- Email: [email protected]
Coagulation factor V exists uncomplexed in bovine plasma
524
Biochimica et Biophysica Acta, 714 (1982) 524-529
Elsevier BiomedicalPress BBA 21061 COAGULATION FACTOR V EXISTS UNCOMPLEXED IN BOVINE PLASMA T. ROY ITTYERAH,RAZIA RAWALAand ROBERTW. COLMAN * Thrombosis Research Center, Temple University Health Science Center, Jones Hall, 3401 North Broad St., Philadelphia, PA 19140 (U.S.A.)
(Received July 29th, 1981) (Revised manuscript received November2nd, 1981)
Key words: Factor V; Immunological assay; Blood clotting; (Bovine plasma)
Bovine coagulation factor V has been examined immunochemically to ascertain whether the coagulant polypeptide (h) with M r = 290 0 0 0 - 3 3 0 000 is complexed in plasma with a second immunoehemically distinct polypeptide (12) of Mr = 400 000. Antiserum containing antibodies to h and 12 detects the 12 polypeptide eluting earlier than the h chain on gel filtration of plasma with either added calcium or EDTA, consistent with the behavior of a higher molecular weight noninteracting species. An immobilized monospecific antibody to 12 removes only the 12 polypeptide from a purified factor V preparation containing both h and 12. Moreover, while a monospecific antibody to the h chain was able to precipitate purified radioactively labelled h chain alone or mixed with plasma, the 12 antibody was unable to precipitate radioactively labelled h chain even after attempted recombination of the h chain with 12 present in plasma. These studies indicate that the 12 polypeptide is not complexed to the h chain in a purified system or in plasma and reinforce the conclusion that factor V is a single polypeptide chain uncomplexed in plasma.
Introduction
Factor V, a plasma glycoprotein, functions as an essential cofactor in blood coagulation, non-enzymatically accelerating the formation of thrombin from its precursor prothrombin by the serine protease factor Xa in the presence of calcium and phospholipid. Since Barton and Hanahan [ 1] and Esnouf and Jobin [2] described preparations 3000-5000-fold purified over native plasma in 1967, many purification procedures have been described. However, much disagreement existed regarding the molecular forms and properties of the protein [3]. One of the major problems in the attempts to isolate the native form of the factor V molecule has been the susceptibility of the molecule to proteolytic fragmentation [4,5] both during blood collection and subsequent purification procedures. Procedural mo* To whom reprint requests should be addressed.
difications to avoid this problem included the use of proteolytic inhibitors [6] and the collection of blood by jugular venipuncture in the living animal [7]. The latter innovation allowed us to isolate bovine factor V with a high specific activity purified 2000-fold over starting plasma. We found that the factor V preparation consisted of a polypeptide of Mr = 290000 designated the h chain which was cleaved by thrombin to a carbohydrate rich fragment of Mr = 190 000 and a polypeptide of 110000 with an associated 18fold increase of activity [7]. Esmon [8] confirmed these findings by isolating a polypeptide of Mr = 290000 which after cleavage by thrombin yielded fragments of 210 000 and 115 000. Additional thrombin and higher temperatures led to further cleavage. Nesheim et al. [9,10] isolated a protein of Mr = 330 000 which on activation by thrombin gave products of Mr = 205000, 150000, 94000 and smaller fragments. All three groups agree that the polypeptide chain cleaved by thrombin has an Mr of
525 2 9 0 0 0 0 - 3 3 0 000 and that thrombin activation produces among other products a polypeptide of 9 4 0 0 0 - 1 1 5 0 0 0 essential for activity. However, in contrast to the later studies of Esmon [8] and Neshelm et al. [9,10], Saraswathi et al. [7] found that the h chain (Mr = 290000) was closely associated with a polypeptide designated the 12 polypeptide of M r = 400 000. On reduction the 12 polypeptide yielded two identical components of Mr = 200 000. In contrast to the h chain, the 12 component was not cleaved by thrombin [7]. Bartlett et al. [11] also reported a high molecular weight form of factor V although its polypeptide composition was more complex. Further investigation of the possible role of the polypeptide of Mr = 400 000 associated with factor V revealed that the h chain and the 12 polypeptide were immunochemically distinct [12]. Activation of factor V by two snake venoms cleaved the polypeptide of Mr = 290000 but not the 12 polypeptide [12,13]. The h chain but not the 12 polypeptide is present in platelets [12,14]. The 12 polypeptide makes if anything, a minor contribution to factor V activity but may alter its thermal stability [12]. Since neither Esmon [8] nor Nesheim et al. [9] find any 12 protein in their purified factor V preparations two hypotheses must be entertained. The first, is that the 12 polypeptide is noncovalently complexed with factor V and was inadverently lost during isolation, perhaps because of exposure to the dissociating solvents of high ionic strength used in both procedures [8,9]. The second, is that the 12 polypeptide represents a fortuitous association of two unrelated proteins which co-purify in some procedures. To resolve this issue we prepared monospecific antibodies against the 12 protein and the h chain and explored the immunochemical properties of both polypeptides in plasma to ascertain their relatonship. Methods
Preparation of factor V. Bovine blood was collected by jugular venipuncture into 0.10 vol. of 0.15M sodium citrate/0.1 M benzamidine hydrochloride/0.1 mM soybean trypsin inhibitor. Plasma was obtained by centrifugation at 2000Xg for 15 rain at 4°C, and factor V was purified as described by Saraswathi et al. [7] which yielded a preparation con-
taining both the h chain and the 12 polypeptide. The specific activity of the preparation was 35 units/rag increasing to 600 units/rag when activated by thrombin. Factor V was also purified as described by Esmon [8] from bovine blood collected by venipuncture into 0.10 vol. of 0.1 M sodium oxalate/0.1 M benzamidine hydrochloride/O.1 mM soybean trypsin inhibitor. This procedure yields only the h chain. The specific activity of the preparation was 52 units/mg increasing to 570 units/mg when activated with thrombin. Monitoring the procedure by immunoelectrophoresis against an antibody directed against the 12 polypeptide indicated that this polypeptide was removed during the calcium oxalate adsorption step of that procedure [8]. Assay of factor V. Factor V activity was assayed by the one-stage method of Lewis and Ware [15]. 1 unit of factor V is defined as the amount in 1 ml normal human plasma (pool of 10 separate plasmas). Preparation ofantisera. A rabbit antiserum to factor V was prepared by the procedure of Greenquist et al. [16] using factor V purified by the method of Saraswathi et al. [7] or that of Esmon [8] as the antigen. An antibody to the 12 polypeptide was prepared by two methods one of which has been described earlier [12]. The second method involved the adsorption of the antiserum containing anti-12 and anti-h prepared against the preparation of Saraswathi et al. [7] with a factor V preparation [8] containing only the 290000 molecular weight polypeptide. The antiserum was incubated with the factor V preparation for 1 h at 37°C and 18 h at 4°C, followed by centrifugation at 12000 ×g for 4 min. The supernatant contained only the anti-12 component since immunoelectrophoresis against plasma revealed only a single arc corresponding to the 12 polypeptide (see Results). Antibody to the h chain was prepared by injecting factor V prepared by the method of Esmon [8]. The antibody was capable of neutralizing factor V activity using purified factor V or factor V in plasma. Inhibition of factor V by 12 antisera. The inhibition of factor V activity in plasma by rabbit 12 antisera was performed at 3~°C. 12 antibody and plasma in either a ratio of 1 : 1 or 5 : 1 were mixed and incubated at 37°C. At specific times aliquots were removed and assayed for factor V coagulant activity.
526 The maximum degree of inhibition was obtained in 10 min. Sera from the same rabbits obtained prior to immunization were tested in parallel and served as control for the experiment. Immunodiffusion. Double immunodiffusion was performed by the method of Ouchterlony [ 18]. Immunoelectrophoresis. Immunoelectrophoresis was performed by a modification of the method described by Scheidegger [17], using 1% agarose in sodium barbital buffer, I = 0.0375, pH 8.6. Preparat&n of h chain. An immunoabsorbent column was prepared by coupling anti-12 to CNBrSepharose 4B as described before [12]. Factor V prepaired by the method of Saraswathi et al. [7] containing both the h and 12 polypeptides was applied to the colunm and the unadsorbed material was assayed for factor V activity and subjected to immunoelectrophoresis against the antiserum containing both anti12 and anti-h. (see Results). Gel filtration of plasma. Bovine blood was collected by venipuncture and anticoagulated with sodium citrate (0.015 M) to which benzamidine hydrochloride, soybean trypsin inhibitor, trasylol, hirudin and diisopropylfluorophosphate had been added to give final concentrations of 10 mM, 0.01 raM, 50 kallikrein inhibitor units/ml, 0.2 units/ml and 2 mM, respectively. Plasma (3 ml) was applied to a 1.5 × 85 cm column of Sepharose 4B (bed vol. 150 ml) equilibrated with Michaelis buffer (0.0265 M sodium acetate/0.0265 M sodium barbital/0.108 M NaCI, pH 7.4) containing 5 mM CaC12 and 10 mM benzamidineHC1. The same buffer was used for elution. The fractions were assayed for factor V activity and the fractions across the activity peak were subjected to immunoelectrophoresis against the antiserum containing both anti-12 and anti-h. The plasma was adsorbed with BaC12 to remove prothrombin and factor X. The BaClz-adsorbed plasma (3 ml) was gel filtered on the same Sepharose 4B column equilibrated with Michaelis buffer, pH 7.4, containing 5 mM CaCI2/10 mM benzaminidine-HCl, and the fractions were assayed for factor V activity. The fractions across the activity peak were subjected to immunoelectrophoresis against the antiserum containing both anti-12 and anti-h. Plasma (3 ml) was gel filtered on the column after it was equilibrated with Michaelis buffer containing 20 mM EDTA. No factor V activity could be de-
tected in the eluted fractions due to inactivation of factor V by EDTA. The fractions were subjected to immunoelectrophoresis against the antiserum containing both anti-12 and anti-h. Thyroglobulin and ferritin (10 mg each)were also applied to the column and the absorbance of the eluted fractions was measured at 280 nm. Preparation of 12sI-factor K Factor V was radioactively labelled using Bolton-Hunter reagent [19]. The free 12sI was removed by the method of Tuszynski et al. [20]. The specific radioactivity of the labelled factor V was 72 cpm/ng. Results and Discussion
Factor V purified by the method of Saraswathi et al. [7] contains two components exhibiting a reaction of nonidentity representing the 12 protein (Mr = 400000) and the h chain associated with factor V activity [7,12]. When this preparation is subjected to electrophoresis against an antibody to 12, a single component is seen which is identical in mobility to the single arc seen in bovine plasma (Fig. la). A reaction of identity was found on double-immunodiffusion between the 12 polypeptide in purified factor V [7] and plasma (Fig. lb). The 12 antibody did not form a precipitin line on immunoelectrophoresis or double-immunodiffusion against factor V prepared by the method of Esmon [8]. However, it is possible that at pH 8.6 and under the conditions of the immunoelectrophoresis the h and 12 chains were dissociated. Therefore, we examined factor V in plasma under conditions more closely approximating those in plasma. The behavior of purified bovine factor V on gel filtration has been well studied. Although its molecular weight is 2 9 0 0 0 0 - 3 3 0 0 0 0 it elutes in plasma anticoagulated with citrate before thyroglobulin (Mr = 670000) because of its high axial ratio [21]. However, it is possible that two noncovalently associated molecules could have been dissociated by removal of calcium in anticoagulated plasma. We gelfiltered bovine plasma (containing inhibitors) and BaC12-adsorbed bovine plasma in the presence of 5 mM Ca 2÷ and found that a single peak of factor V activity eluted just before thyroglobulin (Fig. 2). If factor V were a complex of the h chain and the 12 protein, they might be dissociated by completely removing metal ions since citrated plasma contains about 50
527
Thyrogtobulin Ferritin
t 0.4
E 03
0.2 > t9 O
I,i_
0-1
Oo
2s
|
5o
i
,2s ,5o
ELUTION VOLUME (ml) Fig. 2. Gel filtration of plasma m the presence of Ca2+. Plasma was gel-filtered on a column of Sepharose 4B and the
Fig. 1. a. Immunoelectrophoresis of purified factor V and plasma against an antibody to the 12 polypeptide. The upper well contained 0.025 unit of factor V purified as described by Saraswathi et al. [7]. The lower cell contained plasma with 0.025 unit of factor V activity, b. Double immunodiffusion of purified factor V and plasma against the antibody to 12. The left and right well contain 20 ~tl plasma diluted containing 1 : 16 or 0.002 units factor V activity. The top well contains 20 ul purified factor V (7) (100 units/ml) or 0.31 units factor V activity. The bottom well contains antiserum to 12 diluted 1 : 4.
~M free Ca 2+. Therefore, we gel-filtered plasma in the presence of 20 mM EDTA. Due to inactivation of factor V by EDTA [22], the activity could not be assayed in the eluted fractions. However, using an antiserum containing antibodies against h and 12 we could follow the distribution of the precipitin arcs in fractions representative o f the gel-filtered protein. Immunoelectrophoresis o f the fractions from plasma eluted in the presence o f 5 mM Ca 2÷ showed that the 12 polypeptide eluted early on the ascending limb, both h and 12 overlapped in the peak while the h
eluted fractions were assayed for factor V activity. Plasma adsorbed with BaC12 was also gel-filtered on the same column, o o, factor V activity in fractions from plasma; o. . . . . . o, activity in fractions from BaC12-adsorbed plasma. Thyroglobulin and ferritin were also gel-filtered on the column and their positions are shown.
chain eluted alone on the descending limb (Fig. 3). Immunoelectrophoresis of corresponding fractions from BaC12-adsorbed plasma eluted in the presence o f 5 mM Ca 2÷ and o f fractions eluted in the presence of EDTA exhibited the same pattern. These results clearly show that 12 and h do not co-elute but rather overlap when gel-filtered in the presence o f Ca 2÷ or EDTA, thereby supporting the hypothesis o f two non-associated species. An example of a close noncovalent association o f two immunologically related molecules is the complex o f high molecular weight kininogen and prekallikrein in plasma [23]. Molecules that are noncovalently complexes in plasma may be selectively adsorbed by an antibody to either component. Donaldson et al. [24] has adsorbed the high molecular weight kininogenprekallikrein complex in plasma with an antibody against kininogen or prekallikrein. Accordingly, an affinity column was prepared containing Sepharose
528
Fig. 3. Immunoelectrophoresis of fractions from gel •tration of plasma. The fractions across the activity peak from the gel filtered plasma (Fig. 2) were subjected to immunoelectrophoresis against the antiserum containing anti-12 and anti-h. The top, middle and bottom wells contained 45, 20 and 45 ~ul aliquots of fractions corresponding to 84, 96 and 108 ml, respectively.
4B covalently linked through cyanogen bromide to the monospecific antibody to 12. Purified factor V containing both h and 12 (Fig. 4, top) was passed over the gel and the 12 protein was adsorbed. As seen in Fig. 4, bottom, only the h chain eluted suggesting that it was not bound to the 12 polypeptide, or that it was present in excess. To distinguish between these possibilities and to optimize the conditions for the association of 12 and h, radioactively labelled factor V containing only h chain [8] was incubated with an antibody specific for the h chain. The antibody precipitated the radioactively labelled h chain alone or in the presence of bovine plasma (Table I). In contrast, the monospecific antibody against 12 was unable to precipitate purified radioactively labelled h chain showing the lack of contamination with 12. We then incubated radioactively labelled h chain with plasma to attempt to reform under native conditions the putative 12-h complex. However, the antibody against 12 could not precipitate labelled h chain even in the presence of plasma 12 suggesting that no complex was formed in plasma. Furthermore, the antibody to 12 when incubated at 1 : 1 dilution with bovine plasma showed no inhibition. Even at a antiserum to plasma ratio of 5 : 1 only 15~ of the activity was lost at 37°C in 10 min
Fig. 4. Immunoelectrophoresis of a factor V preparation before and after removal of the 12 polypeptide. Factor V prepared by the method of Sarawathi et al. [7] was passed over an immunoadorbent column prepared by coupling anti12 with cyanogen bromide activated Sepharose 4B. The top well contained the factor V preparation containing both 12 and h. The bottom well contained the material eluted from the column. The factor V activity of the material in each well was 0.02 unit. The troughs were tidied with the antiserum containing both anti-12 and anti-h.
compared to a 5 : 1 ratio of preimmune serum to plasma which demonstrated at 10% loss of factor V activity.
TABLE I The amount of 125 I-labelled h chain (125 I-h) in the incubation mixtures was 10 ~g (648475 cpm). The mixtures were incubated for 1 h at 37°C and 18 h at 4°C, followed by centrifugation at 12 000 Xg for 4 min. The precipitates were washed multiple times with distilled water. The bottom part of the tubes were cut out and counted in a gamma counter. Incubation mixtures
Radioactivity (cpm) in immunoprecipitate
Anti-h (200 ~1) + 12SI-h Anti-I2 (200 #l) + 125I_h Anti-h (200/~1) + plasma (100 td) + 12si_h Anti-12 (200 tad + plasma (100 ~tl) + 12Sl.h Pre-immune rabbit serum (200 ~tl) + 12si.h Pre-immune rabbit serum (200 ~1) + plasma (100 #1) + 12Sl-h
326 648 3940 342447 3708 3578 3 345
529 These studies indicate that the 12 polypeptide is unrelated immunologically to the h chain and is not complexed in purified systems or more importantly in plasma. Thus, in agreement with Esmon [8] and Mann et al. [21] we conclude that factor V is a single chain o f Mr = 2 9 0 0 0 0 - 3 3 0 0 0 0 and the 12 component is an adventitious protein co-purifying under the conditions used in the preparation o f Saraswathi et al.
[7]. Acknowledgements We thank Dr. Linda Knight for radioactive labelling o f factor V and Deborah Morinelli for typing the manuscript. This work was supported by NIH grant HL-18827 and SCOR grant 14217. References 1 Barton, P.G. and Hanahan, D.J. (1967) Biochim. Biophys. Acta 133,506-518 2 Esnouf, M.P. and Jobin, F. (1967) Biochem. J. 102, 660-665 3 Colman, R.W. (1976) Prog. Hemostosis. Thromb. 3,109143 4 Day, W.C. and Barton, P.G. (1972) Biochim. Biophys. Acta 261,567-573 5 Colman, R.W., Moran, J. and Phillip, G. (1970) J. Biol. Chem. 245,5941-5948 6 Smith, C.M. and Hanahan, D.J. (1976) Biochemistry 15, 1830-1838 7 Saraswathi, S., Rawala, R. and Colman, R.W. (1978) J. Biol. Chem. 253, 1024-1029
8 Esmon, C.T. (1979) J. Biol. Chem. 254,964-973 9 Nesheim, M.E., Myrmel, K.M., Hibbard, L. and Mann, K.G. (1979) J. Biol. Chem. 254,508-517 10 Nesheim, M.E. and Mann, K.G. (1979) J. Biol. Chem. 254, 1326-1334 11 Bartlett, S., Latson, P. and Hanahan, D.I. (1980) Biochemistry 19, 273-277 12 Ittyerah, T.R., Rawala, R. and Colman, R.W. (1980) in The Regulation of Coagulation (Mann, K.G. and Taylor, F.B., Jr., eds.), pp. 161-172, Elsevier/North-Holland, Inc., New York 13 Rawala, R., Saraswathi, S., Niewiarowski, S. and Colman, R. (1978) Circulation 58, II, 209 14 lttyerah, T.R., Rawala, R. and Colman, R.W. (1981) Eur. J. Biochem., in the press 15 Lewis, M.L. and Ware, A.G. (1953) Proc. Soc. Exp. Biol. Med. 84,640-643 16 Greenquist, A.C., Weinberg, R.M., Kuo, A. L. and Colman, R.W. (1975) Eur. J. Biochem. 58,213-222 17 Scheidegger, J.J. (1955) Int. Arch. Allergy Appl. Immunol. 7, 103-110 18 Ouchterlony, O. (1958) Prog. Allergy 5, 1-78 19 Bolton, A.E. and Hunter, W.M. (1973) Biochem. J. 133, 529-539 20 Tuszynski, G.P., Knight, L., Piperno, J.R. and Walsh, P.N. (1980) Anal. Biochem. 106,118-122 21 Mann, K.G., Nesheim, M.E. and Tracy, P.B. (1981) Biochemistry 20, 28-33 22 Greenquist, A.C. and Colman, R.W. (1975) Blood 46, 769-782 23 Mandle, R.J., Jr., Colman, R.W. and Kaplan, A.P. (1976) Proc. Natl. Acad. Sci. USA 73, 4179-4183 24 Donaldson, V.H., Kleiniewski, J., Saito, H. and Sayed, J.K. (1977) J. Clin. Invest. 60,571-578
Biochimica et Biophysica Acta, 714 (1982) 524-529
Elsevier BiomedicalPress BBA 21061 COAGULATION FACTOR V EXISTS UNCOMPLEXED IN BOVINE PLASMA T. ROY ITTYERAH,RAZIA RAWALAand ROBERTW. COLMAN * Thrombosis Research Center, Temple University Health Science Center, Jones Hall, 3401 North Broad St., Philadelphia, PA 19140 (U.S.A.)
(Received July 29th, 1981) (Revised manuscript received November2nd, 1981)
Key words: Factor V; Immunological assay; Blood clotting; (Bovine plasma)
Bovine coagulation factor V has been examined immunochemically to ascertain whether the coagulant polypeptide (h) with M r = 290 0 0 0 - 3 3 0 000 is complexed in plasma with a second immunoehemically distinct polypeptide (12) of Mr = 400 000. Antiserum containing antibodies to h and 12 detects the 12 polypeptide eluting earlier than the h chain on gel filtration of plasma with either added calcium or EDTA, consistent with the behavior of a higher molecular weight noninteracting species. An immobilized monospecific antibody to 12 removes only the 12 polypeptide from a purified factor V preparation containing both h and 12. Moreover, while a monospecific antibody to the h chain was able to precipitate purified radioactively labelled h chain alone or mixed with plasma, the 12 antibody was unable to precipitate radioactively labelled h chain even after attempted recombination of the h chain with 12 present in plasma. These studies indicate that the 12 polypeptide is not complexed to the h chain in a purified system or in plasma and reinforce the conclusion that factor V is a single polypeptide chain uncomplexed in plasma.
Introduction
Factor V, a plasma glycoprotein, functions as an essential cofactor in blood coagulation, non-enzymatically accelerating the formation of thrombin from its precursor prothrombin by the serine protease factor Xa in the presence of calcium and phospholipid. Since Barton and Hanahan [ 1] and Esnouf and Jobin [2] described preparations 3000-5000-fold purified over native plasma in 1967, many purification procedures have been described. However, much disagreement existed regarding the molecular forms and properties of the protein [3]. One of the major problems in the attempts to isolate the native form of the factor V molecule has been the susceptibility of the molecule to proteolytic fragmentation [4,5] both during blood collection and subsequent purification procedures. Procedural mo* To whom reprint requests should be addressed.
difications to avoid this problem included the use of proteolytic inhibitors [6] and the collection of blood by jugular venipuncture in the living animal [7]. The latter innovation allowed us to isolate bovine factor V with a high specific activity purified 2000-fold over starting plasma. We found that the factor V preparation consisted of a polypeptide of Mr = 290000 designated the h chain which was cleaved by thrombin to a carbohydrate rich fragment of Mr = 190 000 and a polypeptide of 110000 with an associated 18fold increase of activity [7]. Esmon [8] confirmed these findings by isolating a polypeptide of Mr = 290000 which after cleavage by thrombin yielded fragments of 210 000 and 115 000. Additional thrombin and higher temperatures led to further cleavage. Nesheim et al. [9,10] isolated a protein of Mr = 330 000 which on activation by thrombin gave products of Mr = 205000, 150000, 94000 and smaller fragments. All three groups agree that the polypeptide chain cleaved by thrombin has an Mr of
525 2 9 0 0 0 0 - 3 3 0 000 and that thrombin activation produces among other products a polypeptide of 9 4 0 0 0 - 1 1 5 0 0 0 essential for activity. However, in contrast to the later studies of Esmon [8] and Neshelm et al. [9,10], Saraswathi et al. [7] found that the h chain (Mr = 290000) was closely associated with a polypeptide designated the 12 polypeptide of M r = 400 000. On reduction the 12 polypeptide yielded two identical components of Mr = 200 000. In contrast to the h chain, the 12 component was not cleaved by thrombin [7]. Bartlett et al. [11] also reported a high molecular weight form of factor V although its polypeptide composition was more complex. Further investigation of the possible role of the polypeptide of Mr = 400 000 associated with factor V revealed that the h chain and the 12 polypeptide were immunochemically distinct [12]. Activation of factor V by two snake venoms cleaved the polypeptide of Mr = 290000 but not the 12 polypeptide [12,13]. The h chain but not the 12 polypeptide is present in platelets [12,14]. The 12 polypeptide makes if anything, a minor contribution to factor V activity but may alter its thermal stability [12]. Since neither Esmon [8] nor Nesheim et al. [9] find any 12 protein in their purified factor V preparations two hypotheses must be entertained. The first, is that the 12 polypeptide is noncovalently complexed with factor V and was inadverently lost during isolation, perhaps because of exposure to the dissociating solvents of high ionic strength used in both procedures [8,9]. The second, is that the 12 polypeptide represents a fortuitous association of two unrelated proteins which co-purify in some procedures. To resolve this issue we prepared monospecific antibodies against the 12 protein and the h chain and explored the immunochemical properties of both polypeptides in plasma to ascertain their relatonship. Methods
Preparation of factor V. Bovine blood was collected by jugular venipuncture into 0.10 vol. of 0.15M sodium citrate/0.1 M benzamidine hydrochloride/0.1 mM soybean trypsin inhibitor. Plasma was obtained by centrifugation at 2000Xg for 15 rain at 4°C, and factor V was purified as described by Saraswathi et al. [7] which yielded a preparation con-
taining both the h chain and the 12 polypeptide. The specific activity of the preparation was 35 units/rag increasing to 600 units/rag when activated by thrombin. Factor V was also purified as described by Esmon [8] from bovine blood collected by venipuncture into 0.10 vol. of 0.1 M sodium oxalate/0.1 M benzamidine hydrochloride/O.1 mM soybean trypsin inhibitor. This procedure yields only the h chain. The specific activity of the preparation was 52 units/mg increasing to 570 units/mg when activated with thrombin. Monitoring the procedure by immunoelectrophoresis against an antibody directed against the 12 polypeptide indicated that this polypeptide was removed during the calcium oxalate adsorption step of that procedure [8]. Assay of factor V. Factor V activity was assayed by the one-stage method of Lewis and Ware [15]. 1 unit of factor V is defined as the amount in 1 ml normal human plasma (pool of 10 separate plasmas). Preparation ofantisera. A rabbit antiserum to factor V was prepared by the procedure of Greenquist et al. [16] using factor V purified by the method of Saraswathi et al. [7] or that of Esmon [8] as the antigen. An antibody to the 12 polypeptide was prepared by two methods one of which has been described earlier [12]. The second method involved the adsorption of the antiserum containing anti-12 and anti-h prepared against the preparation of Saraswathi et al. [7] with a factor V preparation [8] containing only the 290000 molecular weight polypeptide. The antiserum was incubated with the factor V preparation for 1 h at 37°C and 18 h at 4°C, followed by centrifugation at 12000 ×g for 4 min. The supernatant contained only the anti-12 component since immunoelectrophoresis against plasma revealed only a single arc corresponding to the 12 polypeptide (see Results). Antibody to the h chain was prepared by injecting factor V prepared by the method of Esmon [8]. The antibody was capable of neutralizing factor V activity using purified factor V or factor V in plasma. Inhibition of factor V by 12 antisera. The inhibition of factor V activity in plasma by rabbit 12 antisera was performed at 3~°C. 12 antibody and plasma in either a ratio of 1 : 1 or 5 : 1 were mixed and incubated at 37°C. At specific times aliquots were removed and assayed for factor V coagulant activity.
526 The maximum degree of inhibition was obtained in 10 min. Sera from the same rabbits obtained prior to immunization were tested in parallel and served as control for the experiment. Immunodiffusion. Double immunodiffusion was performed by the method of Ouchterlony [ 18]. Immunoelectrophoresis. Immunoelectrophoresis was performed by a modification of the method described by Scheidegger [17], using 1% agarose in sodium barbital buffer, I = 0.0375, pH 8.6. Preparat&n of h chain. An immunoabsorbent column was prepared by coupling anti-12 to CNBrSepharose 4B as described before [12]. Factor V prepaired by the method of Saraswathi et al. [7] containing both the h and 12 polypeptides was applied to the colunm and the unadsorbed material was assayed for factor V activity and subjected to immunoelectrophoresis against the antiserum containing both anti12 and anti-h. (see Results). Gel filtration of plasma. Bovine blood was collected by venipuncture and anticoagulated with sodium citrate (0.015 M) to which benzamidine hydrochloride, soybean trypsin inhibitor, trasylol, hirudin and diisopropylfluorophosphate had been added to give final concentrations of 10 mM, 0.01 raM, 50 kallikrein inhibitor units/ml, 0.2 units/ml and 2 mM, respectively. Plasma (3 ml) was applied to a 1.5 × 85 cm column of Sepharose 4B (bed vol. 150 ml) equilibrated with Michaelis buffer (0.0265 M sodium acetate/0.0265 M sodium barbital/0.108 M NaCI, pH 7.4) containing 5 mM CaC12 and 10 mM benzamidineHC1. The same buffer was used for elution. The fractions were assayed for factor V activity and the fractions across the activity peak were subjected to immunoelectrophoresis against the antiserum containing both anti-12 and anti-h. The plasma was adsorbed with BaC12 to remove prothrombin and factor X. The BaClz-adsorbed plasma (3 ml) was gel filtered on the same Sepharose 4B column equilibrated with Michaelis buffer, pH 7.4, containing 5 mM CaCI2/10 mM benzaminidine-HCl, and the fractions were assayed for factor V activity. The fractions across the activity peak were subjected to immunoelectrophoresis against the antiserum containing both anti-12 and anti-h. Plasma (3 ml) was gel filtered on the column after it was equilibrated with Michaelis buffer containing 20 mM EDTA. No factor V activity could be de-
tected in the eluted fractions due to inactivation of factor V by EDTA. The fractions were subjected to immunoelectrophoresis against the antiserum containing both anti-12 and anti-h. Thyroglobulin and ferritin (10 mg each)were also applied to the column and the absorbance of the eluted fractions was measured at 280 nm. Preparation of 12sI-factor K Factor V was radioactively labelled using Bolton-Hunter reagent [19]. The free 12sI was removed by the method of Tuszynski et al. [20]. The specific radioactivity of the labelled factor V was 72 cpm/ng. Results and Discussion
Factor V purified by the method of Saraswathi et al. [7] contains two components exhibiting a reaction of nonidentity representing the 12 protein (Mr = 400000) and the h chain associated with factor V activity [7,12]. When this preparation is subjected to electrophoresis against an antibody to 12, a single component is seen which is identical in mobility to the single arc seen in bovine plasma (Fig. la). A reaction of identity was found on double-immunodiffusion between the 12 polypeptide in purified factor V [7] and plasma (Fig. lb). The 12 antibody did not form a precipitin line on immunoelectrophoresis or double-immunodiffusion against factor V prepared by the method of Esmon [8]. However, it is possible that at pH 8.6 and under the conditions of the immunoelectrophoresis the h and 12 chains were dissociated. Therefore, we examined factor V in plasma under conditions more closely approximating those in plasma. The behavior of purified bovine factor V on gel filtration has been well studied. Although its molecular weight is 2 9 0 0 0 0 - 3 3 0 0 0 0 it elutes in plasma anticoagulated with citrate before thyroglobulin (Mr = 670000) because of its high axial ratio [21]. However, it is possible that two noncovalently associated molecules could have been dissociated by removal of calcium in anticoagulated plasma. We gelfiltered bovine plasma (containing inhibitors) and BaC12-adsorbed bovine plasma in the presence of 5 mM Ca 2÷ and found that a single peak of factor V activity eluted just before thyroglobulin (Fig. 2). If factor V were a complex of the h chain and the 12 protein, they might be dissociated by completely removing metal ions since citrated plasma contains about 50
527
Thyrogtobulin Ferritin
t 0.4
E 03
0.2 > t9 O
I,i_
0-1
Oo
2s
|
5o
i
,2s ,5o
ELUTION VOLUME (ml) Fig. 2. Gel filtration of plasma m the presence of Ca2+. Plasma was gel-filtered on a column of Sepharose 4B and the
Fig. 1. a. Immunoelectrophoresis of purified factor V and plasma against an antibody to the 12 polypeptide. The upper well contained 0.025 unit of factor V purified as described by Saraswathi et al. [7]. The lower cell contained plasma with 0.025 unit of factor V activity, b. Double immunodiffusion of purified factor V and plasma against the antibody to 12. The left and right well contain 20 ~tl plasma diluted containing 1 : 16 or 0.002 units factor V activity. The top well contains 20 ul purified factor V (7) (100 units/ml) or 0.31 units factor V activity. The bottom well contains antiserum to 12 diluted 1 : 4.
~M free Ca 2+. Therefore, we gel-filtered plasma in the presence of 20 mM EDTA. Due to inactivation of factor V by EDTA [22], the activity could not be assayed in the eluted fractions. However, using an antiserum containing antibodies against h and 12 we could follow the distribution of the precipitin arcs in fractions representative o f the gel-filtered protein. Immunoelectrophoresis o f the fractions from plasma eluted in the presence o f 5 mM Ca 2÷ showed that the 12 polypeptide eluted early on the ascending limb, both h and 12 overlapped in the peak while the h
eluted fractions were assayed for factor V activity. Plasma adsorbed with BaC12 was also gel-filtered on the same column, o o, factor V activity in fractions from plasma; o. . . . . . o, activity in fractions from BaC12-adsorbed plasma. Thyroglobulin and ferritin were also gel-filtered on the column and their positions are shown.
chain eluted alone on the descending limb (Fig. 3). Immunoelectrophoresis of corresponding fractions from BaC12-adsorbed plasma eluted in the presence o f 5 mM Ca 2÷ and o f fractions eluted in the presence of EDTA exhibited the same pattern. These results clearly show that 12 and h do not co-elute but rather overlap when gel-filtered in the presence o f Ca 2÷ or EDTA, thereby supporting the hypothesis o f two non-associated species. An example of a close noncovalent association o f two immunologically related molecules is the complex o f high molecular weight kininogen and prekallikrein in plasma [23]. Molecules that are noncovalently complexes in plasma may be selectively adsorbed by an antibody to either component. Donaldson et al. [24] has adsorbed the high molecular weight kininogenprekallikrein complex in plasma with an antibody against kininogen or prekallikrein. Accordingly, an affinity column was prepared containing Sepharose
528
Fig. 3. Immunoelectrophoresis of fractions from gel •tration of plasma. The fractions across the activity peak from the gel filtered plasma (Fig. 2) were subjected to immunoelectrophoresis against the antiserum containing anti-12 and anti-h. The top, middle and bottom wells contained 45, 20 and 45 ~ul aliquots of fractions corresponding to 84, 96 and 108 ml, respectively.
4B covalently linked through cyanogen bromide to the monospecific antibody to 12. Purified factor V containing both h and 12 (Fig. 4, top) was passed over the gel and the 12 protein was adsorbed. As seen in Fig. 4, bottom, only the h chain eluted suggesting that it was not bound to the 12 polypeptide, or that it was present in excess. To distinguish between these possibilities and to optimize the conditions for the association of 12 and h, radioactively labelled factor V containing only h chain [8] was incubated with an antibody specific for the h chain. The antibody precipitated the radioactively labelled h chain alone or in the presence of bovine plasma (Table I). In contrast, the monospecific antibody against 12 was unable to precipitate purified radioactively labelled h chain showing the lack of contamination with 12. We then incubated radioactively labelled h chain with plasma to attempt to reform under native conditions the putative 12-h complex. However, the antibody against 12 could not precipitate labelled h chain even in the presence of plasma 12 suggesting that no complex was formed in plasma. Furthermore, the antibody to 12 when incubated at 1 : 1 dilution with bovine plasma showed no inhibition. Even at a antiserum to plasma ratio of 5 : 1 only 15~ of the activity was lost at 37°C in 10 min
Fig. 4. Immunoelectrophoresis of a factor V preparation before and after removal of the 12 polypeptide. Factor V prepared by the method of Sarawathi et al. [7] was passed over an immunoadorbent column prepared by coupling anti12 with cyanogen bromide activated Sepharose 4B. The top well contained the factor V preparation containing both 12 and h. The bottom well contained the material eluted from the column. The factor V activity of the material in each well was 0.02 unit. The troughs were tidied with the antiserum containing both anti-12 and anti-h.
compared to a 5 : 1 ratio of preimmune serum to plasma which demonstrated at 10% loss of factor V activity.
TABLE I The amount of 125 I-labelled h chain (125 I-h) in the incubation mixtures was 10 ~g (648475 cpm). The mixtures were incubated for 1 h at 37°C and 18 h at 4°C, followed by centrifugation at 12 000 Xg for 4 min. The precipitates were washed multiple times with distilled water. The bottom part of the tubes were cut out and counted in a gamma counter. Incubation mixtures
Radioactivity (cpm) in immunoprecipitate
Anti-h (200 ~1) + 12SI-h Anti-I2 (200 #l) + 125I_h Anti-h (200/~1) + plasma (100 td) + 12si_h Anti-12 (200 tad + plasma (100 ~tl) + 12Sl.h Pre-immune rabbit serum (200 ~tl) + 12si.h Pre-immune rabbit serum (200 ~1) + plasma (100 #1) + 12Sl-h
326 648 3940 342447 3708 3578 3 345
529 These studies indicate that the 12 polypeptide is unrelated immunologically to the h chain and is not complexed in purified systems or more importantly in plasma. Thus, in agreement with Esmon [8] and Mann et al. [21] we conclude that factor V is a single chain o f Mr = 2 9 0 0 0 0 - 3 3 0 0 0 0 and the 12 component is an adventitious protein co-purifying under the conditions used in the preparation o f Saraswathi et al.
[7]. Acknowledgements We thank Dr. Linda Knight for radioactive labelling o f factor V and Deborah Morinelli for typing the manuscript. This work was supported by NIH grant HL-18827 and SCOR grant 14217. References 1 Barton, P.G. and Hanahan, D.J. (1967) Biochim. Biophys. Acta 133,506-518 2 Esnouf, M.P. and Jobin, F. (1967) Biochem. J. 102, 660-665 3 Colman, R.W. (1976) Prog. Hemostosis. Thromb. 3,109143 4 Day, W.C. and Barton, P.G. (1972) Biochim. Biophys. Acta 261,567-573 5 Colman, R.W., Moran, J. and Phillip, G. (1970) J. Biol. Chem. 245,5941-5948 6 Smith, C.M. and Hanahan, D.J. (1976) Biochemistry 15, 1830-1838 7 Saraswathi, S., Rawala, R. and Colman, R.W. (1978) J. Biol. Chem. 253, 1024-1029
8 Esmon, C.T. (1979) J. Biol. Chem. 254,964-973 9 Nesheim, M.E., Myrmel, K.M., Hibbard, L. and Mann, K.G. (1979) J. Biol. Chem. 254,508-517 10 Nesheim, M.E. and Mann, K.G. (1979) J. Biol. Chem. 254, 1326-1334 11 Bartlett, S., Latson, P. and Hanahan, D.I. (1980) Biochemistry 19, 273-277 12 Ittyerah, T.R., Rawala, R. and Colman, R.W. (1980) in The Regulation of Coagulation (Mann, K.G. and Taylor, F.B., Jr., eds.), pp. 161-172, Elsevier/North-Holland, Inc., New York 13 Rawala, R., Saraswathi, S., Niewiarowski, S. and Colman, R. (1978) Circulation 58, II, 209 14 lttyerah, T.R., Rawala, R. and Colman, R.W. (1981) Eur. J. Biochem., in the press 15 Lewis, M.L. and Ware, A.G. (1953) Proc. Soc. Exp. Biol. Med. 84,640-643 16 Greenquist, A.C., Weinberg, R.M., Kuo, A. L. and Colman, R.W. (1975) Eur. J. Biochem. 58,213-222 17 Scheidegger, J.J. (1955) Int. Arch. Allergy Appl. Immunol. 7, 103-110 18 Ouchterlony, O. (1958) Prog. Allergy 5, 1-78 19 Bolton, A.E. and Hunter, W.M. (1973) Biochem. J. 133, 529-539 20 Tuszynski, G.P., Knight, L., Piperno, J.R. and Walsh, P.N. (1980) Anal. Biochem. 106,118-122 21 Mann, K.G., Nesheim, M.E. and Tracy, P.B. (1981) Biochemistry 20, 28-33 22 Greenquist, A.C. and Colman, R.W. (1975) Blood 46, 769-782 23 Mandle, R.J., Jr., Colman, R.W. and Kaplan, A.P. (1976) Proc. Natl. Acad. Sci. USA 73, 4179-4183 24 Donaldson, V.H., Kleiniewski, J., Saito, H. and Sayed, J.K. (1977) J. Clin. Invest. 60,571-578