Rheumatology-1996-VERSCHURE-1044-55

8/12/2019 Rheumatology-1996-VERSCHURE-1044-55

1/12

British Journal of Rheumatology 1996;35:10441055

LOCALIZATION OF INSULIN-LIKE GROWTH FACTOR-1 RECEPTOR INHUMAN NORMAL AND OSTEOARTHRITIC CARTILAGE IN RELATIONTO PROTEOGLYCAN SYNTHESIS AND CONTENTP . J. VERSCHURE, J. VAN MARLE,* L. A. B. JOOSTEN, M. M. A. HELSEN,F. P. J. G. LAFEBERt and W. B. VAN DEN BERG

Department of Rheumatology, University Hospital Nijmegen, G eert Grooteplein Zuid 8, 6525 GA Nijmegen,*Academic M edical C enter, University of Amsterdam, Department of Electron Microscopy, PO Box 22700, 1100DE Amsterdam and ^Department of Rheumatology, University Hospital Utrecht, PO Box 85500, 3508 GA Utrecht,The NetherlandsS U M M A R Y

Insulin-l ike growth factor-1 (IGF-1) plays a key role in the regulat ion of chondrocyte proteoglycan (PG) metabolism. Weinvest igated whether chondrocyte PG synthetic act ivi ty correlates with the presence of chondrocyte IGF-1 receptor in the surface,middle and deeper zones of normal human art icular cart i lage and in cart i lage known to display a shift in chondrocytemetabolism, i .e. cultured cart i lage or osteoarthri t ic (OA) cart i lage. Cart i lage specimens were obtained post mortem from humanknees w ithin 18 h after de ath from d ono rs without kno wn cl inical OA history. The samples were taken from macroscop icallynormal looking regions as well as from damaged regions with osteoarthri t ic appearance, yielding a range of OA grades frommild to moderate and severe OA. We examined chondrocyte PG synthesis by in situ autoradiography of incorpora tedP^Jsulphate and chondrocyte IGF-1 receptor localizat ion by immunohistochemistry, fol lowed by confocal laser scanningmicroscopical (CLSM) analysis in the same cart i lage samples. In normal cart i lage, both the amount of chondrocyte PG synthesisand the level of chondrocyte IGF-1 receptor localizat ion are at low levels in the surface zone chondrocytes, but both are highin middle and deeper zone chondrocytes. Furthermore, after culture, the increase in chondrocyte PG synthesis in the surfacelayer coincides with increase in IGF -1 receptor expression. However, in mild OA part icularly high levels of cho ndro cyte syntheticactivi ty were found in the upp er cart i lage layer, whereas IGF -1 rece ptor expression was low in this layer, suggest ing that factorsother than IGF-1 are involved. High chondrocyte PG synthetic act ivi ty and chondrocyte IGF-1 receptor staining were foundin the upper and deeper layers of moderate OA cart i lage, whereas both low levels of chondrocyte act ivi ty as well as IGF-1receptors were observed in cases of severe OA. Our data indicate that IGF-1 displays cel lular heterogeneity in chondrocytest imulat ion in the various cart i lage zones in normal cart i lage. Gear zonal correlat ion is lost in OA cart i lage, and patterns ofchondrocyte IGF-1 receptor expression and PG synthesis vary with the stage of OA.KEY WORDS: Chondrocyte , Car t i l age , OA, IGF-1 receptor , PG synthes i s , In situ, CLSM, Zonal analys i s .

IN degenerative joint diseases such as osteoarthritis(OA), the equilibrium between the degradationand synthesis of matrix molecules is disturbed [1-3].Increased proteoglycan (PG) metabolism ofchondrocytes in OA cartilage, which is probably anattempt to repair damage, is unable to overcome theincreased losses of PG and leads to focal depletion ofPG, and eventually loss of cartilage and changes inunderlying bone [4]. The biosynthetic function ofchondrocytes is regulated by the complex effects ofcirculating and locally produced growth factors.Understanding the influence of growth factors onchondrocyte metabolism during normal physiologicalregulation of cartilage remodelling, or duringpathological cartilage degradation, will be a basis forsuccessful intervention in cartilage degeneration.Insulin-like growth factor-1 (IGF-1) is an importantanabolic stimulus that regulates chondrocyte PGbiosynthesis during growth and development, cartilage

Subm itted 21 Novemb er 1995; revised version accepted 1 May1996.Correspondence to: P. J. Verschure, Department of Rheumato-logy, University Hospital, Geert Grooteplein Zuid 8, 6525 GANijmegen, The Netherlands.

hypertrophy and repair during adult life [5-10]. IGF-1stimulation of intact articular cartilage results in asignificant increase in [MS]sulphate uptake, i.e.chondrocyte PG synthesis [11-14]. IGF-1 exerts itseffect by high-affinity binding with specific IGF-1 type1 receptors [15-18].Chondrocytes in articular cartilage are not a uniformpopulation of cells. They vary in metabolic activity andfunction in the d ifferent cartilag e zo nes [1923].Moreover, there are indications that chondrocyte PGmetabolism alters in the different cartilage zones incartilage diseases [24-27]. In human OA, high levels ofIGF-1 and IGF-1 m RNA have been found in cartilage,especially in chondrocytes of the surface zone [28].Furthermore, Dore et al. [29] showed abundantimmunohistochemical staining oftheIGF -1 receptor inthe upper two-thirds of human OA cartilage.In the present study, we examined articular cartilageobtained from human knees within 18 h after death,from donors without a clinical history of OA. Thesamples were classified, yielding normal cartilage anda range ofOAgrades from mild via moderate to severeOA. The aim of our study was to investigate whethera correlation exists between chondrocyte syntheticactivity, chondrocyte IGF-1 receptor expression and

1996 British Society for R heum atolog y1044


2/12


3/12

1046 BRITISH JOURNAL OF RHEUMATOLOGY VOL. 35 NO. 11cryostat fitted with a tungsten carbide-tipped knife ata cabinet temperature of -25C [32,33].Serial sectionsfor autoradiography and immunohistochemistry werecollected on poly-L-lysine-coated slides. Sections werekept over silica gel and stored at 25C until furtheruse.Autoradiographic localization ofincorporated labelCryostat sections were air dried for 30 min, fixed for10 min in 4 % formalin dissolved in water andsubsequently washed three times (5 min) with distilledwater. Sections were air dried overnight and sub-sequently covered with a photographic emulsion (K5,Ilford); the emulsion was diluted 1:1 with 6% gelatinand heated to 45C. After exposure for 7-14 days,depending on the autoradiographic blackening ofnormal cartilage, the autoradiographs were developed[34]. For every experiment, i.e. each donor, a normalcartilage sample was analysed together with apathological sample. The autoradiographic stainingwas developed depending on the autoradiographicblackening in the normal samples. The n ormal samplesshowed minor variation in [ S]sulphate uptake (datanot shown).

ImmunohistochemistryCry ostat sections were fixed for 10 min in 4% (w/v)paraformaldehyde (Sigma), dissolved in PBS (pH 7.4)and subsequently washed in PBS containing 10% (w/v)bovine serum albumin (BSA; Sigma, St Louis, MO,USA), 10% (w/v) gelatin and 0.3% (w/v) glycine(Pharmacia, Uppsala, Sweden). Sections were incu-bated for 30 min at 37C in a solution of 2 mg/mltesticular hyaluronidase (Sigma) dissolved in PBS toobtain a comparable penetration of antibody into thematrix of normal and OA cartilage. Subsequently,sections were rinsed in PBS and finally treated with10% (w/v) fetal calf serum and 1% (w/v) BSA in PBSfor 30 min to block aspecific bin ding. Thereafter,section s were incuba ted for 24 h at 4C with mou semonoclonal antibody directed against the extracellulara subun it of hu man IGF -1 type 1 receptor [35, 36] inconcentrations of5/ig/ml in PBS containing 1 BSA(Oncogene Science, Manhasset, NY, USA). Afterincubation with the primary antibody, the slides werewashed three times in PBS. Sections were incubatedwith biotinylated horse anti-mouse secondary antibody(Vector L aboratories, Burlingame, USA) for 2 h in adilution of 1:500 in PBS containing 1% BSA and 1%normal human serum, and biotin was demonstrated byexposing sections to streptavidin-fluorescein isothio-cyanate (FITC) (Boehringer) in a concentration of40 jig/ml in PBS. Sub sequently, sections were m ountedin Vectashield m ountin g medium (Brunschwig Chemie)to avoid fading of fluorescence. As a control foraIG F- 1 receptor antibod y staining, sections were eitherincubated with mouse IgG or stained in the absence ofprimary antibody. Sections were kept at - 25Cuntilevaluation with the use of the CLSM.

Fluorescence CLS M and image analysisThe CLSM collects images that are almost free ofout-of-focus signals and the system allows opticalsectioning of the specimen [37-39]. Immunostainedsections were examined with a Leica CLSM attachedto a Leica Fluovert microscope using the 488 nm lineof an Argon K rypto n laser fitted with a 510 nmdichroic mirror and a 530 nm band pass filter, asdescribed in an earlier study [40]. Inbrief, immunola-belled cryostat sections were subjected to optical serialsectioning with intervals of 0.5IIXD. in the Z direction.

Quantification of immunofluorescence in theconfocal images was performed by area measurementas described previously [40,41].Briefly, a window wasset over the chondrocyte periphery and over theinternal part of the chond rocytes. The distance betweenthe outer rims of the window was kept at ~0.2 /jm. Innormal cartilage, the superficial, middle and deeperzones were easily identified ([25]; Fig. 1). For everynormal cartilage sample, fluorescence intensity wasmeasured in chondrocytes located in the surface,middle and deeper zones in each of two serial sections.In each cartilage zone, the fluorescence intensity wasmeasured in the chondrocytes that were randomlyselected with the use of a microscope occular with ameasuring grid. To assess the contribution ofautofluorescence and aspecific binding, fluoresc encesignals were measured in sections stained with mouseIgGl. For every cartilage sample in each zone, thefluorescence intensity measurement of10chondrocytesin two cryostat sections was averaged. For eachcartilage sample, the highest average value in aparticular zone w as set at 100%. The fluoresc encesignals in the other zones were expressed as apercentage of this value. Data represent themean S.E.M. of 10 different donors. Statisticalevaluation of the experiments was performed with theWilcoxon rank test. APvalue of


4/12

VERSCHURE ET AL:. IGF-1 RECEPTOR IN NORMAL AND OA CARTILAGE 1047

FIG. 1.Human normal cartilage stained with Safranin O and Fast Green (A), zonal distribution of in situ autoradiograpbic staining ofincorporated [ SJsulphate (B) and digital images ofIGF-1receptor immunostaining (C). Forin situautoradiography, the cartilage was labelledwith P'S]sulphate. For IGF-1 receptor localization, cryostat sections were stained immunohistochemically and analysed with CLSM. Therepresentative example of normal human articular cartilage obtained from macroscopically 'norm al' looking cartilage regions exhibited a smoothintact articular surface and the surface zone (S), middle zone (M) a nd deeper cartilage zones (D) could be easily identified (magnification 20x ) .Chondrocytes in the surface zone showed only low levels of P'SJsulphate incorporation, whereas chondrocytes in the middle and deeper zonesof the cartilage displayed distinct amounts of incorporated label (B) (magnification 20x). IGF-1 receptor immunostaining of chondrocytes inthe surface zone (S) was similar to backgroundfluorescenceevels, whereas chondrocytes in the middle (M) and deeper zone (D) showed clearIGF-1 receptor im munoreactivity (Q . The uppermost pa rt of the cartilage is indicated by an arrowhead. Bars* 8.5 /im.

poration. Figure IB shows the autoradiographicstaining of incorporated f'SJsulphate in normalcartilage. Chondrocytes in the surface zone show onlylow levels of sulphate incorporation, whereas chondro-cytes in the middle and deeper zones of the cartilagedisplay distinct amounts of incorporated label (TableII).IGF-1 receptor staining. The IGF-1 receptor

immunostaining pattern of normal cartilage closelycorresponded with the autoradiographic localization ofincorporated [ SJsulphate in norm al cartilage.Immunostaining with the aIR-3 monoclonal antibodyagainst the IGF-1 receptor resulted in a clear stainingof chondrocytes in the middle and deeper zones(Fig. 1C, Table II). Staining of chondrocytes in thesurface zone was similar to ba ckgroun d fluore scence


5/12

1048 BRITISH JOURNAL OF RHEUMATOLOGY VOL. 35 NO. IITABLE II

Chondrocyte PG synthesis, chondrocyte IGF-1 receptor immuno-reactivity and cartilage PG content in human normal cartilage ornormal cartilage after culture for 4 days in the presence of serum aswell as cartilage from different grades of OA. Data in normal orcultured cartilage represent the mean of 10 donors: five donors ofmild OA, three donors of moderate OA and two donors of severeOA. Staining procedures were performed in duplicate on at least 10

sections of each donorPG content PG synthesis IGF-1 receptor

NormalSurface - - -Middle + + + ++ Deep + + + + + +

Nonnal after cultureSurface -I- + + 4-Middle + + + + ++Deep + + + + + +

OA(4)Upper - + +Deeper + + + + + 4-

OA (8)Upper - ++ + + +Deeper ++ + + + + + +

OA (>8)Deeper + +

FIG . 2.Quantification of the fluorescence signal expressed as apercentage in normal human cartilage after incubation withmonoclonal anti-IGF-1 receptor antibody. The fluorescence signalwas examined in cartilage immediately after isolation of the cartilageor after culture for 4 days in the presence of 10% serum. Thefluorescence intensity of each zone was measured in 10 chondrocytesin each of two serial sections. For every normal cartilage sample,fluorescence intensity was measured in chondrocytes located in thesurface, middle and deeper zones in each of two serial sections. Ineach cartilage zone, the fluorescence intensity was measured in thechondrocytes that were randomly selected with the use of amicroscope occular with a measuring grid. To assess the contributionof autofluorescence and aspecific binding, fluorescence signals weremeasured in sections stained with mouse IgGl. For every cartilagesample in each zone, the fluorescence intensity measurement of 10chondrocytes in two cryostat sections was averaged. For eachcartilage sample, the highest average value in a particular zone wasset at 100%. The fluorescence signals in the other zones wereexpressed as a percentage of this value. Data represent themean S.E.M. of 10 different donors. Statistical evaluation of theexperiments was performed with the Wilcoxon rank test APvalueof


6/12

VERSCHURE ETAL:. IGF-1 RECEPTOR IN NORMAL AND OA CARTILAGE 1049middle and deeper zones was slightly increased (TableII).These observations coincide with our findings thatthe immunostaining of the IGF-1 receptor was alsoenhanced in chondrocytes in the surface zone (Fig. 3B,Table II). Quantification of the fluorescence intensityrevealed a significant increase in IGF-1 receptorimmunostaining in chondrocytes in the surface zone ascompared with the fluorescence intensity in chondro-cytes in the surface zone immediately after isolation ofthe cartilage (Fig. 2).OA cartilageSample characterization. Cartilage specimens wereobtained post mortem from donors without knownclinical OA history. The samples were taken fromdamaged regions with an 'osteoarthritic' appearanceand this yielded a wide range of OA grades uponmicroscopic analysis. A summary is shown in Table I,demonstrating that we collected samples with severeOA (n= 2), moderate OA (n= 3), but mainlyspecimens with mild pathology(n= 5). Zonal v ariationin chondrocyte appearance, as found in normalcartilage, was no longer present in OA cartilage.Analysis of OA cartilage was performed in regionsclassified as upper and deeper layers. The characteristicSafranin O staining of the denned OA categories isshown in Fig. 4. A reduction of Safranin O staining inthe upper layer ofth ecartilage was found in all samples(Table II). Since the severity of OA can vary at varioussites of one specimen, special care was taken to analysethe chondrocyte PG synthesis and chondrocyte IGF-1receptor staining in one area. Thereafter, the sampleswere divided into two neighbouring specimens thatwere used for either autoradiography or receptorstaining.Autoradiographic localizationof ( S] sulphateincorpor-ationThe autoradiographic localization of incorporatedPJS]sulphate showed distinct patterns in the three OAcategories (Table II). In the cartilage categorized asrelatively mild OA, autoradiographic localization didnot show the zonal variation as found in normalcartilage. The mild OA cartilage appeared to beuniformly stained. Chondrocytes in the upper anddeeper layer of the cartilage exhibited distinct amountsof incorpo rated [ S]sulphate (Fig. 5A). The samestaining pattern was evident in more severe OAcartilage lesions. The cell clones located in the uppercartilage layer displayed distinct amou nts of au toradio -graphic staining and the chondrocytes in the deeperlayers showed high levels of staining (Fig. 5C).However, OA cartilage categorized as extremely severedemonstrated only very low levels of autoradiographicstaining (Fig. 5E). The upper layer of the cartilagecould no longer be distinguished and chondrocytes inthe deeper layer showed only faint [ S]sulphateincorporation.

IGF-1 receptor staining. IGF-1 receptor immuno-staining of cartilage from the three OA categories

showed a diverse staining pattern between thecategories (Table II). Relatively mild OA cartilageshowed zonal variation in IGF-1 receptor immuno-reactivity. Chondrocytes in the upper layer of thecartilage displayed immunoreactivity similar tobackground staining, whereas the staining in ch ondro-cytes in the deeper layer was obviously high (Fig. 5B).In more severe OA cartilage, the zonal variation waslost; chondrocytes in both the upper and deeper layerexhibited a prominent intense IGF-1 receptor staining(Fig. 5D). In the extreme OA cartilage samples, whenthe upper layer had disappeared, the immunostainingwas indistinguishable from background fluorescencethroughout the whole cartilage sample (Fig. 5F).

DISCUSSIONChondrocyte biosynthetic function plays a key rolein maintaining the integrity of the articular cartilagematrix [38]. IGF-1 is the most important anabolicfactor, regulating the chondrocyte PG synthetic rate bybinding to the type 1 IGF-1 receptor on thechondrocyte membrane [5-18]. Moreover, articularcartilage chondrocytes located in the various cartilagezones exhibit heterogeneity in their metabolic activity[19-22]. We investigated whether chondrocyte PGsynthetic activity, chondrocyte IGF-1 receptorlocalization and cartilage PG content are correlated inthe various cartilage zones of human normal cartilage,and in cartilage know n to display a shift in chon drocytemetabolic activity, i.e. cultured cartilage and OAcartilage. Because OA lesions occur focally throughoutthe cartilage, we analysed chondrocyte activity andchondrocyte IGF-1 receptor localization in the samecartilage samples.

In normal cartilage, chondrocyte PG synthesis,IGF-1 receptor localization and cartilage PG contentwere at low levels in chondrocy tes located in the surfacezone, but were particularly high in chondrocytes in themiddle and deeper zones. These data might indicatethat IG F-1 recep tor expression in the different cartilagelayers reflects heterogeneity in IGF-1 stimulation andmetabolic activity of chondrocytes in these layers.Surface zone chondrocytes have been described toexhibit structural, biochemical and biomechanicalchanges during culture, but also in conditions ofcartilage pathology [43-47]. It remains to be seenwhether IGF-1 itself is a regulatory factor in some ofthese pathological events. After culture of the normalcartilage, high levels of both chondrocyte PG synthesisand chondrocyte IGF-1 receptor localization werestill found, but the distinct zonal distribution ofboth chondrocyte synthetic activity and IGF-1receptor localization disappeared. The shift in thedistribution of chondrocyte metabolic activity has beenreported earlier in bovine as well as human cartilageexplant cultures [20, 24 ,48] . In the present study, wehave shown that this shift in chondrocyte syntheticactivity is correlated with the amount of chondrocyteIGF-1 receptor immunostaining. Therefore, IGF-1stimulation may contribute to upregulation of the


7/12

1050 BRITISH JOURNAL OF RHEUMATOLOGY VOL. 35 NO. 11

B

Fra. 4A, B and C flegend opposite).


8/12

VERSCHURE ET AL.: IGF-1 RECEPTOR IN NORMAL AND OA CARTILAGE 1051

Fio.4.Characteristics of cartilage obtained from damaged cartilage regions with an 'osteoarthritic' appearance stained with Safranin and FastGreen. Th e cartilage samples were classified according to slightly modified Mank in criteria [4,24]. Mild fraying of the articular surface withoutthe presence of cell clones is defined as mild OA (4 on M ankin scale) (A). Advanced fibrillation of the cartilage surface, characterized by deepclefts and the presence ofcellclones, is classified as moderate O A (8 on Mank in scale) (B). Extremefibrillationof the cartilage, the upper layeris almost totally lost and chondrocytes are only found in clusters, is defined as severe OA (> 8on Mankin scale) (C). All OA samples showedreduction of Safranin O staining in the upper layer of the cartilage. Zonal variation in chondrocyte appearance as found in normal cartilagewas not present in OA cartilage. Therefore, the upper and deeper layer were defined as < 150pm from the articular surface and >150pm fromthe articular surface, respectively (magnification 20 x). The uppermost part of the cartilage is identified with an arrowhead.

biosynthetic activity of resting surface zone chondro-cytes during culture.In earlier studies that examined the role ofIGF-1inthe regulation of OA cartilage metabolism, cartilagesamples were taken from patients undergoing kneeathroplasty, representing moderate to severe OAcartilage [28,29]. Our OA samples were obtained postmortem from donors without known clinical OAhistory, ranging from very mild O A lesions to mod erateand extremely severe OA path ology . The very mild OAlesions analysed in the present study are of particularinterest; they display a specific pattern that may occurin the very beginning of OA. In mild OA cartilage, thesynthetic activity of chondro cytes in the upper cartilagelayer was markedly high: the chondrocyte PG syntheticrate was equally distributed over the cartilagespecimen. These findings are in line with observationsin early experimental OA [49]. The particularly highlevels of synthetic activity in the upper cartilage layerindicate that the demand for matrix repair is highest inthis layer. Remarkably, in the mild OA cartilage,chondrocy tes in the u ppe r layer displayed low levels ofIGF-1 receptor immunostaining, whereas chondro-cytes in the middle and deeper layers displayedpronounced levels of IGF-1 receptor immunostaining.These results reveal that the role of IGF-1 in theregulation of chondrocyte PG synthesis in the uppercartilage zone in early OA may be of minorimportance. Of interest, cartilage from OA lesions hasbeen proven to be much more sensitive to stimulationwith transforming growth factor ft (TGF/?) thannormal cartilage [50]. In addition to these in vitrofindin gs, we have recently shown that repeatedintra-articular injections of TGF0 in the murine kneejoint caused marked upregulation of chondrocyte PGsynthesis, a significant rise in cartilage P G content anda clear development of osteophytes [51,52]. Thissituation closely resembles the hypertrophic phase inearly experimental OA, indicating that TGF/? plays animportant role in processes during early OA [53].

Furthermore, in moderate OA cartilage, bothchondrocyte synthetic activity and chondrocyte IGF-1receptor immunostaining were at high levels. The highchondrocyte synthetic activity is known not to reflecta high cartilage PG content in this type of cartilage.This discrepancy in the quality of the cartilage andchondrocyte synthetic activity implies that during thisphase of OA the chondrocytes synthesize aberrant PGsubtypes causing a less stable articu lar cartilage matrix.Regarding the severe OA cartilage, our data show thatboth chondrocyte metabolism and IGF-1 receptor

localization are at low levels. This points to an endstage of OA in which an attempt at increased matrixreplacement no longer occurs. At later stages of OA,focal loss of cartilage, enhanced expression ofproteolytic enzymes and an imbalance with naturalenzyme inhibitors have been described [54, 55].In addition, in a previous study, we have examinedchondrocyte synthetic activity and chondrocyte IGF-1receptor immunostaining in experimental inflammat-ory arthritis [52]. In contrast to the high chondrocytemetabolic activity in OA cartilage, chondrocyte PGsynthesis is markedly inhibited shortly after theinduction of joint inflammation. The cartilage had lostits capacity to respond to IGF-1 stimulation, whereasthe amounts of IGF-1 receptors were at normal levels[11,12].In comparison with the data found in mild OAcartilage, correlation of chondrocyte metabolic activityand chondrocyte IGF-1 receptor immunostaining wasdisturbed, suggesting inadequate or overruled sig-nalling.

The m ethods used in the present study enabledinsitucomparison of chondrocyte PG synthesis, IGF-1receptor immunostaining and the cartilage PG content.The data obtained are still semiquantitative and do notallow concrete statistical correlations. Moreover,IGF-1 receptor immunostaining may at best give anindication of IGF-1 receptor expression. However,quantitative assays will need isolation of the chondro-cytes; analysis of the whole chondrocyte populationwill lead to loss of information about chondrocyteslocated in the different cartilage zones. Therefore, theexperiments performed in the present study are ofadvantage giving information on in situ PG synthesis,PG content and IGF-1 receptor immunostaining.In summary, in normal cartilage, both the amount ofchondrocyte PG synthesis and chondrocyte IGF-1receptor localization are at low levels in surface zonechondrocytes, and high in middle and deeper zonechondrocytes. This indicates that IGF-1 displays

cellular heterogeneity in chond rocyte stimu lation in thevarious cartilage zones. After culture, the shift inchondrocyte PG synthesis is correlated with anincreased IGF-1 receptor immunofluorescence in thesurface zone. Therefore, IGF-1 stimulation contributesto upregulation of resting surface zone cells duringculture. In mild OA, the particularly high levels ofchondrocyte synthetic activity in the upper cartilagelayer did not correlate with amounts ofIGF-1receptorimmunostaining, indicating that other factors play arole in the chondrocyte regulation in this phase of OA.The pronounced high chondrocyte PG synthetic


9/12


VFIG. 5. A-D.


10/12

VERSCHUREETA L :IGF-1 RECEPTOR IN NORMAL AND OA CARTILAGE 1053

Fio. 5.A-D. Distribution of in situ autoradiographic localization of incorporated [ S]sulphate (A, C, E) (magnification 20 x ) and IGF-1receptorstaining (B, D, F) (bars 8.5 /im) in the three defined OA categories: mild OA (A, B), moderate OA (C, D) and severe OA (E, F).Th euppermost part of the cartilage is indicated by an arrowhead. Mild OA cartilage shows no zonal variation in chondrocyte PG synthesis,bu tappears to be uniformly stained (A). Chondrocytes in the upper layer of mild OA cartilage display IGF-1 immunoreactivity similar tobackgroundstaining, whereas staining ofchondrocytesin the deeper layer is obviously high (B). Moderate OA cartilage shows an apparenthighPG synthesis both in the upper and deeper cartilage layer (Q . Chondrocytes in moderate OA cartilage show an intense IGF-1 receptorstainingin the upper and deeper layer (D). Severe OA cartilage exhibits only very low levels of chondrocytePG synthesis (E). Chondrocytesin severe OA cartilage display immunostaining indistinguishable from background fluorescence throughout the whole cartilage sample (F).

activity and chondrocyte IGF-1 receptor staining inmoderate OA cartilage may represent an active attemptat repair, whereas the low levels of chondrocyte activityand IGF-1 receptor localization in severe OA may beindicative of an end stage of OA.ACKNOWLEDGEMENTSThe authors wish to thank Hanneke L. A. M. VanRoy, Henk Van Veen, Jan Peeterse and KarsGravemeijer for excellent technical assistance. Thisstudy was supported by TNO.

REFERENCES1. Gay S, Gay RE, Koopman WJ. Molecular and cellular

mechanisms of joint destruction: two cellularmechanisms explain joint destruction? Ann Rheum Dis1993^2:37-9.2. Dieppe P. Some recent dinical approaches to osteo-arthritis research. Semin Arthritis Rheum 1990;2(k2-ll.3. Hough AR. Pathology of osteoarthritis. In: McCarthyDJ, Koopman WJ, eds. Arthritis and allied conditions.Philadelphia: Lea and Febiger, 1993:1699-721.4. Mankin HJ, Brandt KD. Biochemistry and metabolismof cartilage in osteoarthritis. In: Howell DS, GoldbergVM, Mankin HJ, Moskowitz RW, eds. Osteoarthritisdiagnosis and management. New York: W. B. Saundersand Co., 1984:43-80.5. Hascall VC, Handley CJ, McQuillan DJ, Hascall GK,

Robinson HC, Lowther DA. The effect of serum onbiosynthesis of proteoglycans by bovine articularcartilage in culture. Arch Biochem Biophys 1983^24:206-23 .6. Froesch ER, Schmid C, Schwander J, ZapfJ.Actions ofinsulin-like growth factor. Annu Rev Physiol 1985;47:443-67.7. McQuillan DJ, Handley CJ, Campbell MA, Bolis S,Milway VE, Herington AC. Stimulation of proteoglycansynthesis by serum and insulin-like growth factor-1 incultured bovine articular cartilage. Biochem J 1986^40:424-30.8. Luyten FP, Hacall VC, Nissley SP, Morales TI, ReddiAH. Insulin-like growth factors maintain steady-statemetabolism of proteoglycans in bovine articular cartilageexplants. Arch Biochem Biophys 1988^67:416-25.9. Trippel SB, Chemausek SD, Van Wijk JJ, Moses AC,Mankin HJ. Demonstration of type 1 and type 2somatomedin receptors in bovine growth plate chondro-cytes. / Orthop Res 1988;6:817-26.10. Schalkwijk J, Joosten LAB, Van den Berg WB, Wijk JJ,Van de Putte LBA. Insulin-like growth factor stimulationof chondrocyte proteoglycan synthesis by humansynovial fluid. Arthritis Rheum 1989^2:66-71.11. Schalkwijk J, Joosten LAB, Van den Berg WB, Van dePutte LBA. Chondrocyte nonresponsiveness to insulin-like growth factor 1 in experimental arthritis. ArthritisRheum 1989^2:894-900.12. Verschure PJ, Van Marie J, Joosten LAB, Van den BergWB. Chondrocyte IGF-1 receptor expression and


11/12


responsiveness to IGF-1 stimulation in mouse articularcartilage during various phases of experimentallyinduced arthritis. Ann Rheum Dis1995;54:645-53.13. Verschure PJ, Joosten LAB, Van der Kraan PM, Vanden BergW B.Responsiveness of articular cartilage fromnormal and inflamed knee joints to various growthfactors. Ann Rheum Dis 1994;53:455-60.14. Tyler JA. Insulin-like grow th factor 1 can decreasedegradation and promote synthesis of proteoglycan incartilage exposed to cytokines. Biochem J 1989;260:543-8.15.Le Roith D, Raizada MK.Molecular and cellular biologyof insulin like grow th factors and their receptors. NewYork: Plenum Press, 1989.16. Rechler H M , Nissley SP. Insulin-like growth factors. In:Sporn MB, Robers AB eds. Peptidegrowthfactors andtheir receptors.New York : Springer-Verlag, 1990;263-376.17.Bhaumick B, Bala RM. Differential effects of insulin-likegrowth factor 1 and 2 on growth, differentiation andglucoregulation in differentiating chondrocyte cells inculture. Ada Endocrinol1991;125:201-ll.

18. Neely EK, Beukers MW, Oh Y, Cohen P, Rosenfeld R G.Insulin-like growth factor receptors.A da Paediatr Scand1991;372:116-23.19. Aydelotte MB, Keuttner KE. Heterogeneity of articularchondrocytes and cartilage matrix. In: Woessner JF,Howell DS, eds.Cartilage degradation:basican dclinicalaspects.New Y ork: Raven Press 1991:37.20. Maroudas A, Schneidcrman R, Weinberg C. Compar-ison between effects of serum and insulin on GAGsynthesis in different zones of cultured human articularcartilage. Trans Orthop Res Soc 1990;15:315.21 . Siczkowski M, Watt FM. Subpopulations of chondro-cytes from different zones of pig articular cartilage:Isolation, growth and proteoglycan synthesis in culture./ Cell Sci 1990;97:349-


12/12

VERSCHURE ET AL.: IGF-1 RECEPTOR IN NORM AL AND OA CARTILAGE 1055

48. Korver GHV, Van de Stadt RJ, Van Kampen GP,Kiljan E, Van der Korst JK. Composition ofproteoglycans synthesized in different layers of culturedanatomically intact articular cartilage. Matrix Biol1990;l(h394-401.49. Van der Kraan PM, Vitters EL, Van Beuningen HM,Van de Putte LBA, Van den Berg WB. Proteoglycansynthesis and osteophyte formation in metabolicallyand mechanically induced murine osteoarthritis. Anin vivo autoradiographic study. J Exp Pathol1992;73:335-50.50. Lafeber FPJG, Van der Kraan PM, Huber-Bruning O,Van den Berg WB, Bijlsma JWJ. Osteoarthritichuman cartilage is more sensitive to transforming growthfactor f} than normal cartilage. Br J Rheumatol1993;32:281-6.51. Van Beuningen HM, Van der Kr aan PM , Arntz OJ, Vanden BergWB.Invivo protection against interleukin-1-in-duced articular cartilage damage by transforming growth

factor-/?,: Age related differences.Ann Rheum Dis1994;53:593-600.52. Van Beuningen HM, Van der Kraan PM , Arntz OJ, Vanden BergWB .Transforming growth factor-/?1 stimulatesarticular chondrocyte proteoglycan synthesis and inducesosteophyte formation in the murine knee joint. La bInvest1994;71:279-90.53. Van den BergWB.Cartilage destruction and osteophytesin osteoarthritis: role of transforming growth factor /??Rheumatol Eur1995;snppL 2:60-3.54. Dean DD, Martel-Pelletier J, Pelletier J-P. Evidence formetalloproteinase and metalloproteinase inhibitor imbal-ance in human osteoarthritic cartilage. J Clin Invest1989;84:678-85.55. Martel-Pelletier J, McCollum R, Fujimoto N, Obata K,Cloutier JM, Pelletier JP. Excess of metalloproteinasesover tissue inhibitor of metalloproteinase may contributeto cartilage degradation in osteoarthritis and rheumatoidarthritis. Lab Invest1994;70:807-15.

Documents

Rheumatology-1996-VERSCHURE-1044-55