J. Biochem. 86, 1841-1850 (1979)
Isolation and Characterization of Antithrombin III from Human, Porcine and Rabbit Plasma, and Rat Serum Takehiko KOIDE Department of Biochemistry, Niigata University School of Medicine, 757, Asahimachi-dori, Niigata, Niigata 951 Received for publication, June 25, 1979
1. Human, porcine, rabbit, and rat antithrombin III have been purified by affinity chromatography using heparin-agarose. The amino acid and carbohydrate compositions, aminoterminal sequences, immunological cross-reactivities, and inhibitions of human thrombin were studied. 2. Human, porcine, rabbit, and rat antithrombin III are single-chain glycoproteins containing hexose, glucosamine, and neuraminic acid. 3. The total carbohydrate contents were 17, 16, 14, and 15% for human, porcine, rabbit, and rat antithrombin III, respectively. 4. Molecular weights estimated from the migration in sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis were 59,000, 58,000, 63,000, and 63,000 for human, porcine rabbit, and rat antithrombin HI, respectively. 5. These four proteins have similar amino acid compositions, although some minor differences were noted. 6. Human, porcine, and rabbit antithrombin III have a histidine residue at the aminoterminus, while rat antithrombin III contains an amino-terminal asparagine residue. 7. The amino-terminal sequences up to the first 17 residues showed high homology among the four proteins. 8. Some immunological cross-reactivity was observed only between human and porcine antithrombin III. 9. The apparent dissociation constants (Kt) for the complexes between human thrombin and human, porcine, rabbit, and rat antithrombin III were about 1.2X10~ 1 0 M, 9.5X10~ 9 M, 1.4 x 10"' M, and 2.8 x 10"9 M, respectively.
Antithrombin III is a plasma a2-glycoprotein and is one of the major proteinase inhibitors in mammaHan plasma, having the ability to inhibit a number of serine proteases including trypsin, chymotrypsin (1), plasmin (2), and most coagulation Abbreviations: SDS, sodium dodecyl sulfate; PTH, phenylthiohydantoin. Vol. 86, No. 6, 1979
1841
factors, such as thrombin, factors Xa, IXa, XIa, and Xlla (3), and plasma kallikrein (4-6). It has been reported that individuals with a deficiency of antithrombin III may have recurrent thromboses (1, 7), while an excess of this protein may lead to a tendency for bleeding (8). Therefore, one of the important functions of this protein may be regulation of the coagulation process.
1842
T. KOIDE
The inhibition of an enzyme by antithrombin III is reported to involve the formation of a stable 1 : 1 molar complex between the inhibitor and the enzyme (2, 9-11), with the exception of factor XIa, which bind^s two mol of inhibitor per mol of enzyme (12). It is a unique feature of antithrombin III that the inhibitory activity of this protein is greatly enhanced in the presence of heparin (9, 13,14). Antithrombin III has been purified from several species such as man (9, 15-19), rabbit (20), dog (21), cow (18,19), and horse (19). This paper describes a rapid and simple purification procedure for porcine, rabbit, and rat antithrombin III as well as human antithrombin III, by affinity chromatography on heparin-agarose, and compares their chemical and biological properties. MATERIALS AND METHODS Materials—Human plasma and rat serum were kindly provided by Dr. Tsukada of Niigata City Hospital and by Dr. Ogata of our department, respectively. Porcine and rabbit plasma were obtained from blood collected in 0.1 M sodium oxalate (one-tenth volume of anticoagulant solution). Human thrombin was prepared according to the procedure described by Lundblad et al. (22). The final product was homogeneous as judged by sodium dodecyl sulfate (SDS)-gel electrophoresis and showed a specific activity of 2,300 NIH units/ mg. Rat serum albumin and soybean trypsin inhibitor (Kunitz) used as molecular weight standard proteins were purified in our laboratory. SDS, 2-mercaptoethanol, and N,N,N',N'-tetiamethylethylenediamine were purchased from Nakarai Chemicals, Ltd., Kyoto. Acrylamide, N,N'methylenebisacrylamide, ammonium persulfate, cyanogen bromide, and 4-vinylpyridine were obtained from Wako Pure Chemical Industries, Ltd., Osaka. Heparin sodium salt (Grade I, 170 USP units/mg), fibrinogen (Type IV), dithiothreitol, phosphorylase b, ovalbumin, and bovine carbonic anhydrase were obtained from Sigma Chemical Co., St. Louis, Mo. Bio-Gel A-15m (agarose), 100-200 mesh, was purchased from Bio-Rad Laboratories, Richmond, Calif. Agarose was from Miles Laboratories Inc., Elkhart, Indiana, Freund's adjuvant was from Difco Laboratories,
Detroit, Mich., and bovine serum albumin was from Seikagaku Kogyo Co., Ltd., Tokyo. Reagents used for sequence determination were "sequanal" grade, obtained from Wako Pure Chemical Industries, and were used without further purification. All other chemicals were commercial preparations of the highest quality available. Protein Concentrations—Protein concentrations were determined from absorbance values at 280 nm, taking £ 1 1 ^ n =7.0, 9.3, 7.0, and 6.2 for human, porcine, rabbit, and rat antithrombin III, respectively. The E\%m values were calculated by comparing the protein concentration determined by amino acid analysis with the absorbance at 280 nm. SDS-Gel Electrophoresis—SDS-polyacrylamide gel electrophoresis was performed essentially by the method of Weber and Osborn (23). Samples (7-10 \i% in 10-20 //I) were subjected to electrophoresis at room temperature for 2.5-3 h in 7.5% acrylamide gels at 6mA/gel. Electrophoresis was carried out in 0.1 M Tris-phosphate buffer, pH7.0, containing 0.1% SDS. Gels were stained for protein with Coomassie Brilliant Blue R-250 according to the method of Fairbanks et al. (24). The molecular weights of the proteins were obtained by interpolation from a linear semilogarithmic plot of apparent molecular weight vs. migration distance, using the following proteins as standards: phosphorylase b (Mr=95,000), rat serum albumin (Mr=65,000), ovalbumin ( M r = 45,000), bovine carbonic anhydrase (Mr=29,000), and soybean trypsin inhibitor (Kunitz) ( M r = 20,000). Amino Acid Analyses—Amino acid analyses and preparation of samples were performed by the method of Moore and Stein (25), employing a Hitachi KLA-3B amino acid analyzer. Samples were hydrolyzed in 6 N HC1 at 110°C for 22, 44, and 66 h in evacuated, sealed tubes. The values for threonine and serine were determined by extrapolation to zero time of hydrolysis. Tryptophan was determined after alkaline hydrolysis by the method of Hugh and Moore (26). Halfcystine was determined as cysteic acid by the method of Hirs (27) and as 5-pyridylethylcysteine by the method of Friedman et al. (28). The values for other amino acids are averages of the values at the 3 different hydrolysis times unless otherwise stated.
/ . Biochem.
HUMAN, PORCINE, RABBIT, AND RAT ANTITHROMB1N HI Carbohydrate Determinations—Neutral sugar was determined by the orcinol-sulfuric acid method (29), using a 1 : 1 mixture of mannose and galactose as a standard. Glucosamine and galactosamine were determined by amino acid analysis after hydrolyzing the sample in 4 N HC1 at 100°C for 6 h in evacuated, sealed tubes. Values were corrected for destruction during acid hydrolysis. Neuraminic acid was determined by the thiobarbituric acid method of Warren (30) using JV-acetylneuraminic acid as a standard. Proteins were hydrolyzed in 1 N H 2 SO 4 at 80°C for 1 h prior to analysis. Amino-TerminalSequence—Automated Edman degradations were performed by the method of Edman and Begg (57) using a JEOL JS-47K sequence analyzer. For amino-terminal analysis, samples of about 10 mg of 5-pyridylethyl derivative of antithrombin III prepared according to the procedure of Friedman et al. (28) were employed. The semi-quantitative determination of PTHderivatives was done by measuring the absorbance at 269 nm before thin-layer chromatography. PTH-derivatives were usually identified by thinlayer chromatography on Kiesel gel F-254 sheets (32, 33). PTH-pyridylethylcysteine was identified by the method of Hermodson et al. (34). For the identification of PTH-serine, PTH-leucine, and PTH-isoleucine, and to confirm the results of thin-layer chromatography, PTH-derivatives were regenerated to the amino acids by hydrolysis at 150°C in 6 N HC1 containing 0.3% SnCl2 (35). PTH-arginine and PTH-histidine were further identified by spot tests (36, 37). Antibody Preparation—Antigens in Freund's complete adjuvant were administered subcutaneously four times at intervals of 7 to 14 days. Rabbits were immunized with human, porcine, or rat antithrombin III by the injection of about 300 [ig of the purified protein each time, whereas rats were immunized with rabbit antithrombin III by the injection of about 150 fig of the purified protein each time. Antisera were collected 1 week after the last injection, and saturated ammonium sulfate, adjusted to pH 7.8, was added to give 3 3 % saturation. The pellet obtained by centrifugation was dissolved in half of the original volume of distilled water and extensively dialyzed against 0.15 M NaCl, adjusted to pH 7.8, containing 0.1 % NaN 3 . The /--globulin fractions thus obtained were stored at —20°C until use.
Vol. 86, No. 6, 1979
1843
Immunologic Assay—Precipitin reactions were carried out in 1 % agarose gel by the Ouchterlony method. Assay of Thrombin and Inhibitor Activities— Thrombin was dissolved in 0.05 M Tris-HCl, 0.15 M NaCl, pH 7.4, containing 1 mg/ml of bovine serum albumin. Coagulation activity of thrombin was assayed as described by Lundblad et al. (22). Antithrombin III activity was determined as follows. For the determination of heparin cofactor activity, 50 ^1 of antithrombin III appropriately diluted with 0.05 M Tris-HCl, 0.15 M NaCl, pH 7.4, was mixed with an equal volume of heparin (60300 units/ml) dissolved in the same buffer, and incubated at 37°C for 1 min. Fifty [A of thrombin (100-500 /ig/ml, 2,300 NIH units/mg) was added to the preincubated sample, and at various times of incubation, an aliquot (5 or 10 /al) was removed, diluted, and added to 0.3 ml of the assay mixture (22) containing 0.4 % fibrinogen, then the residual thrombin activity was assayed. For the determination of progressive antithrombin III activity, buffer was substituted for heparin. All coagulation assays were performed at least in triplicate. Preparation of Heparin-Agarose—Heparinagarose was prepared by the cyanogen bromide method of Cuatrecasas (38) as described by Fujikawa et al. (39). Isolation of Antithrombin HI—In this procedure, all operations were performed at 4°C and contact with glass was avoided by employing plastic columns, tubes, and containers. Onethirtieth volume of 1 M BaCI2 was added to human acid-citrate-dextrose plasma, containing 0.1 IDM benzamidine, and the mixture was stirred for 20 min then centrifuged at 7,800 x g for 20 min in a Sorvall RC3 centrifuge. The supernatant was made up to 0.1 mM in ethylenediaminetetraacetic acid and then brought to 50% saturation by the slow addition of solid ammonium sulfate. After stirring for 20 min the precipitate was removed by centrifugation at 7,800xg for 20min. The supernatant was brought to 85% saturation with solid ammonium sulfate, and after stirring for 1 h, the suspension was centrifuged for 30 min at 10,000xg in a Hitachi 18PR-3 centrifuge. The precipitate was redissolved in distilled water and extensively dialyzed against 0.02 M phosphate buffer, pH 6.3, containing 0.4 M NaCl at 4°C.
1844
T. KOIDE
The dialyzed sample was applied to a heparinagarose column (6.8 x l l cm), equilibrated with the same buffer. After extensive washing of the column with the equilibration buffer, followed by 1 M NaCl in 0.02 M phosphate buffer, pH 6.3, antithrombin III was eluted with 2 M NaCl in 0.02 M phosphate buffer, pH 6.3. Porcine, rabbit, and rat antithrombin III were purified by similar procedures. The average yields of pure protein from 1 liter of plasma or serum were 40, 43, 52, and 41 mg for human, porcine, rabbit, and rat antithrombin III, respectively. RESULTS SDS-Polyacrylamide Gel Electrophoresis of Human, Porcine, Rabbit, and Rat Antithrombin III
—SDS-polyacrylamide gel electrophoresis patterns of human, porcine, rabbit, and rat antithrombin III are shown in Fig. 1. Each antithrombin III migrated as a single protein band in the presence and absence of reducing agent, and there was little difference in the migration of the protein before and after reduction. These results are consistent with the view that each antithrombin III consists of a single polypeptide chain. The apparent molecular weights estimated from the migration of reduced protein in 7.5% polyacrylamide gels were 59,000, 58,000, 63,000, and 63,000 for human, porcine, rabbit, and rat antithrombin III, respectively. Amino Acid and Carbohydrate Compositions of Human, Porcine, Rabbit, and Rat Antithrombin HI—The amino acid and carbohydrate compositions of human, porcine, rabbit, and rat anti-
"^ m
95,000 65,000
~
45,000
—
29,000 20,000
1 2
3 4
5 6
7 8
Fig. 1. SDS-polyacrylamide gel electrophoresis of human, porcine, rabbit, and rat antithrombin III. Samples (about 10 fig) were dissolved in 10 ftl of 0.2 M Tris-phosphate buffer, pH 7.0, containing 0.2% sodium dodecyl sulfate, and heated at 100°C for 2 min in the presence or absence of 5 ft] of 2-mercaptoethanol. The samples were applied to a 7.5% gel column (6x95 mm), and electrophoresis was performed as described in " MATERIALS AND METHODS." Gels 1 and 2 are nonreduced and reduced human antithrombin III, respectively. Gels 3 and 4 are nonreduced and reduced porcine antithrombin i n , respectively. Gels 5 and 6 are nonreduced and reduced rabbit antithrombin ID, respectively. Gels 7 and 8 are nonreduced and reduced rat antithrombin III, respectively. The standard gel on the right contains, from top to bottom, phosphorylase b, rat serum albumin, ovalbumin, bovine carbonic anhydrase, and soybean trypsin inhibitor (Kunitz). The anode was at the bottom of the gel.
/. Biochem.
HUMAN, PORCINE, RABBIT, AND RAT ANTITHROMBIN III
1845
TABLE I. Amino acid and carbohydrate compositions of human, porcine, rabbit, and rat antithrombin HI. Values are given as numbers of residues/mol of glycoprotein. Numbers in parentheses are expressed as weight percent. Components
Human
Porcine
Rabbit
Rat
Amino acid (%) Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Half-cystine Valine Methionine Isoleucine Leucine Tyrosine Phenylalanine Tryptophan Lysine Histidine Arginine
(83. 1) 44.2 23.4
(83. 6) 52.3 22.9 30.9 53.3 19.9 18.0 30.2 6.2* 28.3
(85. 8) 53.1 26.6 31.5 47.5 22.8 18.0 31.6 6.4a
(84. 6) 51.8 20.9 29.5 52. 1 24.2 18.1 26.6
29.8 11.2 22.5 48.1 11.3 26.7
34.2
32.8 49.6 23.1 20.0 29. 1 6.U
28.9 10.9 20.5 38.0 10. 1 24. 1
Carbohydrate (%) Hexose Glucosamine Neuraminic acid 1
Determined as cysteic acid.
b
23.7 39. 1 10.0 24.9
11.0 22.0 41.2 8.0 26.4
3.8
5.0
3.8
3.8
32.4
32.5
40.9
33.2
6,2
6.0
7.1
5.8
20.3
23.9
27.3
20.2
(16.9) 18.5 (5.2) 17.0 (6.5) 10.2 (5.2)
(16.4) 17.9 (4.7) 15.4 (5.5) 12.7 (6.1)
(14.2) 11.6 (3.0) 14.6 (5.2) 12.7 (6.0)
(15.4) 11.5 (3.2) 20.0 (7.5)
9.3 (4.7)
Determined as 5-pyridylethylcysteine
thrombin III are shown in Table I. The compositions of the four antithrombin III molecules are very similar, although some minor differences exist. Human, porcine, rabbit, and rat antithrombin III are glycoproteins containing hexose, Af-acetylglucosamine, and N-acetylneuraminic acid. No galactosamine was detected in any of the four proteins. The total carbohydrate contents were 17,16,14, and 15% for human, porcine, rabbit, and rat antithrombin III, respectively. No free sulfhydryl group was detected in any of the antithrombin III molecules by Ellman's procedure, indicating that the six half-cystine residues are present as three disulfide bonds in each protein. Vol. 86, No. 6, 1979
9.2
5.9*>
Amino-Tenninal Sequences of Human, Porcine, Rabbit, and Rat Antithrombin HI—The aminoterminal sequences of the S-pyridylethylated derivatives of human, porcine, rabbit, and rat antithrombin III are shown in Fig. 2. The similar sequences of bovine and horse antithrombin III are also included for comparison. Human, porcine, rabbit, bovine, and horse antithrombin III each contained amino-terminal histidine. Rat antithrombin III, however, contained a'minoterminal asparagine. It is clear from Fig. 2 that the antithrombin III molecules from all six different species have homologous sequences in the aminoterminal region of the molecule. In particular, porcine and horse antithrombin III molecules have
1846
T. KOIDE 1 Rat
5 10 15 Asn Pro Val Asp Asp H e Cys H e Ala Lys Pro Arg Asp H e Pro
Rabbit
His Glu Ser Pro Val Glu Asp H e Cys Thr Ala Lys Pro Arg Asp Phe Pro
Porcine
His Trp Ser Pro Val Glu Asp H e Cys Thr Ala Lys Pro Arg Asp H e Pro
Human
His Gly Ser Pro Val
Asp H e Cys Thr Ala Lys Pro Arg Asp He
Pro
Bovine
His Arg Ser Pro Val Glu Asp Val Cys Thr Ala Lys Pro Arg Asp He
Pro
Horse
His Trp Ser Pro Val Glu Asp H e Cys Thr Ala Lys Pro Arg Asp H e Pro
Fig. 2. Amino-terminal sequences of rat, rabbit, porcine, and human antithrombin III. The sequences of bovine and horse antithrombin III determined by Kurachi et al. (19) are also shown for comparison. Amino acid residues that are identical are shown in blocks. Dash indicates space that has been inserted to bring human antithrombin III into alignment for better homology with the other antithrombin III sequences.
Fig. 3. Ouchterlony double immunodifFusion patterns in 1% agarose gels. The center wells contained: (A) rabbit anti-human antithrombin HI, (B) rabbit anti-porcine antithrombin III, (C) rat anti-rabbit antithrombin III, (D) rabbit anti-rat antithrombin HI. Surrounding wells contained: antithrombin III from man (1), pig (2), • rabbit (3), and rat (4). / . Biochem.
1847
HUMAN, PORCINE, RABBIT, AND RAT ANTITHROMBIN III the same sequence in the first 17 amino-terminal residues. Some minor differences in the sequences of the six proteins, however, are evident. The amino acid residue in the second position was highly variable, following the numbering system shown in Fig. 2. Human antithrombin III contained glycine, porcine and horse antithrombin III contained tryptophan, rabbit antithrombin III contained glutamic acid, and bovine antithrombin III contained arginine. Rat antithrombin III lacked the amino-terminal two residues, and the amino-terminal asparagine in rat antithrombin III replaced serine in position 3 in the other five antithrombin III molecules. Position 6 contained an aspartic acid residue in rat antithrombin III, while glutamic acid was present in porcine, rabbit, bovine, and horse antithrombin III, and in human antithrombin III it was necessary to omit a residue in position 6 in order to bring this protein into alignment for greater homology with the other antithrombin III molecules. Other minor differences were noted in position 8, where isoleucine was replaced by valine in bovine antithrombin III, in position 10, where threonine was replaced by isoleucine in rat antithrombin III, and also in position 16, where isoleucine was replaced by phenylalanine in rabbit antithrombin III. The remainder of the first 17 residues was the same for all six proteins.
Immunological Cross-Reactivity—Immunological cross-reactivity was examined by Ouchterlony's double diffusion method. Figure 3 shows immunodiffusion patterns obtained when rabbit or rat antiserum in the center well was reacted with human, porcine, rabbit, and rat antithrombin III in the surrounding wells. Antisera to rabbit and rat antithrombin III formed a strong immune precipitate only with the corresponding specific antigen, and gave no cross-reactivity with the material from other species. Antiserum to human antithrombin III reacted not only with human antithrombin III, but also with porcine antithrombin III, and antiserum to porcine antithrombin III also reacted strongly with both human and porcine antithrombin III. These results indicate partial immunologic identity between human and porcine antithrombin III. Inhibition of Human Thrombin by Human, Porcine, Rabbit, and Rat Antithrombin 111—The inhibition of the clotting activity of human thromVol. 86, No. 6, 1979
bin by human, porcine, rabbit, and rat antithrombin III was studied at 37°C. The inhibition proceeded time-dependently in the absence of heparin, reaching equilibrium in about 15 min after mixing the inhibitor with thrombin. The inhibition, however, was complete almost instantaneously in the presence of heparin. The inhibition curves of human thrombin by human, porcine, rabbit, and rat antithrombin III in the absence of heparin are shown in Fig. 4. The clotting activity of human thrombin was strongly inhibited by human, porcine, or rat antithrombin III, and less strongly by rabbit antithrombin III. Extrapolation of the inhibition curves in Fig. 4 to 100% inhibition gives an enzymeinhibitor ratio of 1 in each case, indicating that antithrombin III interacts with thrombin to form an equimolar complex. Values for the apparent dissociation constant (Ki) calculated by the method
0
"0
Antithrombin IE Cone. (pM) 4.5 9.0
1.0
2.0
[Antithrombin] / [Thrombin] (mole/mole) Fig. 4. Inhibition of human thrombin by human, porcine, rabbit, and rat antithrombin III. The reaction mixture contained 166 //g/ml of human thrombin (2,300 NIH units/mg) in 0.05 M Tris-HCI, 0.15 M NaCl, pH 7.4, containing 0.33 mg/ml of bovine serum albumin, and various concentrations of either human (#), porcine (O), rabbit (A), or rat ( A ) antithrombin III. The mixture was incubated at 37°C for 15 min, then an aliquot (5 or 10 fi\) was removed, diluted and assayed for residual clotting activity of thrombin.
1848
T. KOIDE
of Green and Work (40) for the complex between human thrombin and human, porcine, rabbit, and rat antithrombin III were calculated to be about
lactosamine in porcine, rabbit, and rat antithrombin III as well as in the human preparation is consistent with the results of Miller-Andersson et al. (17), Nordenman et al. (41), and Danishefsky 1 . 2 X 1 0 - 1 0 M , 9 . 5 X 1 0 - 9 M , 1 . 4 X 1 0 - ' M , and 2.8 x et al. (42). Although rabbit antithrombin III was 10~9 M, respectively. also purified by Yin et al. (20), they did not report the amino acid and carbohydrate compositions but DISCUSSION only gave the contents of hexose (4.1%) and Affinity chromatography on heparin-agarose was neuraminic acid (4.6%). first introduced by Miller-Andersson et al. (17) for Histidine was reported to be the aminothe isolation of human antithrombin III. The terminal residue in human (17-19), bovine (19), same or a similar procedure was successfully used and horse (79) antithrombin III. Amino-terminal for the purification of canine (27), bovine (18, 19, histidine was also found in porcine, and rabbit 41), and horse (79) antithrombin III. In the antithrombin III. Rat antithrombin III, however, present studies, the utility of heparin-agarose was had asparagine as the amino-terminal amino acid. extended to the purification of porcine, rabbit, This residue corresponds to the third position of and rat antithrombin III. other antithrombin III molecules in terms of the Porcine, rabbit, and rat antithrombin III as homology between antithrombin III, which shows well as the previously reported human, bovine, that rat antithrombin III lacks the first two amino and horse antithrombin III have been shown to acid residues at the amino-terminal end. Whether be glycoproteins composed of a single polypeptide this is because the starting material for rat antichain. The amino acid and carbohydrate com- thrombin III was not plasma but serum is not positions of the four antithrombin III prepara- clear. There was no difference in the aminotions reported here are very similar. It is particu- terminal amino acid sequence between rabbit larly noteworthy that the presence of 6 mol of antithrombin III samples isolated from plasma half-cystine as 3 disulfide bridges is common and serum. The amino-terminal sequence of among the four proteins. This value of cystine human antithrombin III was found to be the same was also reported by Kurachi et al. (19) for their as that reported by Kurachi et al. (19) and Magnuhuman, bovine, and horse preparations. Amino sson et al. (43). The remarkable similarity of the acid and carbohydrate compositions of human amino-terminal sequence in human, bovine, and antithrombin III have been reported by several horse antithrombin III shown by Kurachi et al. investigators (16, 17, 19). The amino acid com- (19) was extended to porcine, rabbit, and rat position of our human preparation is in reasonable antithrombin III. Although there are some agreement with those reported by Heimburger differences in amino acid residues at a few posiet al. (16), Miller-Andersson et al. (17), and Kurachi tions, they can be attributed to single base reet al. (19), with the sole exception of the half- placements. No sequence similarity was found cystine content. Miller-Andersson et al. (17) between the amino-terminal region of antithrombin reported no half-cystine in their preparation. The III and those of other plasma proteinase inhibitors presence of cystine in antithrombin III, however, such as alpha-1-antitrypsin (44-46) and alpha-1is clear from the amino-terminal sequence presented antichymotrypsin (46). by Kurachi et al. (19) and in the present studies. Immunological cross-reactivities among antiThe carbohydrate composition of human anti- thrombin III from different species were studied thrombin III is rather inconsistent among investi- by Kurachi et al. (19) and Nordenman et al. (41). gators (16, 17, 19, 41, 42). This may due to the The former investigators reported partial immunorelatively low accuracy of carbohydrate analysis logical cross-reactivity between human and bovine methods available. The major differences are antithrombin III. The latter, however, found no that Miller-Andersson et al. (17) reported the cross-reactivity between the two proteins. In our absence of neuraminic acid in their preparation, preparations immunological cross-reactivity was in contrast to the other reports, including the only observed between human and porcine antipresent studies. The absence of TV-acetyl ga- thrombin III in immunodiffusion experiments, and / . Biochem.
HUMAN, PORCINE, RABBIT, AND RAT ANTITHROMBIN HI no cross-reactivity was observed for rabbit or rat antithrombin III with the other three antithrombin III preparations. The degrees of similarity of the amino-terminal sequences in human, porcine, rabbit, and rat antithrombin III were also consistent with the antigenic similarities among antithrombin III preparations observed by double immunodiffusion, and paralleled the closeness of the taxonomic relationship of the animals used as sources. The inhibitions of human thrombin by human, porcine, rabbit, and rat antithrombin III are almost instantaneous in the presence of heparin and proceed progressively in the absence of heparin, reaching equilibrium about 15 min after mixing the two proteins. These observations are consistent with those of many investigators (9,13,14). Complex formation in a one-to-one molar ratio between human thrombin and human, porcine, rabbit, or rat antithrombin III is also consistent with the many reported results (2, 9-11). Species specificities in the inhibition of coagulation factors by antithrombin III have been less well studied. Kurachi et al. (79) showed that bovine antithrombin III has the highest affinity for bovine factor Xa/3, while that of human antithrombin III is lower, and that of horse antithrombin III is lower still. Nordenman et al. (41) reported that human and bovine antithrombin III inhibit human thrombin to the same extent, although they presented no quantitative data. The present study shows that the affinity of antithrombin III for human thrombin is highest in the case of the human preparation and lowest with the rabbit preparation, while rat antithrombin III shows a higher affinity for human thrombin than the porcine preparation. Thus, the order of species specificity in the inhibitory activity is not as consistent with the closeness of taxonomic relationships of the animals as are the sequence and antigenic similarities among the antithrombin III preparations. The author is grateful to Drs. T. Tsukada and K. Ogata for supplying human plasma and rat serum, respectively. The author is also grateful to Professor T. Ikenaka for his generous help. REFERENCES 1. Abildgaard, U. & Egeberg, O. (1968) Scand. J. Vol. 86, No. 6, 1979
1849
Haematol. 5, 155-157 2. Highsmith, R.F. & Rosenberg, R.D. (1974) /. Biol. Chem. 249, 4335-4338 3. Davie, E.W. & Hanahan, D.J. (1977) in The Plasma Proteins (Putnam, F.W., ed.), Vol. Ill, pp. 421-544, Academic Press, New York 4. Burrowes, C.E., Habal, F.M., & Movat, H.Z. (1975) Thromb. Res. 7, 175-183 5. Vennerod A.M. & Laake K. (1975) Thromb. Res. 7, 223-226 6. Lahiri, B., Bagdasarian, A., Mitchell, B., Talamo, R.C., Colman, R.W., & Rosenberg, R.D. (1976) Arch. Biochem. Biophys. 175, 737-747 7. Egeberg, O. (1965) Thromb. Diath. Haemorrh. 13, 516-520 8. Robinson, A.J., Aggeler, P.M., McNicol, G.P., & Douglas, A.S. (1967) Br. J. Haematol. 13, 510-527 9. Rosenberg, R.D. & Damus, P.S. (1973) / . Biol. Chem. 248, 6490-6505 10. Kurachi, K., Fujikawa, K., Schmer, G., & Davie, E.W. (1976) Biochemistry 15, 373-377 11. Fujikawa, K., Kurachi, K., & Davie, E.W. (1977) Biochemistry 16, 4182-4188 12. Kurachi, K. & Davie, E.W. (1977) Biochemistry 16, 5831-5839 13. Monkhouse, F.C., France, E.S., & Seegers, W.H. (1955) Circ. Res. 3, 397^(02 14. Yin, E.T., Wessler, S., & Stoll, P.J. (1971) /. Biol. Chem. 246, 3703-3711 15. Fagerol, M.K. & Abildgaard, U. (1970) Scand. J. Haematol. 7, 10-17 16. Heimburger, N., Haupt, H., & Schwick, H.G. (1971) in Proc. Int. Res. Conf. Proteinase Inhibitors (Fritz, H. & Tschesche, H., eds.) pp. 1-22, Walter de Gruyter, Berlin 17. Miller-Andersson, M., Borg, H., & Andersson, L-O. (1974) Thromb. Res. 5, 439-452 18. Thaler, E. & Schmer, G. (1975) Br. J. Haematol. 31, 233-243 19. Kurachi, K., Schmer, G., Hermodson, M.A., Teller, D.C., & Davie, E.W. (1976) Biochemistry 15, 368373 20. Yin, E.T., Wessler, S., & Stoll, P.J. (1971) /. Biol. Chem. 1A6, 3694-3702 21. Damus, P.S. & Wallace, G.A. (1974) Biochem. Biophys. Res. Commun. 61, 1147-1153 22. Lundblad, R.L., Kingdon, H.S., & Mann, K.G. (1976) in Methods in Enzymology (Lorand, L., ed.) Vol. XLV, pp. 156-176, Academic Press, New York 23. Weber, K. & Osborn, M. (1969) / . Biol. Chem. 244, 4406-4412 24. Fairbanks, G., Steck, T.L., & Wallach, D.F.H. (1971) Biochemistry 10, 2606-2617
1850 25. Moore, S. & Stein, W.H. (1963) in Methods in Enzymology (Colowick, S.P. & Kaplan, N.O., eds.) Vol. VI, pp. 819-831, Academic Press, New York 26. Hugli, T.E. & Moore, S. (1972) J. Biol. Chem. 241, 2828-2834 27. Hirs, C.H.W. (1967) in Methods in Enzymology (Hirs, C.H.W., ed.) Vol. XI, pp. 197-199, Academic Press, New York 28. Friedman, M., Krull, L.H., & Cavins, J.F. (1970) J. Biol. Chem. 245, 3868-3871 29. Winzler, R.J. (1955) in Methods of Biochemical Analysis (Glick, D., ed.) Vol. 2, pp. 279-311, Interscience Publ. Inc., New York 30. Warren, L. (1959) /. Biol. Chem. 234, 1971-1975 31. Edman, P. & Begg, G. (1967) Eur. J. Biochem. 1, 80-91 32. Brenner, M., Niederwieser, A., & Pataki, G. (1969) in Thin-Layer Chromatography (Stahl, E., ed., Aschworth, M.R.F., translation) pp. 730-786, Springer-Verlag, Berlin 33. Jeppson, J.-O. & Sjoquist, J. (1967) Anal. Biochem. 18, 264-269 34. Hermodson, M.A., Ericsson, L.H., Titani, K., Neurath, H., & Walsh, K.A. (1972) Biochemistry 11, 4493^1502 35. Mendez, E. & Li, C.Y. (1975) Anal. Biochem. 68, 47-53
T. KOIDE 36. Easley, C , Zegers, B.J.M., & DeVijlder, M. (1969) Biochim. Biophys. Ada 175, 211-213 37. Easley, C. (1965) Biochim. Biophys. Ada 107, 386388 38. Cuatrecasas, P. (1970) /. Biol. Chem. 245, 3059-3065 39. Fujikawa, K., Thompson, A.R., Legaz, M.E., Meyer, R.C., & Davie, E.W. (1973) Biochemistry 12, 4938-4945 40. Green, N.M. & Work, E. (1953) Biochem. J. 54, 347-352 41. Nordenman, B., Nystrdm, C , & Bjork, I. (1977) Eur. J. Biochem. 78, 195-203 42. Danishefsky, I., Zweben, A., & Slomiany, B.L. (1978) J. Biol. Chem. 253, 32-37 43. Magnusson, S., Sottrup-Jensen, L., Petersen, T.E., Dudek-Wojciechowska, G., & Claeys, H. (1976) in Proteolysis and Physiological Regulation (Robbins, D.W. & Brew, K., eds.) pp. 203-238, Academic Press, New York 44. Morii, M., Odani, S., Koide, T., & Ikenaka, T. (1978) /. Biochem. 83, 269-277 45. Owen, M.C., Lorier, M., & Carrell, R.W. (1978) FEBS Lett. 88, 234-236 46. Travis, J., Garner, D., & Bowen, J. (1978) Biochemistry 17, 5647-5651
/ . Biochem.
Isolation and Characterization of Antithrombin III from Human, Porcine and Rabbit Plasma, and Rat Serum Takehiko KOIDE Department of Biochemistry, Niigata University School of Medicine, 757, Asahimachi-dori, Niigata, Niigata 951 Received for publication, June 25, 1979
1. Human, porcine, rabbit, and rat antithrombin III have been purified by affinity chromatography using heparin-agarose. The amino acid and carbohydrate compositions, aminoterminal sequences, immunological cross-reactivities, and inhibitions of human thrombin were studied. 2. Human, porcine, rabbit, and rat antithrombin III are single-chain glycoproteins containing hexose, glucosamine, and neuraminic acid. 3. The total carbohydrate contents were 17, 16, 14, and 15% for human, porcine, rabbit, and rat antithrombin III, respectively. 4. Molecular weights estimated from the migration in sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis were 59,000, 58,000, 63,000, and 63,000 for human, porcine rabbit, and rat antithrombin HI, respectively. 5. These four proteins have similar amino acid compositions, although some minor differences were noted. 6. Human, porcine, and rabbit antithrombin III have a histidine residue at the aminoterminus, while rat antithrombin III contains an amino-terminal asparagine residue. 7. The amino-terminal sequences up to the first 17 residues showed high homology among the four proteins. 8. Some immunological cross-reactivity was observed only between human and porcine antithrombin III. 9. The apparent dissociation constants (Kt) for the complexes between human thrombin and human, porcine, rabbit, and rat antithrombin III were about 1.2X10~ 1 0 M, 9.5X10~ 9 M, 1.4 x 10"' M, and 2.8 x 10"9 M, respectively.
Antithrombin III is a plasma a2-glycoprotein and is one of the major proteinase inhibitors in mammaHan plasma, having the ability to inhibit a number of serine proteases including trypsin, chymotrypsin (1), plasmin (2), and most coagulation Abbreviations: SDS, sodium dodecyl sulfate; PTH, phenylthiohydantoin. Vol. 86, No. 6, 1979
1841
factors, such as thrombin, factors Xa, IXa, XIa, and Xlla (3), and plasma kallikrein (4-6). It has been reported that individuals with a deficiency of antithrombin III may have recurrent thromboses (1, 7), while an excess of this protein may lead to a tendency for bleeding (8). Therefore, one of the important functions of this protein may be regulation of the coagulation process.
1842
T. KOIDE
The inhibition of an enzyme by antithrombin III is reported to involve the formation of a stable 1 : 1 molar complex between the inhibitor and the enzyme (2, 9-11), with the exception of factor XIa, which bind^s two mol of inhibitor per mol of enzyme (12). It is a unique feature of antithrombin III that the inhibitory activity of this protein is greatly enhanced in the presence of heparin (9, 13,14). Antithrombin III has been purified from several species such as man (9, 15-19), rabbit (20), dog (21), cow (18,19), and horse (19). This paper describes a rapid and simple purification procedure for porcine, rabbit, and rat antithrombin III as well as human antithrombin III, by affinity chromatography on heparin-agarose, and compares their chemical and biological properties. MATERIALS AND METHODS Materials—Human plasma and rat serum were kindly provided by Dr. Tsukada of Niigata City Hospital and by Dr. Ogata of our department, respectively. Porcine and rabbit plasma were obtained from blood collected in 0.1 M sodium oxalate (one-tenth volume of anticoagulant solution). Human thrombin was prepared according to the procedure described by Lundblad et al. (22). The final product was homogeneous as judged by sodium dodecyl sulfate (SDS)-gel electrophoresis and showed a specific activity of 2,300 NIH units/ mg. Rat serum albumin and soybean trypsin inhibitor (Kunitz) used as molecular weight standard proteins were purified in our laboratory. SDS, 2-mercaptoethanol, and N,N,N',N'-tetiamethylethylenediamine were purchased from Nakarai Chemicals, Ltd., Kyoto. Acrylamide, N,N'methylenebisacrylamide, ammonium persulfate, cyanogen bromide, and 4-vinylpyridine were obtained from Wako Pure Chemical Industries, Ltd., Osaka. Heparin sodium salt (Grade I, 170 USP units/mg), fibrinogen (Type IV), dithiothreitol, phosphorylase b, ovalbumin, and bovine carbonic anhydrase were obtained from Sigma Chemical Co., St. Louis, Mo. Bio-Gel A-15m (agarose), 100-200 mesh, was purchased from Bio-Rad Laboratories, Richmond, Calif. Agarose was from Miles Laboratories Inc., Elkhart, Indiana, Freund's adjuvant was from Difco Laboratories,
Detroit, Mich., and bovine serum albumin was from Seikagaku Kogyo Co., Ltd., Tokyo. Reagents used for sequence determination were "sequanal" grade, obtained from Wako Pure Chemical Industries, and were used without further purification. All other chemicals were commercial preparations of the highest quality available. Protein Concentrations—Protein concentrations were determined from absorbance values at 280 nm, taking £ 1 1 ^ n =7.0, 9.3, 7.0, and 6.2 for human, porcine, rabbit, and rat antithrombin III, respectively. The E\%m values were calculated by comparing the protein concentration determined by amino acid analysis with the absorbance at 280 nm. SDS-Gel Electrophoresis—SDS-polyacrylamide gel electrophoresis was performed essentially by the method of Weber and Osborn (23). Samples (7-10 \i% in 10-20 //I) were subjected to electrophoresis at room temperature for 2.5-3 h in 7.5% acrylamide gels at 6mA/gel. Electrophoresis was carried out in 0.1 M Tris-phosphate buffer, pH7.0, containing 0.1% SDS. Gels were stained for protein with Coomassie Brilliant Blue R-250 according to the method of Fairbanks et al. (24). The molecular weights of the proteins were obtained by interpolation from a linear semilogarithmic plot of apparent molecular weight vs. migration distance, using the following proteins as standards: phosphorylase b (Mr=95,000), rat serum albumin (Mr=65,000), ovalbumin ( M r = 45,000), bovine carbonic anhydrase (Mr=29,000), and soybean trypsin inhibitor (Kunitz) ( M r = 20,000). Amino Acid Analyses—Amino acid analyses and preparation of samples were performed by the method of Moore and Stein (25), employing a Hitachi KLA-3B amino acid analyzer. Samples were hydrolyzed in 6 N HC1 at 110°C for 22, 44, and 66 h in evacuated, sealed tubes. The values for threonine and serine were determined by extrapolation to zero time of hydrolysis. Tryptophan was determined after alkaline hydrolysis by the method of Hugh and Moore (26). Halfcystine was determined as cysteic acid by the method of Hirs (27) and as 5-pyridylethylcysteine by the method of Friedman et al. (28). The values for other amino acids are averages of the values at the 3 different hydrolysis times unless otherwise stated.
/ . Biochem.
HUMAN, PORCINE, RABBIT, AND RAT ANTITHROMB1N HI Carbohydrate Determinations—Neutral sugar was determined by the orcinol-sulfuric acid method (29), using a 1 : 1 mixture of mannose and galactose as a standard. Glucosamine and galactosamine were determined by amino acid analysis after hydrolyzing the sample in 4 N HC1 at 100°C for 6 h in evacuated, sealed tubes. Values were corrected for destruction during acid hydrolysis. Neuraminic acid was determined by the thiobarbituric acid method of Warren (30) using JV-acetylneuraminic acid as a standard. Proteins were hydrolyzed in 1 N H 2 SO 4 at 80°C for 1 h prior to analysis. Amino-TerminalSequence—Automated Edman degradations were performed by the method of Edman and Begg (57) using a JEOL JS-47K sequence analyzer. For amino-terminal analysis, samples of about 10 mg of 5-pyridylethyl derivative of antithrombin III prepared according to the procedure of Friedman et al. (28) were employed. The semi-quantitative determination of PTHderivatives was done by measuring the absorbance at 269 nm before thin-layer chromatography. PTH-derivatives were usually identified by thinlayer chromatography on Kiesel gel F-254 sheets (32, 33). PTH-pyridylethylcysteine was identified by the method of Hermodson et al. (34). For the identification of PTH-serine, PTH-leucine, and PTH-isoleucine, and to confirm the results of thin-layer chromatography, PTH-derivatives were regenerated to the amino acids by hydrolysis at 150°C in 6 N HC1 containing 0.3% SnCl2 (35). PTH-arginine and PTH-histidine were further identified by spot tests (36, 37). Antibody Preparation—Antigens in Freund's complete adjuvant were administered subcutaneously four times at intervals of 7 to 14 days. Rabbits were immunized with human, porcine, or rat antithrombin III by the injection of about 300 [ig of the purified protein each time, whereas rats were immunized with rabbit antithrombin III by the injection of about 150 fig of the purified protein each time. Antisera were collected 1 week after the last injection, and saturated ammonium sulfate, adjusted to pH 7.8, was added to give 3 3 % saturation. The pellet obtained by centrifugation was dissolved in half of the original volume of distilled water and extensively dialyzed against 0.15 M NaCl, adjusted to pH 7.8, containing 0.1 % NaN 3 . The /--globulin fractions thus obtained were stored at —20°C until use.
Vol. 86, No. 6, 1979
1843
Immunologic Assay—Precipitin reactions were carried out in 1 % agarose gel by the Ouchterlony method. Assay of Thrombin and Inhibitor Activities— Thrombin was dissolved in 0.05 M Tris-HCl, 0.15 M NaCl, pH 7.4, containing 1 mg/ml of bovine serum albumin. Coagulation activity of thrombin was assayed as described by Lundblad et al. (22). Antithrombin III activity was determined as follows. For the determination of heparin cofactor activity, 50 ^1 of antithrombin III appropriately diluted with 0.05 M Tris-HCl, 0.15 M NaCl, pH 7.4, was mixed with an equal volume of heparin (60300 units/ml) dissolved in the same buffer, and incubated at 37°C for 1 min. Fifty [A of thrombin (100-500 /ig/ml, 2,300 NIH units/mg) was added to the preincubated sample, and at various times of incubation, an aliquot (5 or 10 /al) was removed, diluted, and added to 0.3 ml of the assay mixture (22) containing 0.4 % fibrinogen, then the residual thrombin activity was assayed. For the determination of progressive antithrombin III activity, buffer was substituted for heparin. All coagulation assays were performed at least in triplicate. Preparation of Heparin-Agarose—Heparinagarose was prepared by the cyanogen bromide method of Cuatrecasas (38) as described by Fujikawa et al. (39). Isolation of Antithrombin HI—In this procedure, all operations were performed at 4°C and contact with glass was avoided by employing plastic columns, tubes, and containers. Onethirtieth volume of 1 M BaCI2 was added to human acid-citrate-dextrose plasma, containing 0.1 IDM benzamidine, and the mixture was stirred for 20 min then centrifuged at 7,800 x g for 20 min in a Sorvall RC3 centrifuge. The supernatant was made up to 0.1 mM in ethylenediaminetetraacetic acid and then brought to 50% saturation by the slow addition of solid ammonium sulfate. After stirring for 20 min the precipitate was removed by centrifugation at 7,800xg for 20min. The supernatant was brought to 85% saturation with solid ammonium sulfate, and after stirring for 1 h, the suspension was centrifuged for 30 min at 10,000xg in a Hitachi 18PR-3 centrifuge. The precipitate was redissolved in distilled water and extensively dialyzed against 0.02 M phosphate buffer, pH 6.3, containing 0.4 M NaCl at 4°C.
1844
T. KOIDE
The dialyzed sample was applied to a heparinagarose column (6.8 x l l cm), equilibrated with the same buffer. After extensive washing of the column with the equilibration buffer, followed by 1 M NaCl in 0.02 M phosphate buffer, pH 6.3, antithrombin III was eluted with 2 M NaCl in 0.02 M phosphate buffer, pH 6.3. Porcine, rabbit, and rat antithrombin III were purified by similar procedures. The average yields of pure protein from 1 liter of plasma or serum were 40, 43, 52, and 41 mg for human, porcine, rabbit, and rat antithrombin III, respectively. RESULTS SDS-Polyacrylamide Gel Electrophoresis of Human, Porcine, Rabbit, and Rat Antithrombin III
—SDS-polyacrylamide gel electrophoresis patterns of human, porcine, rabbit, and rat antithrombin III are shown in Fig. 1. Each antithrombin III migrated as a single protein band in the presence and absence of reducing agent, and there was little difference in the migration of the protein before and after reduction. These results are consistent with the view that each antithrombin III consists of a single polypeptide chain. The apparent molecular weights estimated from the migration of reduced protein in 7.5% polyacrylamide gels were 59,000, 58,000, 63,000, and 63,000 for human, porcine, rabbit, and rat antithrombin III, respectively. Amino Acid and Carbohydrate Compositions of Human, Porcine, Rabbit, and Rat Antithrombin HI—The amino acid and carbohydrate compositions of human, porcine, rabbit, and rat anti-
"^ m
95,000 65,000
~
45,000
—
29,000 20,000
1 2
3 4
5 6
7 8
Fig. 1. SDS-polyacrylamide gel electrophoresis of human, porcine, rabbit, and rat antithrombin III. Samples (about 10 fig) were dissolved in 10 ftl of 0.2 M Tris-phosphate buffer, pH 7.0, containing 0.2% sodium dodecyl sulfate, and heated at 100°C for 2 min in the presence or absence of 5 ft] of 2-mercaptoethanol. The samples were applied to a 7.5% gel column (6x95 mm), and electrophoresis was performed as described in " MATERIALS AND METHODS." Gels 1 and 2 are nonreduced and reduced human antithrombin III, respectively. Gels 3 and 4 are nonreduced and reduced porcine antithrombin i n , respectively. Gels 5 and 6 are nonreduced and reduced rabbit antithrombin ID, respectively. Gels 7 and 8 are nonreduced and reduced rat antithrombin III, respectively. The standard gel on the right contains, from top to bottom, phosphorylase b, rat serum albumin, ovalbumin, bovine carbonic anhydrase, and soybean trypsin inhibitor (Kunitz). The anode was at the bottom of the gel.
/. Biochem.
HUMAN, PORCINE, RABBIT, AND RAT ANTITHROMBIN III
1845
TABLE I. Amino acid and carbohydrate compositions of human, porcine, rabbit, and rat antithrombin HI. Values are given as numbers of residues/mol of glycoprotein. Numbers in parentheses are expressed as weight percent. Components
Human
Porcine
Rabbit
Rat
Amino acid (%) Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Half-cystine Valine Methionine Isoleucine Leucine Tyrosine Phenylalanine Tryptophan Lysine Histidine Arginine
(83. 1) 44.2 23.4
(83. 6) 52.3 22.9 30.9 53.3 19.9 18.0 30.2 6.2* 28.3
(85. 8) 53.1 26.6 31.5 47.5 22.8 18.0 31.6 6.4a
(84. 6) 51.8 20.9 29.5 52. 1 24.2 18.1 26.6
29.8 11.2 22.5 48.1 11.3 26.7
34.2
32.8 49.6 23.1 20.0 29. 1 6.U
28.9 10.9 20.5 38.0 10. 1 24. 1
Carbohydrate (%) Hexose Glucosamine Neuraminic acid 1
Determined as cysteic acid.
b
23.7 39. 1 10.0 24.9
11.0 22.0 41.2 8.0 26.4
3.8
5.0
3.8
3.8
32.4
32.5
40.9
33.2
6,2
6.0
7.1
5.8
20.3
23.9
27.3
20.2
(16.9) 18.5 (5.2) 17.0 (6.5) 10.2 (5.2)
(16.4) 17.9 (4.7) 15.4 (5.5) 12.7 (6.1)
(14.2) 11.6 (3.0) 14.6 (5.2) 12.7 (6.0)
(15.4) 11.5 (3.2) 20.0 (7.5)
9.3 (4.7)
Determined as 5-pyridylethylcysteine
thrombin III are shown in Table I. The compositions of the four antithrombin III molecules are very similar, although some minor differences exist. Human, porcine, rabbit, and rat antithrombin III are glycoproteins containing hexose, Af-acetylglucosamine, and N-acetylneuraminic acid. No galactosamine was detected in any of the four proteins. The total carbohydrate contents were 17,16,14, and 15% for human, porcine, rabbit, and rat antithrombin III, respectively. No free sulfhydryl group was detected in any of the antithrombin III molecules by Ellman's procedure, indicating that the six half-cystine residues are present as three disulfide bonds in each protein. Vol. 86, No. 6, 1979
9.2
5.9*>
Amino-Tenninal Sequences of Human, Porcine, Rabbit, and Rat Antithrombin HI—The aminoterminal sequences of the S-pyridylethylated derivatives of human, porcine, rabbit, and rat antithrombin III are shown in Fig. 2. The similar sequences of bovine and horse antithrombin III are also included for comparison. Human, porcine, rabbit, bovine, and horse antithrombin III each contained amino-terminal histidine. Rat antithrombin III, however, contained a'minoterminal asparagine. It is clear from Fig. 2 that the antithrombin III molecules from all six different species have homologous sequences in the aminoterminal region of the molecule. In particular, porcine and horse antithrombin III molecules have
1846
T. KOIDE 1 Rat
5 10 15 Asn Pro Val Asp Asp H e Cys H e Ala Lys Pro Arg Asp H e Pro
Rabbit
His Glu Ser Pro Val Glu Asp H e Cys Thr Ala Lys Pro Arg Asp Phe Pro
Porcine
His Trp Ser Pro Val Glu Asp H e Cys Thr Ala Lys Pro Arg Asp H e Pro
Human
His Gly Ser Pro Val
Asp H e Cys Thr Ala Lys Pro Arg Asp He
Pro
Bovine
His Arg Ser Pro Val Glu Asp Val Cys Thr Ala Lys Pro Arg Asp He
Pro
Horse
His Trp Ser Pro Val Glu Asp H e Cys Thr Ala Lys Pro Arg Asp H e Pro
Fig. 2. Amino-terminal sequences of rat, rabbit, porcine, and human antithrombin III. The sequences of bovine and horse antithrombin III determined by Kurachi et al. (19) are also shown for comparison. Amino acid residues that are identical are shown in blocks. Dash indicates space that has been inserted to bring human antithrombin III into alignment for better homology with the other antithrombin III sequences.
Fig. 3. Ouchterlony double immunodifFusion patterns in 1% agarose gels. The center wells contained: (A) rabbit anti-human antithrombin HI, (B) rabbit anti-porcine antithrombin III, (C) rat anti-rabbit antithrombin III, (D) rabbit anti-rat antithrombin HI. Surrounding wells contained: antithrombin III from man (1), pig (2), • rabbit (3), and rat (4). / . Biochem.
1847
HUMAN, PORCINE, RABBIT, AND RAT ANTITHROMBIN III the same sequence in the first 17 amino-terminal residues. Some minor differences in the sequences of the six proteins, however, are evident. The amino acid residue in the second position was highly variable, following the numbering system shown in Fig. 2. Human antithrombin III contained glycine, porcine and horse antithrombin III contained tryptophan, rabbit antithrombin III contained glutamic acid, and bovine antithrombin III contained arginine. Rat antithrombin III lacked the amino-terminal two residues, and the amino-terminal asparagine in rat antithrombin III replaced serine in position 3 in the other five antithrombin III molecules. Position 6 contained an aspartic acid residue in rat antithrombin III, while glutamic acid was present in porcine, rabbit, bovine, and horse antithrombin III, and in human antithrombin III it was necessary to omit a residue in position 6 in order to bring this protein into alignment for greater homology with the other antithrombin III molecules. Other minor differences were noted in position 8, where isoleucine was replaced by valine in bovine antithrombin III, in position 10, where threonine was replaced by isoleucine in rat antithrombin III, and also in position 16, where isoleucine was replaced by phenylalanine in rabbit antithrombin III. The remainder of the first 17 residues was the same for all six proteins.
Immunological Cross-Reactivity—Immunological cross-reactivity was examined by Ouchterlony's double diffusion method. Figure 3 shows immunodiffusion patterns obtained when rabbit or rat antiserum in the center well was reacted with human, porcine, rabbit, and rat antithrombin III in the surrounding wells. Antisera to rabbit and rat antithrombin III formed a strong immune precipitate only with the corresponding specific antigen, and gave no cross-reactivity with the material from other species. Antiserum to human antithrombin III reacted not only with human antithrombin III, but also with porcine antithrombin III, and antiserum to porcine antithrombin III also reacted strongly with both human and porcine antithrombin III. These results indicate partial immunologic identity between human and porcine antithrombin III. Inhibition of Human Thrombin by Human, Porcine, Rabbit, and Rat Antithrombin 111—The inhibition of the clotting activity of human thromVol. 86, No. 6, 1979
bin by human, porcine, rabbit, and rat antithrombin III was studied at 37°C. The inhibition proceeded time-dependently in the absence of heparin, reaching equilibrium in about 15 min after mixing the inhibitor with thrombin. The inhibition, however, was complete almost instantaneously in the presence of heparin. The inhibition curves of human thrombin by human, porcine, rabbit, and rat antithrombin III in the absence of heparin are shown in Fig. 4. The clotting activity of human thrombin was strongly inhibited by human, porcine, or rat antithrombin III, and less strongly by rabbit antithrombin III. Extrapolation of the inhibition curves in Fig. 4 to 100% inhibition gives an enzymeinhibitor ratio of 1 in each case, indicating that antithrombin III interacts with thrombin to form an equimolar complex. Values for the apparent dissociation constant (Ki) calculated by the method
0
"0
Antithrombin IE Cone. (pM) 4.5 9.0
1.0
2.0
[Antithrombin] / [Thrombin] (mole/mole) Fig. 4. Inhibition of human thrombin by human, porcine, rabbit, and rat antithrombin III. The reaction mixture contained 166 //g/ml of human thrombin (2,300 NIH units/mg) in 0.05 M Tris-HCI, 0.15 M NaCl, pH 7.4, containing 0.33 mg/ml of bovine serum albumin, and various concentrations of either human (#), porcine (O), rabbit (A), or rat ( A ) antithrombin III. The mixture was incubated at 37°C for 15 min, then an aliquot (5 or 10 fi\) was removed, diluted and assayed for residual clotting activity of thrombin.
1848
T. KOIDE
of Green and Work (40) for the complex between human thrombin and human, porcine, rabbit, and rat antithrombin III were calculated to be about
lactosamine in porcine, rabbit, and rat antithrombin III as well as in the human preparation is consistent with the results of Miller-Andersson et al. (17), Nordenman et al. (41), and Danishefsky 1 . 2 X 1 0 - 1 0 M , 9 . 5 X 1 0 - 9 M , 1 . 4 X 1 0 - ' M , and 2.8 x et al. (42). Although rabbit antithrombin III was 10~9 M, respectively. also purified by Yin et al. (20), they did not report the amino acid and carbohydrate compositions but DISCUSSION only gave the contents of hexose (4.1%) and Affinity chromatography on heparin-agarose was neuraminic acid (4.6%). first introduced by Miller-Andersson et al. (17) for Histidine was reported to be the aminothe isolation of human antithrombin III. The terminal residue in human (17-19), bovine (19), same or a similar procedure was successfully used and horse (79) antithrombin III. Amino-terminal for the purification of canine (27), bovine (18, 19, histidine was also found in porcine, and rabbit 41), and horse (79) antithrombin III. In the antithrombin III. Rat antithrombin III, however, present studies, the utility of heparin-agarose was had asparagine as the amino-terminal amino acid. extended to the purification of porcine, rabbit, This residue corresponds to the third position of and rat antithrombin III. other antithrombin III molecules in terms of the Porcine, rabbit, and rat antithrombin III as homology between antithrombin III, which shows well as the previously reported human, bovine, that rat antithrombin III lacks the first two amino and horse antithrombin III have been shown to acid residues at the amino-terminal end. Whether be glycoproteins composed of a single polypeptide this is because the starting material for rat antichain. The amino acid and carbohydrate com- thrombin III was not plasma but serum is not positions of the four antithrombin III prepara- clear. There was no difference in the aminotions reported here are very similar. It is particu- terminal amino acid sequence between rabbit larly noteworthy that the presence of 6 mol of antithrombin III samples isolated from plasma half-cystine as 3 disulfide bridges is common and serum. The amino-terminal sequence of among the four proteins. This value of cystine human antithrombin III was found to be the same was also reported by Kurachi et al. (19) for their as that reported by Kurachi et al. (19) and Magnuhuman, bovine, and horse preparations. Amino sson et al. (43). The remarkable similarity of the acid and carbohydrate compositions of human amino-terminal sequence in human, bovine, and antithrombin III have been reported by several horse antithrombin III shown by Kurachi et al. investigators (16, 17, 19). The amino acid com- (19) was extended to porcine, rabbit, and rat position of our human preparation is in reasonable antithrombin III. Although there are some agreement with those reported by Heimburger differences in amino acid residues at a few posiet al. (16), Miller-Andersson et al. (17), and Kurachi tions, they can be attributed to single base reet al. (19), with the sole exception of the half- placements. No sequence similarity was found cystine content. Miller-Andersson et al. (17) between the amino-terminal region of antithrombin reported no half-cystine in their preparation. The III and those of other plasma proteinase inhibitors presence of cystine in antithrombin III, however, such as alpha-1-antitrypsin (44-46) and alpha-1is clear from the amino-terminal sequence presented antichymotrypsin (46). by Kurachi et al. (19) and in the present studies. Immunological cross-reactivities among antiThe carbohydrate composition of human anti- thrombin III from different species were studied thrombin III is rather inconsistent among investi- by Kurachi et al. (19) and Nordenman et al. (41). gators (16, 17, 19, 41, 42). This may due to the The former investigators reported partial immunorelatively low accuracy of carbohydrate analysis logical cross-reactivity between human and bovine methods available. The major differences are antithrombin III. The latter, however, found no that Miller-Andersson et al. (17) reported the cross-reactivity between the two proteins. In our absence of neuraminic acid in their preparation, preparations immunological cross-reactivity was in contrast to the other reports, including the only observed between human and porcine antipresent studies. The absence of TV-acetyl ga- thrombin III in immunodiffusion experiments, and / . Biochem.
HUMAN, PORCINE, RABBIT, AND RAT ANTITHROMBIN HI no cross-reactivity was observed for rabbit or rat antithrombin III with the other three antithrombin III preparations. The degrees of similarity of the amino-terminal sequences in human, porcine, rabbit, and rat antithrombin III were also consistent with the antigenic similarities among antithrombin III preparations observed by double immunodiffusion, and paralleled the closeness of the taxonomic relationship of the animals used as sources. The inhibitions of human thrombin by human, porcine, rabbit, and rat antithrombin III are almost instantaneous in the presence of heparin and proceed progressively in the absence of heparin, reaching equilibrium about 15 min after mixing the two proteins. These observations are consistent with those of many investigators (9,13,14). Complex formation in a one-to-one molar ratio between human thrombin and human, porcine, rabbit, or rat antithrombin III is also consistent with the many reported results (2, 9-11). Species specificities in the inhibition of coagulation factors by antithrombin III have been less well studied. Kurachi et al. (79) showed that bovine antithrombin III has the highest affinity for bovine factor Xa/3, while that of human antithrombin III is lower, and that of horse antithrombin III is lower still. Nordenman et al. (41) reported that human and bovine antithrombin III inhibit human thrombin to the same extent, although they presented no quantitative data. The present study shows that the affinity of antithrombin III for human thrombin is highest in the case of the human preparation and lowest with the rabbit preparation, while rat antithrombin III shows a higher affinity for human thrombin than the porcine preparation. Thus, the order of species specificity in the inhibitory activity is not as consistent with the closeness of taxonomic relationships of the animals as are the sequence and antigenic similarities among the antithrombin III preparations. The author is grateful to Drs. T. Tsukada and K. Ogata for supplying human plasma and rat serum, respectively. The author is also grateful to Professor T. Ikenaka for his generous help. REFERENCES 1. Abildgaard, U. & Egeberg, O. (1968) Scand. J. Vol. 86, No. 6, 1979
1849
Haematol. 5, 155-157 2. Highsmith, R.F. & Rosenberg, R.D. (1974) /. Biol. Chem. 249, 4335-4338 3. Davie, E.W. & Hanahan, D.J. (1977) in The Plasma Proteins (Putnam, F.W., ed.), Vol. Ill, pp. 421-544, Academic Press, New York 4. Burrowes, C.E., Habal, F.M., & Movat, H.Z. (1975) Thromb. Res. 7, 175-183 5. Vennerod A.M. & Laake K. (1975) Thromb. Res. 7, 223-226 6. Lahiri, B., Bagdasarian, A., Mitchell, B., Talamo, R.C., Colman, R.W., & Rosenberg, R.D. (1976) Arch. Biochem. Biophys. 175, 737-747 7. Egeberg, O. (1965) Thromb. Diath. Haemorrh. 13, 516-520 8. Robinson, A.J., Aggeler, P.M., McNicol, G.P., & Douglas, A.S. (1967) Br. J. Haematol. 13, 510-527 9. Rosenberg, R.D. & Damus, P.S. (1973) / . Biol. Chem. 248, 6490-6505 10. Kurachi, K., Fujikawa, K., Schmer, G., & Davie, E.W. (1976) Biochemistry 15, 373-377 11. Fujikawa, K., Kurachi, K., & Davie, E.W. (1977) Biochemistry 16, 4182-4188 12. Kurachi, K. & Davie, E.W. (1977) Biochemistry 16, 5831-5839 13. Monkhouse, F.C., France, E.S., & Seegers, W.H. (1955) Circ. Res. 3, 397^(02 14. Yin, E.T., Wessler, S., & Stoll, P.J. (1971) /. Biol. Chem. 246, 3703-3711 15. Fagerol, M.K. & Abildgaard, U. (1970) Scand. J. Haematol. 7, 10-17 16. Heimburger, N., Haupt, H., & Schwick, H.G. (1971) in Proc. Int. Res. Conf. Proteinase Inhibitors (Fritz, H. & Tschesche, H., eds.) pp. 1-22, Walter de Gruyter, Berlin 17. Miller-Andersson, M., Borg, H., & Andersson, L-O. (1974) Thromb. Res. 5, 439-452 18. Thaler, E. & Schmer, G. (1975) Br. J. Haematol. 31, 233-243 19. Kurachi, K., Schmer, G., Hermodson, M.A., Teller, D.C., & Davie, E.W. (1976) Biochemistry 15, 368373 20. Yin, E.T., Wessler, S., & Stoll, P.J. (1971) /. Biol. Chem. 1A6, 3694-3702 21. Damus, P.S. & Wallace, G.A. (1974) Biochem. Biophys. Res. Commun. 61, 1147-1153 22. Lundblad, R.L., Kingdon, H.S., & Mann, K.G. (1976) in Methods in Enzymology (Lorand, L., ed.) Vol. XLV, pp. 156-176, Academic Press, New York 23. Weber, K. & Osborn, M. (1969) / . Biol. Chem. 244, 4406-4412 24. Fairbanks, G., Steck, T.L., & Wallach, D.F.H. (1971) Biochemistry 10, 2606-2617
1850 25. Moore, S. & Stein, W.H. (1963) in Methods in Enzymology (Colowick, S.P. & Kaplan, N.O., eds.) Vol. VI, pp. 819-831, Academic Press, New York 26. Hugli, T.E. & Moore, S. (1972) J. Biol. Chem. 241, 2828-2834 27. Hirs, C.H.W. (1967) in Methods in Enzymology (Hirs, C.H.W., ed.) Vol. XI, pp. 197-199, Academic Press, New York 28. Friedman, M., Krull, L.H., & Cavins, J.F. (1970) J. Biol. Chem. 245, 3868-3871 29. Winzler, R.J. (1955) in Methods of Biochemical Analysis (Glick, D., ed.) Vol. 2, pp. 279-311, Interscience Publ. Inc., New York 30. Warren, L. (1959) /. Biol. Chem. 234, 1971-1975 31. Edman, P. & Begg, G. (1967) Eur. J. Biochem. 1, 80-91 32. Brenner, M., Niederwieser, A., & Pataki, G. (1969) in Thin-Layer Chromatography (Stahl, E., ed., Aschworth, M.R.F., translation) pp. 730-786, Springer-Verlag, Berlin 33. Jeppson, J.-O. & Sjoquist, J. (1967) Anal. Biochem. 18, 264-269 34. Hermodson, M.A., Ericsson, L.H., Titani, K., Neurath, H., & Walsh, K.A. (1972) Biochemistry 11, 4493^1502 35. Mendez, E. & Li, C.Y. (1975) Anal. Biochem. 68, 47-53
T. KOIDE 36. Easley, C , Zegers, B.J.M., & DeVijlder, M. (1969) Biochim. Biophys. Ada 175, 211-213 37. Easley, C. (1965) Biochim. Biophys. Ada 107, 386388 38. Cuatrecasas, P. (1970) /. Biol. Chem. 245, 3059-3065 39. Fujikawa, K., Thompson, A.R., Legaz, M.E., Meyer, R.C., & Davie, E.W. (1973) Biochemistry 12, 4938-4945 40. Green, N.M. & Work, E. (1953) Biochem. J. 54, 347-352 41. Nordenman, B., Nystrdm, C , & Bjork, I. (1977) Eur. J. Biochem. 78, 195-203 42. Danishefsky, I., Zweben, A., & Slomiany, B.L. (1978) J. Biol. Chem. 253, 32-37 43. Magnusson, S., Sottrup-Jensen, L., Petersen, T.E., Dudek-Wojciechowska, G., & Claeys, H. (1976) in Proteolysis and Physiological Regulation (Robbins, D.W. & Brew, K., eds.) pp. 203-238, Academic Press, New York 44. Morii, M., Odani, S., Koide, T., & Ikenaka, T. (1978) /. Biochem. 83, 269-277 45. Owen, M.C., Lorier, M., & Carrell, R.W. (1978) FEBS Lett. 88, 234-236 46. Travis, J., Garner, D., & Bowen, J. (1978) Biochemistry 17, 5647-5651
/ . Biochem.
