Local Changes in Proteoglycan Synthesis During ... - Europe PMC

American Journal ofPathology, Vol. 140, No. 6, June 1992 Copyright C American Association of Pathologists

Local Changes in Proteoglycan Synthesis During Culture Are Different for Normal and Osteoarthritic Cartilage Floris P. J. G. Lafeber,* Peter M. van der Kraan,t Hanneke L. A. M. van Roy,* Elly L. Vitters,t Olga Huber-Bruning,* Wim B. van den Berg,t and Johannes W. J. Bijisma* From the Department of Rheumatology,' University Hospital Utrecht, Utrecht, and the Department of Rbeumatology, University Hospital St. Radboud, Nymegen, The Netherlands

Proteoglycan synthesis of mild-to-moderate osteoarthritic human knee cartilage was compared with that of normal cartilage of the same donor. Immediately after cartilage was obtained, the synthesis rate ofproteoglycans was higher for osteoarthritic cartilage than for normal cartilage. Proteoglycan synthesis was then locatee4 for both normal and osteoarthritic cartilage, in the middle and deep zone. However, after 4 days of culture, proteoglycan synthesis rate was higher for normal cartilage than for osteoarthritic cartilage. The reason for this transition from a lower to a higher proteoglycan synthesis rate was a strong increase in the proteoglycan synthesis in the superficial zone of normal cartilage. This was not observed for the osteoarthritic cartilage. The chondrocytes in the superficial zone of osteoarthritic cartilage, in contrast to normal cartilage, were mainly Joined in cell clusters andproliferating This may explain their inability to contribute to proteoglycan synthesis. (Am J Pathol 1992, 140:1421-1429)

Osteoarthritis is a disease common among the older population; prevalence increases with age. It is a degenerative joint disease that leads, in many cases progressive, to progressive destruction of articular cartilage, which eventually results in a characteristic pathologic picture.1 Healthy articular cartilage is essential for smooth joint movement during load. To enable this, cartilage has an ideal structure and composition; it consists of chondrocytes embedded in an extracellular matrix. This matrix consists mainly of collagen providing tensile strength,

and large aggregates of highly sulphated proteoglycans (PG), which resist compressive forces. The metabolic activity of the chondrocytes determines the turnover of the matrix components.2 Among the many structural alterations that characterize osteoarthritis, the most prominent one is a decrease in the PG content of the matrix. The low content of PGs in osteoarthritic (OA) cartilage compared with normal (N) cartilage results from an imbalance of PG synthesis and PG degradation. An increased proteolytic degradation in OA cartilage has been reported frequently.3 Synthesis of PGs in OA cartilage as compared with N cartilage has mostly been reported to be elevated, possibly as a "repair mechanism" to overcome the increased loss of PGs.4 However, some studies have found unchanged or even a decreased PG synthetic activity in OA cartilage.5'6 We have previously studied the influence of culture time on the PG turnover of human OA and N knee cartilage. The data indicated that culture time affected the PG synthesis rate of the two types of cartilage differently. During culture, a transition from a higher to a lower PG synthesis rate occurred for OA cartilage compared with N cartilage.78 Mild-to-moderate human knee cartilage, as used in this study, can be divided in two halves: a damaged upper layer, with a strong reduction of PGs, with severe tears and with most chondrocytes joined in cell clusters, and a deeper layer that is more or less intact. In this study, we were interested in the contribution of both layers to the PG synthetic activity during culture. More specifically, we were interested in the activity of the, for OA cartilage characteristics cell clusters. Hence, the PG synthetic activity and proliferation of chondrocytes were studied, in the superficial zone containing the cell clusters and in the middle and deep zone, where there are no cell clusters. Local changes in PG synthetic activity during culture were found, which were different for N and OA Supported by the National Reumafonds. Accepted for publication December 31, 1991. Address reprint requests to Dr. F. Lafeber, Department of Rheumatology, F02.223, University Hospital Utrecht, P.O. Box 85500, 3508GA Utrecht, The Netherlands.

1421

1422

Lafeber et al

AJPJune 1992, Vol. 140, No. 6

cartilage. Local differences in proliferation of chondrocytes could be an explanation for the observed differences in PG synthesis.

Materials and Methods

Cartilage N and OA cartilage was obtained from the central part (from anterior to distal side excluding the extreme ends) of human femoral knee condyles within 18 hours after death of the donor. From both knees, OA cartilage was taken from the lateral condyle whereas N cartilage was taken from the medial condyle, or the other way around. Donors (n = 12) ranged in age from 58 to 79 years with a mean age of 66 ± 2 (SEM) years. These donors (9 males and 3 females) were selected for the presence of focal osteoarthritis, identified macroscopically on the basis of a fibrillated surface. A slice of OA cartilage was obtained from the lateral or medial condyle while a slice of N cartilage was obtained from the same spot on the contralateral condyle from the same donor. N and OA cartilage were cut, aseptically, as thick as possible, carefully excluding underlying bone. For N and OA samples, the thickness of the cartilage from the articular surface to the deep zone was comparable. Slices were kept for a maximum of 1 hour in phosphate-buffered saline (PBS)

(pH 7.4).

Light Microscopy Three representative N and OA cartilage samples, from every donor, were fixed in phosphate-buffered 4% formalin (pH 7.0) containing 2% sucrose to assess the histologic grade of the osteoarthritis and to check if the macroscopically N cartilage was healthy control cartilage. Deparaffined sections were stained with safranin-O and fast-green iron hematoxylin. The sections were analyzed and graded according to the criteria described by Mankin et al.9 However, the tide mark between cartilage and bone was not present in our cartilage samples since bone was not included. Also as a result of this dissection method, cartilage samples were not covered with pannus. Therefore, the maximum score that could be obtained was 11 instead of the 14, when all criteria described by Mankin ("pannus," "clefts to calcified zone," and "tidemark crossed by blood vessels") could have been included.

Culture Technique Within 1 hour after dissection, the slices were cut into square pieces, weighed aseptically (5-20 mg, accuracy

0.1 mg) and cultured individually at random in roundbottom 96-well microtiter plates (200 ,ul culture medium/ well, 370C, 5% C02 in air). The culture medium was Dulbecco's Modified Eagle's Medium (DMEM, Gibco, Heutogenbosch, The Netherlands 074-01600; 0.81 mmol/l S042-; 24 mmol/l NaHCO3) supplemented with ascorbic acid (0.85 mmol/l), glutamine (2 mmolA), penicillin (100 lU/ml), streptomycin sulphate (100 lU/ml), and 10% heat-inactivated adult pooled human male AB+ serum. For biochemical determinations, ten N and ten OA cartilage samples of one donor, taken at random and handled individually, were averaged and used as a representative value for a donor. For autoradiographic determinations, the average of five N and five OA samples were taken as a representative value for a donor.

Radioactive Labeling Technique As a measure for the PG synthetic and proliferative activity of the chondrocytes, the sulphate incorporation and the thymidine incorporation were determined, respectively. Sulphate incorporation was determined during a 4-hour period, immediately after the cartilage was obtained and at the end of a 4-day culture period. An amount of 14.8 .104 Bq in 10 ,ul DMEM, Na235SO4 (DuPont Heutogenbosch, The Netherlands NEX-041-H, carrierfree) was added to each 200-,u culture medium. Thymidine incorporation was determined during a 4-day period. An amount of 7.4 1 O8 Bq [Methyl-3H]Thymidine in 10 ,u (Amersham, Breda, The Netherlands TRAl 20) was added to each 200 ,ul culture medium.

Sulphate and Thymidine Incorporation Rate To determine the rate of sulphate incorporation and the rate of thymidine incorporation, procedures were followed as described previously.7 Cartilage samples after labeling (see above) were rinsed two times in 0.5 ml of ice-cold PBS. Part of the samples were cut in the middle, and the articular side (near to the joint space) and bone side (near to the bone) were separated. The sliced samples were rinsed once more and all samples were subsequently frozen at - 200C. For determination of the rate of sulphate incorporation, intact and divided samples were digested in papain (2 hr, 650C). The glycosaminoglycans were precipitated in part of the papain digest by adding cetylpyridinium chloride. The pellet after centrifugation was washed once and subsequently dissolved in 3 mol/l NaCI. The 35SO42radioactivity was measured by liquid scintillation analysis. The remaining part of the digest was used to determine the DNA content as a measure for the amount of cells as we described before.7 Total sulphate incorporation rate

Local OA Cartilage in PG Synthesis

1423

AJP June 1992, Vol. 140, No. 6

was calculated from the 35SO42- incorporation rate and the specific activity of the medium and was normalized to the DNA amount of the sample.7 The sulphate incorporation rate is expressed as nmoles of sulphate incorporated per hour per mg DNA (nmol/hr mg). For thymidine incorporation, samples were washed three times (45 min) in 5% trichloroacetic acid (TCA) at 40C and subsequently rinsed three times (45 min) in diethylether and dried.10 Intact and divided samples were subsequently digested in papain. Incorporation of 3Hthymidine was determined in part of the papain digest by scintillation analysis. The remaining part of the digest was used to determine the DNA content. The thymidine incorporation rate was calculated from the rate of 3Hthymidine incorporated and the specific activity of the medium and normalized to the DNA content of the sample. Thymidine incorporation rate is expressed as pmoles of thymidine incorporated per hour per mg DNA (pmol/ hr mg).

Autoradiographic Localization of Incorporated Label

and OA cartilage was performed by the use of the MannWhitney U test. Significance was accepted for P < 0.05.

Results Histologic Grading For all OA cartilage, the mean grade was 5.1 + 0.2 (SEM; n = 12). The N cartilage had a mean grade of 0.6 ± 0.1 (SEM; n = 12). Figure 1 shows a representative light micrograph of N (Fig 1a) and OA (Figure 1b) cartilage. The N cartilage in this micrograph was scored grade 0; The grade of the OA cartilage shown on the micrograph was scored 5. The damaged articular surface of the OA cartilage was striking (Figures 1 b, 4b, 4d, 7b). In samples of the OA cartilage, cell clusters were present in the superficial zone at the articular surface and close to the tears perpendicular to the surface. These clusters were not observed in the N cartilage. Furthermore, a strong reduction of safranin-O staining at the articular surface of the OA samples was observed. However, this effect is not distinctly visible in the black and white micrographs. The middle and deep zone of the OA samples were more or less intact and histologically comparable to N cartilage.

For autoradiographic examinations, cartilage samples after the 4- hour or 4-day labeling period were rinsed three times (45 min) in 1.5 ml culture medium and washed twice in 0.5 ml ice-cold PBS. Subsequently, they were fixed in phosphate-buffered 4% formalin (pH 7.0) containing 2% sucrose. Standard processing of the tissue in an automatic tissue processing apparatus was followed by embedding the samples in paraffin wax. Histologic sections were prepared and stained as described earlier. Dry deparaffined sections were subsequently covered with a photographic emulsion (K5, Ilford); after exposure for 7 or 14 days, the autoradiographs were developed.1" Autoradiographic staining of the incorporated sulphate or thymidine was independently graded by two persons, for the superficial zone (near to the joint space), the middle, and the deep zone (near to the bone). A score from 0 to 3 was assigned for each zone. The sum of the scores (autoradiographic staining) of the three zones was subsequently put on 100%. The relative score for each zone was calculated and expressed as a percentage. Samples were scored at the same time for the severity of the osteoarthritis by the use of the aforementioned modified Mankin score.

As can be inferred from Figure 2, immediately after cartilage was obtained (day 0), the sulphate incorporation rate was higher for OA cartilage than for N cartilage. After culture, sulphate incorporation had increased for both N and OA cartilage. This increase was much larger for N cartilage than for OA cartilage. This resulted after 4 days of culture in a sulphate incorporation, which was higher for N cartilage than for OA cartilage. The sum of the sulphate incorporation expressed per mg of DNA for the bone and the articular side (A + B) was not significantly different from the sulphate incorporation expressed per mg of DNA for an intact sample. The figure further shows that after culture, a significant increase in sulphate incorporation in N cartilage occurred at the articular side compared with the bone side. OA cartilage showed no change in localization of sulphate incorporation after culture (Figure 2).

Calculations and Statistical Analysis

Autoradiographic Localization of Sulphate Incorporation

Mean values are

obtained

from

several

donors (n)

SEM

given. Statistical evaluation of differences between

N

Sulphate Incorporation Rate

Figure 3 shows the autoradiographic staining in the superficial, middle, and deep zone of N and OA cartilage

1424

Lafeber et al


A

B -k

-

se

t-

W,,

,

I

.

s

v

s4

A

I ..S

M.Al I M.1

M

I

', .a> .0:.

I.it

w

a,S

;.W

ii

D

Figure 1. Lightmicrographsofsafranin 0stained crosssectionofN (a) and OA (b) cartilage. Articularsurfaceon top. Ncartilagewasgraded 0, OA cartilage was graded 5 (original magnification, X50).

samples, immediately after the cartilage was obtained (at day 0) and after 4 days of culture. The autoradiographs (Figure 4a-d) illustrate the findings in Figure 3. Immediately after cartilage was obtained, sulphate incorporation was mainly located in the middle and deep zone, not significantly different for N and OA cartilage (Figure 3, 4a, SO-2 incorporation

4b). Note the apparent discrepancy with the biochemical data. When cartilage samples were sliced for biochemical determinations in two individual halves both sides include part of the middle zone (autoradiographic determinations). After culture for N cartilage, a significant relative increase in the sulphate incorporation occurred in the Figure 2. Sulphate incorporation rate of N and OA cartilage immediately after the cartilage was obtained (day 0) and after 4 days of culture. Mean values + SEM are given, n = 4. 12 A +B sum of the incorporation rate in the articular side (A i) and the bone side (B 0). The white bar behind the hatcbed bar marked A+B represents the incorporation of an intact sample. *Differences with N cartilage are statisticaly signiftcant. * Difference with the bone side ts statLsticaly significant.

(nmol / h-mg)

N

OA day 4


1425


50

Z*SO-2 incorporation

relative

(%)

FA,hFg1

for the bone and the articular side added together (A + B) was not significantly different from the thymidine incorporation expressed per mg of DNA of the whole intact sample. The high thymidine incorporation for OA cartilage is for almost the entire part located at the articular side of the cartilage.

Autoradiographic Localization of Thymidine Incorporation

K 25

00

e .0

a

mm.

M

ulI...

D

-

I

I

//.

I Jr.E

D

N OA dayO

.m

DM -

m

-

D

Jr--m- m

N OA day4

Figure 3. Autoradiographically determined relative 35S-sulphate incorporation in the superficial (S) middle (M) and deep (D) zone. Mean values ± SEM are given, n = 6 The sum of the total autoradiographic staining in the whole sample is put on 100%. * Significant increase compared with the relative sulphate incorporation in the superficial layer at day 0.

superficial zone. This resulted in a sulphate incorporation that was equally spread over the entire sample (Figure 3). This change in localization of sulphate incorporation in N cartilage samples after culture is also clearly illustrated by comparing Figure 4c and 4a. For OA cartilage, however, no significant change in the localization of sulphate incorporation after culture was observed. Sulphate incorporation was still located in the middle and deep zone of the cartilage (Figure 3). Almost no sulphate was incorporated at the superficial zone. This is once more clearly illustrated in the autoradiograph (Figure 4d). The absence of a significant sulphate incorporation was remarkable in almost all of the cell clusters. Nevertheless, cells in these clusters were still viable since, immediately after the cartilage was obtained but also after culture, sulphate incorporation was observed but only to a much lower extent as observed for the single cells in the middle and deep zone of the OA cartilage (Figure 4d, inset).

Thymidine Incorporation Rate Figure 5 shows that in N cartilage thymidine incorporation rate is nearly a fourth of the thymidine incorporation of OA cartilage. The incorporation expressed per mg of DNA

Figure 6 shows the autoradiographic staining in the superficial, middle, and deep zone of N and OA cartilage samples, after a 4-day labeling period with thymidine. For N cartilage thymidine was almost equally incorporated in the middle and superficial layer of the sample. The deep layer did only contribute for a small part to the total thymidine incorporated. This is in contrast to OA cartilage where the total incorporation for the sample originated almost exclusively from the superficial layer. The autoradiograph (Figure 7a) shows a representative of N cartilage. Only a single cell is autoradiographically stained. There were many slices of N cartilage in which no stained cells could be seen at all. In OA cartilage, several cells were autoradiographically stained and the incorporation of thymidine was mainly located at the articular surface (Figure 7b). Most of the stained cells were observed in cell clusters, as is illustrated by Figure 7c-e.

Discussion We compared the PG synthetic activity of mild-tomoderate OA cartilage to healthy N cartilage of the human knee. To avoid local differences in PG synthesis, as has been reported frequently, OA cartilage of one knee was compared with N cartilage from the same spot of the contralateral knee of the same donor. N cartilage of a donor with focal osteoarthritis can be considered as normal cartilage since PG synthesis rate of N cartilage obtained from donors with focal osteoarthritis did not differ from that of N cartilage obtained from donors without any sign of osteoarthritis (data not shown). The OA cartilage samples as used in this study can histologically be divided into two segments. The articular side (superficial layer) close to the joint space is damaged and has a severe reduction of PG content, and most chondrocytes are joined in cell clusters. The side near to the bone (middle and deep zone) seems histologically intact and similar to N cartilage samples. In this segment, cell clusters were not observed. Immediately after the cartilage is obtained, samples of OA cartilage have a higher PG synthesis than N cartilage. These results confirm our earlier findings78 and are in

1426

Lafeber et al


-..,....

.....I

1.

1

iS

'rI

i

it'

4*

*

* ,, , , ;,

E

s

ww....

U

tfWrll2m!v¢eA

ei .9

¢1fX ;1ZX=Z_


1427


Figure 4. Autoradiographs after 35S-sulphate incorporation; (a) N, day 0; (b) OA, day 0; (C) N, day 4; (d) AO, day 4 (magnification x50). Thue superficial (S) middle (M) and deep (D) zone are indicated. The inset in (d) shows a magnification (X250) of the cell cluster.

accordance with those of many others. At this time, the PG synthesis, as determined by autoradiography, was for both N and OA cartilage located for more than 80% in the middle and deep zone. From this, it may be concluded that the increased PG synthetic activity of human OA knee cartilage, the supposed "reparative activity," originates from an enhanced PG synthesis of the middle and deep zone. For artificially induced osteoarthritis in the dog knee, a similar conclusion was made.12 An increase in PG synthesis during culture was much stronger for N cartilage than for OA cartilage and resulted in a transition for OA from a higher to a lower PG synthetic activity, compared with N cartilage. In OA cartilage, the localization of the PG synthesis after culture was similar as before culture. Contribution of the superficial layer of OA cartilage remained insignificant. The strong increase in PG synthesis of N cartilage after culture originated for the larger part from a strong increase in the PG synthesis of the superficial layer - the layer that initially did not contribute to the PG synthesis. The shift in distribution of the PG synthesis during culture for N cartilage has been reported earlier for N bovine sesamoid bones13 and for N human femoral head carti-

100

relat ive 3H-Thymidine incorporation (%)

lage.14 In the latter study, the effect was explained by a different nutrition of the different zones during culture. If this is true, it is still intriguing why OA cartilage does not show such an alteration in local PG synthesis during culture. Differences in matrix composition could result in a different nutrition for OA and N cartilage. However, there is another major difference between OA and N cartilage. The chondrocytes in OA cartilage have an increased proliferation. Enhanced proliferation of chondrocytes isolated from OA cartilage has been reported before.1516 We show that this enhanced proliferation is located in the superficial layer of the OA cartilage. This is in contrast to the chondrocytes at the same location in N cartilage. These proliferating cells are located, for the larger part, in the cell clones which are present in the matrix of depleted PGs.17 Because of this proliferation, these chondrocytes are not able, or to a much lower extent, to synthesize PGs. A reversed relation between cell proliferation and PG synthesis for isolated chondrocytes has been reported before.18 As a result of the increased PG synthesis in the superficial layer of exclusively N cartilage during culture, there is a transition from a higher to a lower PG synthesis. 20

Thymidine incorporation

10

50

I/

L

/

// _

-

M

-Q

MI

--- a

N

a

M

.

a

OA

Figure 5. Thymidine incorporation rate of N and OA cartilage immediately after a 4-day labeling period. Mean values + SEM are given, n = 4. M A +B sum of the incorporation rate in the articular side (A E) and the bone side (B [). The white bar behind the hatched bar markedA +B represents the incorporation ofan intact sample. *Differences with N cartilage are statistically significant. * Difference with the bone side are statistically significant.

0

A.B

A

N

B A.AI

OA

Figure 6. Autoradiographically determined relative 3H-thymidine incorporation in the superficial (S) middle (M) and deep (D) zone. Mean values + SEM are given, n = 4. 7he sum of the total autoradiographic staining in the whole sample is put on 100%. * Statistically significant different from the middle and deep zone.

Lafeber et al

1428


. .-'.. .t'$. .....

*S%e

;'WeSe i,

*'.SW

X@Xsa/s-4s.0

'...a

0~1w * :.. *:

*

,jj,

',b

i(a

1' e.,,

*

'

64UIh.>;, ,,

j-

*" . '.,^ f

4

..

,

_;

_,..

'p

V

to }

S.

t*

.

-'j

*w

'-b

*

4.0

.v

e

e**

Figure 7. Autoradiographs after 3H-tbymidine incotporation; (a) N cartilage; (b) OA cartilage; (c, d, e) magnifications of cell clusters in the supeficial layer of OA cartilage (original magnification, a, b X50; c, d X250).

For isolated chondrocytes, PG synthesis has been reported to be higher for OA than N cartilage after culture.15 '9 However, the absence of the intact characteristic matrix and the localization of the cells makes comparison with our results difficult. Furthermore, a distinctive

isolation with a preference for individual located cells compared with the cells in clusters can not be excluded. We show these cell clusters to be of major importance for the definitive PG synthesis of the entire OA cartilage. We hypothesize that chondrocytes joined in cell clus-


1429


ters in the superficial layer are phenotypically different from chondrocytes in the deeper layer. These cells have an increased proliferation and a decreased PG synthesis compared with chondrocytes of the deeper zone. A changed collagen synthesis compared with N cartilage has also been reported.' Histologic differentiation between the matrix of the superficial and deep layer of OA cartilage may contribute to this hypothesis. It will be interesting to study factors that specifically modulate these proliferating chondrocytes in the superficial layer of OA cartilage. By decreasing the proliferation of these chondrocytes, a stimulation of PG synthesis and possibly also a decrease of PG breakdown could be achieved. In this respect, growth factors, like TGFI3,21 and also intermittent hydrostatic compression8 should be considered. Such factors would contribute significantly to the repair of OA cartilage.

9.

10.

11.

12.

13.

Acknowledgment The authors thank the Department of Pathology of the University Hospitals of Utrecht and Amsterdam for providing the cartilage.

14.

References

15.

1. Bullough PG: The pathology of osteoarthritis. J Educ Info Rheum EULAR 1988, 17:5-8 2. Van Mow C: Molecular structure and function relationships

3.

4.

5. 6.

7.

8.

for articular cartilage. J Educ Info Rheum EULAR 1988, 17:9-13 Dean DD, Martel-Peletier J, Pelletier JP, Howell DS Woesner JF Jr: Evidence for metalloproteinase and metalloproteinase inhibitor imbalance in human osteoarthritic cartilage. J Clin Invest 1989, 84:678-685 Mankin HJ, Johnson ME, Lippiello L: Biochemical and metabolic abnormalities in articular cartilage from osteoarthritic human hips. J Bone Joint Surg 1981, 63:131-140 Lust G, Pronsky WP, Sherman DM: Biochemical and ultrastructural observations in normal and degenerative canine articular cartilage. Am J Vet Res 1972, 33:2429-2440 Dingle JT, Homer A, Shield M: The sensitivity of synthesis of human cartilage matrix to inhibition by IL-1 suggests a mechanism for development of osteoarthritis. Cell Biochem Funct 1991, 9:99-102 Lafeber F, VanRoy H, Wilbrink B, Huber-Bruning 0, Bijlsma JWJ: Human osteoarthritic cartilage is synthetically more active but in culture less vital than normal cartilage. J Education Info Rheum 1992, 19:123-129 Lafeber FPJG, Veldhuijzen JP, Van Roy JLAM, HuberBruning 0, Bijlsma JWJ: Intermittent hydrostatic compres-

16. 17.

18.

19.

20.

21.

sive force stimulates exclusively the proteoglycan synthesis of osteoarthritic cartilage. Br J Rheum 1992, 31 :in press Mankin HJ, Dorfman H, Lippiello L, Zarins A: Biochemical and metabolic abnormalities in articular cartilage from osteoarthritic human hips: II Correlation of morphology with biochemical and metabolic data. J Bone Joint Surg 1971, 53:523-537 Maor G, Silbermann M: In vitro effects of glucocorticoid hormones on the synthesis of DNA in cartilage of neonatal mice. FEBS Lett 1981, 129:256-260 Vanden Berg WB, Vande Putte LBA: Electrical charge of the antigen determines its localisation in the mouse knee joint. Deep penetration of cationic BSA in hyaline articular cartilage. Am J Pathol 1985,121:224-234 Sandy JD, Adams ME, Billingham MEJ, Plaas A, Muir H: In vivo and in vitro stimulation of chondrocyte biosynthetic activity in early experimental osteoarthritis. Arthritis Rheum 1984, 27:388-397 Corver GHV, Vande Stadt RJ, Van Kampen GP, Kiljan E, Vander Korst JK: Bovine sesamoid bones: a culture system for anatomically intact articular cartilage. In Vitro Cell Dev Biol 1989, 25:1099-1106 Maroudas A, Schneiderman R, Weinberg C: Comparison between effects of serum and insulin on GAG synthesis in different zones of cultured human articular cartilage. Trans Orthop Res Soc 1990,15:315 Bulstra SK, Buurman WA, Walenkamp GHIM, Vander Linden AJ: Metabolic characteristics of in vitro cultured human chondrocytes in relation to the histopathologic grade of osteoarthrosis. Clin Orthop 1989, 242:294-302 Malemud CJ, Papay RS: Rabbit chondrocytes maintained in serum-free medium. Exp Cell Res 1986, 167:440-446 Sauren YMHF, Huber-Bruning 0, Lafeber FPJG, Mieremet RHP, Groot CG and Scherft JP: Ultrastructural alterations of proteoglycans in human osteoarthritic cartilage, a study with the anionic dye polyetheleneimine. Calcif Tissue Int 1991, 48:A54. VanKampen GDJ, Veldhuijzen JP, Kuyer R, Vande Stadt RJ, Schipper CA: Cartilage response to mechanical force in high density chondrocyte cultures. Arthritis Rheum 1985, 28:419-424 Teshima R, Treadwell BV, Trahan CA, Mankin HJ: Comparative rates of proteoglycan synthesis and size of proteoglycans in normal and osteoarthritic chondrocytes. Arthritis Rheum 1983, 26:1225-1230 Ronzi6re M-C, Richard-Blum S, Tiollier J, Hartmann DJ, Garrone R, Herbage D: Comparative analysis of collagens solubilized from human foetal, and normal and osteoarthritic adult articular cartilage, with emphasis on type VI collagen. BBA 1990,1038:222-230 Keski-Oja J, Leof EB, Yons RM, Coffey RJ Jr, Moses HL: Transforming growth factors and control of neoplastic cell growth. J Cell Biochem 1987, 33:95-107