THEJOURNA~ OF BIOLOGICAL CHEMIBTRY Vol. 238, No.1, January 1963

Printed in U.S.A.

Kinetics of Conversion of Prothrombin to Thrombin

by Biological Activators

I. DEMONSTRATION OF A PROTHROMBIN DERIVATIVE WHICH REACTS WITH PROTEOLYTIC ENZYME INHIBITORS *

N. RAPHAEL SHULMAN AND JOHN Z. HEARON

From the National Institute of Arthritis and Metabolic Diseases, National Institutes of Health, United States Public Health Service, Bethesda 14, Maryland

(Received for publication, June 21, 1962)

Purified prothrombin can be converted into thrombin in differ- ent ways (2), namely, by the use of biological activators, by the action of high concentrations of sodium citrate, and by treatment with trypsin. Biological activation of prothrombin requires lipo- protein tissue-factor(s) that have prothrombin-converting or thromboplastic activity, protein serum-factor(s) which augment thromboplastic activity, and ionic calcium. Thrombin formed in the presence of these quasi-physiological activators has been referred to as “biothrombin.” Most experimental evidence indi- cates that biothrombin is formed enaymatically from prothrom- bin; but the evidence has been interpreted variously as indicating that the enzymatic process is autocatalytic (3a), first order with respect to prothrombin (4), or involves a more complex sequence of reactions (2, 5).

Prothrombin conversion in citrate (6) resembles autocatalysis in that the time course of thrombin formation is a sigmoid curve and preformed thrombin must be added to initiate the reaction (5, 7, 8). The process of thrombin formation in citrate may in- volve extensive degradation or dissociation of prothrombin, for before thrombin formation, much of the carbohydrate and nitro- gen content of the original prothrombin preparation becomes nonprecipitable in trichloroacetic acid; and prothrombin initially present becomes refractory to biological activators (9). “Ci- trate thrombin,” which subsequently forms, has a lower molecu- lar weight than prothrombin (4). These observations have led to the conclusion that one or more prothrombin derivatives may form before formation of “citrate thrombin.” However, recent studies by Alexander (10) and Goldstein et al. (11) indicate that purified prothrombin that can be converted to thrombin by ci- trate is contaminated with other clotting factors, and therefore the significance of initial changes in physical properties of pro- thrombin preparations exposed to citrate is not clear.

Tryptic digestion of prothrombin results in release of large amounts of acid soluble material and formation of thrombin which has a lower molecular weight than prothrombin (4). It appears that “trypsin-thrombin” is an incidental consequence of nonspecific proteolysis.

There is little, if any, change in the physical properties of pro- thrombin when “biothrombin” forms; for acid soluble material is not released in significant amounts (2) and “biothrombin” ap-

* Part of this work was presented at the Fifth Annual Sym- posium on Blood, January 21, 1956, Detroit (1).

pears to have the same molecular weight and electrophoretic mobility as prothrombin (2, 4). These observations suggest that transformation of prothrombin to “biothrombin” may not involve formation of other prothrombin derivatives.

STP (12) is a crystalline protein that inhibits a number of proteolytic enzymes and also delays coagulation of blood. Glen- dening and Page (13) concluded from their study of the inhibition of “biothrombin” formation by ST1 that the inhibitor does not react with thrombin or with biological activators of prothrombin, but that it interferes with the generation of thrombin by forming a highly dissociable complex with prothrombin. Since ST1 also inhibited “citrate thrombin” formation, they suggested that the inhibitor might form a highly dissociable complex with a pro- thrombin derivative. Alkjaersig, Deutsch, and Seegers (14) in- terpreted results of their study as supporting the latter possi- bility.

Our earlier investigations on the nature of the anticoagulant activity of a proteolytic enzyme inhibitor obtained from human plasma and urine, and of STI, indicated that these inhibitors combine with a derivative of prothrombin formed during the con- version of prothrombin to thrombin by biological activators (1, 15). The detailed kinetic analysis in the present report demon- strates the manner in which these inhibitors complex with a pro- thrombin derivative and elucidates some of the properties of the derivative and of the complex. The mechanism of inhibition is unusual, for it involves an intermediate derivative of a proen- zyme participating in two essentially irreversible, competitive- rate processes.

EXPERIMENTALPROCEDURE

Prothrombin was separated from bovine plasma by the method of Seegers, Loomis, and Vandenbelt (16). The preparation used in this work contained 6.24y0 tyrosine by the Folin-Ciocnlteu method (17) based on dry weight of protein, and its ultraviolet absorption curve in 0.1 N sodium hydroxide had essentially the same configuration as those reported by Seegers et al. (16, 18); the E::,,, at 280 mp was 10.8. Its specific activity, determined by the method of Ware and Seegers (19), was 23,200 units per mg of tyrosine, based on a prothrombin assay that gave 220 to 260 units of prothrombin per ml of human plasma (see below). Thus, this preparation had the characteristics of high quality

‘The abbreviation used is: STI, soybean trypsin inhibitor.

155

by guest on October 20, 2017

http://ww

w.jbc.org/

Dow

nloaded from

156 Kinetics of Prothrombin Conversion Vol. 238, No. 1

MIN. IO 20 30 120

FIG. 1. Conversion of various concentrations of prothrombin to thrombin. Final concentration of reagents in the reaction mix- tures were: prothrombin, 0, 5 units per ml; A, 10 units per ml; 0, 20 units per ml; A, 40 units per ml; accelerator, 2.5 X 10m3 ml per ml; Ca++, 2 X 10-s M; thromboplastin, optimal by previous titration. Imidazole buffer, pH 7.4, in 0.147 M NaCl was used as a diluent. Time measured from addition of Ca++.

purified prothrombin (18). Although some purified prothrombin preparations may be relatively unstable (18), it is noteworthy that the preparation used in this work showed no detectable loss of activity and no spontaneous activation when kept at 5” in the lyophilized state over a g-month period, or when kept at 5” in water solution for several days.

The method used for converting prothrombin to thrombin by biological activators was essentially the “two-stage” procedure devised by Warner, Brinkhous, and Smith (20). Activation was carried out in siliconized tubes to prevent adsorption of thrombin on glass (13). The thromboplastin used was a NaCl suspension of acetone-extracted human brain (21) ; it contained no detectable accelerator or antithrombin activity. Highly diluted bovine serum was used as a source of accelerator activity (19) and is hereafter referred to as accelerator.2 The optimal accelerator concentration, 2.5 x 10e3 ml of bovine serum per ml of reaction mixture, had no detectable antithrombin activity. Although more purified sources of thromboplastic (e.g. (22)) and accelerator (e.g. (3b)) activity have been described, the activities of the ma- terials used in the present work were potent, easily standardized, and consistently reproducible. Optimal concentrations of thromboplastin, accelerator, and ionic calcium are defined in this work as the lowest concentrations that were found by assay (3~) to produce the highest possible yield of thrombin at the most rapid rate of prothrombin conversion. Solutions of purified pro- thrombin in distilled water were heated at 53” for 40 minutes im-

* Accelerator activity in bovine serum has been considered to be the same as Factor V (e.g. (19)) or Factor VII (e.g. (11)) activity in human blood; but actually bovine serum does not appear to contain factors that have the same characteristics as human Factor V or Factor VII. Human Factor V activity disappears rapidly in serum, whereas accelerator activity in bovine serum is stable; and bovine serum, although more potent than human serum as an accelerator in the “two-stage” procedure, contains only approximately 10% of the Factor VII activity of normal human serum when assayed against plasma from a patient con- genitally lacking in Factor VII. Goldstein et al. (11) attributed their anomalous results with bovine serum to a species specificity phenomenon.

mediately before use to destroy the small amount of accelerator activity present in the preparation (23). The only combination of reagents that resulted in thrombin formation was a mixture of prothrombin, thromboplastin, accelerator, and ionic calcium. There was no detectable thrombin formation over periods as long as 6 hours when any one reagent was excluded from the mixture.

Thrombin and prothrombin were assayed in terms of units, by means of methods similar to those described by Wagner et aE. (24), and Ware and Seegers (19). Fibrinogen used in the assay was bovine Fraction I (Armour and Company) further purified by the method of Laki (25), and dissolved at a concentration of 1.25 mg per 0.5 ml in 0.147 M sodium chloride containing imid- azole buffer, pH 7.4. Specifically, 1 unit of thrombin in the present work is that amount which clotted 0.5 ml of a solution of purified fibrinogen in 15 seconds at 2%‘; and 1 unit of prothrombin is that amount which, when fully converted to thrombin by op- timal concentrations of converting factors, produced 1 unit of thrombin. Appropriate aliquots of thrombin solutions were used to permit assay of 0.3 to 1.5 units of thrombin, the standard error of measurement over this range of concentrations being f 12.7 70. In experiments in which final yields of thrombin were measurecl, the mean value of at least five assays was obtained, the standard error of the mean of five assays being ~t5.77~.

The ST13 used, prepared by the method of Kunitz (12), was five times crystallized, and the trypsin3 used, prepared by the method of Northrop and Kunitz (26), was twice crystallized. By means of methods described by Kunitz (27), the activity of 1 pg of trypsin protein was found to be 1.07 X lop2 units; the combining proportion by weight of trypsin with ST1 was 1 part trypsin to 0.84 part STI. Methods of separating a proteolytic inhibitor with anticoagulant activity from human plasma and urine, and the chemical and physical characteristics of the prep- aration from urine which was used in the present work have been reported previously (15). One microgram of trypsin was in- hibited by 2.2 pg of the urine inhibitor. Other trypsin inhibitors used were ovomucoid inhibit.or,3 prepared by the method of Line- weaver and Murray (28)) pancreatic inhibitor,3 crystallized by the method of Kunitz and Northrop (29), and lima bean inhibi- tor,3 prepared by the method of Fraenkel-Conrat et al. (30). The combining proportions by weight of trypsin with the different inhibitors was found to be 1 part trypsin to 1.43 parts ovomucoid, 0.77 part pancreatic, and 0.45 part lima bean inhibitor.

RESULTS

1. Time Course of Prothrombin Conversion and Thrombin Yield in Uninhibited and Inhibited Reactions-When different concen- trations of prothrombin were converted to thrombin at optimal concentrations of thromboplastin, accelerator, and ionic calcium, the final yield of thrombin and the rate of thrombin formation were proportional to the concentration of prothrombin used (Fig. 1). Once the final yield of thrombin was formed, it re- mained constant for at least 6 hours. Even at the most rapid rate of prothrombin conversion, there was an initial lag phase during which little or no thrombin activity was detectable; and this phase could not be eliminated by prior incubation of any three of the four reagents required for thrombin formation. The concentration of thromboplastin could be increased fourfold, ac- celerator sixfold, and ionic calcium concentration threefold above

3 Purchased from Worthington Biochemical Corporation, Free- hold, New Jersey.

by guest on October 20, 2017

http://ww

w.jbc.org/

Dow

nloaded from

January 1963 N. R. Shulman and J. Z. Hearon 157

optimal without changing the thrombin yield or rate of prothrom- bin conversion. These various results agree with those obtained by Owren (3~) but differ somewhat from those obtained by Roka (5).

Ware and Seegers (23) have reported that purified prothrombin in NaCl solution at concentrations in the order of 10,000 units per ml may be inactivated by low concentrations of purified thrombin (in the order of 1 to 10 units per ml) or converted to thrombin by high concentrations of thrombin (in the order of 1000 units per ml). In the present work it was found that “bio- thrombin” from which thromboplastin was removed by sedi- mentation at 100,000 x g (22) neither inactivated nor activated prothrombin when NaCl solutions containing a mixture of as much as 40 units of thrombin per ml and 40 units of prothrom- bin per ml were incubated for as long as 6 hours at 28”. Our findings are similar to results obtained by Owren (3~).

The addition of increasing concentrations of ST1 to mixtures containing a fixed concentration of prothrombin and optimal concentrations of conversion factors produced progressive de- creases in the final yield of thrombin (Fig. 2). The rates of thrombin formation in the inhibited reactions were proportional to the final yields of thrombin and the same as the rates of unin- hibited reactions which produced equivalent final yields of thrombin (Fig. 2). Once the thrombin yield reached a plateau in an inhibited reaction, it remained fixed (Fig. 2) ; and no further formation of thrombin could be induced at this point by adding a surplus of conversion factors. Increasing the initial concen- tration of thromboplastin fourfold, of accelerator sixfold, and of ionic calcium threefold above optimal in mixtures that otherwise contained the same concentration of prothrombin and varying concentrations of ST1 as in Fig. 2, did not change the rates of thrombin formation or final yields of thrombin from those ob- tained with optimal concentrations of conversion factors.

8. Final Yield of Thrombin as Function of Relative Concentra- tions of XTI and Prothrombin-The curves of Fig. 3 show the units of thrombin that were prevented from being formed when various concentrations of ST1 were added to each of five different concentrations of prothrombin, Reduction of thrombin yield

I I i /i-?--i SBTI pg/ml

i

I- . = zo-

L+ A a a 2.5-r i

- I /a

“0 IO 20 30 ” 60

MINUTES

FIG. 2. The effects of various concentrations of ST1 on the rate of thrombin formation and final yield of thrombin. Concentra- tion of conversion factors as in Fig. 1; concentrations of ST1 indi- cated on graph; concentration of prothrombin, 25 units per ml in each case. The symbol, 0, reproduces the points of the curve in Fig. 1 obtained with 10 units of prothrombin per ml without in- hibitor.

1 I I /

40 PROTHROMBIN CONC. Units/ml. -

THROMBIN

SBTI IO 20 30 40

pg/ml.

FIG. 3. Inhibition of thrombin formation, five different fixed concentrations of prothrombin, varying concentrations of STI. Final concentrations of prothrombin and ST1 in reaction mixtures indicated on graph. Conditions otherwise as in Fig. 1. Units of thrombin inhibited obtained by subtracting final yields of throm- bin in inhibited reactions from final yields of thrombin in unin- hibited reactions. Data for prothrombin concentration, 25 units per ml, obtained from Fig. 2.

THROMBIN Umts/ml.

INHIBITED 30

IO

20

5

IO 2.5

0

PROTHROMBIN IO 20 30 40 Units /ml.

FIG. 4. Inhibition of thrombin format,ion, five different fixed concentrations of STI, varying concentrations of prothrombin. The data are the same as those of Fig. 3.

approached a limiting value as the concentration of ST1 was in- creased, and a single concentration of inhibitor (7 pg per ml, vertical broken line) produced approximately 50% reduction in the final yield of thrombin with each concentration of prothrom- bin used. As shown in Fig. 4, the amount of thrombin prevented from being formed by a fixed concentration of ST1 was directly proportional to the concentration of prothrombin used.

3. Dependence of Inhibition on Rate of Prothrombin Conversion -All of the above results were obtained at the maximal rate of prothrombin conversion, i.e. with optimal or above optimal con- centrations of conversion factors. The effect of decreasing the rate of prothrombin conversion on the inhibition of thrombin formation by ST1 is shown in Fig. 5. All curves in Fig. 5 were

by guest on October 20, 2017

http://ww

w.jbc.org/

Dow

nloaded from

158 Kinetics of Prothrombin Conversion Vol. 238, No. 1

2oi4/ /s I ?

0 1” , I / I I

MIN. IO 20 30 100

FIG. 5. Inhibition of thrombin formation at different rates of prothrombin conversion. Final concentration of prothrombin in all reaction mixtures, 33 units per ml. Concentration of conver- sion factors for Curves 1, 0, and 2, l , optimal, as in Fig. 1; for Curves 3, & and 4, A, thromboplastin and Ca++ concentrations optimal, but accelerator concentration reduced to 2.5 X 10m4 ml per ml. Curves 1 and 3 obtained without STI; Curves B and 4 obtained with 1 pg of ST1 per ml. Further explanation in text.

obtained with the same concentration of prothrombin (33 units per ml) in reaction mixtures. Curves 1 and 2 were obtained at an optimal rate of prothrombin conversion, Curve 1 showing throm- bin formation in the uninhibited reaction, and Curve 2 showing thrombin formation in the presence of 1.0 pg of ST1 per ml. At an optimal rate of prothrombin conversion, 1.0 pg of ST1 per ml reduced the thrombin yield by 3.8 units per ml. Curves S and 4 were obtained at a suboptimal rate of prothrombin conversion due to & optimal accelerator concentration. Although the rate of prothrombin conversion in the uninhibited reaction with sub- optimal accelerator concentration (Curve S) was slowed, the final yield of thrombin was the same as at the optimal rate of pro- thrombin conversion (Curve 1). The yield of thrombin obtained in the presence of 1.0 pg of ST1 per ml at the slower rate of prothrombin conversion (Curve 4) was much lower than the yield obtained in the presence of the same concentration of ST1 when the rate of prothrombin conversion was optimal (Curve 2). At the optimal rate of prothrombin conversion, the thrombin yield was reduced by 3.8 units per ml, and at the slower rate by 18 units per ml.

The amount of inhibition produced by a fixed concentration of inhibitor was always inversely related to the rate of prothrom- bin conversion. It did not matter which conversion factor was used to vary the rate. Decreases below optimal of accelerator, thromboplastin, or ionic calcium concentration all increased the inhibition of thrombin formation produced by a given concen- tration of STI. After the final yield of thrombin had been reached in an inhibited reaction carried out in the presence of a suboptimal concentration of any one of the conversion factors, addition of more of that conversion factor did not increase the thrombin yield.

In experiments done as in Fig. 5, but with accelerator decreased to approximately 2 X 1OP ml per ml or thromboplastin reduced to 0.1 optimal, the rate of prothrombin conversion was slowed to the extent that the final yield of thrombin in uninhibited reac- tions was reduced by 20 to 30%. Decreased thrombin yield at very slow rates of prothrombin conversion is a well known phe- nomenon (e.g. (3~)). At these very slow rates of prothrombin conversion, ST1 at a concentration of 0.20 pg per ml further re-

duced the thrombin yield by as much as 20 units per ml when the initial concentration of prothrombin was 60 units per ml. In these experiments, increasing prothrombin concentration above 60 units per ml in the presence of 0.20 pg of ST1 per ml produced relatively small increases in the amount of thrombin formation inhibited, i.e. inhibition was not directly proportional to prothrombin concentration, as it was in Fig. 4.

/,. Indications that STI Combines with Prothrombin Derivative- That ST1 does not act on any of the conversion factors had been shown previously (13), and was also demonstrated in the above experiments in which the amount of inhibition produced by a fixed concentration of ST1 was the same when the concentration of conversion factors was varied from optimal to far above op- timal (Section 1). Although ST1 does not inactivate thrombin itself (13, 15), the amount of thrombin formation inhibited by a fixed concentration of ST1 was directly proportional to the con- centration of prothrombin used in conversion mixtures (Fig. 4). Because the only identified substances in conversion mixture with ST1 were prothrombin, conversion factors, and thrombin, these various findings suggested that ST1 might combine with prothrombin. However, this possibility could not be reconciled with results of the following experiments in which the order of adding reagents was varied.

In one case, prothrombin and ST1 were incubated together for 30 minutes before conversion factors were added; and in the other case, ST1 and conversion factors were added simultaneously to prothrombin. A range of ST1 concentrations and prothrom- bin concentrations shown in Fig. 3, and the optimal and sub- optimal concentrations of conversion factors shown in Fig. 5, were used in these experiments. Regardless of the order of add- ing reagents, both the thrombin yield and the rate of thrombin formation was the same at a given ratio of ST1 to prothrombin and a given concentration of conversion factors.

The finding that the order of adding reagents had no effect on the rate of thrombin formation or final thrombin yield in in- hibited reactions indicated that any hypothetical combination between ST1 and prothrombin would have to occur so rapidly a.s to be completed before conversion factors, at any concentration, could act on prothrombin. Since it was evident from Figs. 2 and 5 that the substance that combined with ST1 was no longer available for transformation into thrombin by conversion factors, if prothrombin were the substance that combined with STI, then simply changing the rate of prothrombin converion by varying the concentration of conversion factors should not have affected the yield of an inhibited reaction. The fact that inhibition of thrombin formation varied inversely with the rate of prothrom- bin conversion (Fig. 5) therefore was inconsistent with the possi- bility that the inhibitor combined with prothrombin. Since ST1 apparently did not act on conversion factors, thrombin or pro- thrombin, a remaining possibility was that the inhibitor com- bined with an intermediate derived from prothrombin during the formation of thrombin.

5. Nature of Prothrombin Derivative that Combines with STI- Kunitz (27) has shown that trypsin and ST1 react together al- most instantaneously to form a highly stable, chemically inert compound, the dissociation constant of which is in the order of 1OVO M (31). In view of the rapidity of formation and high stability of the STI-trypsin complex, it was considered that tryp- sin might displace ST1 from a possibly less stable STI-derivative complex. Results of adding trypsin to the reaction mixture at different times after the final yield of thrombin had been reached

by guest on October 20, 2017

http://ww

w.jbc.org/

Dow

nloaded from

January 1963 N. R. Shulman and J. 2. Hearon

THROMBIN

0 ‘IO 20 30 40 50 60 70 80

MINU’TES

159

FIG. 6. Effects of adding trypsin to an inhibited reaction. Curve 1 was obtained with 25 units of prothrombin per ml without inhibitor; Curve 6 was obtained with the same concentration of prothrombin in the presence of 20 pg of ST1 per ml; optimal con- centrations of conversion factors used for both curves. A large volume of reaction mixture containing reagents of Curve 6 was divided into aliquants and treated as follows. Curve 2 was ob- tained by adding 24 /*g of trypsin per ml after 16 minutes; Curve S, by adding 24 pg of trypsin per ml after 40 minutes; Curve 4, by adding 30 pg of trypsin per ml after 40 minutes, and Curve 6, by adding 20 pg of trypsin per ml after 40 minutes. Trypsin was added over a period of 10 to 15 seconds with vigorous stirring, the

in an inhibited reaction are shown in Fig. 6. dfter addition of an amount of trypsin equimolar with respect to ST1 (Curves 1 and S), thrombin formation again took place and the subsequent final yield of thrombin approached that of the uninhibited reac- tion. Further formation of thrombin could be stopped at any point after addition of trypsin by again adding an excess of ST1 (Fig. 6, Curve ‘7). Large amounts of the chemically inert STI- trypsin complex added to reaction mixtures before initiating prothrombin conversion had no effect on the rate of thrombin formation or on the final yield of thrombin. It appeared, there- fore, that trypsin could competitively displace ST1 from com- bination with a prothrombin derivative, the apparent result being formation of the highly stable STI-trypsin complex and release of a prothrombin derivative which could then form thrombin.

The following experiments indicated that conversion factors are required to form thrombin from the derivative. No further formation of thrombin took place if trypsin was added to an in- hibited reaction (as in Curve 2, Fig. 6) along with sufficient sodium citrate to bind ionic calcium present in the reaction mixture. Further formation of thrombin could be stopped at any point after addition of trypsin by removing ionic calcium with citrate or any other calcium binding agent (Fig. 6, Curve 7) and started again if ionic calcium was re-added to the reaction mixture. If trypsin was added to neutralize the inhibitor after the final yield of an inhibited reaction was reached in the presence of a subop- timal concentration of accelerator or thromboplastin, the rate of

volume of trypsin solution added being & volume of the aliquant Aliquants were brought to 0” just before trypsin was added and returned to 22” immediately afterwards. Controls for Curves 2 through 6 were aliquants of Mixture 6 to which the different amounts of trypsin were added along with sodium citrate to give a final 0.013 M citrate concentration. There was no thrombin for- mation in the controls for Curves 2, S, and 5, but in the control for Curve 4, approximately 4 units of thrombin per ml formed within 10 minutes. Curve 7 was obtained by adding 40 pg of ST1 per ml to the mixture used to obtain Curve 2 ten minutes after trypsin was added, or by adding sodium citrate to a similar mixture to produce a 0.013 M citrate concentration.

formation of additional thrombin was much slower than the rate of formation of an equal amount of additional thrombin after ST1 was neutralized in the presence of optimal concentrations of conversion factors. Further formation of thrombin after neu- tralization of the inhibitor with trypsin could be prevented if thromboplastin was removed from the reaction mixture by sedi- mentation at 30,000 X g before trypsin was added. Thus it appeared that formation of thrombin from the derivative re- quired the presence of the conversion factors that are necessary to convert prothrombin into thrombin.

In view of the fact that trypsin is capable of enzymatically converting prothrombin directly into thrombin in the absence of ionic calcium or biological converting factors (32)) it should be emphasized that trypsin, in the amounts used to reverse an in- hibited reaction (Fig. 6, Curves 2, S, and 5), was not producing thrombin by direct enzymatic activity. I f trypsin had acted to produce thrombin enzymatically, thrombin formation would not have been dependent on the concentration of ionic calcium or other conversion factors. Moreover, the amount of thrombin that formed when trypsin was used to reverse an inhibited reac- tion was much more than the amount of thrombin that could be formed from the same prothrombin preparation by the enzymatic action of trypsin. The maximal yield of thrombin that could be obtained by converting prothrombin to thrombin with trypsin alone was found to be approximately 25% of the yield obtained when prothrombin was converted to thrombin by biological ac- tivators (Fig. 7). In contrast, when trypsin was used to reverse

by guest on October 20, 2017

http://ww

w.jbc.org/

Dow

nloaded from

160 Kinetics of Prothrombin Conversion Vol. 238, No. 1

THROMBIN Units I ml.

30 0 “do I

0 I 20 30 40 50

MINUTES

FIG. 7. Comparison of thrombin yields produced by conversion factors and by trypsin alone. Curve 1 was obtained with 31 units of prothrombin per ml in a reaction mixture containing optimal concentrations of conversion factors. For Curves 2 through 5, reaction mixtures contained 31 units of prothrombin per ml and varying concentrations of trypsin in imidazole-buffered saline, pH 7.4, at 25”; trypsin concentration, Curve 2,8.4 pg per ml; Curve S, 4.25 pg per ml;-&rue 4,1.1 pg per ml, and Curve 5,0.55 fig per ml. The same vields of thrombin were obtained when 0.013 M sodium citrate was present in the mixtures of prothrombin plus trypsin. The fibrinogen solution used for assaying units of thrombin con- tained 10 pg of ST1 per 0.5 ml.

TABLE I

Effects of various proteolytic enzyme inhibitors on thrombin formation

Inhibitor Final concent+on,of inhibitor Thrombin m reactmn Imxture formation

Inhibited*

Soy bean. Urine (plasma)

Lima bean, Ovomucoid. Pancreatic.

6% r&,/ml Mt U?dS/?d

1.2 2.5 1.25 X lo-’ 6.0 0.46 6.6 1.25 X 10-7 9.0 2.2 50.0 4.63 X 10-C 0 0.7 150.0 4.38 x 10-C 0 1.3 80.0 4.34 x 10-G 0

* Reaction mixture for measuring inhibition of thrombin for-

mation contained 25 units of prothrombin per ml and optimal concentrations of conversion factors.

t Molar concentration based on trypsin-inhibiting activity, assuming a trypsin to inhibitor ratio, l:l, and molecular weight

of trypsin, 24,000.

an inhibited reaction, the subsequent yield of thrombin ap- proached the maximal yield that could be obtained with bio- logical activators (Fig. 6, Curves 2 and 3). Trypsin added to solutions of preformed “biothrombin” in the absence of inhibitor produced only inactivation of the thrombin which was already present; and biological activators added to solutions of pre- formed “trypsin-thrombin” did not increase the thrombin yield. It was clear, therefore, that formation of thrombin after the addi- tion of trypsin as shown in Fig. 6, Curves 9, S, and 5, was not due to the direct enzymatic action of trypsin, but rather could be ex- plained by release of a prothrombin derivative from combination with ST1 and conversion of the derivative into thrombin by bio- logical activators.

In the experiments shown in Fig. 6, approximately 80% of the

possible yield of thrombin was obtained when the amount of trypsin added to the inhibited reaction was equimolar with re- spect to STI. Smaller amounts of trypsin than this could be added to inhibited reactions to produce further formation of thrombin, but small decreases in the amount of added trypsin produced sharp decreases in the increments of further formation of thrombin and sharper decreases in the rate at which it formed (e.g. Fig. 6, Curve 5). The least amount of trypsin that produced any further formation of thrombin was approximately 0.8 the amount required to neutralize STI. Trypsin added in amounts greater than equimolar with respect to ST1 resulted in additional formation of thrombin, but the maximal yield of thrombin was progressively lower and there was subsequent inactivation of thrombin which had formed (e:g. Fig. 6, Curve 4). Results of experiments in which more than equimolar amounts of trypsin were added were no doubt due to the combined effects of neu- tralizing ST1 and the direct tryptic digestion of thrombin that formed. Relationships between the molar ratios of trypsin to STI, and additional amounts of thrombin formed, are considered further under “Discussion.”

6. Anticoagulant Activity of Various Proteolytic Enzyme In-

hibitors--The purified proteolytic inhibitor obtained from human urine was substituted for ST1 in the series of experiments shown in Figs. 2, 3, 4, and 5; and in each case results with the urine inhibitor were qualitatively similar to those obtained with STI. When the two inhibitors were compared in terms of their trypsin- inhibiting activity (see “Experimental Procedure”), the amounts of urine inhibitor required for inhibition of thrombin formation,

although similar, were not quantitatively the same as the amounts of ST1 required. For example, when the concentration

of prothrombin and conversion factors was the same as in Fig. 2, an amount of urine inhibitor (13.2 pg) sufficient to inhibit the ac- tivity of 6 ng of trypsin prevented formation of 19 units of throm-

bin per ml, whereas an amount of ST1 (5 pg) sufficient to inhibit 6 fig of trypsin prevented formation of 12 units of thrombin per ml. The initial slope of curves plotted as in Fig. 3 (with abscissa changed to trypsin-inhibiting activity) were steeper, and asymp- totic values were higher with the urine inhibitor than with STI; but otherwise the curves had the same characteristics as those in Fig. 3. Just as in Fig. 4, the amount of thrombin formation inhibited by a fixed concentration of urine inhibitor was directly proportional to the concentration of prothrombin used. In ex- periments done as in Fig. 5, decreases in the rate of prothrombin

conversion had less influence on the degree of inhibition obtained with the urine inhibitor than with STI. For example, with the decreased rate of prothrombin conversion in Fig. 5, there was approximately a fivefold increase in the inhibition produced by a fixed concentration of STI, but a two- to threefold increase in the inhibition produced by a fixed concentration of urine inhibitor.

The other inhibitors tested were the pancreatic trypsin in- hibitor, lima bean trypsin inhibitor, and ovomucoid trypsin in- hibitor. These inhibitors did not inhibit thrombin formation when used at concentrations at least 30-fold greater, based on trypsin-inhibiting activity, than the concentrations of ST1 and urine inhibitor which markedly inhibited thrombin formation (Table I).

DISCUSSION

Analysis of Asymptotic Properties of System-The following analysis of the characteristics of inhibited and uninhibited pro- thrombin conversion reactions is based mainly on those proper-

by guest on October 20, 2017

http://ww

w.jbc.org/

Dow

nloaded from

January 1963 N. R. Xhulman and J. Z. Hearon 161

ties of the system that are related to the&& yields of thrombin. For simplicity, STI, or inhibitors that act like STI, prothrombin, and thrombin are denoted by S, P, and T, respectively; the term “factors” refers collectively to Ca++, thromboplastin, and ac- celerator, any one of which is designated as a factor. The amount of inhibition is defined as the difference between the final yields of T in the absence and presence of S; the degree of inhibition is defined as the amount of inhibition divided by the initial concentration of P. There are several distinctive proper- ties of the system which must be accounted for by any proposed reaction scheme or kinetic model. They may be summarized as follows. (a) For varying initial concentrations of P in the absence of S, the final yield of T is proportional to the initial concentration of P (Fig. 1). (b) At a given concentration of fac- tors, the presence of S decreases not only the rate of formation of T, but also the final yield (Fig. 2). (c) The degree of inhibi- tion produced by a given concentration of S is independent of the initial concentration of P; for example, in Fig. 3, at one con- centration of S the degree of inhibition is one-half for each con- centration of P. Therefore, the amount of inhibition produced by a given concentration of S is proportional to the initial con- centration of P (Fig. 4). (d) For given initial concentrations of P and S, the degree of inhibition is a function of the concentration of the factors; specifically, the degree of inhibition is increased as the rate of conversion of P to T is decreased by lowering the level of one or more of the factors (Fig. 5).

It appears necessary to assume that X enters whatever reac- tions in which it participates in an essentially irreversible man- ner. For, if the overall P-to-T conversion is irreversible (and there is no evidence to indicate that T, in the presence of factors, is converted to P or any other species), no reversible combination can account for the decreased final yield in the presence of X. The evidence indicates (Section 4) that X combines with P or some derivative or intermediate formed from P, and not with T or any of the factors. The evidence further indicates that it is not P itself that combines with X. It is demonstrable (Section 5) that in the presence of S there is stored or sequestered, in a form not convertible to T, a species which can be regenerated by the addition of trypsin and converted to T in the presence of the biological conversion factors. On the basis of the above ob- servations it is possible to construct relatively simple reaction schemes which predict properties (a) through (d) above and which are consistent with the qualitative findings of Section 4. When the analysis is confined to the study of final yields there is a fairly broad class of models, any one of which is consistent with the observed properties of the system. In a forthcoming paper,4 alternative models are extensively analyzed. In what follows we present a minimal model which exhibits properties (a) through (d) and which has additional implications that are discussed.

Consider the model

P&d+ T

Is

where X is an intermediate, C is a complex between X and X, k is a formal first order rate constant, and r is the second order

4 In a forthcoming paper by J. Z. Hearon and N. R. Shulman, it is shown that this assumption can be validated.

rate constant for the combination of X and X. The rate equa- tions for T and C are written as

dT - = kS dt

dC - = rsx dt

(2)

Assuming S has an essentially constant value, So, Equations 1 and 2 can be integrated, under the conditions T = C = 0 at t = 0, to give Equations 3 and 4 below. The essential constancy of X can be validated4 both in terms of the time course of the yield of thrombin and the final yield, but we remark in passing that it is all but evident that no model can lead to property (c) without this condition. Given this condition, we have

T(m) = k/l (3)

C(m) = rSw4 (4)

where rl = s,” X(t) dt. But clearly if PO is the initial value of P

T(m)+C(m)= ,$I+$$]=& (5)

which with Equations 3 and 4 gives

T(m) = & a (6)

where o( = r/k. The amount of inhibition, I = PO - T( m) is then

POSO I=--..--

1 + so

(7)

01

Equation 6 gives the final yield as a decreasing function of So. It predicts that T( =a) = PO/2 when &!& = 1 and from Equation 7 the corresponding degree of inhibition is one-half when So = I/ar and this point of one-half inhibition is independent of PO (cf. Fig. 3). The amount of inhibition is proportional to PO and according to Equation 7, the slope of I against PO increases with 80, approaching unity as SO becomes large (cf. Fig. 4). Finally, from Equation 7 the degree of inhibition for given So increases as cy increases. Therefore, since cz = r/k, any change in the system which decreases k should increase the degree of inhibition at a fixed X0. This is in accord with experimental fact if k is an increasing function of the concentration of each factor. That k is indeed such a function is supported by the facts (Sec- tion 5) that in an inhibited system that has attained its final yield, additional T is formed, after X is neutralized with trypsin, only in the presence of Ca++; and given the presence of Ca++ the rate of production of T after neutralization of X is an increasl ing function of accelerator and thromboplastin. Therefore, the above reaction scheme correctly predicts all of the asymptotic properties of the system, i.e. the final yield as a function of PO and SO, as well as the rate dependence of the degree of inhibition as it is influenced by the concentration of the factors.

In the above scheme, the functional form of Equation 6 and hence of Equation 7 is a consequence of the assumption of a mole- for-mole combination of S with X. The data on the final yields of thrombin are clearly consistent with this assumption. In a

by guest on October 20, 2017

http://ww

w.jbc.org/

Dow

nloaded from

162 Kinetics of Prothrombin Conversion Vol. 238, No. 1

forthcoming paper,4 it is shown that the data on the time course of thrombin production are inconsistent with any scheme that supposes a different stoichiometry. In view of the evidence that S and X combine mole for mole, the following apparent stoichiometric relationships have significant implications.

Implications Concerning Prothrombin and Its Derivatiue.s-The best preparations of purified prothrombin are capable of yielding approximately 2000 units of thrombin per mg under optimal con- ditions of conversion (18) ; the molecular weight of prothrombin is approximately 63,000 (4) ; and the molecular weight of ST1 is 20,000 (31). In experiments done as in Fig. 5, but with higher prothrombin concentrations and slower rates of prothrombin conversion, the thrombin yield could be reduced by at least 20 units per ml with ST1 at a concentration of 0.20 pg per ml (Sec- tion 3). Assuming that in these experiments ST1 was not in surplus, that prothrombin was 100% pure, and that all prothrom- bin molecules were changed to a derivative in the pathway of thrombin formation, then the apparent molar ratio of derivative to ST1 in the complex would be a minimum of 16:l. At even slower rates of prothrombin conversion in 25% sodium citrate, the apparent molar ratio of prothrombin to ST1 required for in- hibition was found to be a minimum of 4O:l by Alkjaersig, Deutsch, and Seegers (14). Since all of the characteristics of inhibitory reactions can be explained only on the basis of a 1: 1 irreversible combination between ST1 and derivative, one could conclude that the molecular species that can be converted to thrombin by biological activators (or by 25% sodium citrate) accounts for no more than 670 (and possibly as little as 2 %) of a purified prothrombin preparation. considering the physical homogeneity of purified prothrombin (4), it is unlikely that this percentage represents the degree of purity. It may reflect, rather, the fraction of a single molecular species that is converted by biological activators into thrombin. Trypsin is an example of a prothrombin activator that appears to convert even a smaller fraction of prothrombin molecules into thrombin; for the yield of “trypsinthrombin” is approximately 25 70 the yield of throm- bin obtained with biological activators (Fig. 7). Thus, trypsin appeers to act on prothrombin in more than one way; and for every molecule that is changed to a form that has thrombin ac- tivity, apparently many more are altered in such a way that they have no thrombin activity. Biological activators (and 25% sodium citrate), which convert more prothrombin to thrombin than does trypsin, nevertheless may change a large proportion of prothrombin into derivatives that do not yield thrombin. This would suggest that if thrombin could be physically separated from other prothrombin derivatives, or if there were a means of converting every prothrombin molecule to thrombin, then throm- bin would have manyfold the specific activity ascribed to it so far.

Approximation of Association Constant of Derivative-STI Com- plex-In experiments in which trypsin was added to inhibited reactions to produce further formation of thrombin (Section 5, Fig. B), amounts of trypsin equimolar with respect to ST1 pro- duced approximately 80% the maximal thrombin yield, amounts less than 0.8 equimolar produced no appreciable additional thrombin, and effects of amounts greater than equimolar were complicated by the presence of free trypsin. Since trypsin re- versed inhibition apparently by displacing ST1 from the deriva- tive-ST1 complex, the amount of bound derivative freed at a particular molar ratio of trypsin to ST1 would be a function of the relative affinities of trypsin and the derivative for STI. Given the association constant of the trypsin-ST1 complex, an

order of magnitu.de estimate of the association constant of the derivative-ST1 complex was obtained in the following manner.

After the addition of trypsin, the competing equilibria are

x+sec1 (8)

E+SzCz (9)

where X, as before, is the prothrombin derivative, Cl is the com- plex previously denoted by C, E is trypsin, and Cz is the trypsin-S complex. In the absence of trypsin, Reaction 8 is exceedingly far to the right with a vanishingly small concentration of X, for it has been seen that stable yields of thrombin obtained in a matter of minutes remain constant for many hours. If the asso- ciation constants for Reactions 8 and 9 are K1 and KS, respec- tively, then the mass action expressions are

Cl/X = K$’ (10)

Cz/E = KeS (11)

The totality conditions are

Xo = X + CI = X(1 + KIS) (12)

so = s + Cl + c2 03)

Eo = E + Cz (14)

Since the data on final yields are consistent with the assumption that S is essentially constant (see above), the term Cl in Equa- tion 13 can be neglected with respect to SO. If CL is negligible to good approximation in the absence of E, it is certainly negligi- ble when E is competing with X for S. Neglecting Cl in Equa- tion 13, Equations 10 through 14 give

s,=s+g& 2

(15)

where fz, is the dissociation constant, i.e. K& = 1. Solution of Equation 15 for S in terms of the molar ratio f = Eo/Xo gives

28 = (1 - f)& - l?s + [((f - I)& + &)z + ~&S’O]‘~~ (16)

For a fixed value of X0 and given & S can be computed from Equation 16 for all values off. Then

X/X, = l/(1 + K&‘) (17)

gives the proportion of the derivative freed under these conditions for any assigned value of KL In the experiments being discussed (Section 5, Fig. 6), the value of 80 was 20 pg per ml = 10F6 M.

With this value of 80 and 22 E lo-lo M (31), values of X were computed for a range of values off and the corresponding values of X/X0 were calculated for various assignments of K1. These calculated values (expressed as per cent, i.e. 100 X/X,) are shown for K1 = lo8 liters per mole in Fig. 8. Approximately 50% of the derivative is freed when f = 1. The release of bound deriva- tive is an exceedingly sensitive function off in the neighborhood off = 1. When f = 1.05, approximately 84% of the derivative is freed; when f = 0.8, somewhat less than 5% is freed. These results are in fair agreement with the experimental results, and such is not the case when values either higher or lower than 108 are assigned to K1. For example, for K1 = 107, f = 0.8, and f = 1 lead, respectively, to 33% and 99% freed; for K1 = 109, the corresponding values are 0.5% and 1.0%. We therefore conclude that K1 is of the order of 108. This relatively firm asso- ciation between derivative and ST1 suggests that it may be

by guest on October 20, 2017

http://ww

w.jbc.org/

Dow

nloaded from

January 1963 N. R. Shulman and J. Z. Hearon 163

possible to separate physically the complex from other sub- stances in a conversion mixture.

Of the various proteolytic enzyme inhibitors that were used (Section 6), only ST1 and the urine inhibitor prevented thrombin formation. Since inhibition of thrombin formation is a com- petitive-rate reaction, effectiveness of an inhibitor in reducing the final thrombin yield would depend on the rate constant of its reaction with derivative, rather than on the association constant of a potential derivative-inhibitor complex which may fcrm. ST1 and the urine inhibitor have the properties in common of combining instantaneously with trypsin to form relatively irre- versible complexes (15, 27). The complex between pancreatic inhibitor and trypsin has an association constant equal to that of the STI-trypsin complex (31) ; but pancreatic inhibitor re- quires a period of time to attain equilibrium with trypsin (31). The complex between ovomucoid (31) or lima bean (30) inhibitor and trypsin has a lower association constant than the STI-trypsin complex. Since inhibitors that prevent thrombin formation are those that combine most rapidly and firmly with trypsin, it would appear that the prothrombin derivative and trypsin may contain a similar reactive group which is responsible for combination with proteolytic inhibitors.

Physiological Implications-The proteolytic inhibitor with anticoagulant activity obtained from plasma (and urine) (15) may play a homeostatic role in blood coagulation. Because in- hibition of thrombin formation by proteolytic inhibitors is mark- edly rate-dependent, inhibition would be most effective at the slow rates of prothrombin conversion which may occur intra- vascularly, but would lose its effectiveness at rapid rates of pro- thrombin conversion which occur extravascularly. However, since almost all of the trypsin inhibiting activity of plasma is due to several different inhibitors (33) which do not act as anticoagu- lants (15), and since it is difficult to separate physically plasma inhibitors, it is not known with certainty whether the plasma in- hibitor which acts as an anticoagulant is present normally in amounts that would influence physiological blood coagulation

(15). We have confirmed the observation of Biggs, Douglas, and

Macfarlane (34) that ST1 interferes with the generation of throm- boplastic activity from blood elements; but we have not been able to demonstrate that ST1 inactivates preformed blood throm- boplastin. In view of the results reported herein, and of the body of evidence that indicates that minute amounts of thrombin accelerate thromboplastin formation from blood elements (e.g. 35-37)) it appears that the effects of ST1 on blood thromboplastic activity are due primarily to inhibition of thrombin formation and consequent suppression of thromboplastin formation, rather than to inactivation of preformed thromboplastin. Proteolytic inhibitor anticoagulants not only inhibit thrombin formation, but also appear to interfere with other coagulation reactions, such as thromboplastin generation, which may be catalyzed by thrombin or some other prothrombin derivative.

At very slow rates of prothrombin conversion, the final yield of thrombin is decreased for reasons which are not yet clear (e.g. 3~). The stoichiometry of the STI-derivative complex (dis- cussed above) suggests that prothrombin may give rise to deriva- tives which do not yield thrombin. If derivatives outside the pathway of thrombin formation are formed preferentially at slow rates of prothrombin conversion, this may represent still another regulatory process in blood coagulation.

Although the usual increases in trypsin inhibition which occur

04 0.8 1.2 1.6

MOLAR RATIO TRYPSIN:INHIBITOR

FIG. 8. Calculated amount of bound intermediate freed at different molar ratios of trypsin to STI. Association constant of STI-intermediate complex is assumed to be lo8 liters per mole. See text.

in various disease states (33) are not due to an inhibitor with anticoagulant activity (15), there is a possibility that physio- logically significant increases in a proteolytic inhibitor antico- agulant may occur. The mechanism of inhibition presented herein appears to be relevant to the problem of anticoagulants which arise occasionally in association with lupus erythematosus. These anticoagulants are puzzling because they inhibit thrombin formation to a degree that is dependent on thromboplastin concentration (e.g. (3%40)), yet they do not appear to combine directly with prothrombin or thromboplastin. Since the ac- tivity of anticoagulants associated with lupus is influenced by the rate of prothrombin conversion, the possibility exists that these anticoagulants combine with the prothrombin derivative which reacts with proteolytic inhibitors.

SUMMARY

Kinetic analysis of the effects of soybean trypsin inhibitor and other proteolytic enzyme inhibitors on the conversion of pro- thrombin to thrombin by biological activators indicated that a prothrombin derivative formed during the conversion process combines with proteolytic inhibitors. The rate of thrombin production from the derivative was found to be dependent on the concentration of biological activators and to compete with the rate of formation of an essentially irreversible derivative-in- hibitor complex. Degree of association between derivative and inhibitor in the complex and the nature of the conversion of free derivative to thrombin was determined by displacing inhibitor from the complex with trypsin. Stoichiometry of complex formation indicated either that processes of prothrombin ac- tivation convert only a small fraction of prothrombin molecules

by guest on October 20, 2017

http://ww

w.jbc.org/

Dow

nloaded from

lG4 Kinetics of Prothrombin Conversion Vol. 238, No. 1

into the derivative which gives rise to thrombin, or that current methods of purifying prothrombin yield a very impure product. A minimal reaction model deduced from experimental results is described by linear kinetic equations. Physiological implications of the various findings, in particular the possible regulatory effects of plasma proteolytic inhibitors in blood coagulation, are dis- cussed.

21. Acknowledgment-We greatly appreciate the assistance of Miss

Merilyn C. Hiller in this work. 22.

REFERENCES 23.

1. SHULMAN, N. R., J. Mich. State Med. Sot., 55, 968 (1956). 2. SEEGERS, W. H., AND ALKJAERSIG, N., Arch. Biochem. Bio-

phys., 61,1 (1956).

24.

3. OWREN, P. A., The coagulation of blood, J. Chr. Gundersen, Boktrykkeri, Oslo, 1947, (a) pp. 214-219, (b) pp. 76-90, (c) pp. 174-226.

25. 26.

4. LAMY, F., AND WAUGH, D. F., Physiol. Revs., 34, 722 (1954). 5. ROKA, L., Transactions of the fourth international congress of

biochemistry, Vol. 10, Pergamon Press, Ltd., London, 1958, p. 195.

27. 28.

29. 6. SEEGERS, W. H., MCCLAUGHRY, M. S., AND FAHEY, J. L.,

Blood, 5, 421 (1950). 30. 7. MILLER, K. D., AND VAN VUNAKIS, H., J. Biol. Chem., 223,

227 (1956). 8. SCHULTZE, H. E., AND SCHWICK, G., Hoppe-Seyl. Z., 289, 26

(1951).

31. 32.

9. LORAND, L., ALKJAERSIG, N., AND SEEGERS, W. H., Arch. Biochem. Biophys., 45, 312 (1953).

10. ALEXANDER, B., Transactions of the fourth international con- gress of biochemistry. Vol. 10, Pergamon Press, Ltd., London, 1958, p. 37.

33. 34.

11. GOLDSTEIN, R., LEBOLLOC’H, A., ALEXANDER, B., AND ZON- DERMAN, E., J. Biol. Ghem., 234, 2857 (1959).

12. KUNITZ, M. M., J. Gen. Physiol., 29, 149 (1946). 13. GLENDENING, M. B., AND PAGE, E. W., J. Clin. Invest., 30,

1298 (1951).

35.

36.

37.

38. 39.

14. ALKJAERSIG, N., DEUTSCH, E., AND SEEGERS, W. H., Am. J. Physiol., 180, 367 (1955). 40.

15. SHULMAN, N. R., J. Biol. Chem., 213, 655 (1955).

16.

17. 18.

19.

20.

SEEGERS, W. H., LOOMIS, E. C., AND VANDENBELT, J. M., Arch. Biochem.. 6, 85 (1945).

FOLIN, O., AND &O&ALT&J, ‘ii., J. Biol. Chem., 73, 627 (1927). SEEGERS, W. H., AND ALEJAERSIG, N., Am. J. Physiol., 172,

731 (1953). WARE, A. G., AND SEEGERS, W. H., Am. J. Clin. Path., 19,

471 (1949). WARNER, E. D., BRINKHOUS, K. M., AND SMITH, H. P., Arch.

Path., 18, 587 (1943). ALEXANDER, B., in TOCANTINS, L. M. (Editor), The coagulation

of blood, Grune and Stratton, New York, 1955, p. 90. CHARGAFF, E., BENDICH, A., AND COHEN, S. S., J. Biol. Chem.,

156, 161 (1944). WARE, A. G., AND SEEGERS, W. H., J. BioZ. Chem., 174, 565

(1948). WAGNER, R. H., GRAHAM, J. B., PENICK, G. D., AND BRINK-

HOUS, K. M., in TOCANTINS, L. M. (Editor), The coagulation of blood, Grune and Stratton, New York, 1955, p. 105.

LAKI, K., Arch. Biochem. Biophys., 32, 317 (1951). NORTHROP, J. H., AND KUNITZ, M. M., J. Gen. Physiol., 16,

267 (1932). KUNITZ, M. M., J. Gen. Physiol., 30,291 (1947). LINEWEAVER, H., AND MURRAY, C. W., J. BioZ. Chem., 171,

565 (1947). KUNITZ, M. M., AND NORTHROP, J. H., J. Gen. Physiol., 30,

291, 311 (1947). FRAENKEL-CONRAT, H., BEAN, R. S.. DUCAY. E. D.. AND

OLCOTT, H. S., A&h. kochem. Biophys., 37, 393 (1952). GREEN. N. M.. J. BioZ. Chem.. 206. 535 (1953). EAGLE,‘H., AN; HARRIS, T. N.,‘J. &en. PiysiiZ., 20, 543 (1937). SHULMAN, N. R., J. Exptl. Med., 95,593 (1952). BIGGS, R., DOUGLAS, A. S., AND MACFARLANE, R. G., J.

Physiol. (London), 122, 554 (1953). BIGGS, R., DOUGLAS, A. S., AND MACFARLANE, R. G., J.

Physiol. (London), 119, 89 (1953). MIALE. J. B.. AND WILSON. M. P.. Am. J. Clin. Path.. 26.

984 (1956). I I

CONNOR, W. E., WARNER, E. D., AND CARTER, J. R., J. CZin. Invest., 40, 13 (19611.

FRICK, P’. Gi, Blood, i0, 691 (1955). RAPAPORT, S. I., AMES, S. B., DUVALL, B. J., Blood, 16, 212

(1960). PERKINS, H. A., AND ACRA, D. J., Thromb. Diath. haem., 6,

250 (1961).

by guest on October 20, 2017

http://ww

w.jbc.org/

Dow

nloaded from

N. Raphael Shulman and John Z. HearonREACTS WITH PROTEOLYTIC ENZYME INHIBITORS

I. DEMONSTRATION OF A PROTHROMBIN DERIVATIVE WHICH Kinetics of Conversion of Prothrombin to Thrombin by Biological Activators:

1963, 238:155-164.J. Biol. Chem. 

  http://www.jbc.org/content/238/1/155.citation

Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/238/1/155.citation.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on October 20, 2017

http://ww

w.jbc.org/

Dow

nloaded from