INTRODUCTION

Chondroitin sulfate proteoglycans, consisting of a core protein with at least one covalently attached glycosaminoglycan (GAG)¹ chain, are distributed on the surfaces of most cells and the extracellular matrix in virtually every tissue (for reviews, see Refs. 1 and 2). Despite the ubiquity of this family of molecules, a wide variety of chondroitin sulfate proteoglycans with characteristic sulfated GAG chains exhibit tissue-specific and developmentally regulated expression (3), and have been implicated in the regulation and maintenance of cell proliferation, cytodifferentiation, and tissue morphogenesis (4). The structures of cartilage chondroitin sulfate GAGs change with normal embryonic development and growth or aging (4-8). In neural development, chondroitin sulfate governs developmentally significant events such as cellular adhesion, migration, and neurite outgrowth (9, 10). The molecular basis for the developmentally regulated and tissue-specific synthesis of chondroitin sulfate, as well as other GAGs, has yet to be clarified.

Chondroitin sulfate has a linear polymer structure that possesses repetitive, sulfated disaccharide units containing glucuronic acid and GalNAc (1, 2). Since GAG structures are largely determined by the specificities of the glycosyltransferases and sulfotransferases responsible for their synthesis, it is presumed that the differential expression of these enzymes is the key for the controlled synthesis of GAGs. However, few reports have systematically investigated the degree to which specific glycosyltransferases and sulfotransferases are differentially expressed in normal tissues or how such expression might be regulated in a tissue-specific and development-dependent manner.

The major chondroitin sulfate found in the mammalian tissues bears sulfate groups at position 4 or 6 of GalNAc residues. It was reported that the ratio of 4-sulfation/6-sulfation changed during development of chicken and human epiphyseal cartilage and rat skin (4-6, 11). Considering the fact that the extracellular matrix of cartilage resembles that of the brain, it is rich in chondroitin sulfates, lecticans, and hyaluronic acid, but unlike the brain matrix, it contains abundant fibrous collagen (12), and it is of particular interest to determine the changes in the sulfation profile of the chondroitin sulfate chains during brain development and to further elucidate the factors responsible for the specific sulfation. In this report, we present evidence that the ratio of 4-sulfation/6-sulfation in the embryonic chick brain changes with development and that relative levels of the specific sulfotransferase activities are closely coordinated with relative changing levels of the specific chondroitin sulfate structures. The results support the concept that expression of sulfotransferases is a predominant factor regulating the sulfation profile of chondroitin sulfate structures.

EXPERIMENTAL PROCEDURES

[³⁵S]PAPS and unlabeled PAPS were purchased from NEN Life Science Products and Sigma, respectively. Fertile White Leghorn chicken eggs were purchased from a local poultry farm and incubated at 37 °C under a humidified atmosphere, until the desired developmental stage was reached according to Hamburger and Hamilton (13). Chondroitin (a chemically desulfated derivative of whale cartilage chondroitin sulfate A), five unsaturated standard disaccharides derived from chondroitin sulfate, ^4,5HexA1-3GalNAc (Di-0S), ^4,5HexA1-3GalNAc(4-O-sulfate) (Di-4S), ^4,5HexA1-3GalNAc(6-O-sulfate) (Di-6S), ^4,5HexA1-3GalNAc(4,6-O-disulfate) (Di-diS_E), and ^4,5HexA(2-O-sulfate)1-3GalNAc(6-O-sulfate) (Di-diS_D), chondroitin ABC lyase (EC 4.2.2.4), chondro-4-sulfatase (EC 3.1.6.9), and chondro-6-sulfatase (EC 3.1.6.10) were purchased from Seikagaku Corp., Japan. HITRAP^TM desalting columns were obtained from Pharmacia Biotech Inc., Sweden. All other reagents and chemicals were of the highest quality available.

Most chondroitin sulfate proteoglycans can be extracted from brain using physiological buffers without detergent, and approximately 70% of total chondroitin sulfate proteoglycans are reportedly extracted by phosphate-buffered saline (14). Thus, soluble chondroitin sulfate proteoglycan fractions from brain were prepared as described previously (14). In brief, embryonic chick brains from various developmental stages were homogenized with a tight-fitting Potter glass homogenizer in 5 volumes (w/v) of ice-cold phosphate-buffered saline containing 20 mM EDTA and 2 mM phenylmethylsulfonyl fluoride. The homogenate was centrifuged at 16,000 × g for 40 min at 4 °C. The pellet was subjected to rehomogenization in phosphate-buffered saline. After centrifugation, both supernatant fluids were combined, concentrated, and then washed twice with 50 mM Tris-HCl buffer (pH 8.0) containing 50 mM sodium acetate using a Centricon-10 concentrator (Amicon Inc.). The protein concentration of the proteoglycan fractions was determined using the BCA protein assay kit (Pierce) with bovine serum albumin as a standard.

The proteoglycan fractions prepared above (about 100 µg of protein each) were first digested using 5 mIU of chondroitin ABC lyase as described elsewhere (15), and evaporated to dryness. The digests were derivatized with 2-aminobenzamide according to the manufacturer's instructions (SIGNAL^TM labeling kit, Oxford GlycoSystems). The labeled disaccharides were analyzed by HPLC on an amine-bound silica PA03 column (4.6 × 250 mm; YMC Co., Kyoto, Japan) as described previously (16). The HPLC was performed in an LC-10AS system (Shimadzu Co., Kyoto, Japan) using a linear gradient from 16 to 798 mM NaH₂PO₄ over a 60-min period at a flow rate of 1.0 ml/min at room temperature. Eluates were monitored using an RF-535 fluorometric detector (Shimadzu Co.) with excitation and emission wavelengths of 330 and 420 nm, respectively. The identification and quantification of the resulting disaccharide units were accomplished by comparison with the standard chondroitin sulfate-derived unsaturated disaccharide labeled with 2-aminobenzamide as described elsewhere (17) and by enzymatic digestion using chondro-4- and 6-sulfatases as reported elsewhere (15).

All procedures were carried out at 4 °C. Embryonic chick brains from various developmental stages were homogenized with a tight-fitting Potter glass homogenizer in 5 volumes (w/v) of 0.15 M imidazole-HCl (pH 6.8) containing 1% Triton X-100 and 1 mM phenylmethylsulfonyl fluoride. The homogenates were used as the enzyme source. The protein concentration of the homogenates was determined as described above.

In a preliminary study, assay conditions for brain chondroitin 4-sulfotransferase (C4ST) and chondroitin 6-sulfotransferase (C6ST) were established by examining various factors such as buffers, metal ions, substrate concentrations, and inhibitors for PAPS degradation (data not shown). The assay mixture contained 120 µg of chondroitin, 10 µM PAPS (about 2.5 × 10⁵ cpm), 24 µg of polylysine, 50 mM imidazole-HCl (pH 6.8), 5 mM EDTA, 5 mM 2,3-dimercaptopropan-1-ol, 10 mM MgCl₂, and enzymes in a total volume of 60 µl. Incubations were carried out at 37 °C for 2 h, and the reactions were terminated by boiling for 1 min. After centrifugation at 16,000 × g for 5 min at 4 °C, the supernatant was subjected to gel filtration on a column of HITRAP^TM desalting to isolate ³⁵S-labeled chondroitin as described previously (18), and the radioactivity was measured with a liquid scintillation counter. The radioactive peak corresponding to the ³⁵S-labeled chondroitin was pooled and evaporated to dryness. For determining the transfer of sulfate to positions 4 and 6 of GalNAc residues, one-third of each ³⁵S-labeled chondroitin fraction was digested using 5 mIU of chondroitin ABC lyase, and the resulting unsaturated disaccharides were identified by HPLC as reported elsewhere (15). To confirm the disaccharide structure, chondro-4-sulfatase or 6-sulfatase digestion of chondroitin ABC lyase digest was conducted with the remainder of each ³⁵S-labeled chondroitin fraction as described elsewhere (15). From the incorporation into Di-4S and Di-6S, activities of C4ST and C6ST were determined, respectively. Under the established incubation conditions for the C4ST and the C6ST, [³⁵S]sulfate incorporation into polymer chondroitin was proportional to the incubation time for up to 4 h.

Total RNA from chick embryo tissues at various developmental stages was isolated using the QuickPrep® total RNA extraction kit (Pharmacia LKB Biotechnology, Uppsala, Sweden) according to the manufacturer's protocol. RNA integrity and concentration were assessed by electrophoresis in a denaturing formaldehyde-agarose gel (1%) and ethidium bromide staining.

The RT-PCR reactions were performed according to the manufacturer's instructions (Takara RNA LA PCR kit, Kyoto, Japan), using 1 µg of total RNA as a template. To amplify an equivalent quantity of cDNA for each sample, we determined the exact amount of cDNA of each sample required to obtain equal levels of amplification of the glyceraldehyde-3-phosphate dehydrogenase, whose transcript is always present in the tissues at the same level. The amplification reaction was carried out in a total volume of 50 µl using the 5 primer, 5-ACCACTGTCCATGCCATCAC-3, and the 3 primer, 5-TCCACAACACGGTTGCTGTA-3, by 20 cycles of 95 °C for 45 s, 55 °C for 45 s, and 72 °C for 90 s. Of the amplified products, 10 µl were visualized by electrophoresis on a 1.5% agarose gel containing ethidium bromide. Using normalized cDNA input, we then performed amplification of a C6ST transcript. The experiment was performed with a serial number of cycles (25-30-35) to find the conditions for a semiquantitative amplification, since a low number of cycles resulted in no amplification and a high number of cycles resulted in an overamplification. The best results were obtained by carrying out 30 cycles of 95 °C for 45 s, 52 °C for 45 s, and 72 °C for 90 s using the 5 primer, 5-CGAGAAGGAAAACAACTTCA-3 (the nucleotide sequence corresponding to 321-340 of the cDNA for chick C6ST) (19), and the 3 primer, 5-CTCGGGCGCTGGTGAGAT-3 (the nucleotide sequence corresponding to 769-786), which have been designed to span the intron in the C6ST gene² to discriminate a PCR product amplified from cDNA from, if any, one amplified from contaminating genomic DNA. PCR products were then visualized by electrophoresis on a 1.5% agarose gel containing ethidium bromide. To confirm that the amplified DNAs were derived from the C6ST mRNA, the amplified fragments were gel-purified, subcloned into the SrfI site of the pCR-Script cloning vector (Stratagene), and sequenced. The nucleotide sequences of the amplified DNAs were identical to that of the chick C6ST cDNA demonstrated by Fukuta et al. (19) (data not shown). The sequencing reaction with a thermal cycler was performed with a dye terminator cycle sequencing FS ready reaction kit (PE Applied Biosystems) using a T3 or T7 primer (Stratagene). Gel electrophoresis and analysis of the data were performed with a 377 DNA sequencer (PE Applied Biosystems).

RESULTS

To study the changes in the amounts of the specific chondroitin sulfate chains in embryonic chick brain during development, disaccharide analysis of the chondroitin ABC lyase digest of the brain chondroitin sulfate at each developmental stage was carried out. As shown in Fig. 1A, the amounts of all unsaturated disaccharide units detected, Di-0S, Di-4S, Di-6S, Di-diS_D, and Di-diS_E (on a per protein basis) were the highest at stage 29, the first point examined, although those of the Di-diS_D and the Di-diS_E were relatively small (0.70 and 0.20 pmol/µg of protein, respectively). Thereafter, the amount of each disaccharide unit sharply declined with development.

[View Larger Version of this Image (11K GIF file)]

Previous studies have indicated that the ratio of 6-sulfation/4-sulfation of GalNAc residues in chondroitin sulfate chains changes during the development of epiphyseal cartilage (4, 6). Thus, proportions of these disaccharide units during brain development were calculated. As shown in Fig. 1B, the ratio of Di-6S to Di-4S decreased gradually. Specifically, Di-6S accounted for about 60% of the total disaccharide units at stage 29, and its proportion decreased gradually with development. Similarly, Di-0S accounted for about 10% of the total disaccharide units at stage 29, and its proportion reached a maximum around stage 33 then declined. In contrast, Di-4S accounted for about 30% of the total disaccharide units at stage 29, and its proportion increased progressively with development. In addition, the proportion of disulfated disaccharide units, Di-diS_D and Di-diS_E, increased slightly as the development reached completion.

To evaluate the key regulatory factor controlling the developmental changes in the sulfation profile, we investigated developmental profiles of the C4ST and the C6ST activities responsible for the synthesis of 4-O-sulfated GalNAc and 6-O-sulfated GalNAc, respectively. Activities of the C4ST and the C6ST were measured in the brain homogenates at various stages of embryonic development. As can be seen in Fig. 2A, the specific activities of these two enzymes were the highest at stage 29, and then sharply declined with development. Remarkably, the ratio of the C6ST activity to the C4ST activity decreased gradually with development (Fig. 2B), which correlated closely with the decreasing ratio of Di-6S to Di-4S as described above (see Fig. 1B). These findings suggested that expression of sulfotransferases is a predominant factor regulating the sulfation profile of chondroitin sulfate structures.

[View Larger Version of this Image (9K GIF file)]

Of several sulfotransferases responsible for chondroitin sulfate biosynthesis, only C6ST cDNA has so far been cloned (19). To directly examine the relationship between the C6ST gene expression and the C6ST activity in embryonic chick brain during development, the relative proportion of C6ST mRNA was determined in chick brain at various stages of embryonic development. Since the level of the C6ST mRNA expression is relatively low (19, 20), the abundance of the C6ST mRNA in the RNA samples was estimated by RT-PCR analysis. The relative gene expression levels, which were normalized as to transcription of the glyceraldehyde-3-phosphate dehydrogenase gene as described under "Experimental Procedures," are shown in Fig. 3. A single amplified DNA of the expected size (466 base pairs) was obtained from each RNA preparation of stages 29, 33, 35, and 40. From stage 29 to stage 40, no significant developmental change was detected in the amount of the C6ST mRNA in the brain, with the overall expression remaining at a low level. At stage 43, a detectable decrease occurred, and no transcripts were detected in the adult brain. Notably, the signal was not detected in the adult brain after 40 cycles of PCR, even when analyzed by Southern blotting (data not shown). Qualitative comparison of the results shown in Fig. 2A with the RT-PCR analysis in Fig. 3 shows a correlation between the levels of the C6ST activity and the levels of the C6ST mRNA in the brain.

[View Larger Version of this Image (26K GIF file)]

DISCUSSION

In this study, we demonstrated that the sulfation profile of chondroitin sulfate chains and the ratio of the sulfotransferase activities forming the specific sulfation profile changed markedly with development in the chick embryo brain, and these alterations were precisely coordinated. The importance of the differential expression of sulfotransferases has not been fully evaluated. Several sulfotransferases responsible for GAG synthesis appear to compete for common acceptor substrates (1, 2). Thus, the type of sulfotransferases expressed by a cell should influence the sulfation pattern found on the GAGs it produces. For example, the C6ST and the C4ST studied here compete for GalNAc residues in the (4GlcA1-3GalNAc1-)_n sequence of chondroitin sulfate chains to form 6-O-sulfated GalNAc and 4-O-sulfated GalNAc, respectively. Structural analysis shows that chondroitin sulfate chains of various animal tissues have highly variable sulfation profiles, and some have exclusively 4-O-sulfated GalNAc residues, whereas most others have a hybrid structure of 4-O-sulfated and 6-O-sulfated GalNAc residues with various ratios in the repeating disaccharide region (1, 2, 21, 22). Although smaller proportions of disulfated disaccharide units (Di-diS_D and Di-diS_E) may indicate the existence of multiple C6ST and C4ST isoenzymes and genes, the observations seem to suggest that the gross sulfation pattern results from the ratio of mainly C6ST and C4ST expressed by a cell. This concept is supported by the present results obtained from the examination of the developmental expression of the C4ST and the C6ST in chick embryo brain. As observed in Figs. 1B, 2B, and 3, the developmental changing patterns of both the expression ratio of the C6ST to the C4ST and of the C6ST gene corresponded closely to the present finding that the relative amount of chondroitin 6-sulfate as compared with 4-sulfate progressively decreased with development of chicken brain. Therefore, the developmentally regulated expression of sulfotransferases is a predominant factor regulating the sulfation profile of chondroitin sulfate structures.

Although the exact biological significance of the developmental changes in the sulfation profile is presently unclear, it is feasible that they affect at least some developmentally significant events such as cellular adhesion, migration, and neurite outgrowth (9, 10). In this respect, it should be noted that the structure of chondroitin sulfate chains on 6B4 proteoglycan/phosphacan changes during development of the brain (23, 24). In the early developmental stages, substantial amounts of chondroitin 6-sulfate are found, but later, the chondroitin sulfate chains of 6B4 proteoglycan are virtually composed of only chondroitin 4-sulfate (23, 24). Very recently, Maeda et al. (25) have reported that 6B4 proteoglycan binds pleiotrophin and chondroitin ABC lyase digestion of 6B4 proteoglycan decreases the binding affinity. They also showed that chondroitin 6-sulfate was a potent inhibitor of 6B4 proteoglycan-pleiotrophin binding as well as heparan sulfate, whereas chondroitin 4-sulfate was a poor inhibitor (25). Therefore, the developmental change in the sulfation profile of chondoritin sulfate chains of 6B4 proteoglycan may regulate the binding affinity of the proteoglycan to pleiotrophin.

In addition, in chick embryo epiphyseal cartilage and rat skin, there has been observed a progressive decrease in the ratio of chondroitin-6-sulfate to 4-sulfate with development, chondroitin 4-sulfate being predominant with the progress of tissue development (5, 6, 11). Thus, it seems to be a general phenomenon that immature tissues containing proliferating and differentiating cells synthesize more chondroitin 6-sulfate than adult tissues composed of quiescent mature cells. Hence, it is worth investigating whether chondroitin 6-sulfate itself plays a role in the stimulation of cell division.

The carbohydrate moieties of glycoconjugates on the surfaces of cells are also known to undergo various changes during the malignant transformation of cells. Many of these carbohydrate structures are described as onco/fetal antigens because they are most abundant in early fetal development, and their expression is developmentally regulated (26, 27). Indeed, the relative amount of chondroitin 6-sulfate as compared with 4-sulfate progressively increases with malignant transformation of human colon carcinoma cells (28). In view of the present results that the expression ratio of the C6ST to the C4ST was highest during early embryonic stages, then decreased as the development reached completion, it is of particular interest to evaluate the C6ST and C4ST expression during malignant transformation of human colon carcinoma cells.

Earlier studies have indicated that the sulfation of GAG chains ordinarily proceeds together with polymerization at a single Golgi site and that there appears to be close interrelationships between sulfation and polymer elongation/termination (for reviews, see Refs. 29 and 30). In fact, our recent findings have revealed that specific sulfate groups have either stimulatory or inhibitory effects on GalNAc transfer, and consequently sulfation reactions indeed play important roles in chain elongation and termination (31, 32). Moreover, there is an additional possibility that the regulated expression of glycosyltransferases involved in chondroitin sulfate biosynthesis could in part control the chain length of chondroitin sulfate. In this regard, it is of interest to note that chondroitin sulfate chains become shorter with age (7) and that the serum levels of glucuronyltransferase activity involved in chondroitin sulfate biosynthesis change developmentally, being markedly higher at the middle prenatal than at the late prenatal stage (33). Therefore, although only two sulfotransferase activities (the C6ST and the C4ST) and one gene (the C6ST) have been examined in this study among the 10 or more which are thought to be required to form known chondroitin sulfate isoform structures with various sulfation profiles, it is expected that the regulated expression of other sulfotransferases and even glycosyltransferases involved in chondroitin sulfate biosynthesis are equally important in establishing the cell type-specific and developmentally regulated expression of chondroitin sulfate isoforms. As the glycosyltransferase cDNAs and additional sulfotransferase cDNAs become available, it will be of considerable interest to establish the degree to which the expression of these enzymes is regulated in differentiating and developing cells. Such information will be required for understanding the control mechanism of cellular recognition events mediated by chondroitin sulfate chains during development and differentiation (3, 9, 10, 24).