Charge isomers of urinary bikunin (trypsin inhibitor)

298

Biochimica et Biophysica Acta, 1203(1993) 298-303 © 1993 Elsevier SciencePublishers B.V. All rights reserved 0167-4838/93/$06.00

BBAPRO 34631

Charge isomers of urinary bikunin (trypsin inhibitor) Yoshikazu Yuki a,,, Keiichi Nomura a, Motoko Kirihara a, Maki Shimomura a, Hajime Hiratani a, Ryuichiro Nishimura b and Kazuo Kato a a Biochemistry Research Laboratories, JCR Pharmaceuticals Co., Ltd., Takatsuka-dai, Nishi-ku, Kobe 651-22 (Japan) and b Hyogo Institute for Research in Adult Diseases, 13-70, Kitaoji-cho, Akashi 673 (Japan)

(Received 25 May 1993) (Revised manuscript received 2 August 1993)

Key words: Urinary bikunin; Trypsininhibitor; Charge isomer; Glycosaminoglycan;Chondroitin sulfate It was observed that the purified urinary bikunin (trypsin inhibitor) consisted of four major isomers with different electric charges which could be separated by HPLC using a Mono Q column. These isomers revealed the same antitrypsin activity and did not show any differences in the apparent molecular weight by SDS-PAGE, amino-acid composition, N-terminal amino-acid sequence (1-40) and C-terminal amino acid (Leu). The contents of sialic acid and uronic acid were also identical among these isomers. However, analysis of chondroitin sulfate revealed all the glycosaminoglycan chains of these isomers were undersulfated, comprising nonsulfated and 4-sulfated disaccharide units, and 4-sulfated disaccharide unit ratio varied among these isomers. After the chondroitin ABC lyase digestion, all the isomers were eluted at the same position on a Mono Q column chromatography. These results indicated that charge isomers of urinary bikunin was attributed to the difference on suifation ratio in a glycosaminoglycan chain.

Introduction

Human urine contains many kinds of plasma proteins excreted passively through the glomerulus [1]. Although the concentration of urinary proteins is less than 0.01% of plasma proteins, biologically active proteins, such as hormones, enzymes and inhibitors, are secreted in native states a n d / o r in certain modified forms in the urine [2]. Some of these urinary proteins have been purified and used for the clinical therapy. Urinary bikunin is one of these bioactive proteins with inhibitory activity on trypsin, chymotrypsin, plasmin and leukocyte elastase [3-6]. The nomenclature of urinary bikunin is used in this report for customarily used urinary trypsin inhibitor according to the recommendation of Gebhard et al. [7]. The urinary bikunin, originally reported as HI-30 by Hochstrasser et al. [8], is a glycoprotein containing about 35% carbohydrate [9]. The protein moiety con* Corresponding author. Fax:+81 78 9914465. Abbreviations: SDS-PAGE, sodium dodecyi sulfate-polyacrylamide gel electrophoresis; FPLC, fast protein liquid chromatography; HPLC, high-performance liquid chromatography; ADi-0S, 2acetamido-2-deoxy-3-O-(fl-o-gluco-4-enepyranosyluronic acid)-Dgalactose; ADi-6S,2-acetamido-2-deoxy-3-O-(fl-D-gluco-4-enepyranosyl uronic acid)-6-O-sulfo-D-galactose;ADi-4S,2-acetamido2-deoxy-3-O-(/3-D-gluco-4-enepyranosyluronicacid)-4-O-sulfo-D-galactose.

sists of 143 amino acids [10] and its sequence was shown to be identical with that of the light chain of inter-a-inhibitor in human blood [7]. The bikunin was found to contain two sugar chains attached through glycosidic linkage to Ser-10 and to Asn-45, respectively [11]. It was also shown that Ser-10 carried a glycosaminoglycan chain with chondroitin sulfate [12,13]. The glycan is probably important to maintain physicochemical, physiological and biological properties of urinary bikunin in vitro and in vivo. However, these properties and carbohydrate structure of this glycan have not been characterized. In this report, it will be shown that urinary bikunin has four major charge isomers which differ each other in difference of sulfation ratio in a glycosaminoglycan chain. Materials and Methods Materials. Human urinary bikunin was highly purified by the method of Proksch et al. [14] from flesh urine pools from healthy men and women individually, followed by gel filtration on a Superose 12 column, respectively. Mono Q 5 / 5 HR, Mono Q 10/10 HR, Superose 12 and PD-10 packed columns were purchased from Pharmacia LKB Biotechnology, Uppsala, Sweden. Amine-bound Silica PA03 column (4.6 x 250 ram) was purchased from YMC, Kyoto. Benzoyl-L-

299 arginine p-nitroanilide was obtained from the Peptide Institute, Osaka, Japan. Bovine pancreatic trypsin was obtained from Boehringer-Mannheim, Mannheim, Germany. Chondroitin ABC lyase (protease free) and standard unsaturated disaccharides were obtained from Seikagaku Kogyo, Tokyo. Serine carboxypeptidase was obtained from Oriental Yeast, Tokyo. Sialidase was purchased from Nacalai Tesque, Tokyo. Other chemicals were purchased from Wako Pure Chemical Industries, Osaka. All of them were of analytical grade. Isolation of isomers. Isolation of isomers from purified human urinary bikunin was performed with a Pharmacia FPLC system consisting of two pumps, a UV detector and a controller. Purified bikunin was applied to a Mono Q 10/10 column equilibrated with 0.075 M NaCI containing 0.02 M glycine-HCl (pH 3.4) and eluted with a linear gradient of 0.075-0.5 M NaC1 at a flow rate of 3 m l / m i n . The elution pattern was monitored by measuring absorbance at 280 nm. The fractions containing each isomer were pooled, rechromatographed on the same column, followed by gel filtration on PD-10 column with distilled water for desalting. Analytical procedure. Antitrypsin activity of bikunin and each of its isomers were measured by the method

of Kassell [15] using benzoyl-L-arginine p-nitroanilide as substrate. One unit of bikunin is defined as the amount inhibiting 1 /~g of trypsin. SDS-PAGE was carried out by the method of Laemmli [16]. The sialic acid content was measured by the thiobarbituric acid assay, after the treatment of 0.1 M sulfuric acid for 30 min at 80°C as described by Aminoff [17]. Uronic acid was determined by the carbazole method [18] using o-glucuronic acid as a standard. Amino-acid analysis. Bikunin and its isomers were hydrolyzed by using a Pico Tag Workstation (Waters, Milford, MA) with 6 N HCI containing 1% (v/v) phenol in the vapor phase at 150°C for 1 h. The hydrolysates were analyzed with a Beckman 6300 amino-acid analyzer (Beckman Instruments, Fullerton, CA). The protein concentration was determined based on the result of amino-acid analysis. N-terminal sequence analysis. The sequence of 40 amino-acid residues of bikunin and each of its isomers in their N-terminal regions were determined by using a Model 473A p r o t e i n / p e p t i d e sequencer (Applied Biosystems, Foster, CA). The amount of sample applied to the sequencer was about 500 pmol. C-terminal analysis. Serine carboxypeptidase digestion was carried out in 50 mM phosphate buffer (pH

D

A 2

E O cO o,I

=o m o

,<

0

1'0

15

2'0

0

5

10

15

20

25

Retention Time (min) Fig. 1. Elution profiles of bikunin, its isomers and those digested by chondroitin ABC lyase by FPLC on a Mono Q column. (A) Purified urinary bikunin; (B) isomer 1; (C) isomer 2; (D) isomer 3; (E) isomer 4; (F) isomer 3 digested by chondroitin ABC lyase (solid line) and that followedby sialidase (dotted line). Elution profiles of other isomers after chondroitin ABC lyase digestion and both enzymes' digestion individuallywere similar to those shown in F, solid line and dotted lines, respectively.Numbers of isomers represent the elution order of these isomerson FPLC as shown in A.

300 6.0), containing 0.1% SDS for various periods at 37°C. The liberated amino acids were directly analyzed with amino-acid analyzer. Enzymatic digestion. Bikunin and each of its isomer (1 mg as polypeptide) were digested with 0.1 U of chondroitin ABC lyase in 0.05 M Tris-HC1 buffer, p H 7.9 for 2 h at 37°C [19]. After that, aliquots (0.1 mg, each) were used for determination of contents of chondroitin sulfate by H P L C as described below. Bikunin and each of its isomer (1 mg as polypeptide) was digested with 0.25 U of sialidase in 0.1 M acetate buffer (pH 5.0) for 12 h at 37°C [20]. Analysis of chondroitin sulfate. Chondroitin disaccharides produced by digestion of bikunin and of each of its isomer with chondroitin ABC lyase were analyzed according to the modified method of Sugahara et al. [21]. In brief, each digestion mixtures by chondroitin ABC lyase were precipitated with ethanol and left for 3 h at 4°C. The clear supernatants obtained by centrifugation were dried under a stream of nitrogen, and each of the residue dissolved in 10/xl of 16 m M N a H 2 P O 4 (5-50 nmol as disaccharides) was applied to the HPLC. H P L C (LC4A HPLC system, Shimadzu, Kyoto) was performed on an amine-bound Silica PA03 column using a linear gradient from 16 to 320 m M N a H 2 P O 4 over 30 min period at a flow rate of 1.0 m l / m i n at room temperature. Eluates were monitored by measuring absorbance at 232 nm. HPLC analysis. H P L C was performed on a Mono Q 5 / 5 column in a FPLC system using a linear gradient from 0.02 M glycine-HCl/0.075 M NaCI (pH 3.4), to 0.02 glycine-HC1/0.5 M NaC1 ( p H 3.4), over a 15 min period at a flow rate of 0.5 m l / m i n at room temperature. Eluates were monitored by measuring absorbance at 280 nm. Samples were diluted to the concentration of 100/xg polypeptide in 5 0 - 2 0 0 / z l of 0.075 M NaCI containing 0.02 M glycine-HC1 (pH 3.4). Results

Urinary bikunin was purified from healthy-men urine and major isomers (isomer 1, 2, 3 and 4) were separated on a Mono Q column as shown in Fig. 1 A. The total recovery by absorbance at 280 nm were about 85%. The percentual ratio of recovered isomers 1, 2, 3 and 4 was 16:38:32:14. The ratio was almost the same with that of the urinary bikunin from healthy-women urine (isomer 1 : 15%, 2 : 40%, 3 : 31%, 4 : 14%). Therefore, the isomers from men urine were analysed in the subsequent experiments. Each isomer was isolated and rechromatographed on a Mono Q column. Each of the isolated isomers was eluted on a Mono Q column as a symmetrical peak (Fig. 1B-E). The specific antitrypsin activities of bikunin and its isomers were essentially identical (original purified bikunin: 3342; isomer 1: 3415; isomer 2: 3597; isomer 3:

TABLE I Amino-acid compositions of urinary bikunin and each of its isomers Amino-acid compositions were expressed as mol per mol. Cysteine and tryptophan were not determined. The amino-acid components from complete amino-acid sequence, Ref. 10 was included in the table. Amino acid

Bikunin

Asx Thr Ser G~ Pro G~ Ala Cys Val Met Ile ~u ~r His Arg Phe ~s Trp

11.2 6.7 5.6 20.1 6.6 ~.9 8.1 . 8.5 3.9 1.9 8.6 5.4 0.1 4.8 7.0 7.1 .

.

.

Isomer 1 2 3 11.4 11.0 10.9 6.8 6.7 6.6 6.3 5.7 5.2 20.3 20.2 19.7 6.6 6.7 6.8 21.7 20.6 20.3 8.1 7.9 7.9 . . . 8.3 8.4 8.4 4.1 3.9 4.0 2.1 1.8 1.8 8.7 8.4 8.4 5.7 5.5 5.6 0.2 0.1 0.1 5.2 4.8 4.7 7.0 7.0 7.0 7.2 7.0 6.9 . . .

Ref. 10 4 11.2 6.8 5.9 20.5 6.9 21.0 7.9 8.6 4.0 2.0 8.5 5.5 0.1 4.8 7.0 7.2

11 8 6 21 6 20 8 12 9 4 2 9 7 0 5 7 7 1

3541; isomer 4 : 3 5 0 5 I U f m g of polypeptide, respectively). On SDS-PAGE, bikunin and each of its isomer migrated as a relatively broad band (data not shown). However, no significant difference on S D S - P A G E between bikunin and each isomer was found. All the apparent molecular weights were shown to be 44 000 Da. The amino-acid compositions of bikunin and each isomer are listed in Table I and these compositions were identical to those reported by Wachter et al. [10]. No significant differences in amino-acid composition were found among bikunin and its isomers. 40 N-terminal amino-acid residues of bikunin and all of its isomers coincided one another completely. Disappearance of P T H - a m i n o acid at the 10th cycles of bikunin and each isomer showed the glycosylation sites at the same positions as reported by Hochstrasser et al. [11]. The same C-terminal amino-acid residue, Leu, was found from the bikunin and each isomer. These results were consistent with the sequence of urinary bikunin reported by Wachter et al. [10]. Sialic acid and uronic acid contents of bikunin and each isomer were indicated in Table II. No significant difference of their contents among bikunin and each isomer was found. Compositions of chondroitin sulfate of bikunin and its isomers were estimated by H P L C analysis of the disaccharide units released by digestion with chondroitin ABC lyase. All disaccharide units from each isomer consisted of two peaks as shown in Fig. 2. The

301 TABLE II

TABLEIII

Sialic acid and uronic acid contents of urinary bikunin and its four

major isomers

Composition of unsaturated disaccharide of urinary bikunin and its isomers produced by chondroitin ABC lyase digestion

Contents were expressed as /zg per mg of polypeptide calculated from amino-acid analysis.

Recoveries (%) were calculated from the area of each disaccharide on a chromatogram on a PA03 column.

Sugar

Bikunin

Disaccharide

Bikunin

Isomer 1

2

3

4

Sialic acid Uronic acid

51.4 192.6

ADi-0S ADi-4S ADi-6S

69 31 0

82 18 0

73 27 0

66 34 0

58 42 0

Isomer 1

2

3

4

48.9 170.6

49.1 188.7

46.7 182.7

48.2 188.8

first and second peaks corresponded to nonsulfated (ADi-0S) and 4-sulfated disaccharide (ADi-4S) units, respectively. No peak corresponded to 6-sulfated disaccharide (ADi-6S) unit was found. Compositions of nonsulfated and 4-sulfated disaccharide units were listed in Table III. Total disaccharide units (ADi-0S + ADi-

1

1

3

E cO O,I

1

3 1

o

_3 i_

O (,0

F

2 1

0

10

20

30

10

20

30

RetentionTime (min) Fig. 2. HPLC analysis of the disaccharide units of bikunin and its isomers produced by chondroitin ABC lyase digestion. (A) Chondrodisaccharides from purified urinary bikunin; (B) chondro-disaecharides from isomer 1; (C) chondro-disaecharides from isomer 2; (D) chondro-disaecharides from isomer 3; (E) chondro-disaccharidesfrom isomer 4; (F) standard unsaturated chondro-disaceharides (5 nmol each). 1, ADi-0S; 2, ADi-6S; 3, ADi-4S.

4S) by measuring the sum of areas of these peaks on the chromatogram were almost the same per protein concentration among all the isomers (Data not shown). The results of the total disaccharide units and uronic acid contents showed that numbers of the disaccharide units are equivalent among all the isomers. However, there were significant differences in the ratios of 4sulfated disaccharide units among the isomers, as shown in Table III. Furthermore, the sulfation ratio of each isomer was considered to reflect the order in the elution pattern on a Mono Q column (Table III, Fig. 1B-E). To investigate the effect of glycosaminoglycan chain a n d / o r sialic acid on the elution position, each isomer was analyzed by FPLC on a Mono Q column after chondroitin ABC lyase digestion or both chondroitin ABC lyase and sialidase digestion, respectively. All the chondroitin ABC lyase digested isomers showed the same pattern at the same position on a Mono Q column (Fig. IF, solid line). All the both enzymes digested isomers shifted at the same position as almost single peak (but at lower concentration than that of only chondroitin ABC lyase digestion) on a Mono Q column (Fig. 1F, dotted line). To characterize the digested samples used, N-terminal amino-acid sequence (1-40) and C-terminal residue of these samples were compared. It was found that they were all identical with those of the intact isomers. The uronic acid content in all the digested samples were about 15% of that in each of the intact isomers. The result suggested that about 85% of the glycosaminoglycan chain in each isomer were removed by chondroitin ABC lyase treatment. Sialic acid content in both enzyme digested isomers were less than 1.5/zg/mg polypetide. However, elution pattern of the bikunin digested by sialidase only (sialic acid amounts after the digestion: less than 1.5 ~ g / m g polypeptide) was almost the same as compared with that of the intact bikunin (data not shown). Discussion

The urinary bikunin purified in this study was practically identical with HI-30 reported by Wachter et al.

302 [10] in regard of the activity as a trypsin inhibitor, apparent molecular weight on SDS-PAGE, the aminoacid composition, the amino-acid sequence of N-terminal regions and C-terminal amino acid. Although the bikunin showed a single profile on SDS-PAGE and size exclusion HPLC on Superose 12 (data not shown), a distinct heterogeneity was observed in the electric charge. In fact the four major isomers of urinary bikunin could be separated by ion-exchange FPLC using a Mono Q column. Since bikunin is a proteoglycan [12], such charge heterogeneity might be attributed to some different charge of its amino-acid a n d / o r carbohydrate moieties. In protein moiety, however, the amino acid composition, the amino-acid sequences in the N-terminal region (1-40) and the C-terminal amino-acid residue (Leu) of all the four isomers were completely identical with those of the original purified urinary bikunin. These results suggested that the protein moieties were essentially the same among these isomers. The bikunin was known to contain two sugar chains with a glycosaminoglycan and a N-linked glycan at positions Ser10 and Asn-45, respectively [11,13]. Sialic acids, uronic acids and chondroitin sulfates, all of which could give a strong negative charge, were present in these carbohydrate structures. The present results indicated that the contents of sialic acid,uronic acid and total disaccharide units (ADi-0S + zlDi-4S) were almost the same among the four isomers. However, all the glycosaminoglycan chains of these isomers were undersulfated and the ratio of sulfation varied among these four isomers. In addition, the sulfation ratio of each isomer reflected the difference of the retention time on a Mono Q column. These results indicated the charge heterogeneity of urinary bikunin was attributed to the difference in sulfation ratio in a glycosaminoglycan chain. However, there is also a possibility that observed charge heterogeneity of urinary bikunin was attributed to different amide contents in protein moity (different ratio of A s p / A s n a n d / o r G l u / G i n ) of the four major isomers. To confirm this possibility, we investigated the effect of glycosaminoglycan chain removal on elution positon on a Mono Q column of each four isomer. These isomers shifted to the same elution position on a Mono Q column after the chondroitin ABC lyase digestion without changing peptide structure. The result indicated the observed chrarge heterogeneity in this study was attributed not to the difference in protein moity but to the difference in a glycosaminoglycan. It was found that the digestion of bikunin and its isomers by chondroitin ABC lyase resulted in new charge heterogeneity (Fig. l-F, solid line). Sialic acids in a N-linked sugar chain probably cause the microheterogeneity. Indeed, all the isomers digested by both chondroitin ABC lyase and sialidase without changing peptide structure was eluted at the lower salt concen-

tration as almost single peak on the same Mono Q column (Fig. l-F, dotted line). Inter-a-inhibitor in blood, which is thought to be the source of urinary bikunin, consists of two distinct heavy chains and one light chain. It has been reported that amino-acid sequence and carbohydrate attachment sites of light chain of this inhibitor are identical with those of urinary bikunin [7]. Balduyck et al. [22] has shown that the light chain is a proteoglycan related to that of the urinary bikunin. Enghild et al. [23] have recently found a novel trypsin inhibitor in blood, and have named it as pre-a-inhibitor, which composes of one heavy and one light chain. They have suggested that a light chain of this inhibitor is identical with urinary bikunin. Enghild et al. [24] have also reported that the heavy and light chains of pre-a-inhibitor are covalently cross-linked by a chondroitin-4 sulfate glycosaminoglycan from the light chain. It has, therefore, been deduced that inter-a-inhibitor might contain the cross-link with two distinct heavy chains by attaching with the glycosaminoglycan chain. Although the glycosaminoglycan chain is important on assembly of these inhibitors, their carbohydrate structures have not been well known yet. Enghild et al. [24] have observed that inter-a-inhibitor or pre-a-inhibitor are extremely heterogeneous in charge. Heterogeneity of these inhibitors may be attributed to difference in chondroitin 4-sulfate ratio like urinary bikunin as shown in this paper. Analysis of glycosaminoglycans of these inhibitors will reveal whether the sulfate ratios of the glycosaminoglycan of urinary bikunin isomers are different from those of the bikunins in the blood.

Acknowledgements The authors are grateful to Mr. Hiroshi Matumoto, Mr. Toshiyuki Kotani and Mr. Masakazu Ota for their skillful assistance.

References 1 Bergard, I. (1961) Clin. Chim. Acta 6, 413-429. 2 Keutel, M.J., Hereman, G. and Light, W. (1959) Clin. Chim. Acta 4, 665-673. 3 Told, N., Maehara, S., Hayashi, M.,Ukai, R. and Sumi, H. (1980) Invest. Uroi. 17, 465-469 4 Hochstrasser, K., Albrecht, G.J., Schonberger, O.L., Rasche, B. and Lempart, K. (1981) Hoppe-Seyler's Z. Physiol. Chem. 362, 1369-1375. 5 Sumi, H., Toki, N., Takasugi, S., Maehara, S., Maruyama, M., Akazawa, K., Matsuo, O. and Mihara, H. (1982) Thromb. Haemostas. 47, 14-18. 6 Jonsson, B.-M., Loftier, C. and Ohlsson, K. (1982) Hoppe-Seyler's Physiol. Chem. 363, 1167-1175. 7 Gebhard, W., Hochstrasser, K., Fritz, H., Enghild, J.J., Pizzo, S.V. and Salvesen, G. (1990) Biol. Chem. Hoppe-Seyler Suppl. 371, 13-22. 8 Hochstrasser, K., Bretzel, G., Feuth, H., Hilla, W. and Lempart, K. (1976) Hoppe-Seyler's Z. Physiol. Chem. 357, 153-162.

303 9 Balduyck, M., H., Kerckaert, J.P., Mizon, C. and Mizon, J. (1982) Biochem. Biophys. Res. Commun. 109, 1247-1255. 10 Wachter, E. and Hochstrasser, K. (1981) Hoppe-Seyler's Z. Physiol. Chem. 362, 1351-1355. 11 Hochstrasser, K., Schonberger, O.L., Rossmanith, I. and Wachter, E. (1981) Hoppe-Seyler's Z. Physiol. Chem. 362, 1357-1362. 12 Balduyck, M., Mizon, C., Loutfi, H., Richet, C., Roussel, P. and Mizon, J. (1986) Eur. J. Biochem. 158, 417-422. 13 Chirat, F., Balduyck, M., Mizon, C., Lavoul. S., Mizon. J. and Sautiere, P. (1991) Int. J. Biochem. 231201-1203. 14 Proksch, G.J. and Routh, J.J. (1972) J. Lab. Clin. Med. 79, 491-499 15 Kassell, B. (1970) Methods Enzymol. 19, 844-852. 16 Laemmli, U.K. (1970) Nature 227, 680-682.

17 Aminoff, D. (1961) Biochem. J. 81, 384-392. 18 Bitter, M. and Muir, H.M. (1962) Anal. Biochem. 4, 330-334. 19 Yoshida, K., Miyauchi, S., Kikuchi, H., Tawada, A. and Tokuyasu, K. (1989) Anal. Biochem. 177, 327-332. 20 Nishimura, R., Ide, K., Utsunomiya, T., Kitajima, T., Yuki, Y. and Mochizuki, M. (1988) Endocrinology 123, 420-425. 21 Sugahara, K., Okumura, Y. and Yamashina, I. (1989) Biochem. Biophys. Res. Commun. 162, 189-197. 22 Balduyck, M., Laroui, S., Mizon, C. and Mizon, J. (1989) Biol. Chem. Hoppe-Seyler 370, 239-336. 23 Enghild, J.J., Thogersen. I.B., Pizzo. S.V. and Salvesen. G. (1989) J. Biol. Chem. 264, 15975-15981. 24 Enghild, J.J., Salvesen. G., Hefta, S.A., Thogersen. I.B., Rutherfurd. and Pizzo, S.V. (1991) J. Biol. Chem. 266, 747-751.