Intra- and extracellular localization of hyaluronic acid and proteoglycan constituents (chondroitin sulfate, keratan sulfate, and protein core) in articular cartilage of rabbit tibia.

0022-1554/92/$3.30 The Joumal of Histochemistry and Cytochemistry Copyright 0 1992 by The Histochemical Society, Inc.

Vol. 40, No. 11. pp. 1693-1703, 1992 Printed in UXA.

Original Article

Intra- and Extracellular Localization of Hyaluronic Acid and Proteoglycan Constituents (Chondroitin Sulfate, Keratan Sulfate, and Protein Core) in Articular Cartilage of Rabbit Tibia AKIRA ASARI,' SATOSHI MIYAUCHI, KYOSUKE MIYAZAKI, AKIO HAMAI, KATSUYUKI HORIE, TOYOMI TAKAHASHI, TOMOKO SEKIGUCHI, AKEMI MACHIDA, KUNIO KOHNO, and YASUO UCHIYAMA Tokyo Research Institute, SeiRagaku Corporation, Tokyo, Japan (AA,SM,KM,AH,KH,TPS,AM); Department of Anatomy, Institute of Basic Medical Sciences, University of Tsukuba, Trukuba, Ibaraki-Ken, Japan (AA,KK); and Department of Cell Biology and Neuroanatomy, School of Medicine, Iwate Medical University, Morioka, Iwate-Ken, Japan flu).

Received for publication September 12, 1991 and in revised form February 12, 1992 and June 9, 1992; accepted June 24, 1992 (1A2449).

I

To demonstrate the intra- and extracellular localization of hyaluronic acid (HA) in articular cartilage of the rabbit tibia, biotinylated HA binding region, which specifically binds to the HA molecule, was applied to the tissue. In comparison with the localization of HA, that of chondroitin sulfate (CS), keratan sulfate (KS),and the protein core (PC) of the proteoglycan was examined by immunohistochemistry. Strong positive staining for HA was detected in chondrocytes located in the transition between the superficial and middle zones of the tissue. Pre-treatment with chondroitinase ABC, keratanase II, or trypsin enhanced the stainability for HA in peri- and intercellular matrices. Immunohistochemistry with or without enzymatic pre-treatment demonstrated that immunoreactivity for CS,KS, and PC was distinctly discerned

Introduction Hyaluronic acid (HA) is a linear high molecular weight polysaccharide in connective tissue. The HA binding region (HABR) of proteoglycan specifically binds to HA molecules in articular cartilage (Hascall and Heineghrd, 1974a,b,c; Hardingham and Muir, 1972). but not to dextran sulfate, sodium alginate, DNA, or chondroitin sulfate (Hardingham and Muir, 1972). Therefore, HABR has been used to detect HA in chick embryo and rat brain (Ripellino et al., 1985,1988) and rat fibrosarcoma cells (Knudson and Toole, 1985). In addition to HABR, hyaluronectin has been used to demonstrate HA in some tissues (Girard et al., 1986). Until the present, however, no specific antibody to HA has been prepared. Immunohistochemical and -microchemical analyses have demon-

'

Correspondence to: Akira Asari, Tokyo Research Institute, Seikagaku Corporation. Tateno 3-1253, Higashiyamato, Tokyo 207, Japan.

in chondrocytes and in the extracellular matrix located in the middle and deep zones. In particular, the immunoreactivity for KS and PC was augmented by pre-treatment with chondroitinaseABC not only in chondrocytes but in the extracellular matrix located in the middle and deep zones. hiiaobiochemical analysis corresponded well with histochemical and immunohistochemicalresults. These results suggest that HA is abundantly synthesized and secreted in chondrocytes located in the transition between the superficialand middle zones. (JHistdem Cytochem40:1693-1703,1992) KEYWORDS: Rabbit articulac cartilage; Hyaluronic acid; Hyaluronic

acid binding region; Chondroitin sulfate; Keratan sulfate; Protein core of proteoglycan.

strated that chondroitin sulfate (CS) and keratan sulfate (KS) are abundantly localized in the middle and deep zones of articular cartilage (Ratcliffeet al., 1984; Bayliss et al., 1983; Franzen et al., 1981; Maroudas et al., 1969; Stockwell and Scott, 1967). In articular cartilage, however, little is known about the precise localization of HA. The intracellular pathway of synthesis and secretion of HA has been studied in various cells by many researchers (Goldberg and Toole, 1983; Mitchell and Hardingham, 1982;Barland et al., 1968). From the experiments on monensin inhibition of HA synthesis in rat fibrosarcomacells and human articular chondrocytes, Goldberg and Toole have speculated that there exist two putative pathways of HA synthesis and secretion: one is via the Golgi apparatus and the other is by an unknown alternative pathway which may or may not involve the Golgi. To further understand HA synthesis and secretion in articular chondrocytes, it is necessary to examine the intracellular localization of HA. The present study examined the intra- and extracellular local-

1693 Downloaded from jhc.sagepub.com at UNIVERSITE DE MONTREAL on June 23, 2015

ASARI, MlYAUCHI, M I Y A W U . HAMAI. HORIE. TAKAHASHI. SEKIGUCHI, MACHlDA, KOWNO, UCHlYAMA

16%

1

HA

0 "*D

r.

I

'*

'

U

la *1

r -

HA*

HA

i

i

1f

4

Downloaded from jhc.sagepub.com at UNIVERSITE DE MONTREAL on June 23, 2015

--

'

LOCALIZATION OF HYALURONATE IN ARTICULAR CARTILAGE

1695

HA

Figure 1. Histochemical demonstration of hyaluronic acid (HA) in (a) a section fixed with acetic acid-ethanol, and (b-h) in sections fixed with paraformaldehyde-glutaraldehyde.HA staining is completely abolished by pre-treatment with hyaluronidase (d), whereas it is enhanced by pre-treatment with chondroitinase ABC (e) or trypsin (1). (g) No stainability for HA is detected in the section pre-treated with chondroitinaseABC followed by hyaluronidase.(h) HA staining is detected in the chondrocytesof tissue that was treated first with hyaluronidase,then fixed with a paraformaldehyde-glutaraldehyde solution, cryosectioned. and incubated with biotinylated HABR. Original magnification: a.b.d-h x 230; c x 700. Bars = 20 pm.

ization of HA, CS, KS, and t h e protein core (PC) of proteoglycan i n rabbit articular cartilage with biotinylated HABR a n d specific antibodies t o CS. KS. a n d PC. Concentrations of these glycosaminoglycans were also analyzed by microbiochemistry.

Materials and Methods Preparation of Biotinylated HABR HABR was prepared from bovine nasal cartilage proteoglycan with affinity chromatography on immobilized HA by the procedure of Tengblad (1979). Purified HABR was biotinylated with biotinyl-N-hydroxy-succinimide ester in the presence of complcxed HA to protect the binding site from biotinylation.

Antibodies Monoclonal antibodies to CS (CS56; ICN Immunobiologicals. Rehovoth. Israel), KS (5D4; Seikagaku, Tokyo, Japan), and PC (2B1; Seikagaku) were used to examine localization of CS, KS, and PC. respectively. CS56 is specific for the glycosaminoglycan portion of the native CS proteoglycan. reacting specifically with chondroitin-4- and -6-sulfate, but not with dermatan sulfate (Avnur and Benjamin, 1984). Enzyme immunoassay analyses indicate that 5D4 recognizes KS in native proteoglycan from hyaline cartilage, but not dermatan sulfate, heparan sulfate, or HA (Caterson e t al., 1983). 2B1 reacts specifically with the intact molecules and the chondroitinase ABCtreated core molecules of CS proteoglycan (Sobue et al., 1989).

Procedures for Histochemistry and Immunohistochemistry. Light Microscopy. Full-depth slices of articular cartilage of tibia condyle were dissected f r o m p rabbits (about 3 kg). Some cartilage slices were sectioned at 5 pm with a cryostat. The sections mounted on glass slides

were fixed with an acetic acid-ethanol solution at 4°C for 10 min. Other cartilage slices were immersed in 2% glutaraldehyde-2% paraformaldehyde buffered with 0.1 M phosphate buffer, pH 7.4, at 4'C for 4 hr. Fixed cartilage slices were then sectioned at 5 pm with the cryostat. Thesections were treated with 2% normal rabbit serum for 15 min and then with the following reaction media: biotinylated HABR (2 pglml). CS56 (1:lOO). 5D4 (1:100), and 2B1 (1:lOO) at 37°C for 1 hr. The sections treated with the primary antibodies for CS. KS, and PC were further incubated with biotinylated rabbit anti-mouse IgG and IgM (Techniclone International; Tustin, CA) at 37°C for l hr. Then all sections were treated with a strepravidin-peroxidase solution (Techniclone International). After each step, sections were rinsed thoroughly in 0.01 M phosphate buffer solution, pH 7.4. Staining for peroxidasewas performed with 0.03% 3,3'-diaminobenzidine (DAB) and 0.02% H202 in 0.05 M 7%-HC1 buffer, pH 7.6, for 10 min. Electron Microscopy. Cryosections fixed with 2% glutaraldehyde-2% paraformaldehyde and processed for HA staining were used for the electron microscopic samples of the pre-embedding method. Before incubation with biotinylated HABR, the sections were digested with chondroitinase ABC (pH 8.0) (Seikagaku) for 1 hr. Then these sections were further fixed with 1% glutaraldehyde buffered with 0.1 M phosphate buffer, p H 7.4, for 10 min. After rinsing with the buffer, they were incubated with 0.03% DAB and 0.02% H202 in 0.05 M %is-HCl buffer, p H 7.6. for 10 min. Then they were post-fixed with 1% os04 for 10 min. dehydrated with graded alcohols, and embedded in Quetol 812. For the post-embedding method, cartilage slices were fixed with 2% paraformaldehyde-2% glutaraldehyde buffered with 0.1 M phosphate buffer, p H 7.4, at 4% for 4 hr. After rinsing with the same buffer. they were dehydrated with graded glycometacrylate and embedded in Quetol812. Thin sectionswere cut with an ultramicrotome (Reichert, Ultracut N) and mounted on nickel g:ids. They were treated with 1% H202 and with chondroitinase ABC (pH 8.0) for 1 hr. After treatment with 5 % normal rabbit serum for 15 min. they were incubated with biotinylatcd HABR (2 pglml) in 0.1 M phosphate buffer, pH 7.4, at 4'C for 48 hr. and with streptavidin-colloidal gold (1:lOO. 10 nm) (Zymed; Burlingame, CA) in 0.1 M phosphate buffer, p H 7.4, at 4°C for 24 hr. After staining with uranyl acetate, they were observed with a Hitachi H-7000 electron microscope.


ASARI. MIYAUCHI, MIYAZAKI, HAMAI, HORE, TAKAHASHI, SEKIGUCHI, MACHIDA, KOHNO, UCHIYAMA

- --

- --.

..~ ~

~

G

G

., . -

1 L

2b

- .

-

Control Experiments Control sections were pretreated with specific enzymes to HA, CS, and KS for 1 hr before the following staining procedures: for HA with 200 TRUlml of hyaluronidase derived from Streptomyceshydumfyticus (Amano Pharmaceutical; Nagoya,Japan) in 100 mM sodium acetate buffer. pH 6.0, at 60'C; for CS with 5 Ulml of chondroitinase ABC (Seikagaku) in 100 mM sodium acetate buffer. p H 8.0. at 37'C; and for KS with 0.1 Ulml of kcratanase I1 (Seikagaku) in 100 mM sodium acetate buffer, p H 6.0, at 37'C. Pre-incubations with each reaction buffer were also performed. For the HA staining, some sections were incubated with biotinylated HABR adsorbed by HA (Seikamku) . . or with non-biotinylatcd HABR, followed by strcptavidin-peroxidase. For the CS. KS. or PCstining, some sectionswere incubated with non-immune rabbit serum diluted to 1:1000,followed by

rabbit anti-mouse immunoglobulin and streptavidin-peroxidase, or directly incubated with the second antibody without any preceding primary antibody step. Some control sections processed for HA staining were used for electron microscopy. Thin sections for electron microscopy were incubated with the adsorbed biotinylated HABR or with non-immune rabbit serum diluted to 1:200.followed by streptavidin-colloidal gold. Some thin sections were directly incubated with streptavidin-colloidal gold without pretreatment with the first reaction solution.

Enzymatic Pre-treatments In addition to the control study. staining of HA, CS, KS, and PC was also examined by prc-treatments with hyaluronidase, chondroitinasc ABC (pH



Figure 2. Electron micrographsshowing cytochemical localization of hyaluronic acid (HA) in (a-c) chondrocytes and in (d) extracellular matrix. (a) Reaction products for HA are seen in the ground cytoplasm of chondrocytes but not in cisternal, vesicular, or vacuolar components of the Golgi complex (G) (b). (c) Some vesicular structures near the plasma membraneare positivelystained (arrows). (d) Dot-like reaction products are regularly demonstrated beside collagen fibrils (arrowheads). Materialswere incubated with biotinylated HABR before embedding Ouetol 812. M. mitochondrion; N, nucleus; r, rough endoplasmic reticulum. Original magnifications: a-c x 30,000; d x 45.000. Bars = 0.2 um.

-.,

1697

-

- - -

8.0). keratanase 11, 0.1% trypsin (Type 111. bovine pancreas; Sigma, St Louis, MO) in 0.05 M Tris-HCI buffer, p H 7.6, at 37'C. or with combinations of these enzymes. To examine the alteration in the stainabilitv for HA after abolishment of HA in the extracellular matrix. raw cartilage was treated first with hyaluronidase at 37'C for 1 hr. Then the tissue was fixed with 2 % paraformaldehydc-2% glutaraldehyde buffered with 0.1 M phosphate buffer. pH 7.4, at 4'C for 2 hr. cryosectioned, and stained with biotinylatcd HABR as mentioned above.

Preparation of Articular Cartilage of Tibiafor Microbiochemistry Full-depth slices of tibial cartilage were dissected f r o m y rabbits (about 3 kg). The tissue slices were embedded in Tissue-Tek (Miles; Elkhart. IN) and immediately frozen at - 8O'C. Serial sections of the tissue were cut by the cryostat at 10 pm parallel to the cartilage surface. The serial sections were separated into three groups according to the depth from the cartilage


1698

ASARI, MIYAUCHI, MIYAZAKI, HAMAI, HORIE, TAKAHASHI, SEKIGUCHI, MACHIDA, KOHNO, UCHIYAMA

surface: superficial zone (0-100 pm), middle zone (100-200 pm), and deep zone (200-300 pm).

Biochemical Analysis Concentrations of HA, CS, and KS were determined by analysis of disaccharides with high-performance liquid chromatography (HPLC)as previously described by Toyoda et al. (1988) and Yoshida et al. (1989). Briefly, cartilage sections were lyophilized after thorough washing with PBS, pH 7.2, to remove the Tissue-Tek, and dried samples were weighed. Then they were digested with Actinase E (Karen Pharmaceutical: Tokyo,Japan), and the glycosaminoglycan fraction was prepared from the digestive solutions by anion exchange chromatography (Q-Sepharose column: 9.2 mm I.D. x 4.5 cm) (Pharmacia LKB Biotechnology: Tokyo, Japan). They were further digested with each glycosaminoglycanase.By HPLC, disaccharides in each sample were eluted by a gradient of 0-100 mM sodium sulfate for 60 min at a flow rate of 0.5 ml/min. To the eluent from the column, 100 mM sodium tetraborate buffer (pH 9.0) containing 10 mg/ml of 2-cyanoacetamide was added at a flow rate of 0.5 ml/min, and the mixture was passed through a polyetheretherketone (PEEK) reaction coil (0.23 mm 1.D. x 20 m) set in a dry reaction bath maintained at 140'C. The effluent was monitored by a spectrofluorometer set at excitation 331 nm and emission 383 nm,and the area of each peak corresponding with each disaccharide of the glycosaminoglycan was calculated by the integrator. Final data are expressed by nmol/mg (dry weights of the tissue belonging to each zone), and each value obtained is the mean of the two experiments.

Eflects of Chondroitinase ABC on the Preservation of H A in Cartilage Tissue Fifty cryosections of cartilage tissue fixed with 2% paraformaldehyde-2 % glutaraldehyde solution were incubated with chondroitinase ABC in 100 mM sodium acetate buffer, pH 8.0, at 37'C for 1 hr, or with the buffer only. Concentrations of HA in both treated tissue and reaction medium were determined by the biochemical analysis described above.

Results Histochemistry and Immunohistochemistry In the present study we used two kinds of fixatives: acetic acid-ethanol and paraformaldehyde-glutaraldehyde. With acetic acid-ethanol fixation histochemical or immunohistochemical reaction products appeared mainly in the extracellular matrix, whereas with paraformaldehyde-glutaraldehyde fixation these products were detected mainly in chondrocytes. In the enzymatic digestion study we used sections fixed with paraformaldehyde-glutaraldehyde. Hyaluronic Acid. In sections fixed with acetic acid-ethanol, HA was positivelystained mainly in the extracellular matrix of the middle

and deep zones (Figure la). In sections fixed with paraformaldehyde-glutaraldehyde, HA-positive deposits were demonstrated in chondrocytes located in all zones and in lamina splendens (Figures 1b and IC).In particular, chondrocytes located in the transition between superficial and middle zones were intensely stained by HABR. Weak staining for HA, however, was detected in the extracellular matrix. After pre-treatment with hyaluronidase, no staining appeared in any zone (Figure Id). Pre-treatment with chondroitinase ABC enhanced the staining for HA in extracellular matrix, particularly in the pericellular matrix of the transition between the su-

perficial and middle zones (Figure le). Pre-treatment with trypsin also enhanced the staining in the extracellular matrix in all zones, particularly the transition between the superficial and middle zones, although that in chondrocytes was slightly reduced (Figure If). Pretreatment with keratanase I1 enhanced the staining for HA in the extracellularmatrix of the deep zone. After pre-treatment with chondroitinase ABC, keratanase 11, or trypsin followed by hyaluronidase, no staining for HA appeared in the tissue (Figure 1g). To confirm the staining for HA in chondrocytes, raw cartilage samples were treated first with hyaluronidase for 1 hr. Then they were fixed with a paraformaldehyde-glutaraldehyde solution, cryosectioned, and stained with biotinylated HABR; positive HA staining was seen in chondrocytes but not in the extracellular matrix (Figure Ih). Moreover, as stated above, HA stained positively in the cartilage tissue after treatment with chondroitinase ABC. To confirm this result, the cartilage fixed with paraformaldehyde-glutaraldehyde was first treated with chondroitinase ABC and then the concentration of HA was determined in both treated tissue and reaction medium; the ratio of HA in the tissue and medium was 92:8. The same result was also obtained when the tissue was treated with the buffer only. By electron microscopy with the pre-embedding method, HA reaction product was seen in the ground cytoplasm of chondrocytes (Figure 2a). Positive deposits were also detected in small vesicular structures near the plasma membrane, but the rough endoplasmic reticulum (rER), Golgi complex, dense bodies, and mitochondria were not stained by HABR (Figures 2b and 2c). In the intercellular matrix, reaction products were detected forming dot-like structures along collagen fibrils (Figure 2d); each dot-like reaction product was laid at a certain distance on collagen fibrils (72.5 k 14.4 nm, mean distance k SD; n = 24). By the post-embedding method, gold particles indicating HA were localized mainly in vesicular structures near the Golgi complex of chondrocytes but not in rER (Figure 3a). In the ground cytoplasm of the cells, a few gold particles showing HA were detected. The gold particles were demonstrated on collagen fibrils located in the intercellular space (Figure 3b).

Chondroitin Sulfate. Immunodeposits for CS were demonstrated in both chondrocytes and extracellular matrix of the middle and deep zones in sections fixed with both acetic acid-ethanol and paraformaldehyde-glutaraldehyde (Figures 4a and 4b). Immunoreactivity for CS became intense corresponding to the depth from the superficial zone of the cartilage. As seen in HA staining (Figures la and Ib), the immunoreactivity for CS was prominent in the intercellular matrix in sections fixed with acetic acid-ethanol, whereas it was intense in chondrocytes in sections fixed with paraformaldehyde-glutaraldehyde. In the superficial zone, the immunoreactivity for CS was weak in both intra- and extracellular components. In particular, no immunodeposits were detected in lamina splendens. After pre-treatment with chondroitinase ABC, the immunoreactivity for CS was reduced considerably in both cellular and extracellular components of the cartilage (Figure 4c). Pretreatment with hyaluronidase or keratanase I1 caused no changes in CS immunoreactivity (Figure 4d). With trypsin pre-treatment, the immunoreactivity for CS was enhanced in the extracellular matrix of the middle zone close to the superficial zone, but it was reduced in chondrocytes (data not shown).



Figure 3. Electron micrographs showing cytochemical localization of hyaluronic acid (HA) (a) in chondrocytes and (b) in the extracellular matrix. (a) Colloidal gold particles indicating HA are distinctly localizedin vesicular and vacuolar structures (V) near the Golgi complex (G). (b)Colloidal gold particles demonstrating HA are localized in electron-lucent collagen fibrillar structures. r, rough endoplasmic reticulum. Original m a g nifications: a x 30,000; b x 45.000. Bars = 0.2 pm,

-

Keratan Sulfate. In sections fixed with acetic acid-ethanol, diffuse immunodeposits for KS were detected in all zones including lamina splendens, but the immunoreactivity was relatively weak (Figure 5a). In sections fixed with paraformaldehyde-glutaraldehyde, dense immunodeposits for KS were demonstrated in chondrocytes and the pericellular matrix located mainly in the middle and deep zones (Figure 5b). Lamina splendens and the most superficially located chondrocytes were positively immunostained by 5D4. After pre-treatment with keratanase 11, the KS immunoreac-

1699

.

.

...

tivity was considerably reduced, but weak immunoreactivity remained in both intra- and extracellular components (Figure 5c). Pre-treatment with hyaluronidase did not alter the immunoreactivity for KS. Pre-treatment with chondroitinase ABC enhanced the immunoreactivity in chondrocytes and the extracellular matrix, particularly in the pericellular matrix (Figure 5d). Therefore, we further tried to pre-treat the sections with a combination of chondroitinase ABC and hyaluronidase and of chondroitinase ABC and trypsin. The former combination abolished KS immunoreac-


1700

ASAN, MIYAUCHI,MIYAWCI, HAMAI. HORE, TAKAHASHI. SEKIGUCHI. MACHTDA, KOHNO, UCHIYAMA

cs

cs

Figure 4. lmmunohistcchemicaldemonstration of chondroitin sulfate (CS) in e &ion fixed with (a) acetic acid-ethanol end in &ions fixed with (b-d)parafonneldehyde-glutaraldehyde. Immunoreactivityfor CS is completely abolished by the pre-treatment with chondmitinese ABC (c). whereas no clear-cut change is detected in the immunoreactivity after treatment with hyaluronidase (d). Original magnification x 230. Bar I 20 pm.

tivity in all z o n a (Figure Se), whereas with the latter combination it was augmented in the intercellular matrix of the middle and deep zones but decreased in chondrocytes (data not shown). Protein Core. Immunodeposits for PC were seen in chondrocyte located in all zona. the intercellular matrix of the middle and deep zones, and lamina splendens in sections Tied with acetic acid-ethanol (Figure 61). In sections fixed with paraformaldehyde-glutaraldehyde. PC immunostaining was intense in chondrocyta in a11 zones with 2B1 (Figure 6b). The extracellular matrix, and particularly the intercellular matrix of the middle and deep zona. also showed immunoreachvity for PC. whereas weak immunoreactivity was demonstrated in lamina splendens. Enzymatic pre-treatments demonstrated that chondroitinase ABC slightly in-

creased the immunoreactivity for PC in the extracellular matrix of all zones (Figure 6c). whereas no clear-cut alteration was detected by hyaluronidase or kcratanax 11. The combined pre-treatment with chondroitinase ABC and hyaluronidase or trypsin markedly decreased the immunoreactivity for PC (Figure 6d). Alterations in the immunoreactivity for HA, CS. KS. and PC after various enzymatic treatments are summarized in Table 1. Control thin and cryosections incubated with the adsorbed biotinylated HABR. or with non-immune rabbit antiserum as a first reaction medium, showed no specific reaction deposits in articular cartilage tissue. Moreover. control cryosections incubated directly with the biotinylated second antibody also exhibited no specific reaction product.


LOCAUWllON OF HYALURONATE IN ARTICULAR CARTILAGE

?K'S

! h

1701

KS I

$i i i

5a KS

5c KS

Figure 5. Immunohistochemicaldemonstrationof keratansulfate (KS)in a section fixed with (a) acetic acid-ethanol and in seclions fixed with (b-0)pambrmaldehydeglutaraldehyde.Immunoreactivityfor KS is decreasedby pre-treatment with keratanase I1 (c). whereas it is increased by pre-treatmentwith chondroitinase ABC (d). Combined pre-treatmentwith chondroitinaseABC and hyaluronidase abolishes the immunoreactivity (e). Original magnification x 230.Bar 20 pm.

-

5e


1702

ASARI. MWAUCHI, MIYAWKI. HAMAI. HORIE. TAKAHASHI, SEKIGUCHI, MACHIDA. KOHNO, UCHlYAMA

*

i

.

.- . . l5

I

b

:

..L a

-.

6a

* ' ._

*

PC

Figure 6. Immunohistochemicaldemonstration of protein core (PC) of the proteoglycan in a section fixed with (a) acetic acid-ethanol and in sections fixed with (b-d) paraformaldehyde-glutaraldehyde. Immunoreactivityfor PC is enhanced by pre-treatment with chondroilinase ABC (c). whereas it is abolished by combined pre-treatment with chondroitinase ABC and hyaluronidase (d). Original magnification x 230. Bar I 20 vm.

Microbiochemistry Concentrations of HA, CS. and KS according to the depth from the surface of the articular cartilage were determined by HPLC analysis (Table 2). Concentrations of these glycosaminoglycansdiffered in each zone. The HA concentration was high in the superficial zone, whereas the CS concentration was high in the middle zone. The KS concentration became higher corresponding to the depth from the superficial to the deep zones.

Discussion The present study utilizing biotinylated HABR and monoclonal antibodies to CS. KS. and PC demonstrated that HA, CS. KS. and PC are localized in rabbit articularcartilage. Concentrations of HA,

CS, and KS determined by HPLC were shown to differ according to the depth from the surface of articular cartilage. It is interesting that the distribution patterns of each glycosaminoglycan and PC examined apparently differ with the futatiws used. Staining for these glycosaminoglpans in the present study was completely abolished or reduced in cryosections fixed with each fixative after the glycosaminoglycanase digestion. These results s u g gest that the positive products shown in each histochemical or immunohistochemical reaction demonstrate distinct localization of these substances in articular cartilage tissue. At present, it remains unknown why histochemical and immunohistochemical stainability differs between cartilage fixed with acetic acid-ethanol and paraformaldehyde-glutaraldehyde solutions. To confirm the HA staining within chondrocytes, we f i m treated raw cartilagemples with hyaluronidase. Then the tissue was fixed with a parafor-



1703

Table 1. Changes in stainability of glycosaminoglycans ana’ protein core ofproteoglycan aftr tregtment with various enzymesa

The intracellular pathway for synthesis and secretion of HA has been previously investigated. Mitchell and Hardingham (1982) have shown that monensin, which interferes with the intracellularpathways of secretory proteins in many different cell types, inhibits secretion of proteoglycan but not of HA in chondrocytes. On the other hand, Goldberg and Toole (1983) have demonstrated that monensin inhibits HA synthesis in rat fibrosarcoma cells, and these authors suggested that there are two possible pathways for synthesis and secretion of HA: one similar to chondroitin sulfate side chains of proteoglycansmade by chondrocytes, i.e., via the Golgi apparatus, and one by an unknown alternative route which may or may not involve the Golgi. In the present study, we had contradictoryresults on the localization of HA in chondrocytes with the pre-embedding method; reaction products for HA were seen in the ground cytoplasm and in small vesicular structures near the plasma membrane. The post-embedding method demonstrated that gold particles indicating HA are localized in small vesicles; this agrees with the results of Ratcliffe et al. (1984). who have shown that the hyaluronate binding region of proteoglycan is immunocytochemically localized in the Golgi complex and in membrane-bound vesicles of chondrocytes in porcine articular cartilage. At present, our result on the localization of HA in the ground cytoplasm of chondrocytes remains unsettled. With biotinylated HABR as a specific histochemicalprobe, Ripellino et al. (1988) have shown that HA-positive products are detected in the granule cell cytoplasm and axoplasm in adult rat cerebellum. Consideringthese results, it may be that the localization of HA in the ground cytoplasm of chondrocytes demonstrates an alternative pathway of synthesis and secretion. Moreover, it has been shown that HA is anchored by a phosphate ester bridge to the membrane during synthesis (Prehm, 1983). As stated above, we could not find reaction products indicating HA bound to membranes of chondrocytes, including the plasma membrane. This may indicate that synthetic sites of HA differ depending on the cell type. The present electron microscopy study shows that HA-positive products are laid on collagen fibrils with a periodicity of 72.5 nm. This agrees with the results of Poole et al. (1982). who have shown that the immunoreactivity for PC is detected as particulate deposits spaced at regular intervals (72 nm) along collagen fibrils in bovine articular cartilage. The periodic arrangementof HA-positivedeposits along collagen fibrils suggests that proteoglycan aggregates are associated with collagen fibrils by HA. This is also supported by the fact that combined pre-treatment of hyaluronidase with chondroitinase ABC decreases the immunoreactivity for KS and PC. The microbiochemical method used in the present study directly measured saturated or unsaturated disaccharidesto determine the amount of CS and KS in rabbit articular cartilage; the concentration of CS was high in the middle zone, whereas that of Ks high in the deep zone. These results are compatible with those reported by many researchers (Ratcliffe et al., 1984; Bayliss et al., 1983; Franzen et al., 1981; Maroudas et al., 1969; Stockwell and Scott, 1967). In the present digestion study, the immunoreactivity for Cs was not enhanced by pre-treatment with hyaluronidase or ketatanase 11. This may be attributed to the fact that CS is a major constituent of proteoglycans in articular cartilage (Campo and Tourtellotte, 1967). After pre-treatment with chondroitinase ABC, the immunoreactivity for KS was augmented in the middle and deep zones,

Changes in stainability Enzymatic me-trratment

Ks

cs

HA

PC

C

M

C

M

C

M

C

M

A NC NC D A D A

A I I I A I A

NC D NC D

NC D NC I

NC I D A D

NC 1 D A I

NC I NC D D D -

NC 1 NC D D D -

~~

HAase CHase Keratanase I1 Trypsin CHase-HAase CHase-trypsin Trypsin-HAase

~~

HA, hyaluronic acid; CS. chondroitin sulfate; KS, keratan sulfate; PC, protein core of proteoglycan; C, chondrocytes; M. cxuacellular ma&; HAase, hyaluronidase; CHase, chondroitinase ABC; A, abolished; I . increased; D, decreased; NC, no change.

maldehyde-glutaraldehyde solution, cryosectioned, and stained with biotinylated HABR. With this treatment HA staining was still detected in chondrocytes but not in the extracellular matrix, indicating that the staining of HA in chondrocytes after paraformaldehyde-glutaraldehyde fixation does not result from disarrangement of HA during the fixation procedure. With biotinylated HABR, HA was distinctly demonstrated in chondrocyteslocated in all zones, particularly in the transition between the superficial and middle zones. It is well known that chondrocytes in the superficial zone have a flat or spindle-likeform and possess fibroblast-likecytoplasmicorganelles. Chondrocytes in the transition between the superficial and middle zones have welldeveloped rER and Golgi complexes, whereas those in the deeper zones are large and have poorly developed cytoplasmic organelles. From the evidence, it seems likely that chondrocytes in the superficial zone and the transition between the superficial and middle zones abundantly produce and secrete HA. This coincides well with the present biochemical result showing that the HA concentration in the superficial zone is high as compared with that of other zones. Moreover, these results were also confirmed by the present enzymatic digestion study indicating that chondroitinase ABC or trypsin strongly enhances the stainability for HA in the inter- and pericellular matrices located in the transition between the superficial and middle zones. We confirmed that treatment with chondroitinase ABC does not affect the presence of HA in the cartilage tissue by measuring its amount in both treated tissue and chondroitinase ABC reaction medium. The present enzymatic study also suggests that at least the HA of proteoglycan aggregates is masked by CS, and by KS in the extracellular matrix.

Table 2 . Microbiochemical analysis of hyaluronic acid, chondroitin su(fte, and ieratan sulfitea Zone

HA

cs

Ks

Superficial zone Middle zone Deep zone

2.71 1.77 1.63

1.13 x lo2 1 . 3 0 x lo2 1.15 x 102

3.96 x 10 4 . 5 2 x 10 4 . 8 3 x 10

HA, hyaluronic acid; CS, chondroitin sulfate; KS. kcratan sulfate. Units are expressed as nmol/mg. Each value is the mean of the two experiments.


1704

ASARI, MIYAUCHI, MIYAZAKI, HAMAI, HORIE, TAKAHASHI, SEKIGUCHI, MACHIDA, KOHNO, UCHIYAMA

whereas that of PC was slightly enhanced in all zones. These results suggest that not only HA but also KS and PC are masked by CS in rabbit articular cartilage. From the results mentioned above, biotinylated HABR is available for the histochemical demonstration of HA in articular cartilage. By use of this HABR and microbiochemistry,our results suggest that HA is abundantly synthesized and secreted in chondrocytes located in the transition between the superficialand middle zones. Moreover, histochemistry and immunohistochemistry combined with microbiochemistrysuggest that the organization of proteoglycan aggregates occurs mainly in the middle and deep zones.

Literature Cited Avnur 2, Benjamin G (1984): Immunocytochemicallocalization of native chondroitin-sulfate in tissues and cultured cells using specific monoclonal antibody. Cell 38:811 Barland P, Smith C, Hamerman D (1968): Localization of hyaluronic acid in synovial cells by radioautography.J Cell Biol 37:13 Bayliss MT, Venn M, Maroudas A. Ali YS (1983): Structure of proteoglycans from different layers of human articular cartilage. Biochem J 209387 Campo RD, Tourtellotte CD (1967): The composition of bovine cartilage and bone. Biochim Biophys Acta 141:614 Caterson B, ChristnerJE, Baker JR (1983): Identification of a monoclonal antibody that specifically recognizes corneal and skeletal keratan sulfate. J Biol Chem 258:8848 Franzen A, Inerot S, Hejderup S-0, Heinegkd D (1981): Variations in the composition of bovine hip articular cartilagewith distance from the articular surface. Biochem J 195:535 Girard N, Delpech A, Delpech B (1986): Characterization of hyaluronic acid on tissue sections with hyaluronectin.J Histochem Cytochem 34:539 Goldberg RL., Toole BP (1983): Monensin inhibition of hyaluronate synthesis in rat fibrosarcoma cells. J Biol Chem 258:7041 Hardingham TE, Muir H (1972): The specific interaction of hyaluronic acid with cartilage proteoglycans. Biochim Biophys Acta 279:401

Hascall VC, Heinegird D (1974~):Aggregation of cartilage proteoglycans. 111. Characteristics of the proteins isolated from trypsin digests of aggregates. J Biol Chem 249:4250 Knudson CB, Toole BP (1985): Fluorescent morphological probe for hyaluronate. J Cell Biol 100:1753 Maroudas A, Muir H, Wingham J (1969): The correlation of fixed negative charge with glycosaminoglycan content of human articular cartilage. Biochim Biophys Acta 177492 Mitchell D, Hardingham T (1982): Monensin inhibits synthesis of proteoglycan, but not of hyaluronate, in chondrocytes. Biochem J 202:249 Poole AR, Pidoux I, Reiner A, Rosenberg L (1982): An immunoelectron microscope study of the organization of proteoglycan monomer, link protein, and collagen in the matrix of articular cartilage. J Cell Biol 93:921 Prehm P (1983): Synthesis of hyaluronate in differentiated teratocarcinoma cells. Characterization of the synthase. Biochem J 211:181 Ratcliffe A, Fryer PR, HardinghamTE (1984):The distributionof aggregating proteoglycans in articular cartilage: comparison of quantitative immunoelectron microscopy with radioimmunoassay and biochemical analysis. J Histochem Cytochem 32:193 Ripellino]A, Klinger MM, Margolis RU, Margolis RK (1985): The hyaluronic acid binding region as a specific probe for the localization of hyaluronic acid in tissue sections. Application to chick embryo and rat brain. J Histochem Cytochem 33:1060 Ripellino JA, Bailo M, Margolis RU, Margolis RK (1988): Light and electron microscopic studies on the localization of hyaluronic acid in developing rat cerebellum. J Cell Biol 106:845 Sobue M. Nakashima N, Fukatsu T, Nagasaka T, Fukata S, Ohiwa N. Nara

Y,Ogura T, Katoh T, Ekeuchi J (1989):Production and immunohistochemical characterizationof a monoclonal antibody raised to proteoglycan purified from a human yolk sac tumour. Histochem J 21:455 Stockwell RA, Scott JE (1967): Distribution of acid glycosaminoglycans in human articular cartilage. Nature 215:1376 Tengblad A (1979): Affinity chromatographyon immobilized hyaluronate and its application to the isolation of hyaluronate binding proteins from cartilage. Biochim Biophys Acta 578:281

Hascall VC, Heinegird D (1974a): Aggregation of cartilage proteoglycans. I. The role of hyaluronic acid. J Biol Chem 2494232

Toyoda H, Shinomiya K, Yamanashi S, Koshiishi I, Imanari T (1988): Microdeterminationof unsaturated disaccharidesproduced from chondroitin suifate in rabbit plasma by high performance liquid chromatographywith fluorometric detection. Anal Sci 4381

Hascall VC, Heinegkd D (1974b): Aggregation of cartilage proteoglycans. 11. Oligosaccharide competitors of the proteoglycan-hyaluronic acid interaction. J Biol Chem 249:4242

Yoshida K, Miyauchi S, Kikuchi H, Tawada A, Tokuyasu K (1989):Analysis of unsaturated disaccharides from glycosaminoglycuronan by high-performance liquid chromatography. Anal Biochem 177:327


Affinity binding of the cartilage proteoglycan protein-keratan sulfate core to immobilized hyaluronic acid.

Co-localization of hyaluronic acid and chondroitin sulfate proteoglycan in rat cerebral cortex.

Localization of hyaluronic acid, chondroitin sulfate, and CD44 in rabbit cornea.

Analysis of the chondroitin sulfate proteoglycan core protein (CSPGCP) gene in achondroplasia and pseudoachondroplasia.

dermatan sulfate, hyaluronic acid, heparan sulfate, and keratan sulfate by liquid chromatography-electrospray ionization-tandem mass spectrometry.

Localization of heparan sulfate glycosaminoglycan and proteoglycan core protein in aged brain and Alzheimer's disease.

Detection of keratan sulfate by immunological methods in commercial chondroitin sulfate preparations.

Selective removal of keratan sulfate in chondroitin sulfate samples by sequential precipitation with ethanol.

Biocompatibility assessment of novel collagen-sericin scaffolds improved with hyaluronic Acid and chondroitin sulfate for cartilage regeneration.

Detection of chondroitin sulfate proteoglycan 4 (CSPG4) in melanoma.

Tough and elastic hydrogel of hyaluronic acid and chondroitin sulfate as potential cell scaffold materials.

Comparison of chondroitin sulfate and hyaluronic Acid doped conductive polypyrrole films for adipose stem cells.

dermatan sulfate from the lumpsucker fish, C. lumpus.

Undersulfated chondroitin sulfate in the cartilage matrix of brachymorphic mice.

Distribution of chondroitin sulfate in cartilage proteoglycans under associative conditions.

Silk fibroin-chondroitin sulfate scaffold with immuno-inhibition property for articular cartilage repair.

dermatan sulfate proteoglycan.

Identification of a developmentally regulated keratan sulfate proteoglycan that inhibits cell adhesion and neurite outgrowth.

Identification of chick corneal keratan sulfate proteoglycan precursor protein in whole corneas and in cultured corneal fibroblasts.

The synthesis of hyaluronic acid by sheep and rabbit articular cartilage in vitro.

Presence of an endo-beta-galactosidase degrading the linkage region between the chondroitin sulfate chain and core peptide of proteoglycan.

Investigation of chondroitin sulfate D and chondroitin sulfate E as novel chiral selectors in capillary electrophoresis.

Quantitation of chondroitin 4-sulfate and chondroitin 6-sulfate in pathologic joint fluid.

Identification of keratan sulfate disaccharide at C-3 position of glucuronate of chondroitin sulfate from Mactra chinensis.