrecan bands have the same mobility as those present in CBA cartilage, as does the third band, a fast migrat ing C4S proteo- glycan. No additional proteoglycan species were detected in the STR/Ort extracts. Moreover, Rostand et al. [7] showed no quali tat ive differences in the size of proteoglycan monomer or gly- cosaminoglycan chains between the STR/1N mouse and controls. This suggests tha t the high level of glycosaminoglycan in STR/Ort tibial carti- lage is due to increased deposition in the tissue of proteoglycans of normal structure. Increased deposition could occur due to increased synthetic activity or because of reduced catabolism in the tissue. Prel iminary experiments indicated tha t carti lage from 12-week male STR/Ort mice which had been pulse labelled in vivo with 35S-sulfate incorporated more isotope into their GAG chains t h a n age-matched controls. Many investigators have reported tha t OA cartilage is metabolically more active than normal cartilage, as measured by the incorporat ion of ~SS-sulfate [22-27]. However,
1C6 2B6
SM SL SM SL
FIG. 6. Identification of aggrecan core protein by West- ern blot analysis. Western blot analysis was performed as described in Fig. 5. In agreement with the 24 week data, the lateral (L) and medial (M) tibial articular cartilage from 48 week STR/Ort (S) mice probed with mab 2B6 showed three C4S-rich proteoglycan species. When the same blot was re-probed with mab 1C6 only the two bands of slower mobility contained epitopes consistent with the aggrecan core protein.
other groups have not found this and suggest tha t there is either no difference or even a slight decrease in act ivi ty in OA cartilage [28-31]. One factor affecting the results of such invest igat ions may be the age of the animal and the stage of the OA.
The high level of cart i lage CS found in young STR/Ort mouse carti lage undergoes a marked reduction in the period between 19 and 25 weeks. This coincides with the time when histological lesions of OA first appear in the medial tibial plateau in some mice. Following this there is an increase in the amount of proteoglycan in mice of 31 weeks or older. One speculation is tha t a general increase in proteoglycan synthesis leads to in- creased deposition of the macromolecules in young cartilage. A relatively sudden decrease in proteo- glycan and the appearance of OA lesions may correspond to a period of enhanced catabolism, a response by the tissue perhaps to earlier excessive deposition. Finally, following this, the late phase increase in proteoglycan content of the tissue may be the result of those chondrocytes in undamaged parts of the carti lage increasing synthesis over catabolism as they respond to signals ini t ia ted by local matrix depletion or damage.
The major proteoglycan present in both mouse strains is aggrecan. Despite the marked reduct ion in proteoglycan content in STR/Ort cart i lage between 19-24 weeks we found no evidence in tissue extracts of bands representing proteoglycan degradation products (Figs 5 & 6). This is not surprising since large aggrecan products arising from proteolytic cleavage of the hya lu ronan binding region (G1 domain) would not necessari ly separate from the normal population of in tac t molecules on agarose-acrylamide gels, whilst small fragments may be rapidly lost by either diffusion to the synovial fluid or by catabolism by chondro- cytes. Moreover, other studies have shown tha t the majori ty of aggrecan monomers in OA cart i lage
102 Gaffen et a l : P r o t e o g l y c a n s in o s t e o a r t h r i t i c m i c e
are fully functional with respect to binding hyaluronan [7, 32].
In agreement with Rostand et al. [7], our data show that C4S is the predominant isomer at all ages. Measurement of the 4:6 CS ratio on both sides of the joint in the CBA mice showed tha t it was greater in the lateral than the medial tibial cartilage, especially in younger animals. This could reflect a difference in the distribution or activity of sulfotransferase enzymes in the tissue, but the functional significance of the isomer ratio, if any, is not current ly understood. However, no such age-related differences in the ratio occurred between the lateral and the medial sides in the STR/Ort mice. They all had a higher proportion of C6S in their carti lage on both sides of the joint. This clear difference between the isomer ratio in the normal CBA mouse and the STR/Ort mouse at an early age and before OA lesions appear in the latter raises the question as to whether it relates in any way to the development of the disease.
The relevance of these observations to human disease is not known, part icularly since the pre- dominant isomer in normal human adult ar t icular cartilage is C6S [33]. It has been reported tha t in human OA cartilage there is a concomitant de- crease in this isomer relative to C4S [34], and Shinmei et al. [35] found a relative increase in the C4S content of synovial fluids from the knee joints of patients with OA. Such changes may be due to the OA chondrocytes synthesizing a more immature, C4S-rich form of aggrecan, such as is found in fetal cartilage, and possibly to enhanced catabolism of decorin, which carries 4-sulfated chains [36, 37]. Similar changes should not necess- arily b e expected in murine OA cartilage. The murine aggrecan core protein sequence differs from the human sequence [38]. C4S is the predom- inant isomer at all ages and, unlike the human proteoglycan, it carries no kera tan sulfate chains [39]. Thus, direct comparisons between the two species are unlikely to be useful in understanding the significance of the different CS isomer ratios in CBA and STR/Ort cartilage, which remains elusive.
Due to the very small size of the mouse joints, it was not possible to carry out histological and biochemical investigations on the same joint. How- ever, a separate histological study revealed that overt changes were seen in the STR/Ort mice from about 20 weeks onwards [19]. No comparable OA changes were seen in CBA mice [10]. McDevitt and Muir [3] suggested tha t the 'metabolism of articu- lar cartilage as a whole appears to be affected in OA--not merely focal areas where the cartilage surface is damaged and lesions develop'. Our
results may support this view since we found biochemical differences between CBA and STR/Ort carti lage at an age prior to the appearance of OA lesions in the latter. Moreover, in contrast to the biochemical differences which occurre~d in both sides of the joint, the histological OA changes were confined almost exclusively to the medial tibial cartilage. It is not known whether the biochemical changes are primary or secondary to the disease process, but ~ restrict ion of the histological lesion to the medial side suggests that a specific factor(s) acts on that carti lage to induce lesions. One possibility is tha t the mechanical forces to which the medial and lateral cartilages" are exposed may differ.
In humans, OA is more prevalent in females. In contrast, in many mouse strains, including STR/Ort mice, spontaneous disease is more com- mon in males [40]. However, in the closely related STR/1N mouse, OA does not depend on male sex hormones, and castrat ion has little or no effect on the incidence of the disease [41]. STR/1N and STR/Ort mice have several abnormalities in addition to OA. Both have an increased incidence of obstructive geni tourinary disease [41-43] and in the STR/1N strain this was prevented by castrat ion [42]. Growth plate abnormalities were reported in STR/1N, including irregulari ty of the chondrocyte columns, abnormal vascularization and focal necrosis [44, 45]. These changes were not seen in STR/Ort mice [5]. A high incidence of spontaneous hepatomas has also been reported in the STR/1N strain [46]. STR/Ort mice have a high incidence of patellar displacement and, in some colonies, calcification of ligaments occurs [5, 43 47]. How these various abnormalities relate to OA is not known. They may be unrelated to one another or, al ternat ively some, or all, may be secondary mani- festations of an underlying metabolic defect. Nev- ertheless, animal models may hold the key to our eventual understanding of the processes tha t lead to cartilage degeneration. Fur ther studies are required before the value of the STR/Ort model can be assessed.
R e f e r e n c e s
1. Pond MJ, Nuki G. Experimentally induced osteoar- thritis in the dog. Ann Rheum Dis 1973;32:387-8.
2. McDevitt CA, Gilbertson EMM, Muir H. An exper- imental model of osteoarthritis; early morphologi- cal and biochemical changes. J Bone Joint Surg [Br ] 1977;5:24-35.
3. McDevitt CA, Muir H. Biochemical changes in the cartilage of the knee in experimental and natural osteoarthritis in the dog. J Bone Joint Surg [Br] 1976;58:94-101.
4. Sandy JD, Adams ME, Billingham MEJ, Plaas A, Muir H. In vivo and in vitro stimulation of chon- drocyte biosynthetic activity in early experimental osteoarthritis. Arthritis Rheum 1984;27:388-97.
5. Walton M. Degenerative joint disease in the mouse knee: histological observations. J Pathol 1977;123:109-22.
7. Rostand KS, Baker JR, Caterson B, Christner JE. Articular cartilage proteoglycans from normal and osteoarthri t ic mice. Arthritis Rheum 1986;29:95-105.
8. Altman FP. A metabolic dysfunction in early murine osteoarthritis. Ann Rheum Dis 1981;40:303-6.
9. Dunham J, Chambers MG, Jasani MK, Bitensky L, Chayen J. Altered organization of matrix proteo- glycans in natural murine osteoarthritis. Arthritis Rheum 1988;31:73.
10. Dunham J, Chambers MG, Jasani M, Bitensky L, Chayen J. Changes in oxidative activities of chron- drocytes during the early development of natural murine osteoarthritis. Br J Exp Path 1988;69:845-53.
11. Chambers MG, Jasani MK, Bitensky L, Cheyen J. Chondrocytic monoamine oxidase activity in the development of natural murine osteoarthritis. Int J Exp Path 1992;73:115-23.
12. Morales TI, Hascall VC. Correlated metabolism of proteoglycans and hyaluronic acid in bovine carti- lage organ cultures. J Biol Chem 1988;263:3632-8.
13. Gaffen JD, Price FM, Bayliss MT, Mason RM. A ruthenium-103 red dot blot assay specific for nanogram quantities of sulfated glycosaminogly- cans. Anal Biochem 1994;218:124-30.
14. Carney SL, Osborne DJ. The separation of chon- droitin sulfate disaccharides and hyaluronan oligosaccharides by capillary zone electrophoresis. Anal Biochem 1991;195:132-40.
15. Carney SL, Bayliss MT, Collier JM, Muir H. Electro- phoresis of 85S-labeled proteoglycans on polyacryl- amide-agarose composite gels and their visualization by. fluorography. Anal Biochem 1986;156:38-44.
16. Caterson B, Christner JE, Baker JR, Couchman JR. Production and characterization of monoclonal antibodies directed against connective tissue pro- teoglycans. Fed Proc 1985;44:386-93.
17. Stevens JW, Oike Y, Handley CJ, Hascall V, Hamp- ton A, Caterson B. Characteristics of the core protein of the aggregating proteoglycan from the Swarm rat chondrosarcoma. J. Cell Biochem 1984;26:247-59.
18. Caterson B, Calabro T, Hardingham TE. Markers of osteoarthritis: immunological aspects. In: Brandt KD, Ed. Cartilage changes in osteoarthritis. India- nopolis, IN: Indiana University Press 1990:89-97.
19. Chambers M, Gaffen JD, Gleave S, Bayliss M, Mason RM. Two colonies o f STR/Ort mice develop osteoarthrit is of differing severity: histological and biochemical evidence. Trans Orthop Res Soc 1994;19:343.
20. Lohmander LSS, Kimura JH. Biosynthesis of carti- lage proteoglycan. In: Kuet tner KE, Schleyerbach R, Hascall VC, Eds. Articular cartilage biochem- istry. New York: Raven Press 1986:93-111.
21. Thomas GJ, Davies M Harper K, B/tyliss MT, Mason
RM. Glomerular mesangial cells in vitro synthesize an aggregating proteoglycan immunologically related to versican. Biochem J 1994;302:49-56.
22. Collins DH, McElligott TF. Sulfate uptake by chondrocytes in relat ion to histological changes in osteoarthrit ic human art icular cartilage. Ann Rheum Dis 1960;9:318-22.
23. Jacoby RK, Jason MIV. The organ culture of adult human art iular carti lage from patients with osteoarthrosis. Rheumatol Rehabil 1976;15:116-22.
24. Mankin HJ, Johnson ME, Lippiello L. Biochemical and metabolic abnormalities in art icular cart i lage from osteoarthrit ic human hips. HI. Distribution and metabolism of amino sugar-containing macromelecules. J Bone Joint Surg [Am] 1981r,57:76-9.
25. McDevitt CA, Billingham MEJ, Muir H. In vivo metabolism of proteoglycans in experimental os- teoarthri t ic and normal canine articular cart i lage and the intervertebral disc. Semin Arthritis Rheum 1981;11 (Suppl 1):17-18.
26. Teshima R, Treadwell BV, Trehan CA, Mankin HA. Comparative rates of proteoglycan synthesis and size of proteoglycans in normal and osteoarthrotic chondrocytes Arthritis Rheum 1983;26:1225-30.
27. Thompson RC, Oegema TR. Synthesis of proteogly- cans in osteoarthrit ic human art icular cartilage. J Bone Joint Surg [Am] 1979;61:407-17.
28. McKenzie LS, Horsburgh BA, Ghosh P, Taylor TKF. Sulfated glycosaminoglycan synthesis in normal and osteoarthrit ic hip cartilage. Ann Rheum Dis 1977;36:369-73.
29. Maroudas A. Glycosaminoglycan turnover in articu- lar cartilage. Philos Trnas R Soc London [Biol] 1975;271:293-313.
30. Byers PD, Maroudas A, Oztop F, Stockwell RA, Venn MF. Histological and biochemical studies on cartilage from osteoarthrotic femoral heads with special reference to surface characteristics. Con- nect Tissue Res 1977;5:41-9.
31. Brocklehurst R, Bayliss MT, Maroudas A et al. The composition of normal and osteoarthrit ic ar t icular cartilage from human knee joints. J Bone Joint Surg [Am] 1984;66:95-106.
32. Bayliss MT. Metabolism of animal and human osteoarthrit ic cartilage. In: Kuet tner KE, Schleyerbach R, Hascall VC, Eds. Articular carti- lage and osteoarthritis. New York: Raven Press 1992:487-500.
33. Mourao PAS. Distribution of chondroitin 4-sulfate and chondroitin 6-sulfate in human art icular and growth cartilage. Arthritis Rheum 1988;31:1028-33.
34. Michelacci YM, Mourao PAS, Laredo J, Dietrich CP. Chondroitin sulfates and proteoglycans from normal and arthrosic human cartilage. Connect Tissue Res 1979;7:29-36.
35. Shinmei M, Miyauchi S, Machida A, Miyazaki K. Quanti tat ion of chondroitin 4-sulfate and chon- droitin 6-sulfate in pathologic joint fluid. Arthritis Rheum 1992;35:13044.
36. Witsch-Prehm P, Miehlke R, Cresse H. Presence of small proteoglycan fragments in normal and arthrit ic human cartilage. Arthritis Rheum 1992;35:1042-52.
37. Choi HU, Johnson TL, Pal S, Tang L-H, Rosenberg L. Characterization of the dermatan sulfate proteo- glycans, DS-PGI and DS-PGII, from bovine articu-
104 Gaffen et a l : Proteog lycans in os teoarthr i t ic mice
lar carti lage and skin isolated by octyl-Sepharose chromatography. J Biol Chem 1989;264:2876-84.
38. Walcz E, De~k F, Erhardt P e t al. Complete coding sequence, deduced primary structure, chromosomal localization, and structural analysis of marine aggrecan. Genomics 1994;22:364-71.
39. Venn G, Mason RM. Biosynthesis and metabolism in vivo of intervertebral disc proteoglycans in the mouse. Biochem J 1983;215:217-25.
40. Sokoloff I. Jay GF. Natural history of degenerative joint disease in small laboratory animals. 2. Epi- physeal maturat ion and osteoarthrit is of the knee of mice of inbred strains. Arch Pathol 1956;62:129-35.
41. Sokoloff L. Failure of orchiectomy to affect degener- ative joint disease in STR/1N mice. Proc Soc Exp Biol Med 1961;106:792-3.
42, Sokolff L, Barile MF. Obstructive genitourinary disease in male STR/1N mice. Am J Pathol 1962;41:233-46.
43. Walton M. Studies of degenerative joint disease in an inbred strain of mouse. Ph.D. Thesis, University of London, 1975.
44. Silberberg M, Silberberg R. Dyschondrogenesis and osteoarthritis in mice. Arch Patho11964;77:519-524.
45. Sokoloff L, Varney DA, Scott JF. Sex hormones, bone changes and osteoarthrit is in DBA/2JN mice. Arthritis Rheum 1965;6:1027-37.
46. Sokoloff L, Lyman B, Crittenden LB, Yamamoto RS, Jay GE JR. The genetics of degenerative joint disease in mice. Arthritis Rheum 1962;5:531-45.
47. Walton M. Degenerative joint disease in the mouse knee: radiological and morphological obser- vations. J Pathol 1977;123:97-107.
