- Email: [email protected]
Characterization of glycosaminoglycans from human normal and scoliotic nasal cartilage with particular reference to dermatan sulfate
Biochimica et Biophysica Acta 1528 (2001) 81^88
www.bba-direct.com
Characterization of glycosaminoglycans from human normal and scoliotic nasal cartilage with particular reference to dermatan sulfate Achilleas D. Theocharis a , Marina E. Tsara b , Nikoletta Papageorgakopoulou a , Demitrios H. Vynios a , Dimitrios A. Theocharis b; * a
Laboratory of Biochemistry, Department of Chemistry, School of Medicine, University of Patras, 261 10 Patras, Greece b Laboratory of Biological Chemistry, School of Medicine, University of Patras, 261 10 Patras, Greece Received 18 April 2001; received in revised form 27 June 2001; accepted 28 June 2001
Abstract The composition and the distribution of glycosaminoglycans (GAGs) present in normal human nasal cartilage (HNNC), were examined and compared with those in human scoliotic nasal cartilage (HSNC). In both tissues, hyaluronan (HA), keratan sulfate (KS) and the galactosaminoglycans (GalAGs) ^ chondroitin sulfate (CS) and dermatan sulfate (DS) ^ were identified. The overall GAG content in HSNC was approx. 30% higher than the HNNC. Particularly, a 114% increase in HA, and 46% and 86% in KS and DS, respectively, was recorded. CS was the main type of GAG in both tissues with no significant compositional difference. GalAG chains in HSNC exhibited an altered disaccharide composition which was associated with significant increases of non-sulfated and 6-sulfated disaccharides. DS, which was identified and quantitated for the first time in HNNC and HSNC, contained low amounts of iduronic acid (IdoA), 18% and 28% respectively. In contrast to other tissues, where IdoA residues are organized in long IdoA rich repeats, the IdoA residues of DS in human nasal cartilage seemed to be randomly distributed along the chain. DS chains in HSNC were of larger average molecular size than those from HNNC. These results clearly indicate the GAG content and pattern in both HNNC and HSNC and demonstrate that scoliosis of nasal septum cartilage is related to quantitative and structural modifications at the GAG level. ß 2001 Elsevier Science B.V. All rights reserved. Keywords : Scoliosis; Nasal cartilage ; Glycosaminoglycan ; Proteoglycan
1. Introduction Hyaline cartilage is a highly specialized connective tissue with biomechanical, developmental and other biological functions. Cartilages from several anatomical sites exhibit di¡erent functional properties and are characterized by di¡erent macromolecular composition [1,2]. Collagen and glycosaminoglycans (GAGs)/proteoglycans (PGs) are the major components of hyaline cartilage. PGs are highly anionic macromolecules consisting of a protein core onto which sulfated GAG chains are covalently bound. In contrast, hyaluronan (HA) is a high molecular weight nonsulfated GAG, which does not occur as a PG but in free form. The common GAGs include the galactosaminoglycans (GalAGs) (chondroitin sulfate (CS) and dermatan sulfate (DS)) and glucosaminoglycans (HA, heparan sulfate, heparin and keratan sulfate (KS)) [3,4]. The GAGs are intrinsic constituents and, particularly because of their
* Corresponding author. Fax: +30-61-99-76-90.
complex polyanionic nature, have been implicated in a number of biological processes and functions [5]. Although GAGs from various types of human hyaline cartilage such as articular [2], costal, tracheal and bronchial [6] as well as those from sheep [7^10] and bovine [1,11] have been extensively studied, no information has been reported about the GAGs found in human nasal cartilage. Scoliosis of the nasal septum cartilage is the most frequent deformity of the nose in man. The deformity of the nasal septum cartilage causes di¤culty in breathing and is implicated in a number of episodes of rhinitis, pharyngitis, otitis as well as in the sphenopalatine neuralgia which is known as Sluder's syndrome [12,13]. The present study was undertaken to determine the chemical composition, the GAG content and distribution of human normal septum cartilage and to compare them with those of scoliotic nasal septum cartilage of the same age. This study apparently constitutes the ¢rst study of GAG content, composition and structure of human nasal cartilage and provides a baseline for future studies in this ¢eld.
0304-4165 / 01 / $ ^ see front matter ß 2001 Elsevier Science B.V. All rights reserved. PII: S 0 3 0 4 - 4 1 6 5 ( 0 1 ) 0 0 1 7 3 - 8
BBAGEN 25214 18-10-01
82
A.D. Theocharis et al. / Biochimica et Biophysica Acta 1528 (2001) 81^88
exhibited the typical characteristics of hyaline cartilage that were observed in the normal specimens.
2. Materials and methods 2.1. Chemicals Sepharose CL-6B and DEAE-Sephacel were obtained from Pharmacia (Uppsala, Sweden) and Bio Gel P-10 from Bio-Rad. Papain twice crystallized (EC 3.4.22.2), chondroitinase ABC (EC 4.2.2.4, Proteus vulgaris), chondroitinase B (Flavobacterium heparinum), keratanase (EC 3.2.1.103, Pseudomonas species), GAG standards and standard preparations of chondroitin disaccharides, vDi-4S, vDi-6S and vDi-0S were from Sigma (St. Louis, MO, USA). ECTEOLA-cellulose (ET-11) and cellulose powder (CF-11) were purchased from Whatman (W.pR. Balston, UK). All other chemicals used were of the best grade commercially available. 2.2. Analytical methods Uronic acid was determined by a modi¢ed carbazole reaction [14]. Analysis for glucosamine and galactosamine was performed on a Beckman amino acid analyzer [15]. Hexoses were measured by the anthrone reaction, using galactose as standard [16]. The amounts of unsaturated disaccharides released by the chondroitinase ABC were determined by high performance liquid chromatography (HPLC) [17]. The quanti¢cation of each v-oligosaccharide pool produced by the action of chondroitinase B and isolated by gel chromatography, besides hexuronic acid determination, was based on their absorbance at 232 nm (E232 = 5500 M31 cm31 ) owed to the C-4 double bond of the uronic acid at the non-reducing end. The accuracy of the assay was tested using internal standards of v-disaccharides. Glucuronic acid (GlcA) and iduronic acid (IdoA) were determined by the method of Karamanos et al. [18]. In brief, the galactosaminoglycans were reacted with 1ethyl-3-(3-dimethylaminopropyl)carbodiimide and then subjected to reduction of their hexuronic acid to the respective alditols. Finally, the samples were hydrolyzed and the monosaccharides were liberated, the alditols were puri¢ed using ion exchange chromatography, and transformed to the respective perbenzoyl derivatives, which were separated and quantitated by HPLC. Collagen content was calculated via determination of hydroxyproline [19] as described elsewhere [7] in papain digests of the tissues. 2.3. Tissue source ^ patients Human normal nasal septum cartilages (n = 5, 25^35 years old) were obtained by autopsy and frozen at 320³C. Specimens (n = 12, 25^35 years old) from patients with scoliosis of the nasal septum cartilage were taken directly from operating rooms as surgical discard tissues. The scoliotic specimens were free from any ¢brous structure and on the basis of biochemical and histological data
2.4. Isolation of GAGs Intrabatch di¡erences in GAG content were examined in specimens of normal human nasal cartilage (HNNC ; n = 5) and human scoliotic nasal cartilage (HSNC ; n = 12) which have been separately isolated and analyzed for their GAG content. Particularly, known amounts of wet weight of both tissues were exhaustively digested with papain and then the GAGs were precipitated by addition of 5 vols. of ethanol. Subsequently, GAG chains were released from their protein core/peptide fragments by treatment with 0.05 M NaOH/1 M sodium borohydride at 45³C for 48 h under vacuum [20]. The solutions were brought to pH 5.0 with glacial acetic acid and then precipitated with 5 vols. of ethanol. The GAG content was estimated in each sample by measuring the uronic acid, hexoses, glucosamine and galactosamine in the precipitates. 2.5. Fractionation and characterization of GAGs According to pilot experiments (see Section 3) all normal specimen isolated GAGs were mixed in a group, and the scoliotic specimen isolated GAGs were mixed in a second group. Fractionation of various types of GAGs was performed as described elsewhere [7,9,21]. Brie£y, the isolated GAGs from each group were chromatographed on cetylpyridinium chloride (CPC)-cellulose column [22] and eluted with 2 bed vols. of 1% CPC/0.005 M Na2 SO4 , and 4 M NaCl/0.05% CPC. The 1% CPC fractions (containing KS) were freed from CPC by extraction with isoamyl alcohol, desalted and reduced in volume by dialysis through Amicon UM-2 membrane and then KS was puri¢ed further by chromatography on ECTEOLAcellulose [23]. The column was eluted with 2 bed vols. of 0.02 M HCl and 2 M NaCl. KS was eluted with 2 M NaCl. The 4 M NaCl/0.05% CPC fractions of CPC-cellulose column (containing the other types of GAGs) were desalted and reduced in volume by dialysis and then fractionated on a DEAE-Sephacel column [21]. The column was eluted stepwise with 3 vols. of 0.1 M NaCl and a linear gradient ranging from 0.1 M to 1.2 M NaCl (10 vols.). This ion exchange chromatography separated the retarded uronic acid positive material into two populations in each tissue. Populations I and II, which were eluted with 0.35 M and 0.7 M NaCl respectively, were pooled, quanti¢ed in terms of uronic acid and concentrated on PM-10 membranes. The recovery of hexosamine from the combined chromatographies was 94^97%. Identi¢cation and quantitation of each GAG fraction was assessed by cellulose acetate electrophoresis [24], hexosamine analysis and speci¢c enzymatic treatment. Disaccharide composition and sulfation pattern of GalAGs were esti-
BBAGEN 25214 18-10-01
A.D. Theocharis et al. / Biochimica et Biophysica Acta 1528 (2001) 81^88
mated following digestion with chondroitinase ABC and analyses of the digests by HPLC. 2.6. Enzymatic degradation Digestion with papain was performed at 60³C for 18 h using twice crystallized papain (0.2 mg of papain/g wet weight of tissue) in 0.1 M Tris^HCl, pH 7.2, containing 0.01 M disodium EDTA and 0.005 M cysteine-HCl. For chondroitinase ABC digestion, solutions of GAGs (1 mg of uronic acid/ml) in 0.1 M Tris-acetate, pH 7.3, were incubated at 37³C for 24 h in the presence of 0.5 units of chondroitinase ABC per ml. Digestion with chondroitinase B was performed in 0.05 M Tris^HCl, pH 8.0, containing 0.05% (w/v) bovine serum albumin, at 37³C for 24 h using 0.3 units enzyme per mg of uronic acid. For keratanase digestion, samples containing KS were incubated at 37³C for 24 h in 0.1 M Tris^HCl, pH 7.4, containing 0.1 U of keratanase per Wmole of galactose. All digestions were terminated by heating the solutions at 100³C for 3 min. 2.7. Gel chromatography Gel chromatography of GAG chains was performed on analytical (110U0.6 cm) Sepharose CL-6B column, equilibrated and eluted with 0.5 M sodium acetate, pH 7.0, before and after treatment with speci¢c degradative enzymes. The average molecular weight (Mr ) of the GAG chains was estimated according to Wasteson [25] using CS molecular weight standards for calibrating the Sepharose CL-6B column as previously described [24]. In particular, the Mr of the DS chains was calculated by the construction of the secondary pro¢le as described in a recent study [26]. In brief, the isolated GalAGs were digested with chondroitinase B and the pro¢le, upon gel chromatography on Sepharose CL-6B, was compared to that of the initial undigested sample. The absolute amounts or percentage of uronic acid measured in each fraction of the chromatographies before and after treatment with chondroitinase B were subtracted, fraction by fraction, and the di¡erences were obtained. The positive di¡erences were used to construct the secondary pro¢le, which represented the uronic acid content of the fractions removed after the enzymatic digestion, i.e. the elution pro¢le of intact DS chains, and thus the size distribution of DS was obtained and its Mr calculated. The size of DS-derived v-oligosaccharides produced by chondroitinase B was determined by chromatography on Bio Gel P-10 (160U0.6 cm) in 1.0 M ammonium acetate pH 7.0. The column was calibrated using standard preparations of v-oligosaccharides [26]. The void (V0 ) and total volumes (Vt ) of each column were determined with sheep nasal cartilage proteoglycan aggregates and tritiated water, respectively.
83
2.8. Statistical analysis Statistical signi¢cance was evaluated by t-test using Microcall origin software (version 3.2). 3. Results 3.1. Chemical composition of human nasal cartilage To examine for intrabatch and interbatch di¡erences in total GAG and collagen contents chemical analyses were carried out on isolated total GAGs from known amounts of tissue from each specimen (normal and scoliotic cartilage) as described in Section 2. The results of the chemical composition (Table 1) indicated that no signi¢cant intrabatch di¡erences were found in the overall GAG contents (expressed either as uronic acid or hexosamine or hexoses) and collagen contents (expressed as hydroxyproline), between the various HNNC (n = 5) and HSNC (n = 12) examined. For these reasons, further analyses were performed on pooled scoliotic and normal cartilages. The overall GAG content in scoliotic cartilage (13.8 þ 0.7 mg of hexosamine/g wet weight (w.wt.) of tissue) was found to be increased (P90.001), about 30%, as compared to normal cartilage (10.6 þ 0.6 mg/g w.wt. of tissue). The marked changes in the molar ratio of galactosamine/glucosamine indicate signi¢cant di¡erences in the GAG composition. The collagen content in the various normal and scoliotic cartilages (216 þ 12 and 208 þ 11 mg/g w.wt. of tissue, respectively) was similar, showing no signi¢cant intrabatch and interbatch di¡erences (P90.001). 3.2. Fractionation and characterization of GAGs The purity of isolated and separated KS from both HNNC and HSNC by combined chromatography (CPCcellulose and ECTEOLA-cellulose), was assessed by electrophoresis on cellulose acetate membrane. KS migrated as a single band with a mobility identical to that of standard KS and completely degraded after digestion with Table 1 Chemical composition of HNNC and HSNC Compound
HNNC
HSNC
Total hexosamine GalN/GlcN (molar ratio) Uronic acid Hexoses Collagena
10.6 þ 0.6 4.8 þ 0.21 9.1 þ 0.54 3.6 þ 0.14 216 þ 12
13.8 þ 0.7 3.0 þ 0.12 11.3 þ 0.62 5.8 þ 0.28 208 þ 11
Chemical analyses were carried out on isolated GAGs and papain digests from known amounts of tissues from each specimen of HNNC and HSNC. The results are expressed as mg per g wet weight of tissue. Number of samples analyzed: n = 12 for scoliotic cartilage and n = 5 for normal cartilage. All values represent the mean þ S.D. a Calculated in papain digests from hydroxyproline U11. GalN, galactosamine ; GlcN, glucosamine.
BBAGEN 25214 18-10-01
84
A.D. Theocharis et al. / Biochimica et Biophysica Acta 1528 (2001) 81^88
Fig. 1. Electrophoresis of isolated GAGs on cellulose acetate membranes using 0.1 M pyridine^0.47 M formic acid, pH 3.1, of (A) KS of HNNC and HSNC isolated and separated by combined chromatography (CPC-cellulose and ECTEOLA-cellulose) and (B) fractionated populations I and II from normal (N) and scoliotic (S) nasal cartilage, obtained by DEAE-Sephacel. Stds, standards of HA, DS, KS, and CS.
keratanase (Fig. 1A). The purity of the other GAG fractions obtained by combined chromatography (CPC-cellulose and DEAE-Sephacel) was also assessed by electrophoresis and hexosamine analysis. The electrophoresis of populations I and II obtained by DEAE-Sephacel from both tissues (Fig. 1B) revealed that populations I migrated as standard HA, whereas populations II appeared as a broad band located between the migration positions of DS and CS. Hexosamine analysis of populations I and II from both tissues showed that these populations contained only glucosamine and galactosamine, respectively, con¢rming their HA and DC/CS nature, respectively. The composition of GAGs in HNNC and HSNC, as it resulted from the quantitative determinations of the identi¢ed GAG fractions, is presented in Table 2. HNNC and HSNC contained mainly GalAGs (i.e. DS and CS) ac-
counting for 8.1 þ 0.38 and 9.0 þ 0.42 mg galactosamine/g w.wt. of tissue, corresponding to 80% and 72% of total GAGs, respectively. The molar ratio of GlcA/IdoA in GalAGs for both HNNC and HSNC was found to be 63 and 24, respectively, by HPLC analysis [26] and revealed that the GalAGs in scoliotic cartilage contained relatively higher proportion of iduronic acid (2.5 times) in comparison to those of normal nasal cartilage. The presence of iduronic acid in GalAGs indicated the occurrence of DS in both tissues, since the iduronic acid consists a of characteristic structural component of DS. According to the obtained results, HNNC contained also HA (0.7 mg/g w.wt.) and KS (1.3 mg/g w.wt.), corresponding to 7% and 13% of total GAGs, respectively. In HSNC the HA (1.5 mg/g w.wt.) and KS (1.9 mg/g w.wt.) contents were higher, corresponding to a signi¢cant increase of 114% and 46% (P90.001), respectively, as compared to HNNC. The GalAG content of HSNC was only slightly increased (11%, P90.01). 3.3. Sulfation pattern of GalAGs GalAGs from both tissues were digested with chondroitinase ABC and their disaccharide composition was determined by HPLC (Fig. 2). In HNNC the 4-sulfated disaccharides predominated (53%), whereas in HSNC the sulfation pattern is signi¢cantly altered. Particularly, the amounts and relative proportion of 6-sulfated disaccharides and non-sulfated disaccharides were signi¢cantly increased by 30% and 75%, respectively (P90.001). These results demonstrated the predomination of 6-sulfated disaccharides in GalAG chains of HSNC, which also contained increased proportions of non-sulfated disaccharides. 3.4. Characterization, quantitation and estimation of molecular size of GalAG chains Known amounts of samples of the puri¢ed GalAGs
Table 2 Composition of GAGs obtained from HNNC and HSNC Type of GAG
HNNC
Keratan sulfate Hyaluronan GalGAGs GlcA/IdoAa Chondroitin sulfateb Dermatan sulfatec
1.3 þ 0.04 0.7 þ 0.03 8.1 þ 0.30 63 7.4 þ 0.27 0.7 þ 0.02
HSNC (13) (7) (80) (73) (7)
1.9 þ 0.08 1.5 þ 0.05 9.0 þ 0.32 24 7.7 þ 0.30 1.3 þ 0.04
(16) (12) (72) (62) (10)
Quantitative determination of the identi¢ed GAG fractions from the combined chromatographies. The results are expressed as mg of hexosamine per g wet weight of tissue. All values represent the mean þ S.D. Percent values are given in parentheses. a Glucuronic acid (GlcA) to iduronic acid (IdoA) molar ratios in GalAGs. b CS was determined as the resistant c DS as the degradable material by chondroitinase B.
Fig. 2. Disaccharide composition of GalAG chains from HNNC and HSNC. GalAG chains were digested with chondroitinase ABC and their disaccharide composition was determined by HPLC analysis. The results are expressed as mg uronic acid/g wet weight of tissue. The values are the mean þ S.D. of three separate determinations. Percent values are given in parentheses.
BBAGEN 25214 18-10-01
A.D. Theocharis et al. / Biochimica et Biophysica Acta 1528 (2001) 81^88
85
Fig. 3. Gel chromatography of GalAGs isolated from normal cartilage (A) and scoliotic cartilage (B) on an analytical Sepharose CL-6B column (110U0.6 cm i.d.), before (b) and after (a) digestion with chondroitinase B. The column was eluted with 0.5 M sodium acetate, pH 7.0. Fractions of 0.9 ml were collected and analyzed for uronic acid. V0 and Vt indicate the void and total volume, respectively. (Insets) Secondary pro¢le of dermatan sulfate, obtained from the positive di¡erences of uronic acid in each fraction of the sample before and after digestion with chondroitinase B. For details see text.
from both HNNC and HSNC were subjected to exhaustive digestion with chondroitinase B, which cleaves the L(1C4) galactosaminyl^iduronic acid bonds. The extent of the digestion was estimated by measuring the A232 , from which the nmoles of liberated unsaturated oligosaccharides were calculated and found to correspond to 95^ 98% of the IdoA content of the sample analyzed by the HPLC method of Karamanos et al. [18], suggesting that the digestion was completed. The GalAGs from both HNNC and HSNC were chromatographed on an analytical column of Sepharose CL-6B before and after digestion with chondroitinase B (Fig. 3A,B). The enzyme degraded appreciable amounts of the GalAGs, con¢rming the presence of DS in the samples of both tissues, whereas the majority of the sample, being CS, remained unaltered.
The chondroitinase B resistant material which represented CS was eluted as a single peak with Kav = 0.69 in both tissues, corresponding to a Mr of 17 000. The secondary pro¢le of DS from HNNC and HSNC was constructed (Fig. 3A and B insets, respectively) as described in Section 2. From these secondary pro¢les the relative amounts and Mr of dermatan sulfate chains were estimated. Thus, the amounts of DS and CS in both tissues were calculated from the relative proportions of degradable and resistant material by chondroitinase B (Table 2). These results showed that the amount of DS in HSNC (1.3 mg/g w.wt.) was increased about 86% (P90.001) as compared to HNNC (0.7 mg/g w.wt.), while the amounts of CS remained almost unchanged (no signi¢cant di¡erence). The chains of DS in HSNC were of larger molecular
BBAGEN 25214 18-10-01
86
A.D. Theocharis et al. / Biochimica et Biophysica Acta 1528 (2001) 81^88
Fig. 4. Gel chromatography of GalAGs isolated from normal cartilage (A) and scoliotic cartilage (B), on Bio Gel P-10 before (b) and after (a) digestion with chondroitinase B. The column was eluted with 1.0 M ammonium acetate, pH 7.0, and fractions of 1 ml were collected and their uronic acid was measured by the borate^carbazole reaction. The individual peaks (a, b, c etc.) of the oligosaccharides were pooled as designated by the horizontal bars and their A232 measured. For details see text. (Insets) Distribution of hexuronic acids in each v-oligosaccharide pool produced by the action of chondroitinase B on galactosaminoglycans isolated from human nasal cartilage. Iduronic acid content was calculated from A232 of the chondroitinase B digest. Filled bars, uronic acid; open bars, iduronic acid.
size (Kav = 0.5 corresponding to a Mr of 33 000) as compared to those of HNNC (Kav = 0.58 corresponding to a Mr of 22 000). Aliquots from the same samples, i.e. before and after digestion with chondroitinase B, were also chromatographed on an analytical column of Bio Gel P-10 and the elution pro¢les are shown in Fig. 4A,B. All of the degradation products were retarded by the gel, since no absorbance at 232 nm was measured in the V0 fractions. The relative contents of CS and DS in the samples were estimated from the amounts of uronic acid in the V0 fractions and in the various v-oligosaccharide pools. The proportions of CS and DS were found to be 91% and 9% for HNNC and 85% and 15% for HSNC, respectively, in accordance to those estimated by the secondary pro¢le. The
IdoA content of the degraded material (i.e. DS) was determined via the double bonds from their absorbance at 232 nm, using E232 = 5500 M31 cm31 , with as internal standard the reference v-disaccharides. It was found to be 18% and 28% in HNNC and HSNC, respectively, of the total DS uronic acid. In order to provide enough digested material and quantitate each pool of v-oligosaccharides derived from the digestion of DS with chondroitinase B, fractions from four identical chromatographies were pooled. In each pool of v-oligosaccharides the total uronic acid and IdoA were determined. Total uronic acid was measured by the borate^carbazole assay. IdoA and GlcA in intact dermatan sulfate (as GalAG mixture) were quantitated by HPLC. IdoA in DS-oligosaccharides derived after diges-
BBAGEN 25214 18-10-01
A.D. Theocharis et al. / Biochimica et Biophysica Acta 1528 (2001) 81^88
tion with chondroitinase B was estimated from the absorbance at 232 nm. It should be noted that the sum of nmoles of IdoA, measured in the oligosaccharide pools from the double bond content, was equal to the nmoles of the respective sugar measured in intact DS chains by HPLC. The content of total uronic acid and IdoA was expressed as percentage of the respective total amounts in the degradable material. It was found that the action of chondroitinase B on DS chains from HNNC produced eight types of v-oligosaccharides, i.e. v-20-saccharides down to v-6-saccharides, with more predominant the v8-saccharides (Fig. 4A inset). Furthermore, the chondroitinase B on DS chains of HSNC produced also eight types of v-oligosaccharides (v-16-saccharides down to v-2-saccharides). Neither v-20-saccharides nor v-18-saccharides were produced in this case (Fig. 4B inset). The absence of v-disaccharides suggested the absence of IdoA-containing clusters in HNNC. 4. Discussion The primary purpose of this study was to determine, for the ¢rst time, the amount, type and structure of GAGs in HNNC and to compare them with those from HSNC. The overall content of GAGs in the HSNC increased about 30% in comparison to that from HNNC, while the collagen content in both tissues remained constant. GAGs from human nasal cartilage showed qualitative similarities in comparison with those from the sheep nasal cartilage [7^10]. On the contrary, signi¢cant quantitative variations were observed, since HNNC and HSNC contained lower amounts of GAG than sheep nasal cartilage. The concentration of GAGs and collagen of HNNC reported in the present investigation was in agreement with those demonstrated by other investigators for other non-weight-bearing human cartilages such as tracheal [6] and tracheobronchial [27] cartilage and articular [1^3] cartilage as well. Both HNNC and HSNC contained HA and KS as was expected. HA and KS were markedly increased, in HSNC 114% and 46%, respectively. GalAGs predominated in both tissues and were only slightly increased in HSNC. CS represented the major proportion of GalAGs, whereas a lower but signi¢cant proportion of DS was also present in both tissues. The presence of DS in human nasal cartilage is demonstrated for the ¢rst time and was estimated by its susceptibility to degradation with chondroitinase B and further analytical gel chromatography. Although the absolute amount of CS remained almost constant in both tissues, that of DS was signi¢cantly increased in HSNC. The documentation of the presence of DS in hyaline cartilage from various sources has interested several investigators [11,26,28^32]. So far, only articular cartilage was found to contain DS in the form of biglycan and decorin, of about 40% IdoA content [28]. The data reported in the present study established the presence of small amounts of
87
DS of low IdoA content (18^28%) in nasal cartilage. These results were in agreement with those of a recent study from our laboratory [26] and of a previous study [29], which reported that a small amount of the side chains of biglycan isolated from bovine nasal cartilage was sensitive to chondroitinase B digestion. The large di¡erences in IdoA content between the nasal and articular cartilage DS may be related to the di¡erent functions of the respective tissues. The extent and the distribution of uronosyl epimerization in DS chains seem to be tissue speci¢c. DS derived from bovine mucosa appears to contain 60% [26] of its IdoA in long IdoA-GalNAc repeats, the remaining being rather randomly distributed. Similarly, the majority of IdoA in DS from pig skin also appears in IdoA-GalNAc repeats [33,34]. DS of both HNNC and HSNC seemed to have almost all of its IdoA randomly distributed, since all of the IdoA in the sample was recovered in oligosaccharides ranging from di- to eicosasaccharides long. Similar results were also obtained for DS of sheep nasal cartilage [26]. The random distribution of IdoA in DS chains of nasal cartilages might be due to the expression of the C5-epimerase activity at long intervals in this tissues. The di¡erences in IdoA distribution may be elucidated by studying the turnover and kinetic properties of the C5-epimerase in this tissues. GalAGs derived from HSNC also exhibited signi¢cant structural alterations concerning their disaccharide composition and molecular size. GalAGs showed an altered sulfation pattern, which was characterized by a signi¢cant increase of non-sulfated and 6-sulfated disaccharides, with a simultaneous decrease of 4-sulfated disaccharides. In HSNC a signi¢cant increase in the length of DS chains was also observed. Taking into account the increase in the chain size and the increase in the amount of DS in HSNC, we suggest that the increase in DS content is mainly due to the increase in the chain length. Our results, which concern the sulfation patterns of GalAGs and the molecular size of DS, revealed altered biosynthesis of these GAG types by HSNC chondrocytes and suggested that scoliosis of the nasal septum cartilage is related to quantitative and structural modi¢cations at the GAG level.
References [1] D. Heinegard, A. Oldberg, FASEB J. 3 (1989) 2042^2051. [2] T.E. Hardingham, A.J. Fosang, J. Dudhia, Eur. J. Clin. Chem. Clin. Biochem. 32 (1994) 249^257. [3] S.L. Carney, H. Muir, Physiol. Rev. 68 (1988) 858^910. [4] L. Kjellen, U. Lindahl, Annu. Rev. Biochem. 60 (1991) 443^475. [5] R.V. Iozzo, Annu. Rev. Biochem. 67 (1998) 609^652. [6] C.R. Roberts, P.D. Pare, Am. J. Physiol. 261 (1991) L92^L101. [7] D.A. Theocharis, Int. J. Biochem. 17 (1985) 155^160. [8] D.A. Theocharis, C.P. Tsiganos, Int. J. Biochem. 17 (1985) 479^484. [9] D.A. Theocharis, D.L. Kalpaxis, C.P. Tsiganos, Biochim. Biophys. Acta 841 (1985) 131^134.
BBAGEN 25214 18-10-01
88
A.D. Theocharis et al. / Biochimica et Biophysica Acta 1528 (2001) 81^88
[10] D.D. Dziewiatkowski, J.L.A. Valley, A.G. Beaudoin, Connect. Tissue Res. 19 (1989) 103^120. [11] D. Heinegard, M. Paulsson, S. Inerot, S. Carlstrom, Biochem. J. 197 (1981) 355^366. [12] J.J. Ballenger, in: Diseases of the Nose, Throat and Ear, 11th edn., Lea and Febinger, Philadelphia, PA, 1969. [13] G.W. Bruyn, in: F. Cli¡ord Rose (Ed.), Handbook of Clinical Neurology, Vol. 4 (48), Headache, Elsevier Science Publishers, Amsterdam, 1986, pp. 475^482. [14] T. Bitter, H. Muir, Anal. Biochem. 4 (1962) 330^334. [15] M.L. Hayes, F.J. Castellino, J. Biol. Chem. 246 (1971) 430^435. [16] T. Scott, E.H. Melvin, Anal. Chem. 25 (1953) 1656^1661. [17] N.K. Karamanos, A. Syrokou, P. Vanky, M. Nurminen, A. Hjerpe, Anal. Biochem. 221 (1994) 189^199. [18] N.K. Karamanos, A. Hjerpe, T. Tsegenidis, B. Engfeldt, C.A. Antonopoulos, Anal. Biochem. 172 (1988) 410^419. [19] J.F. Woessner, Arch. Biochem. Biophys. 93 (1961) 440^446. [20] D.M. Carlson, J. Biol. Chem. 243 (1968) 616^626. [21] A.D. Theocharis, I. Tsolakis, T. Tsegenidis, N.K. Karamanos, Atherosclerosis 145 (1999) 359^368. [22] C.A. Antonopoulos, S. Gardell, J.A. Szirmai, E.P. De Tyssonsk, Biochim. Biophys. Acta 83 (1964) 1^19.
[23] A. Anseth, C.A. Antonopoulos, A. Bjelle, L.-A. Fransson, Biochim. Biophys. Acta 215 (1970) 522^526. [24] M.E. Tsara, N. Papageorgacopoulou, D.D. Karavias, D.A. Theocharis, Anticancer Res. 15 (1995) 2107^2112. [25] A. Wasteson, J. Chromatogr. 59 (1971) 87^97. [26] D.A. Theocharis, N. Papageorgacopoulou, D.H. Vynios, S.Th. Anagnostides, C.P. Tsiganos, J. Chromatogr. B Biomed. Sci. Appl. 754 (2001) 297^309. [27] R.M. Mason, F.S. Wusteman, Biochem. J. 120 (1970) 777^785. [28] H.U. Choi, T.L. Johnson, S. Pal, L.-H. Tang, L. Rosenberg, P.J. Neame, J. Biol. Chem. 264 (1989) 2876^2884. [29] F. Cheng, D. Heinega®rd, A. Malmstro«m, A. Schmidtchen, K. Yoshiî . Fransson, Glycobiology 4 (1994) 685^696. da, L.-A [30] O. de Sampaio, M.T. Bayliss, T.E. Hardingham, H. Muir, Biochem. J. 254 (1988) 757^764. [31] P.J. Roughley, R.J. White, Biochem. J. 262 (1989) 823^827. [32] P.J. Roughley, R.J. White, J. Orthop. Res. 10 (1992) 631^637. [33] N.K. Karamanos, P. Vanky, A. Syrokou, A. Hjerpe, Anal. Biochem. 225 (1995) 220^230. [34] A. Malmstro«m, J. Biol. Chem. 259 (1984) 161^165.
BBAGEN 25214 18-10-01
www.bba-direct.com
Characterization of glycosaminoglycans from human normal and scoliotic nasal cartilage with particular reference to dermatan sulfate Achilleas D. Theocharis a , Marina E. Tsara b , Nikoletta Papageorgakopoulou a , Demitrios H. Vynios a , Dimitrios A. Theocharis b; * a
Laboratory of Biochemistry, Department of Chemistry, School of Medicine, University of Patras, 261 10 Patras, Greece b Laboratory of Biological Chemistry, School of Medicine, University of Patras, 261 10 Patras, Greece Received 18 April 2001; received in revised form 27 June 2001; accepted 28 June 2001
Abstract The composition and the distribution of glycosaminoglycans (GAGs) present in normal human nasal cartilage (HNNC), were examined and compared with those in human scoliotic nasal cartilage (HSNC). In both tissues, hyaluronan (HA), keratan sulfate (KS) and the galactosaminoglycans (GalAGs) ^ chondroitin sulfate (CS) and dermatan sulfate (DS) ^ were identified. The overall GAG content in HSNC was approx. 30% higher than the HNNC. Particularly, a 114% increase in HA, and 46% and 86% in KS and DS, respectively, was recorded. CS was the main type of GAG in both tissues with no significant compositional difference. GalAG chains in HSNC exhibited an altered disaccharide composition which was associated with significant increases of non-sulfated and 6-sulfated disaccharides. DS, which was identified and quantitated for the first time in HNNC and HSNC, contained low amounts of iduronic acid (IdoA), 18% and 28% respectively. In contrast to other tissues, where IdoA residues are organized in long IdoA rich repeats, the IdoA residues of DS in human nasal cartilage seemed to be randomly distributed along the chain. DS chains in HSNC were of larger average molecular size than those from HNNC. These results clearly indicate the GAG content and pattern in both HNNC and HSNC and demonstrate that scoliosis of nasal septum cartilage is related to quantitative and structural modifications at the GAG level. ß 2001 Elsevier Science B.V. All rights reserved. Keywords : Scoliosis; Nasal cartilage ; Glycosaminoglycan ; Proteoglycan
1. Introduction Hyaline cartilage is a highly specialized connective tissue with biomechanical, developmental and other biological functions. Cartilages from several anatomical sites exhibit di¡erent functional properties and are characterized by di¡erent macromolecular composition [1,2]. Collagen and glycosaminoglycans (GAGs)/proteoglycans (PGs) are the major components of hyaline cartilage. PGs are highly anionic macromolecules consisting of a protein core onto which sulfated GAG chains are covalently bound. In contrast, hyaluronan (HA) is a high molecular weight nonsulfated GAG, which does not occur as a PG but in free form. The common GAGs include the galactosaminoglycans (GalAGs) (chondroitin sulfate (CS) and dermatan sulfate (DS)) and glucosaminoglycans (HA, heparan sulfate, heparin and keratan sulfate (KS)) [3,4]. The GAGs are intrinsic constituents and, particularly because of their
* Corresponding author. Fax: +30-61-99-76-90.
complex polyanionic nature, have been implicated in a number of biological processes and functions [5]. Although GAGs from various types of human hyaline cartilage such as articular [2], costal, tracheal and bronchial [6] as well as those from sheep [7^10] and bovine [1,11] have been extensively studied, no information has been reported about the GAGs found in human nasal cartilage. Scoliosis of the nasal septum cartilage is the most frequent deformity of the nose in man. The deformity of the nasal septum cartilage causes di¤culty in breathing and is implicated in a number of episodes of rhinitis, pharyngitis, otitis as well as in the sphenopalatine neuralgia which is known as Sluder's syndrome [12,13]. The present study was undertaken to determine the chemical composition, the GAG content and distribution of human normal septum cartilage and to compare them with those of scoliotic nasal septum cartilage of the same age. This study apparently constitutes the ¢rst study of GAG content, composition and structure of human nasal cartilage and provides a baseline for future studies in this ¢eld.
0304-4165 / 01 / $ ^ see front matter ß 2001 Elsevier Science B.V. All rights reserved. PII: S 0 3 0 4 - 4 1 6 5 ( 0 1 ) 0 0 1 7 3 - 8
BBAGEN 25214 18-10-01
82
A.D. Theocharis et al. / Biochimica et Biophysica Acta 1528 (2001) 81^88
exhibited the typical characteristics of hyaline cartilage that were observed in the normal specimens.
2. Materials and methods 2.1. Chemicals Sepharose CL-6B and DEAE-Sephacel were obtained from Pharmacia (Uppsala, Sweden) and Bio Gel P-10 from Bio-Rad. Papain twice crystallized (EC 3.4.22.2), chondroitinase ABC (EC 4.2.2.4, Proteus vulgaris), chondroitinase B (Flavobacterium heparinum), keratanase (EC 3.2.1.103, Pseudomonas species), GAG standards and standard preparations of chondroitin disaccharides, vDi-4S, vDi-6S and vDi-0S were from Sigma (St. Louis, MO, USA). ECTEOLA-cellulose (ET-11) and cellulose powder (CF-11) were purchased from Whatman (W.pR. Balston, UK). All other chemicals used were of the best grade commercially available. 2.2. Analytical methods Uronic acid was determined by a modi¢ed carbazole reaction [14]. Analysis for glucosamine and galactosamine was performed on a Beckman amino acid analyzer [15]. Hexoses were measured by the anthrone reaction, using galactose as standard [16]. The amounts of unsaturated disaccharides released by the chondroitinase ABC were determined by high performance liquid chromatography (HPLC) [17]. The quanti¢cation of each v-oligosaccharide pool produced by the action of chondroitinase B and isolated by gel chromatography, besides hexuronic acid determination, was based on their absorbance at 232 nm (E232 = 5500 M31 cm31 ) owed to the C-4 double bond of the uronic acid at the non-reducing end. The accuracy of the assay was tested using internal standards of v-disaccharides. Glucuronic acid (GlcA) and iduronic acid (IdoA) were determined by the method of Karamanos et al. [18]. In brief, the galactosaminoglycans were reacted with 1ethyl-3-(3-dimethylaminopropyl)carbodiimide and then subjected to reduction of their hexuronic acid to the respective alditols. Finally, the samples were hydrolyzed and the monosaccharides were liberated, the alditols were puri¢ed using ion exchange chromatography, and transformed to the respective perbenzoyl derivatives, which were separated and quantitated by HPLC. Collagen content was calculated via determination of hydroxyproline [19] as described elsewhere [7] in papain digests of the tissues. 2.3. Tissue source ^ patients Human normal nasal septum cartilages (n = 5, 25^35 years old) were obtained by autopsy and frozen at 320³C. Specimens (n = 12, 25^35 years old) from patients with scoliosis of the nasal septum cartilage were taken directly from operating rooms as surgical discard tissues. The scoliotic specimens were free from any ¢brous structure and on the basis of biochemical and histological data
2.4. Isolation of GAGs Intrabatch di¡erences in GAG content were examined in specimens of normal human nasal cartilage (HNNC ; n = 5) and human scoliotic nasal cartilage (HSNC ; n = 12) which have been separately isolated and analyzed for their GAG content. Particularly, known amounts of wet weight of both tissues were exhaustively digested with papain and then the GAGs were precipitated by addition of 5 vols. of ethanol. Subsequently, GAG chains were released from their protein core/peptide fragments by treatment with 0.05 M NaOH/1 M sodium borohydride at 45³C for 48 h under vacuum [20]. The solutions were brought to pH 5.0 with glacial acetic acid and then precipitated with 5 vols. of ethanol. The GAG content was estimated in each sample by measuring the uronic acid, hexoses, glucosamine and galactosamine in the precipitates. 2.5. Fractionation and characterization of GAGs According to pilot experiments (see Section 3) all normal specimen isolated GAGs were mixed in a group, and the scoliotic specimen isolated GAGs were mixed in a second group. Fractionation of various types of GAGs was performed as described elsewhere [7,9,21]. Brie£y, the isolated GAGs from each group were chromatographed on cetylpyridinium chloride (CPC)-cellulose column [22] and eluted with 2 bed vols. of 1% CPC/0.005 M Na2 SO4 , and 4 M NaCl/0.05% CPC. The 1% CPC fractions (containing KS) were freed from CPC by extraction with isoamyl alcohol, desalted and reduced in volume by dialysis through Amicon UM-2 membrane and then KS was puri¢ed further by chromatography on ECTEOLAcellulose [23]. The column was eluted with 2 bed vols. of 0.02 M HCl and 2 M NaCl. KS was eluted with 2 M NaCl. The 4 M NaCl/0.05% CPC fractions of CPC-cellulose column (containing the other types of GAGs) were desalted and reduced in volume by dialysis and then fractionated on a DEAE-Sephacel column [21]. The column was eluted stepwise with 3 vols. of 0.1 M NaCl and a linear gradient ranging from 0.1 M to 1.2 M NaCl (10 vols.). This ion exchange chromatography separated the retarded uronic acid positive material into two populations in each tissue. Populations I and II, which were eluted with 0.35 M and 0.7 M NaCl respectively, were pooled, quanti¢ed in terms of uronic acid and concentrated on PM-10 membranes. The recovery of hexosamine from the combined chromatographies was 94^97%. Identi¢cation and quantitation of each GAG fraction was assessed by cellulose acetate electrophoresis [24], hexosamine analysis and speci¢c enzymatic treatment. Disaccharide composition and sulfation pattern of GalAGs were esti-
BBAGEN 25214 18-10-01
A.D. Theocharis et al. / Biochimica et Biophysica Acta 1528 (2001) 81^88
mated following digestion with chondroitinase ABC and analyses of the digests by HPLC. 2.6. Enzymatic degradation Digestion with papain was performed at 60³C for 18 h using twice crystallized papain (0.2 mg of papain/g wet weight of tissue) in 0.1 M Tris^HCl, pH 7.2, containing 0.01 M disodium EDTA and 0.005 M cysteine-HCl. For chondroitinase ABC digestion, solutions of GAGs (1 mg of uronic acid/ml) in 0.1 M Tris-acetate, pH 7.3, were incubated at 37³C for 24 h in the presence of 0.5 units of chondroitinase ABC per ml. Digestion with chondroitinase B was performed in 0.05 M Tris^HCl, pH 8.0, containing 0.05% (w/v) bovine serum albumin, at 37³C for 24 h using 0.3 units enzyme per mg of uronic acid. For keratanase digestion, samples containing KS were incubated at 37³C for 24 h in 0.1 M Tris^HCl, pH 7.4, containing 0.1 U of keratanase per Wmole of galactose. All digestions were terminated by heating the solutions at 100³C for 3 min. 2.7. Gel chromatography Gel chromatography of GAG chains was performed on analytical (110U0.6 cm) Sepharose CL-6B column, equilibrated and eluted with 0.5 M sodium acetate, pH 7.0, before and after treatment with speci¢c degradative enzymes. The average molecular weight (Mr ) of the GAG chains was estimated according to Wasteson [25] using CS molecular weight standards for calibrating the Sepharose CL-6B column as previously described [24]. In particular, the Mr of the DS chains was calculated by the construction of the secondary pro¢le as described in a recent study [26]. In brief, the isolated GalAGs were digested with chondroitinase B and the pro¢le, upon gel chromatography on Sepharose CL-6B, was compared to that of the initial undigested sample. The absolute amounts or percentage of uronic acid measured in each fraction of the chromatographies before and after treatment with chondroitinase B were subtracted, fraction by fraction, and the di¡erences were obtained. The positive di¡erences were used to construct the secondary pro¢le, which represented the uronic acid content of the fractions removed after the enzymatic digestion, i.e. the elution pro¢le of intact DS chains, and thus the size distribution of DS was obtained and its Mr calculated. The size of DS-derived v-oligosaccharides produced by chondroitinase B was determined by chromatography on Bio Gel P-10 (160U0.6 cm) in 1.0 M ammonium acetate pH 7.0. The column was calibrated using standard preparations of v-oligosaccharides [26]. The void (V0 ) and total volumes (Vt ) of each column were determined with sheep nasal cartilage proteoglycan aggregates and tritiated water, respectively.
83
2.8. Statistical analysis Statistical signi¢cance was evaluated by t-test using Microcall origin software (version 3.2). 3. Results 3.1. Chemical composition of human nasal cartilage To examine for intrabatch and interbatch di¡erences in total GAG and collagen contents chemical analyses were carried out on isolated total GAGs from known amounts of tissue from each specimen (normal and scoliotic cartilage) as described in Section 2. The results of the chemical composition (Table 1) indicated that no signi¢cant intrabatch di¡erences were found in the overall GAG contents (expressed either as uronic acid or hexosamine or hexoses) and collagen contents (expressed as hydroxyproline), between the various HNNC (n = 5) and HSNC (n = 12) examined. For these reasons, further analyses were performed on pooled scoliotic and normal cartilages. The overall GAG content in scoliotic cartilage (13.8 þ 0.7 mg of hexosamine/g wet weight (w.wt.) of tissue) was found to be increased (P90.001), about 30%, as compared to normal cartilage (10.6 þ 0.6 mg/g w.wt. of tissue). The marked changes in the molar ratio of galactosamine/glucosamine indicate signi¢cant di¡erences in the GAG composition. The collagen content in the various normal and scoliotic cartilages (216 þ 12 and 208 þ 11 mg/g w.wt. of tissue, respectively) was similar, showing no signi¢cant intrabatch and interbatch di¡erences (P90.001). 3.2. Fractionation and characterization of GAGs The purity of isolated and separated KS from both HNNC and HSNC by combined chromatography (CPCcellulose and ECTEOLA-cellulose), was assessed by electrophoresis on cellulose acetate membrane. KS migrated as a single band with a mobility identical to that of standard KS and completely degraded after digestion with Table 1 Chemical composition of HNNC and HSNC Compound
HNNC
HSNC
Total hexosamine GalN/GlcN (molar ratio) Uronic acid Hexoses Collagena
10.6 þ 0.6 4.8 þ 0.21 9.1 þ 0.54 3.6 þ 0.14 216 þ 12
13.8 þ 0.7 3.0 þ 0.12 11.3 þ 0.62 5.8 þ 0.28 208 þ 11
Chemical analyses were carried out on isolated GAGs and papain digests from known amounts of tissues from each specimen of HNNC and HSNC. The results are expressed as mg per g wet weight of tissue. Number of samples analyzed: n = 12 for scoliotic cartilage and n = 5 for normal cartilage. All values represent the mean þ S.D. a Calculated in papain digests from hydroxyproline U11. GalN, galactosamine ; GlcN, glucosamine.
BBAGEN 25214 18-10-01
84
A.D. Theocharis et al. / Biochimica et Biophysica Acta 1528 (2001) 81^88
Fig. 1. Electrophoresis of isolated GAGs on cellulose acetate membranes using 0.1 M pyridine^0.47 M formic acid, pH 3.1, of (A) KS of HNNC and HSNC isolated and separated by combined chromatography (CPC-cellulose and ECTEOLA-cellulose) and (B) fractionated populations I and II from normal (N) and scoliotic (S) nasal cartilage, obtained by DEAE-Sephacel. Stds, standards of HA, DS, KS, and CS.
keratanase (Fig. 1A). The purity of the other GAG fractions obtained by combined chromatography (CPC-cellulose and DEAE-Sephacel) was also assessed by electrophoresis and hexosamine analysis. The electrophoresis of populations I and II obtained by DEAE-Sephacel from both tissues (Fig. 1B) revealed that populations I migrated as standard HA, whereas populations II appeared as a broad band located between the migration positions of DS and CS. Hexosamine analysis of populations I and II from both tissues showed that these populations contained only glucosamine and galactosamine, respectively, con¢rming their HA and DC/CS nature, respectively. The composition of GAGs in HNNC and HSNC, as it resulted from the quantitative determinations of the identi¢ed GAG fractions, is presented in Table 2. HNNC and HSNC contained mainly GalAGs (i.e. DS and CS) ac-
counting for 8.1 þ 0.38 and 9.0 þ 0.42 mg galactosamine/g w.wt. of tissue, corresponding to 80% and 72% of total GAGs, respectively. The molar ratio of GlcA/IdoA in GalAGs for both HNNC and HSNC was found to be 63 and 24, respectively, by HPLC analysis [26] and revealed that the GalAGs in scoliotic cartilage contained relatively higher proportion of iduronic acid (2.5 times) in comparison to those of normal nasal cartilage. The presence of iduronic acid in GalAGs indicated the occurrence of DS in both tissues, since the iduronic acid consists a of characteristic structural component of DS. According to the obtained results, HNNC contained also HA (0.7 mg/g w.wt.) and KS (1.3 mg/g w.wt.), corresponding to 7% and 13% of total GAGs, respectively. In HSNC the HA (1.5 mg/g w.wt.) and KS (1.9 mg/g w.wt.) contents were higher, corresponding to a signi¢cant increase of 114% and 46% (P90.001), respectively, as compared to HNNC. The GalAG content of HSNC was only slightly increased (11%, P90.01). 3.3. Sulfation pattern of GalAGs GalAGs from both tissues were digested with chondroitinase ABC and their disaccharide composition was determined by HPLC (Fig. 2). In HNNC the 4-sulfated disaccharides predominated (53%), whereas in HSNC the sulfation pattern is signi¢cantly altered. Particularly, the amounts and relative proportion of 6-sulfated disaccharides and non-sulfated disaccharides were signi¢cantly increased by 30% and 75%, respectively (P90.001). These results demonstrated the predomination of 6-sulfated disaccharides in GalAG chains of HSNC, which also contained increased proportions of non-sulfated disaccharides. 3.4. Characterization, quantitation and estimation of molecular size of GalAG chains Known amounts of samples of the puri¢ed GalAGs
Table 2 Composition of GAGs obtained from HNNC and HSNC Type of GAG
HNNC
Keratan sulfate Hyaluronan GalGAGs GlcA/IdoAa Chondroitin sulfateb Dermatan sulfatec
1.3 þ 0.04 0.7 þ 0.03 8.1 þ 0.30 63 7.4 þ 0.27 0.7 þ 0.02
HSNC (13) (7) (80) (73) (7)
1.9 þ 0.08 1.5 þ 0.05 9.0 þ 0.32 24 7.7 þ 0.30 1.3 þ 0.04
(16) (12) (72) (62) (10)
Quantitative determination of the identi¢ed GAG fractions from the combined chromatographies. The results are expressed as mg of hexosamine per g wet weight of tissue. All values represent the mean þ S.D. Percent values are given in parentheses. a Glucuronic acid (GlcA) to iduronic acid (IdoA) molar ratios in GalAGs. b CS was determined as the resistant c DS as the degradable material by chondroitinase B.
Fig. 2. Disaccharide composition of GalAG chains from HNNC and HSNC. GalAG chains were digested with chondroitinase ABC and their disaccharide composition was determined by HPLC analysis. The results are expressed as mg uronic acid/g wet weight of tissue. The values are the mean þ S.D. of three separate determinations. Percent values are given in parentheses.
BBAGEN 25214 18-10-01
A.D. Theocharis et al. / Biochimica et Biophysica Acta 1528 (2001) 81^88
85
Fig. 3. Gel chromatography of GalAGs isolated from normal cartilage (A) and scoliotic cartilage (B) on an analytical Sepharose CL-6B column (110U0.6 cm i.d.), before (b) and after (a) digestion with chondroitinase B. The column was eluted with 0.5 M sodium acetate, pH 7.0. Fractions of 0.9 ml were collected and analyzed for uronic acid. V0 and Vt indicate the void and total volume, respectively. (Insets) Secondary pro¢le of dermatan sulfate, obtained from the positive di¡erences of uronic acid in each fraction of the sample before and after digestion with chondroitinase B. For details see text.
from both HNNC and HSNC were subjected to exhaustive digestion with chondroitinase B, which cleaves the L(1C4) galactosaminyl^iduronic acid bonds. The extent of the digestion was estimated by measuring the A232 , from which the nmoles of liberated unsaturated oligosaccharides were calculated and found to correspond to 95^ 98% of the IdoA content of the sample analyzed by the HPLC method of Karamanos et al. [18], suggesting that the digestion was completed. The GalAGs from both HNNC and HSNC were chromatographed on an analytical column of Sepharose CL-6B before and after digestion with chondroitinase B (Fig. 3A,B). The enzyme degraded appreciable amounts of the GalAGs, con¢rming the presence of DS in the samples of both tissues, whereas the majority of the sample, being CS, remained unaltered.
The chondroitinase B resistant material which represented CS was eluted as a single peak with Kav = 0.69 in both tissues, corresponding to a Mr of 17 000. The secondary pro¢le of DS from HNNC and HSNC was constructed (Fig. 3A and B insets, respectively) as described in Section 2. From these secondary pro¢les the relative amounts and Mr of dermatan sulfate chains were estimated. Thus, the amounts of DS and CS in both tissues were calculated from the relative proportions of degradable and resistant material by chondroitinase B (Table 2). These results showed that the amount of DS in HSNC (1.3 mg/g w.wt.) was increased about 86% (P90.001) as compared to HNNC (0.7 mg/g w.wt.), while the amounts of CS remained almost unchanged (no signi¢cant di¡erence). The chains of DS in HSNC were of larger molecular
BBAGEN 25214 18-10-01
86
A.D. Theocharis et al. / Biochimica et Biophysica Acta 1528 (2001) 81^88
Fig. 4. Gel chromatography of GalAGs isolated from normal cartilage (A) and scoliotic cartilage (B), on Bio Gel P-10 before (b) and after (a) digestion with chondroitinase B. The column was eluted with 1.0 M ammonium acetate, pH 7.0, and fractions of 1 ml were collected and their uronic acid was measured by the borate^carbazole reaction. The individual peaks (a, b, c etc.) of the oligosaccharides were pooled as designated by the horizontal bars and their A232 measured. For details see text. (Insets) Distribution of hexuronic acids in each v-oligosaccharide pool produced by the action of chondroitinase B on galactosaminoglycans isolated from human nasal cartilage. Iduronic acid content was calculated from A232 of the chondroitinase B digest. Filled bars, uronic acid; open bars, iduronic acid.
size (Kav = 0.5 corresponding to a Mr of 33 000) as compared to those of HNNC (Kav = 0.58 corresponding to a Mr of 22 000). Aliquots from the same samples, i.e. before and after digestion with chondroitinase B, were also chromatographed on an analytical column of Bio Gel P-10 and the elution pro¢les are shown in Fig. 4A,B. All of the degradation products were retarded by the gel, since no absorbance at 232 nm was measured in the V0 fractions. The relative contents of CS and DS in the samples were estimated from the amounts of uronic acid in the V0 fractions and in the various v-oligosaccharide pools. The proportions of CS and DS were found to be 91% and 9% for HNNC and 85% and 15% for HSNC, respectively, in accordance to those estimated by the secondary pro¢le. The
IdoA content of the degraded material (i.e. DS) was determined via the double bonds from their absorbance at 232 nm, using E232 = 5500 M31 cm31 , with as internal standard the reference v-disaccharides. It was found to be 18% and 28% in HNNC and HSNC, respectively, of the total DS uronic acid. In order to provide enough digested material and quantitate each pool of v-oligosaccharides derived from the digestion of DS with chondroitinase B, fractions from four identical chromatographies were pooled. In each pool of v-oligosaccharides the total uronic acid and IdoA were determined. Total uronic acid was measured by the borate^carbazole assay. IdoA and GlcA in intact dermatan sulfate (as GalAG mixture) were quantitated by HPLC. IdoA in DS-oligosaccharides derived after diges-
BBAGEN 25214 18-10-01
A.D. Theocharis et al. / Biochimica et Biophysica Acta 1528 (2001) 81^88
tion with chondroitinase B was estimated from the absorbance at 232 nm. It should be noted that the sum of nmoles of IdoA, measured in the oligosaccharide pools from the double bond content, was equal to the nmoles of the respective sugar measured in intact DS chains by HPLC. The content of total uronic acid and IdoA was expressed as percentage of the respective total amounts in the degradable material. It was found that the action of chondroitinase B on DS chains from HNNC produced eight types of v-oligosaccharides, i.e. v-20-saccharides down to v-6-saccharides, with more predominant the v8-saccharides (Fig. 4A inset). Furthermore, the chondroitinase B on DS chains of HSNC produced also eight types of v-oligosaccharides (v-16-saccharides down to v-2-saccharides). Neither v-20-saccharides nor v-18-saccharides were produced in this case (Fig. 4B inset). The absence of v-disaccharides suggested the absence of IdoA-containing clusters in HNNC. 4. Discussion The primary purpose of this study was to determine, for the ¢rst time, the amount, type and structure of GAGs in HNNC and to compare them with those from HSNC. The overall content of GAGs in the HSNC increased about 30% in comparison to that from HNNC, while the collagen content in both tissues remained constant. GAGs from human nasal cartilage showed qualitative similarities in comparison with those from the sheep nasal cartilage [7^10]. On the contrary, signi¢cant quantitative variations were observed, since HNNC and HSNC contained lower amounts of GAG than sheep nasal cartilage. The concentration of GAGs and collagen of HNNC reported in the present investigation was in agreement with those demonstrated by other investigators for other non-weight-bearing human cartilages such as tracheal [6] and tracheobronchial [27] cartilage and articular [1^3] cartilage as well. Both HNNC and HSNC contained HA and KS as was expected. HA and KS were markedly increased, in HSNC 114% and 46%, respectively. GalAGs predominated in both tissues and were only slightly increased in HSNC. CS represented the major proportion of GalAGs, whereas a lower but signi¢cant proportion of DS was also present in both tissues. The presence of DS in human nasal cartilage is demonstrated for the ¢rst time and was estimated by its susceptibility to degradation with chondroitinase B and further analytical gel chromatography. Although the absolute amount of CS remained almost constant in both tissues, that of DS was signi¢cantly increased in HSNC. The documentation of the presence of DS in hyaline cartilage from various sources has interested several investigators [11,26,28^32]. So far, only articular cartilage was found to contain DS in the form of biglycan and decorin, of about 40% IdoA content [28]. The data reported in the present study established the presence of small amounts of
87
DS of low IdoA content (18^28%) in nasal cartilage. These results were in agreement with those of a recent study from our laboratory [26] and of a previous study [29], which reported that a small amount of the side chains of biglycan isolated from bovine nasal cartilage was sensitive to chondroitinase B digestion. The large di¡erences in IdoA content between the nasal and articular cartilage DS may be related to the di¡erent functions of the respective tissues. The extent and the distribution of uronosyl epimerization in DS chains seem to be tissue speci¢c. DS derived from bovine mucosa appears to contain 60% [26] of its IdoA in long IdoA-GalNAc repeats, the remaining being rather randomly distributed. Similarly, the majority of IdoA in DS from pig skin also appears in IdoA-GalNAc repeats [33,34]. DS of both HNNC and HSNC seemed to have almost all of its IdoA randomly distributed, since all of the IdoA in the sample was recovered in oligosaccharides ranging from di- to eicosasaccharides long. Similar results were also obtained for DS of sheep nasal cartilage [26]. The random distribution of IdoA in DS chains of nasal cartilages might be due to the expression of the C5-epimerase activity at long intervals in this tissues. The di¡erences in IdoA distribution may be elucidated by studying the turnover and kinetic properties of the C5-epimerase in this tissues. GalAGs derived from HSNC also exhibited signi¢cant structural alterations concerning their disaccharide composition and molecular size. GalAGs showed an altered sulfation pattern, which was characterized by a signi¢cant increase of non-sulfated and 6-sulfated disaccharides, with a simultaneous decrease of 4-sulfated disaccharides. In HSNC a signi¢cant increase in the length of DS chains was also observed. Taking into account the increase in the chain size and the increase in the amount of DS in HSNC, we suggest that the increase in DS content is mainly due to the increase in the chain length. Our results, which concern the sulfation patterns of GalAGs and the molecular size of DS, revealed altered biosynthesis of these GAG types by HSNC chondrocytes and suggested that scoliosis of the nasal septum cartilage is related to quantitative and structural modi¢cations at the GAG level.
References [1] D. Heinegard, A. Oldberg, FASEB J. 3 (1989) 2042^2051. [2] T.E. Hardingham, A.J. Fosang, J. Dudhia, Eur. J. Clin. Chem. Clin. Biochem. 32 (1994) 249^257. [3] S.L. Carney, H. Muir, Physiol. Rev. 68 (1988) 858^910. [4] L. Kjellen, U. Lindahl, Annu. Rev. Biochem. 60 (1991) 443^475. [5] R.V. Iozzo, Annu. Rev. Biochem. 67 (1998) 609^652. [6] C.R. Roberts, P.D. Pare, Am. J. Physiol. 261 (1991) L92^L101. [7] D.A. Theocharis, Int. J. Biochem. 17 (1985) 155^160. [8] D.A. Theocharis, C.P. Tsiganos, Int. J. Biochem. 17 (1985) 479^484. [9] D.A. Theocharis, D.L. Kalpaxis, C.P. Tsiganos, Biochim. Biophys. Acta 841 (1985) 131^134.
BBAGEN 25214 18-10-01
88
A.D. Theocharis et al. / Biochimica et Biophysica Acta 1528 (2001) 81^88
[10] D.D. Dziewiatkowski, J.L.A. Valley, A.G. Beaudoin, Connect. Tissue Res. 19 (1989) 103^120. [11] D. Heinegard, M. Paulsson, S. Inerot, S. Carlstrom, Biochem. J. 197 (1981) 355^366. [12] J.J. Ballenger, in: Diseases of the Nose, Throat and Ear, 11th edn., Lea and Febinger, Philadelphia, PA, 1969. [13] G.W. Bruyn, in: F. Cli¡ord Rose (Ed.), Handbook of Clinical Neurology, Vol. 4 (48), Headache, Elsevier Science Publishers, Amsterdam, 1986, pp. 475^482. [14] T. Bitter, H. Muir, Anal. Biochem. 4 (1962) 330^334. [15] M.L. Hayes, F.J. Castellino, J. Biol. Chem. 246 (1971) 430^435. [16] T. Scott, E.H. Melvin, Anal. Chem. 25 (1953) 1656^1661. [17] N.K. Karamanos, A. Syrokou, P. Vanky, M. Nurminen, A. Hjerpe, Anal. Biochem. 221 (1994) 189^199. [18] N.K. Karamanos, A. Hjerpe, T. Tsegenidis, B. Engfeldt, C.A. Antonopoulos, Anal. Biochem. 172 (1988) 410^419. [19] J.F. Woessner, Arch. Biochem. Biophys. 93 (1961) 440^446. [20] D.M. Carlson, J. Biol. Chem. 243 (1968) 616^626. [21] A.D. Theocharis, I. Tsolakis, T. Tsegenidis, N.K. Karamanos, Atherosclerosis 145 (1999) 359^368. [22] C.A. Antonopoulos, S. Gardell, J.A. Szirmai, E.P. De Tyssonsk, Biochim. Biophys. Acta 83 (1964) 1^19.
[23] A. Anseth, C.A. Antonopoulos, A. Bjelle, L.-A. Fransson, Biochim. Biophys. Acta 215 (1970) 522^526. [24] M.E. Tsara, N. Papageorgacopoulou, D.D. Karavias, D.A. Theocharis, Anticancer Res. 15 (1995) 2107^2112. [25] A. Wasteson, J. Chromatogr. 59 (1971) 87^97. [26] D.A. Theocharis, N. Papageorgacopoulou, D.H. Vynios, S.Th. Anagnostides, C.P. Tsiganos, J. Chromatogr. B Biomed. Sci. Appl. 754 (2001) 297^309. [27] R.M. Mason, F.S. Wusteman, Biochem. J. 120 (1970) 777^785. [28] H.U. Choi, T.L. Johnson, S. Pal, L.-H. Tang, L. Rosenberg, P.J. Neame, J. Biol. Chem. 264 (1989) 2876^2884. [29] F. Cheng, D. Heinega®rd, A. Malmstro«m, A. Schmidtchen, K. Yoshiî . Fransson, Glycobiology 4 (1994) 685^696. da, L.-A [30] O. de Sampaio, M.T. Bayliss, T.E. Hardingham, H. Muir, Biochem. J. 254 (1988) 757^764. [31] P.J. Roughley, R.J. White, Biochem. J. 262 (1989) 823^827. [32] P.J. Roughley, R.J. White, J. Orthop. Res. 10 (1992) 631^637. [33] N.K. Karamanos, P. Vanky, A. Syrokou, A. Hjerpe, Anal. Biochem. 225 (1995) 220^230. [34] A. Malmstro«m, J. Biol. Chem. 259 (1984) 161^165.
BBAGEN 25214 18-10-01