Distribution of Keratan Sulfate in Cartilage Proteoglycans *

After chondroitinase digestion of bovine nasal and tracheal cartilage proteoglycans, subsequent treatment with trypsin or trypsin followed by chymotrypsin yielded two major types of polypeptide-glycosaminoglycan fragments which could be separated by Sepharose 6B chromatography. One fragment, located close to the hyaluronic acid-binding region of the protein core, had a high relative keratan sulfate content. This fragment contained about 60% of the total keratan sulfate, but less than 10% of the total chondroitin sulfate present in the original proteoglycan preparation. The weight average molecular weight of the keratan sulfateenriched fragment was 122,000, as determined by sedimentation equilibrium centrifugation. The chemical and physical data indicate that this fragment contains an average of 10 to 15 keratan sulfate chains, if the average molecular weight of individual chains is assumed to be about 8,000, and about 5 chondroitin sulfate chains attached to a peptide of about 20,000 daltons. The other population of fragments was derived from the other end of the proteoglycan molecule, the chondroitin sulfate-enriched region, and contained mainly chondroitin sulfate chains. About 90% of the total chondroitin sulfate, but only 20 to 30% of the total keratan sulfate was recovered in these fragments. On the average, approximately 5 chondroitin sulfate chains and 1 keratan sulfate chain could be linked to the same peptide. Another 10 to 20% of the total keratan sulfate, originally found in or near the hyaluronic acid-binding region, was not separated from the chondroitin sulfate-enriched fragments. Hydroxylamine could be used to liberate a large molecular size, chondroitin sulfate-enriched fragment (K,, 0.54 on Sepharose 2B) from the proteoglycan aggregates. The remainder of the protein core, containing the keratan sulfate-enriched region, was bound to hyaluronic acid with the link proteins and recovered in the void volume on the Sepharose 2B column.

After chondroitinase digestion of bovine nasal and tracheal cartilage proteoglycans, subsequent treatment with trypsin or trypsin followed by chymotrypsin yielded two major types of polypeptide-glycosaminoglycan fragments which could be separated by Sepharose 6B chromatography. One fragment, located close to the hyaluronic acid-binding region of the protein core, had a high relative keratan sulfate content. This fragment contained about 60% of the total keratan sulfate, but less than 10% of the total chondroitin sulfate present in the original proteoglycan preparation. The weight average molecular weight of the keratan sulfateenriched fragment was 122,000, as determined by sedimentation equilibrium centrifugation. The chemical and physical data indicate that this fragment contains an average of 10 to 15 keratan sulfate chains, if the average molecular weight of individual chains is assumed to be about 8,000, and about 5 chondroitin sulfate chains attached to a peptide of about 20,000 daltons. The other population of fragments was derived from the other end of the proteoglycan molecule, the chondroitin sulfate-enriched region, and contained mainly chondroitin sulfate chains. About 90% of the total chondroitin sulfate, but only 20 to 30% of the total keratan sulfate was recovered in these fragments. On the average, approximately 5 chondroitin sulfate chains and 1 keratan sulfate chain could be linked to the same peptide. Another 10 to 20% of the total keratan sulfate, originally found in or near the hyaluronic acid-binding region, was not separated from the chondroitin sulfate-enriched fragments. Hydroxylamine could be used to liberate a large molecular size, chondroitin sulfate-enriched fragment (K,, 0.54 on Sepharose 2B) from the proteoglycan aggregates. The remainder of the protein core, containing the keratan sulfate-enriched region, was bound to hyaluronic acid with the link proteins and recovered in the void volume on the Sepharose 2B column.
The current model for the structure of proteoglycan monomer molecules isolated from hyaline cartilages is that of a central protein core, with an average molecular weight of 200,000, to which approximately 100 chondroitin sulfate and 50 keratan sulfate side chains are covalently attached (1,2). About one-third of the protein, located at one end of the core, contains few or no polysaccharide side chains (3). This portion of the protein (the hyaluronic acid-binding region) exhibits a specific, noncovalent interaction with hyaluronic acid (3-5). This interaction is essential for the formation of proteoglycan aggregate structures. Such aggregates appear to be the predominant way in which the proteoglycans are organized in cartilage extracellular matrices (6). Available data indicate that the chondroitin sulfate chains are attached to the protein core in clusters which contain from 1 to as many as 10 individual, closely spaced chains (7). The peptide sequences between the chains within the clusters are short, whereas the peptide sequences separating the clusters are considered to be longer. The peptide sequences within the clusters are not hydrolyzed by treatment with a combination of trypsin and chymotrypsin, while those between the clusters are.
Little is known about the distribution of the keratan sulfate side chains along the protein core, although some evidence has been presented which indicates that the region of the core protein isolated with the hyaluronic acid-binding region is enriched in keratan sulfate chains relative to chondroitin sulfate chains (3). However, many details about the keratan sulfate chains, their chemical structure, their mode of attachment to the protein core, and their average chain size remain to be determined. The available evidence suggests that most of the keratan sulfate chains are attached to the protein through glycosidic bonds between galactosamine and the hydroxyl groups of serine and threonine residues (8, 91, although another type of linkage to glutamic acid or glutamine has also been proposed (9,10). The galactosamine moiety often appears to be substituted on position 3 with a neuraminylgalactosyl disaccharide and on position 6 with the characteristic keratan sulfate chain (9) which consists of about 10 to 15 repeat units (2, 11) of the disaccharide (P-1,3-galactose P-1,4-2-deoxy-2acetamidoglucose 6-sulfate). The keratan sulfate-peptide fragments isolated from papain digests of proteoglycans are polydisperse, with molecular weights of 5 to 10 x 10" (2, 9), whereas keratan sulfate chains isolated from alkali-treated proteoglycans are considerably smaller (2). ' Previously it was shown that at least 50% of the keratan sulfate chains are present in the chondroitin sulfate peptide fractions isolated by cetylpyridinium chloride precipitation of trypsinichymotrypsin-digested proteoglycans (7). These keratan sulfate frag-  2) -The hyaluronic acid-binding region of the proteoglycan (3) was prepared from proteoglycan aggregates (the Al fraction1 by a modification of the procedure described previously (3). A 10 mglml solution of the Al fraction was incubated at 37" with diphenylcarbamyl chloridetreated trypsin (4 pg/mg of Al) for 8 h. The pH ofthe digest was then adjusted to 5.8 with acetic acid and solid CsCl was added to adjust the density to 1.66 g/ml. The sample was centrifuged for 66 h at 34,000 rpm in a MSE angle rotor (8 x 25 ml) at 15". The tubes were emptied into eight fractions of 2 ml. Fractions were analyzed for density by pycnometry using a 200-~1 constriction pipette and for contents of uranic acid and protein with the carbazole procedure and absorbance at 280 nm, respectively.
The top two fractions (4 ml) contained more than half of the protein in the gradient. These two fractions were pooled, dialyzed against water, and freeze-dried. 3. Fragmentation of proteoglycans with hydroxylamine followed by enzymic degradation.  9). It has been observed, however, that extensive treatment of keratan sulfate with alkali will yield molecules of considerably lower molecular weight (2, 10). Therefore, the ratio of the number of keratan sulfate chains to chondroitin sulfate chains in the original peptides were calculated from the chemical data given in this paper. It has been shown that chondroitinase ABC digestion of cartilage proteoglycans will leave a minimum of 1 disaccharide unit from the chondroitin sulfate still attached to the protein.
Consequently, in the peak 1 component the galactosamine derived from chondroitin sulfate would equal the xylose contents. The ratio of the remaining galactosamine to glucosamine would be 0.207. Assuming 1 galactosamine/keratan sulfate chain (9),' the M,L for the keratan sulfate chain would be about 2500. This would give a ratio of the number of keratan sulfate chains to the number of chondroitin sulfate chains in the original peptides of about 4 to 5. The chemical composition of the peak 2 component indicates that most of the chondroitin sulfate chains in the undigested sample were originally present in these peptides. Further, the high contents of glycine, serine, and glutamic acid very much resemble that of chondroitin sulfate isolated after trypsin, chymotrypsin, or papain digestion of cartilage or of proteoglycans (7, 21). The ratio of glucosamine to xylose of about 1.651 indicates that these peptides contain few keratan sulfate chains, approximately 1 for every 5 to 10 chondroitin sulfate chains (assuming a M,. of about 8000 for keratan sulfate), that is an average of no more than one keratan sulfate chain for every cluster of chondroitin sulfate chains (7, 22). Alternatively, the peptides may contain one keratan sulfate chain with a M, of about 2500 for every two to three chondroitin sulfate chains. The apparent molecular weight of the peak 1 fraction was determined at several solute concentrations by sedimentation equilibrium centrifugation as described under "Experimental Procedures." The graphs of In c against r2/2 yielded straight lines for each solute concentration (Fig. 4). However, this apparent ideal behavior appears to be the result of a balance between the nonideality from charge effects of the polyanion solute molecules and the inherent polydispersity of molecular weights present in the sample. This is indicated by the fact  Aspartic  acid  21  64  Threonine  60  44  Serine  127  154  Glutamic  acid  207  141  Proline  239  88  Glycine  71  160  Alanine  45  61  Cysteine  Valine  31  77  Methionine  Isoleucine  26  40  Leucine  37  88  Tyrosine  7  7  Phenylalanine  78  23  Lysine  40  3  Histidine  1  19  Arginine  2  17 that the apparent M, determined from the slopes of the graphs were different for each solute concentration. The ideal weight average M, was estimated to be about 122,000 from the plot of the reciprocal of the apparent M, values against solute concentration (inset in Fig. 4). If the average M, of a keratan sulfate chain is either 8,000 (2, 9) or 2,500 as calculated above, the chemical composition and the M, of the peak I fraction indicate that each peptide would have an average M, of about 20,000, and be substituted with about 12 or 30 to 40 keratan sulfate chains, respectively. Since the material in the fractions was recovered quantitatively, the data could be used for determining the stoichiometry of the fractions. It can be calculated from the figures given in Tables I and II that 50 to 60% of the keratan sulfate in the proteoglycan Al-D1 occurred in the larger size peak 1 fragment, while the remainder was found in the smaller peak 2 peptides. The chondroitin sulfate, as estimated from the xylose values, was originally primarily linked to the peptides recovered in peak 2, which contained about 90% of the total xylose.
Another sample of nasal cartilage proteoglycan monomers (Al-Dl) was digested with chondroitinase and then only with diphenylcarbamyl chloride-treated trypsin. The digest was chromatographed on a preparative Sepharose 6B column. Two major polysaccharide-peptide peaks were obtained, I (K,, = 0.32) and II (K,, = 0.69) (Fig. 5). The elution positions of the corresponding peaks observed when both trypsin and chymotrypsin were used are indicated by arrows 1 (K,, = 0.35) and2 (K,, = 0.81) in Fig. 5. In addition to the two major peaks, a minor component chromatographed between peaks I and II. The compositions (Table III) indicate that peak I contains the keratan sulfate-enriched fragment, since the contents of glucosamine, glutamic acid, and proline are high. The elution position indicates that the keratan sulfate-enriched fragments liberated by trypsin alone are somewhat larger than those liberated by the combined action of trypsin and chymotrypsin. Both the glucosamine and the galactosamine to protein ratios were lower in keratan sulfate-enriched fragments liberated by trypsin, compared with those isolated after trypsinlchymotrypsin digestion. It is possible, then, that the chymotrypsin treatment primarily liberates peptide-rich fragments from the keratan sulfate-enriched trypsinized peptides. The composition of the peak II material is very similar to that of the chondroitin sulfate-enriched peptides (Tables II and  III). The fragments obtained after digestion with trypsin containing no chymotryptic activity were considerably larger than the corresponding peptides isolated from the trypsin/ chymotrypsin digests. Also the glucosamine to protein ratio was considerably lower in the trypsinized peptides (II), suggesting that subsequent chymotrypsin digestion not only splits the peptide to yield smaller oligosaccharide-peptide fragments, but also liberates some unsubstituted peptides. The minor peak between peaks I and II, Fig. 5 products produced when proteoglycan aggregates, Al fraction from nasal cartilage, were treated with hydroxylamine as described under "Experimental Procedures." When an aliquot of hydroxylamine-treated Al was chromatographed on Sepharose 2B, the elution pattern shown in Fig. 6 was obtained. Fractions 1 and 2 indicated on the figure were isolated. The void volume peak (Fraction 1) contained the link proteins as shown by electrophoresis on sodium dodecyl sulfate-polyacrylamide gels (data not shown). Proteoglycan fragments and hyaluronic acid were also present in this fraction. Therefore, this sample was chromatographed on Sephadex G-200 in 4 M guanidinium chloride and the elution profile is shown in Fig.  7. The link proteins were recovered in the included peak and all the proteoglycan fragments and hyaluronic acid were recovered in the void volume peak. The excluded material was analyzed for chemical composition and the data are shown in Table IV. Its keratan sulfate and protein contents were high, with keratan sulfate representing about 65% of the total present in the original Al preparation.
The high relative proportion of glutamic acid and proline is also indicative of the presence of keratan sulfate, since these amino acids are predominant in keratan sulfate peptide fragments. The contents of serine and glycine, which are usually associated with chondroitin sulfate, on the other hand were low. The material EFFLUENT VOLUME  Fig. 6) when hydroxylamine-treated nasal cartilage Al was chromategraphed. Glucosamine (% of hexosamines) Galactosamine (% of hexosamines) 26.4 (63%") 5.4 (37%") 2 73.6 (22%") 94.6 (78%b) i? 9 03 Protein (% of dry weight) 17.9 7.9 fl Percentage of total glucosamine in the original proteoglycan. ' Percentage of total galactosamine in the original proteoglycan.  Fig. 6, however, had a low keratan sulfate content; the ratio of glucosamine to galactosamine was 0.057. Approximately 80% of the chondroitin sulfate in the aggregate was recovered in this fraction, and its elution position (K,,. = 0.55) was more retarded than that of a nasal cartilage Al-D1 fraction (K,,. = 0.21) on Sepharose 2B. Fraction 2 had a very low protein content with the amino acid composition shown in Table IV. It contained a large relative proportion of serine, glutamic acid, and glycine, amino acids which occur close to the chondroitin sulfate-protein linkage. The data suggest that the proteoglycan monomers in the aggregates were cleaved somewhere along their lengths to produce chondroitin sulfate-enriched fragments of relatively small size (Fraction 21, while leaving the hyaluronic acidbinding region and a large proportion of the keratan sulfate still associated with the hyaluronic acid and the link proteins (Fraction 1).  (Fig. 8). Support for this hypothesis was obtained by analyses of fractions recovered from associative CsCl density gradients of hydroxylamine-treated proteoglycan aggregates. The material was distributed in the gradient as shown in Fig. 8. The top fraction contained a large proportion of material excluded on Sepharose 2B (Fig. 9). The link proteins were present in this fraction. Analyses showed that the contents of glutamic acid and proline were high, as was the ratio of glucosamine to galactosamine (Table V). This material, then, is similar to Fraction 1 isolated from the Sepharose 2B chromatography. The bottom fraction, on the other hand, which was included on Sepharose 2B (K,,. = 0.54), Fig. 9, did not contain detectable amounts of the link proteins, and had a very high galactosamine to glucosamine ratio ( Table V).
The different fractions obtained by gel chromatography and by CsCl density gradient centrifugation of the hydroxylaminetreated Al were digested with chondroitinase followed by trypsin. The digests were then chromatographed on Sepharose 6B, yielding the chromatograms shown in Figs. 10 and 11. The general pattern described above for the chondroitinase-trypsin digest of proteoglycan monomers can be identified. The gel chromatograms show that the fragments capable of interacting with hyaluronic acid (recovered from Sepharose 2B, Fraction 1, or from the top of the density gradient) contained a large proportion of the large size keratan sulfate-enriched peptide. The smaller sized fragments (recovered from Sepharose 2B, Fraction 2, or the bottom of the density gradient), on the other hand, only contained small amounts of this keratan sulfate-enriched peptide. These latter fragments, then, which contained most of the chondroitin sulfate chains, would be located on portions of the protein core of proteoglycan mono- mers distal to the hyaluronic acid-binding region. Conversely, the keratan sulfate-enriched peptide, which was recovered primarily from fragments which were associated with hyaluronic acid and the link proteins, would be located closer to the hyaluronic acid-binding region of the core protein. Some 20% of the chondroitin sulfate, however, is recovered with the fraction bound to hyaluronic acid; and the exact position of these chondroitin sulfate chains cannot be identified presently.

Keratan
Sulfate in Hyaluronic Acid-binding Region -The relationship of the keratan sulfate-enriched region to the keratan sulfate known to be present in the hyaluronic acid-binding region of proteoglycans (3) was studied. Proteoglycans (Al) were digested with diphenylcarbamyl chloride-treated trypsin and fractionated in a CsCl density gradient. The hyaluronic acid-binding region was isolated from the top fraction by Sepharose 2B and Sephadex G-200 chromatography, as described under "Experimental Procedure." The isolated hyaluronic acid-binding region was digested with chondroitinaseltrypsin and chromatographed on an analytical Sepharose 6B column. The elution patterns, Fig. 12, show that the keratan sulfate peptides (indicated by anthrone reactivity) were eluted in 2 to 3 peaks, at about the position of the chondroitin sulfate peptides, peak 2, and much later than the keratan sulfate-enriched peptide, peak 1. It is likely then that the keratan sulfate located in the hyaluronic acid-binding region is recovered with the peak 2 chondroitin sulfate peptides when proteoglycan monomers are digested with chondroitinase and trypsin. The hyaluronic acid-binding region contains 10 to 20% of the total keratan sulfate in a proteoglycan monomer (3). Subtracting this 10 to 20% of the keratan sulfate from the 40% of the total keratan sulfate, which is recovered in peak 2, one can estimate that a more accurate value for the keratan sulfate content of the chondroitin sulfate-enriched region is 20 to 30% of the total.  (Fig. 7). The material originated from the void volume of a Sepharose 2B chromatogram (Fig. 6) of hydroxylamine-treated nasal septum cartilage Al. Bottom, Sepharose 6B chromatogram of chondroitinase and trypsin-digested small molecular size proteoglycan fragment which was isolated from the included peak when the hydroxylamine-treated nasal cartilage Al was chromatographed on Sepharose 2B (Fig. 6, peak 2).
The presence of the keratan sulfate-enriched region in the bottom fraction of the CsCl gradient of the trypsin digest (Al-T-Al) was established. This fraction, which does not contain the hyaluronic acid-binding region, as discussed above, has a glucosamine to galactosamine ratio of 0.061. A sample was digested with chondroitinase ABC and chromatographed on Sepharose 6B (Fig. 13). The first peak represents the keratan sulfate-enriched region, as could be established by amino acid analysis and determination of the hexosamine ratio. The material had a high content of glutamic acid and proline (18.9% and 20.0% of the total amino acids, respectively) and the glucosamine to protein ratio was 0.75, while the galactosamine to protein ratio was 0.25. The major protein peak (peak 3) eluting at about 600 ml had the same K,, as the chondroitin sulfate-enriched peptides isolated from trypsin digests of proteoglycan monomers. The material in this peak had high contents of serine, glutamic acid, and glycine (16.7%, 14.9%, and 16.7% of the total amino acids, respectively). The galactosamine to protein ratio was 0.32, while the glucosamine to protein ratio was 0.11. This latter value is about half that of the value from the corresponding material isolated from trypsin digests of Al-D1 (-6B2, Table III). The difference is most likely due to the absence of the keratan sulfate from the hyaluronic acid-binding region in the Al-T-Al preparation. Therefore, all of the keratan sulfate recovered with the chondroitin sulfate-enriched peptides is probably derived from the chondroitin sulfate-enriched region of the proteoglycan. The small amount of keratan sulfate were isolated (lo), in accord-cal assistance by Annika Bjorne-Persson and Ewa Ljungberg ance with the data presented in this report.
is gratefully acknowledged. It has been claimed that small molecular weight proteoglycans with very low keratan sulfate and protein contents can be isolated from cartilage by extraction with 0.15 M potassium acetate or potassium chloride (5, 26). The very low keratan sulfate contents of 2 to 4% of the glycosaminoglycans in these proteoglycans strongly indicate that they do not contain the keratan sulfate-enriched peptide region discussed above. The low protein content, the chemical composition, and the Sepharose 2B K,, values of the low salt-extracted proteoglycans, K,, 0.54 (51, are very similar to that of the low molecular weight fragment liberated from proteoglycan aggregates by the action of hydroxylamine as is discussed above. Therefore it is possible that the low salt-extracted proteoglycan is a degradation product of the normal proteoglycan monomer. In support of such a hypothesis, as is discussed in a subsequent report (27), the low salt-extracted proteoglycans cannot be identified in nasal cartilage guanidinium chloride extracts when procedures which minimize proteolysis are used.
The data discussed are consistent with the model shown in Fig. 14. In this model the keratan sulfate-enriched region occupies a part of the protein core very near the portion of the proteoglycan interacting with the hyaluronic acid. This keratan sulfate-enriched region contains about 60% of the keratan sulfate in the proteoglycan monomer, but only 10% of the chondroitin sulfate. The chondroitin sulfate-enriched region is located at the other end of the monomer, away from the hyaluronic acid-binding region. The chondroitin sulfate-enriched region contains 90% of the chondroitin sulfate chains, but only about 20% of the total number of keratan sulfate chains in the proteoglycan.
The remainder of the keratan sulfate chains are attached to the protein recovered with the hyaluronic acid-binding region.