Analysis of sialidase and N-acetylneuraminate pyruvate-lyase substrate specificity by high-performance liquid chromatography

ANALYTICAL

BIOCHEMISTRY

158,

158- 164 ( 1986)

Analysis of Sialidase and N-Acetylneuraminate Pyruvate-Lyase Substrate Specificity by High-Performance Liquid Chromatography’ ASHOK

K. SHUKLA

AND ROLAND

SCHAUER*

Biochemisches Institut, Christian-Albrechts-Universitcit, Olshausenstrasse 40, D-2300 Kiel, Federal Republic of Germany Received February 10, 1986 A rapid and sensitive assay by high-performance liquid chromatography for determination of the activity and substrate specificity of sialidase (EC 3.2.1.18) and N-acetylneuraminate lyase (EC 4.1.3.3) is described. Sialic acids were separated on a strong anion-exchange resin using 0.75 mM sodium sulfate as elution medium. This method allows the determination of a minimum amount of 200 pg (0.6 pmol) of sialic acid. Usually the enzyme mixtures were directly applied to the column without prior purification of substrates and products. The action of sialidase was studied either by the decrease of sialyllactose concentration or by the amount of sialic acid liberated. The relative hydrolysis rates of N-acetylneuraminyl-cr(2-3)-lactose, N-glycolylneuraminyl-a(2-3)-lactose. N-acetylneuraminyl-o(2-6)-lactose, N-acetyl-9-0-acetylneuraminyl-a(2-3)-lactose, and Nacetyl-4-0-acetylneuraminyl-cy(2-3)Jactose by Vibrio cholerae sialidase were 100,88,25, 12, and 0, respectively. The activity of N-acetylneuraminate lyase from Clostridium perfringens was determined by measuring the rate of disappearance of siahc acids or the formation of acylmannosamines, which is possible in the same chromatogram. Relative cleavage rates of N-acetylneuraminic acid, N-glycolylneuraminic acid, N-acetyl-9-O-acetylneuraminic acid, N-acetyl-7-O-acetylneuraminic acid, and N-acetyl-4-O-acetylneuraminic acid were found to be 100, 67, 24, 3, and 0, respectively. Comparison of the substrate specificities shows that substituents on the neuraminic acid molecule influence the reactions of both enzymes in a similar way. Q 1986 Academic press, IX. KEY WORDS: sialidase; N-acetylneuraminate pyruvate-lyase; sialic acids; sialyllactose; highperformance liquid chromatography.

Information about the nature and quantity of sialic acids can be obtained by calorimetry (l), fluorimetry (2), thin-layer chromatography (I), high-performance liquid chromatography (3-8), and gas-liquid chromatography (9). The structures of more than 30 naturally occurring sialic acids have been elucidated by the combination of GLC (10) or HPLC (11) with mass spectrometry and by ‘H-NMR spectroscopy ( 12). All these analytical procedures are timeconsuming and in most casesrequire extensive purification of sialic acids before analysis, which may result in loss of sialic acids or their

labile substituents. HPLC of sialic acids on ion-exchange resin (7) has advantages over these methods, as it allows direct determination of sialic acids within a short time ( 10 min). It therefore appears especially suited for the observation of enzyme reactions of sialic acid metabolism, requiring only small amounts of substances and enzymes and frequently giving information about both substrates and reaction products in one chromatographic run. MATERIAL

Copyright 0 1986 hy Academic Press. Inc. All rights of reproduction in any form reserved.

METHODS

Sialic acids were isolated from bovine, porcine, and equine submandibular gland glycoproteins as described in Refs. (1,13); a(23)- and cr(2-6)-sialyllactose isomers were purified from bovine colostrum (14); N-acetyl9-O-acetylneuraminyl-a(2-3)-lactose was iso-

’ Thanks are due to Deutsche Forschungsgemeinschaft (Grant Scha 202/10-4) and the Fonds der Chemischen Industrie. 2 To whom correspondence should be addressed. 0003-2697/86 $3.00

AND

158

STUDY OF SIALIC ACID CATABOLIC

lated from rat urine (15) N-glycolylneuraminyl-cr(2-3)-lactose was prepared from rat erythrocyte gangliosides ( 16), and N-acetyl-4O-acetylneuraminyl-c(2-3)-lactose from echidna (Tachyglossus aculeatus) milk was a gift from Dr. M. Messer, Sydney. CT- and Csanalogues of N-acetylneuraminic acid were prepared as described in Ref. ( 17). Bovine and porcine submandibular gland mucins were isolated from fresh glands (18). N-Acetyl2-deoxy-2,3-didehydroneuraminic acid was purchased from Boehringer Mannheim. Mbrio cholerae sialidase (EC 3.2.1.18) was obtained from Behringwerke and Arthrobacter ureafaciens sialidase was from Nakarai Chemicals Japan. N-Acetylneuraminate pyruvate-lyase (EC 4.1.3.3) from Clostridium perfringens was purchased from Sigma. Solvents and chemicals of analytical grade were products of E. Merck. HPLC analysis was performed on a SpectraPhysics SP 8000 apparatus as described in detail in Ref. (7). Sialic acid samples (0.01-2 pg) dissolved in 20 ~1 water or buffer (enzyme incubation mixtures) were applied to the HPLC stainless steel column (40 X 4.6 mm, filled with Aminex A-29, Bio-Rad) containing a lo~1 sample loop. Sialic acids and sialyllactoses were eluted isocratically by 0.75 mM sodium sulfate solution at a flow rate of 0.5 ml/min at 15 bar and monitored by a photometer (Spectroflow SF 773, Kratos) at 200 nm. Detector sensitivity was between 0.05-0.00 1 a.u.f.s. and attenuation varied between 5 and 40 mV. A minimum amount of 200 pg (0.6 pmol) sialic acid can be identified by this method. The reactions of VCS3 with sialyllactoses and mucin (BSM) were carried out in 50 mM 3 Abbreviations used: NeuSAc, N-acetylneuraminic acid, NeuSGc, N-glycolylneuraminic acid; Neu4,5Ac2, N-acetyl4-0-acetylneuraminic acid; Neu5,7Acz, N-acetyl-7-0acetylneuraminic acid; Neu5,9Acz, N-acetyl-9-O-acetylneuraminic acid; NeuSAQen, N-acetyl-2-deoxy-2,3-didehydroneuraminic acid; BSM, bovine submandibular gland mucin; AUS, Arthrobacfer ureufuciens sialidase; VCS, Vibrio cholerue sialidase; HPLC, high-performance liquid chromatography; TLC, thin-layer chromatography: GLC, gas-liquid chromatography.

ENZYMES

BY HPLC

159

sodium acetate buffer of pH 5.5, containing 154 mM sodium chloride and 9 mM calcium chloride (1) at 37°C. These substrates were also hydrolyzed with AUS in 50 mM sodium phosphate buffer of pH 7.2 at 37°C or by acid (5 mM H2S04) at 90°C. The incubation mixtures of 100 ~1 contained about 25 pg of glycosiditally bound sialic acid and I-5 mU of sialidase. After different incubation times (O-60 min in the case of sialyllactoses and O-24 h in the case of BSM) lo-p1 samples were applied to the column. The incubation mixtures with mucin were diluted 1: 10 with water before injection. Free sialic acids were cleaved by N-acetylneuraminate pyruvate-lyase in 50 mM sodium phosphate buffer of pH 7.2. The incubation mixtures of 100 ~1 contained about 20 pg free sialic acids as a mixture or as individual compounds and 1-5 mU enzyme. After different incubation times (O-60 min) at 37°C lo-111 samples were applied to the column to estimate the amount of residual sialic acids or of acylmannosamines formed. For control, sialic acid concentrations after sialidase or lyase treatments were measured in some cases by microadaptions of the periodic acid/thiobarbituric acid or orcinol/Fe3+/HC1 assays ( 1). RESULTS

Sialidase

In the HPLC system applied not only sialic acids can be separated and identified, but also the different sialyllactoses studied here. The retention times are indicated in Table 1. In Fig. 1 the separation of substrate (sialyllactose), product NeuSAc, and inhibitor (NeuSAc2en) for the sialidase reaction is demonstrated as an example, showing that the quantity and nature of substances involved in sialidase action can be obtained in the same chromatographic run. Based on such analyses, the rate of hydrolysis of different sialyllactoses by VCS was investigated. Figure 2 shows that NeuSAc-cY(2-6)-lactose is hydrolyzed at about a fourfold lower rate than the cu(2-3)-isomer.

160

SHUKLA

AND SCHAUER

TABLE 1 RETENTION TIMES OF DIFFERENT SIALYLLACTOSES AND SIALIC ACIDS ON HPLC USING AMINEX A-29

Compound N-Acetylmannosamine N-Glycolylmannosamine N-Acetylneuraminyl-a(2-6)-lactose N-Acetyl-4-O-acetylneuraminylcu(2-3)-lactose N-Acetylneuraminyl-a(2-3)-lactose N-Acetyl-9-O-acetylneuraminylcY(2-3)-lactose N-Glycolylneuraminyl-a(2-3)lactose N-Acetyl-7-O-acetylneuraminic acid N-Acetylneuraminic acid N-Cilycolylneuraminic acid N-Acetyl-4Gacetylneuraminic acid N-Acetyl-9-O-acetylneuraminic acid N-Acetyl-2deoxy-2,3didehydroneuraminic acid Note. For details see Material Ref. (7).

Retention time (s) 43 49 137 147 173 205 210 316 361 480 510 530

proportions of NeuSGc, which is not well separated from Neu5,9Acz under the conditions of the experiment, but occurs only in low quantities in BSM (19). Di- and tri-o-acetylated sialic acids also known to occur in BSM (13,19) were present in the eluates only in small amounts. Acid hydrolysis of the same substrate gives a qualitatively similar but quantitatively different picture (Fig, 3a). Similar to enzymic hydrolysis, at the beginning of the reaction a pronounced Neu5,7Acz peak is visible, which disappears on further incubation (over 20 min) in favor of Neu5,9Ac2 and NeuSAc peaks. Small peaks representing diand tri-0-acetylated sialic acids were also seen. In contrast to AUS, VCS hydrolyzes O-acetylated sialic acids at slower rates, as was observed with BSM during prolonged incubation (4 days).

605 and Methods and 0.03

0-Acetylation on the side chain at O-9 also appreciably reduces the action of VCS (over 80% lower rate when compared with the non0-acetylated compound) and 0-acetylation at O-4 makes sialyllactose completely resistant against sialidase action. VCS releases NeuSGc from NeuSGc-a(2-3)-lactose at about a 10% slower rate than from the corresponding Neu5Ac derivative. To prove the applicability of HPLC for the study of sialidase action on a complex mixture of glycosidically bound sialic acids, BSM was incubated with AUS and aliquots were withdrawn from the incubation mixture after short time intervals. The results are shown in Figs. 3b and Fig. 4. At the beginning of incubation mainly Neu5,7Ac2 and oligo-0-acetylated sialic acids, besides small amounts of NeuSAc and Neu5,9Ac2, are liberated. On prolonged incubation over 5 h the Neu5,7Ac2 peak decreases while the amount of Neu5,9Acz increases. It has to be noted that the Neu5,9Ac2 peak may also contain small (< 10%) relative

0.01

I

0

16

1 min

FIG. I. Separation by HPLC of N-acetylneuraminyla(2-3)Jactose (I ), N-acetylneuraminic acid (2), N-glycolylneuraminic acid (3), 2-deoxy-2,3didehydro-N-acetylneuraminic acid (4) and 2-deoxy-2,3-didehydro-N-glycolylneuraminic acid (5) on strong anion-exchange resin. For details see Material and Methods.

STUDY OF SIALIC ACID CATABOLIC

ENZYMES

100

ti

, 10

.

M

30

LO

50

60

70

a0

ml"

FIG. 2. Enzymic hydrolysis of different sialyllactoses by cholerae sialidase followed by HPLC. Symbols: n , N-acetylneuraminyl-c(2-3)-lactose; q , N-glycolylneuraminyl-cu(2-3)Jactose; A, N-acetylneuraminyl-cu(2-6)-lactose; 0, N-acetyl-9-O-acetylneuraminyl-ff(2-3)-lactose; and 0, N-acetyl-4~O-acetylneuraminyl-a(2-3)-lactose. For experimental details see Material and Methods. Vibrio

N-Acetylneuraminate

Pyruvate-lyase

The enzymic reaction was followed by observing either the disappearance of sialic acid

2L h

2Omm &y

h p-y-,

1Omln

omn

200

161

BY HPLC

300

LOO

500

600

ILL0

In,"

FIG. 4. The nature of free sialic acids liberated from bovine submandibular gland glycoproteins by Arthrobucter ureufuciens sialidase in the course of time. The values are calculated from Fig. 3b. A, N-acetylneuraminic acid, 0, N-acetyl-7-O-acetylneuraminic acid; n, N-acetyl-90acetylneuraminic acid containing about 10% NeuSGc.

or the formation of acylmannosamine. However, when higher amounts of proteins other than the lyase were present in the incubation mixture (10 mg/ml), only the disappearance of sialic acid was taken into account for the study of enzyme reactions, as acylmannosamine and protein peaks, both eluting just after the void volume, overlap (Fig. 5). In this figure a comparison of the rates of cleavage of NeuSAc and Neu5Gc is shown by HPLC profiles. It demonstrates clearly that NeuSAc is a better substrate for this enzyme than NeuSGc. The initial rates of cleavage of different sialic acids by the lyase are depicted in Fig. 6, showing that sialic acid modification strongly influences the activity of this enzyme. While acetylation at U-4 makes sialic acid completely and at O-7 almost completely (about 3% of the cleavage rate of NeuSAc) resistant against the action of this enzyme, acetylation at O-9 makes the corresponding Neu5,9Ac2 fairly cleavable, although at only one-fourth of the rate of NeuSAc. Substitution of the N-acetyl group by a N-glycolyl residue reduces the lyase activity by about 30%.

2 L 6 0 IO min

a

b

FIG. 3. HPLC profiles of sialic acids released from bovine submandibular gland glycoproteins in dependence on time, by (a) acid hydrolysis (5 mM H2S04) and (b) enzymic hydrolysis (Arthrobacter ureafaciens sialidase). Peaks: (1) Nacetyl-7-O-acetylneuraminic acid, (2) N-acetylneuraminic acid; (3) N-acetyl-9-O-acetylneuraminic acid and N-glycolylneuraminic acid, and (4) mainly higher O-acetylated sialic acids.

DISCUSSION

The periodic acidlthiobarbituric acid assay for free sialic acids is commonly used for the study of sialidase action. A disadvantage of this method is, however, that different sialic acids have different extinction coefficients and substances like 2deoxyribose and unsaturated fatty acids interfere ( 1,19), which may lead to

162

SHUKLA

AND

wrong sialic acid and enzyme activity values. In the case of the presence of various kinds of sialic acids in substrate molecules, differentiation of sialic acids after hydrolysis by colorimetry is not possible. Distinction would only be possible by TLC or GLC. These methods, however, require extensive purification of sialic acids and relatively large amounts (20100 pg) of these sugars. The periodic acid/ thiobarbituric acid assay requires sialic acids in microgram quantities. The HPLC assay described here requires a much smaller amount of sialic acid (ng range) and overcomes most of the problems involved in the calorimetric assay discussed. Usually an aliquot of the enzyme incubation mixture can be injected directly into the column without any prior purification or concentration of sialic acids. Due to the short time required for analysis, and the possibility of direct identification of reaction products and/or substrate molecules, this method is especially suited for the follow up of enzymic reactions. The results obtained by HPLC for sialidase and N-acetylneuraminate pyruvate-lyase confirm earlier observations that both enzymes

SCHAUER

act on 0-acetylated sialic acids at slower rates when compared with the non-0-acetylated sugars (2 l-24). With the HPLC assay a much slower cleavage rate of Neu5,9Ac2 and Neu5,7Acz by the lyase was observed than was described before (22) and Neu4,5Ac2 revealed to be completely resistant towards the action of this enzyme (Fig. 6). The hydrolysis rate of Neu5,9Ac2 by sialidase from sialyllactose was also found to be much slower ( 12- 15% the rate of NeuSAc, Fig. 2) than was anticipated from earlier experiments with the less defined substrate bovine submandibular gland mucin and calorimetric sialic acid assays. This shows that direct and fast determination of the reaction products (and substrates) by HPLC is advantageous when compared with the indirect, calorimetric techniques applied earlier. The results of the calorimetric assays may be influenced by saponification and migration (25) of 0-acetyl groups due to the time and chemical conditions required for the analyses. Furthermore, the calorimetric assays are sensitive for interfering substances, often present in enzymic assays, a fact which was found to be not as critical for the HPLC assay. Corre-

X

Jill 11

1

2

12

.X

4

I!

1

0

-----iis 0

min

10min

16

IllI ---ii0 30 min

----ii 8 LO min

.9

16

SO min

FIG. 5. HPLC elution profiles after different times of the action of N-acetylneuraminate pyruvate-lyase from Clostridium perfringens on a nearly equimolar mixture of N-acetylneuraminic acid (1) and N-glycolylneuraminic acid (2). The peak of N-acylmannosamines formed is also visible (X). For details see Material and Methods.

STUDY

IO

OF SIALIC ACID CATABOLIC

20 ml”

6. Rates of cleavage of sialic acids by N-acetylneuraminate pyruvate-lyase from Clostridium perfringens followed by HPLC. Symbols: A, N-acetylneuraminic acid; n , N-glycolylneuraminic acid; 0, N-acetyl-9-O-acetylneuraminic acid; A, N-acetyl-70acetylneuraminic acid, and l , N-acetyl4O-acetylneuraminic acid. For experimental details see Material and Methods. FIG.

spondingly, HPLC was successfully applied to a variety of enzyme reactions of sialic acid metabolism, e.g., sialyltransferase (5), CMPsialate synthase, CMP-sialate hydrolase, and sialate-O-acetylesterase (26). Application of HPLC also enabled the elaboration of conditions suited for the isolation of Neu5,7Ac2 required for the lyase studies made here and for investigation of the migration of the 0-acetyl group from O-7 to O-9 (25). It is shown in Fig. 3 that at the beginning of both enzymic and acid hydrolysis of bovine submandibular gland mucin, Neu5,7Ac2 is one of the predominant sialic acids released. To obtain good yields of this sialic acid, acid hydrolysis for about 30 min is recommended, although not all sialic acids are released from mucin within this short time. Preparations of Neu5,7Ac2 and Neu4,5Ac2, respectively, can be further enriched in these sialic acids by cleavage of other, e.g., non-0-acetylated or 90-acetylated sialic acids by the lyase, which leaves Neu5,7Ac2 almost and Neu4,5Ac2 completely intact. If this reaction is carried out at about pH 5.5, migration of the 0-acetyl residue from O-7 to O-9 can be kept at a minimum (25). A comparison of the substrate specificity of sialidase and N-acetylneuraminate pyruvatelyase with regard to N- and O-substitution of sialic acid shows the following striking similarities:

ENZYMES

BY HPLC

163

O-Acetylation on the glycerol side chain of sialic acid decreases the action of both enzymes strongly and to a similar extent (cf. Figs. 2 and 6). 0-Acetylation in the pyranose ring, at C-4, makes sialic acid and its glycosides even completely resistant towards the action of these enzymes. A further similarity is the reduction of the activity of both enzymes with sialic acid, the side chain of which was reduced in length by treatment with periodate and borohydride. This observation, summarized in Ref. (23), was confirmed in the present studies. Substitution of the N-acetyl group of sialic acid by a N-glycolyl residue also reduces the reaction rates of sialidase and lyase. This was observed here and is in agreement with earlier observations obtained with other methods for the bacterial lyase (22) and various sialidases (27). The negative charge of the carboxyl group of sialic acid is required for the action of sialidase and the lyase. If the carboxyl group of sialic acid is reduced to an alcohol group or converted to an amide or ester, these sialic acids are no longer substrates for sialidases (24,28) and the lyase (24). Both enzymes act on the cu-anomer of sialic acids. This has been known for long for sialidase (23,24) and was recently elucidated for the lyase from Clostridium perfiingens (29). These observations suggest that sialidase and lyase have some similarities in the binding of substrates to their active centers. The only marked difference can be obtained with NeuSAden, an inhibitor of sialidase (30) but not of the lyase. This may be due to the fact that sialidase acts on a-glycosidically bound sialic acids, i.e., sialic acids existing in the pyranose ring form, while the lyase is assumed to interact with sialic acid having an open chain structure, resulting in binding to the amino group of a lysine residue at the beginning of the enzyme reaction (24,29,31). As NeuSAQen has ring structure it may inhibit sialidase reaction, as does free, saturated N-acetylneuraminic acid, which also exists mainly

164

SHUKLA

AND

in ring form (23). These structural features may be the reason why NeuSAQen cannot inhibit the lyase action. All these observations and considerations may stimulate suitable experiments to get insight into the mode of interaction of different sialic acids and NeuSAQen with the catabolic enzymes investigated here and also to increase the knowledge of the conformation including hydrogen bonding (12) of differently substituted, natural sialic acids in aqueous environment.

SCHAUER

and Function (Schauer, R., ed.), pp. 127-172, Springer-Verlag, Vienna. 13. Reuter, G., Pfeil. R., Stall, S., Schauer, R., Kamerling, J. P., Versluis, C., and Vhegenthart, J. F. G. (1983) Eur. J. Biochem. 134, 139-143. 14. Veh, R. W., Michalski, J.-C., Corfield, A. P., Sander, M., Gies, D., and Schauer, R. ( 198 1) J. Chromatogr. 212,313-322. 15.

REFERENCES 1. Schauer, R. (1978) in Methods in Enzymology 50, 64-89. 2. Shukla, A. K., and Schauer, R. (1982) Hoppe Seylers Z. Physiol.

Chem.

363, 255-262.

3. Krantz, M. J., and Lee, Y. C. (1975) Anal. Biochem. 63,464-469.

Silver, H. K. B., I&rim, K. A., Gray, M. J., and Salinas, P. A. (198 1) J. Chromatogr. 244, 38 I-390. 5. Bergh, M. L. E., Koppen, P., and Van den Eijnden, D. H. (1981) Carbohydr. Res. 94,255-259. 6. Shukla, A. K., Scholz, N., Reimerdes, E. H., and Schauer. R. (1982) Anal. Biochem. 122,78-82. 7. Shukla, A. K.. and Schauer, R. (1982) J. Chromatogr. 4.

244,81-89.

Shukla. A. K., Schauer, R., Unger, F. M., Zahringer, U., Rietschel, E. T., and Brade, H. (1985) Carbohydr. Rex 140, l-8. 9. Kamerling, J. P., and Vliegenthart, J. F. G. (1982) in Sialic Acids-Chemistry, Metabolism and Function (Schauer, R., ed.). pp. 95-125, Springer-Verlag, Vienna. 10. Schauer, R., Schroder, C., and Shukla, A. K. (1984) in Ganghoside Structure, Function and Biomedical Potential (Ledeen, R. W., Yu, R. K., Rapport, M. M., and Suzuki, K., eds.), pp. 75-86, Plenum Press, New York. Il. Shukla, A. K., Schauer, R., Schade, U., Mall, H., and Rietschel, E. T. (1985) J. Chromtogr. 337, 2318.

238.

12. Vliegenthart, J. F. G., Dorland, L., and Van Halbeek, H. (1982) in Sialic Acids-Chemistry, Metabolism

Biophys.

Acta 252,472-

480.

16. Corfield, A. P., Veh, R. W., Wember, M., Michalski, J.-C., and Schauer, R. (198 1) Biochem. J. 197,293299.

17. McLean, R. L., Suttajit, M., Beider, J., and Winzler, R. J. (1971) J. Biol. Chem. 246,803-809. 18. Tettamanti, G., and Pigman, W. (1968) Arch.

ACKNOWLEDGMENT The authors thank Sabine Stoll for excellent technical assistance.

Maury, P. (1971) Biochim.

Biochem.

Biophys.

124,41-50.

19. Buscher, H.-P., Casals-Stenzel, J., and Schauer, R. (1974)

Eur. J. Biochem.

50,71-82.

20. Schauer, R., and Cortield, A. P. (1982) in Sialic Acids-Chemistry, Metabolism and Function (Schauer. R.. ed.), pp. 77-94, Springer-Verlag. Vienna. 2 I. Schauer, R.. and Faillard, H. (1968) Hoppe-Seylers Z. Physiol.

Chem.

349, 96 l-968.

22. Schauer, R., Wember, M., Wirtz-Peitz, F., and Ferreira do Amaral, C. (197 1) Hoppe-Seyler’s Z. Physiol. Chem. 352, 1073-1080. 23. Corheld, A. P., and Schauer, R. (1982) in Sialic Acids-Chemistry, Metabolism and Function (Schauer, R., ed.), pp. 195-26 I, Springer-Verlag, Vienna. 24. Schauer, R. (1982) Adv. Carbohydr. Chem. Biochem. 110, 131-234.

25. Kamerling, J. P.. Van Halbeek, H., Vliegenthart, J. F. G.. Pfeil, R., Shukla, A. K., and Schauer, R. (1983) in Glycoconjugates, Proc. 7th Int. Symp. (Chester, M. A., Heinegard, D., Lundblad, A., and Svensson, S., eds.), pp. 160-16 1, Rahms, Lund. 26. Shukla, A. K., and Schauer, R. (1983) in Structural Carbohydrates in the Liver (Popper, H., Reutter, W., Gudat, F., and Kottgen, E., eds.), pp. 565-576, MTP Press, Lancaster. 27. Corfield, A. P., Veh, R. W., Wember, M., Michalski, J.-C., and Schauer, R. (198 1) Biochem. J. 197,293299.

Brossmer. R., and Holmquist, L. (197 1) Hoppe-Seyler’s Z. Physiol. Chem. 352, 1715-1719. 29. Deijl, C. M., and Vliegenthart, J. F. G. (1983) 28.

Biochem.

Biophys.

Res. Commun.

111,688-674.

30. Meindl, P., and Tuppy, H. (1969) Hoppe-Seylers Z. Physiol. Chem. 350, 1088-1092. 3 1. Schauer, R., and Wember, M. (I 97 1) Hoppe-Seyler’s Z. Physiol. Chem. 352, 1517-1523.