The Conversion of Prothrombin to Thrombin II. DIFFEREKTIATION BETWEEN THROMBIN-AND FACTOR X,CATALYZED PROTEOLYSES*

SUMMARY Prothrombin is activated by activated Factor X f (Factor V and phospholipid) in the presence of diisopropylfluorophosphate which preferentially inhibits thrombin. Under these conditions a single activation fragment, designated Fragment 1.2, is released which represents the complete nonthrombin portion of the prothrombin molecule. The same fragment is seen as a transient species in the absence of diisopropylfluorophosphate when the activation reaction is monitored by disc electrophoresis in 8 M urea. The NH2-terminal residue of both prothrombin and Fragment 1.2 is alanine, suggesting that Fragment 1.2 is derived from the NH&erminal end of the prothrombin molecule. Fragment 1.2 is not cleaved by activated Factor X f (Factor V and phospholipid), but is cleaved by thrombin to form two fragments which are identical by electrophoresis, ion exchange chromatography, and amino acid composition to Fragment 1, the fragment released from prothrombin on incubation with thrombin, and Fragment 2, the fragment released by activated Factor X when Intermediate 1 is converted to Intermediate 2 (Paper I; OWEN,

. It is proposed that the formation of Intermediate 1 during prothrombin activation is solely the result of thrombin-catalyzed proteolysis of prothrombin.
Rapid activation of prothrombin requires four components in addition to prothrombin itself: two proteins, Factor X,' and Factor V, phospholipid, and Ca2+. Of these components, only X, catalyzes the proteolytic conversion of prothrombin to thrombin in the absence of the other components (l-4).
Studies in both our laboratory (5) and other laboratories (6612) have resulted in the following description of this process (5).
As indicated by Equation 1, thrombin cleaves prothrombin to form two polypeptide products.
These products, which can be produced in the absence of X,, are indistinguishable from the Intermediate 1 and Fragment 1 isolated after partial activation of prothrombin with X, (5). Mechanisms in which Intermediate 1 is assumed to be on the kinetic path of X,-catalyzed prothrombin activation have been proposed (13,14). Since the concentration of thrombin greatly exceeds that of X, after very little prothrombin has been activated, it is difficult to demonstrate unambiguously that X, actually catalyzes proteolysis of prothrombin to form Intermediate 1 and Fragment 1. While the present study was in progress, Stenn and Hlout (11) reported results which they interpreted to indicate that X, does not cleave the Intermediate l-Fragment 1 peptidc bond of prothrombin. If the proposal of Stenn and Ulout is correct, then 011 the basis of the chemical equations above and the knowledge that thrombin comes from the COOH-terminal end of prothrombin (15), an activation fragment which consist,s of Fragments 1 and 2 still covalently linked is predicted.
As thrombin cleaves the peptide bond linking Fragment 1 and Intermediate 1 (5), successful demonstration of this hypothetical product, Fragment 1.2, would require rapid inhibition of the thrombin formed during activation.
Stenn and Hlout (11) used diisopropylfluorophosphate to prevent degradation of a transient band which was observed by sodium dodecyl sulfate gel electrophoresis and which provided the basis for their proposal.
The same approach was employed in this investigation, although the activations reported here have been carried out under more nearly physiological conditions, with the complete prothrombin-activating system as well as with X, alone. This communication reports the isolat,ion and characterization of the proposed product, Fragment 1.2, and demonstrates that thrombin, but not Factor X,, is responsible for the formation of the separate activation Fragments 1 and 2.
is abbreviated [X,, V, phospholipid, Cat+]. All activations were performed in 0.1 M NaCl, 0.01 M CaClz, 0.02 nf Tris-HCl, pH 7.5. Specific conditions for activation are given in the figure legends. Reaction products from prothrombin activation mixtures were separated on columns of QAE-(quaternaryaminoethyl) Sephadex Q-50 (5) and Bio-Gel P-100. Disc gel electrophoresis was performed by the technique of Ornstein (16) and Davis (17). Sodium dodecyl sulfate gel electrophoresis was carried out by the procedure of Laemmli (18) except that the separating gels contained 0.27 To N , N'-methylenebisacrylamide.
Urea disc gel electrophoresis was performed by the addition of 0.48 g of urea to each milliliter of separating or stacking gel solution prior to polymerization, No sample gel was used. Samples were prepared by heating aliquots of the reaction mixture at 100" for 1 to 2 min in 6 to 8 M urea. The sample was layered directly on top of the stacking gel. The gels were rapidly fixed and stained at 70". Soybean trypsin inhibitor (type II-S, Sigma Chemical Co., St. Louis, MO.) was linked to Sepharose 413 with cyanogen bromide by the method of Cuatrecasas (19). Amino acid analyses were performed on protein samples after hydrolysis in 6 N HCl in sealed evacuated ampules for 24 and 72 hours at 110". Single column analyses employed Durrum DC-l resin (Durrum Chemical Co., Palo Alto, Calif.) and a Beckman 120C analyzer. Diisopropylfluorophosphate (Sigma Chemical Co.) was used as a 1 M solution in anhydrous 2-propanol.
NHn-terminal amino acids were determined by the procedure of Gray (20). The dansylamino acids were identified by chromatography on polyamide sheets (21).  [X,,V,phospholipid,and Ca2+] in the presence and absence of 10 rnM iPr2PF.
Prothrombin was incubated with [X,, V, phospholipid, Ca*+] for 12 min, an interval known from previous experiments to be sufficient to convert all the starting prothrombin to thrombin in the absence of iPr2PF (5). After incubation, the reaction mixtures were chromatographed on separate analytical QAE-Sephadex columns.
Each column had 2 cm of soybean trypsin inhibitor-Sepharose over the QAE-Sephadex gel bed to remove X, from the reaction mixtures.
Otherwise, in the process of chromatography, small amounts of thrombin were formed from remaining prothrombin or reaction intermediates ( Fig. 1). As iPrpPF also inhibits X, (22), although much less effectively than it inhibits thrombin (23), the X, activity is decreasing continually during the activation reaction.
As a consequence, activation in the presence of iPrJ'F results in small amounts of prothrombin and Intermediate 1 remaining after 12 min. In contrast to activation without iPr2PF, these components are detected both in the chromatogram and in a sodium dodecyl sulfate electrophoresis gel of the reaction mixture that contained iPr,PF.
The sodium dodecyl sulfate electrophoresis gels shown above the peaks of the two chromatograms differ only in the appearance of an additional band in the Fragment 2 peak. This new product has an apparent molecular weight by sodium dodecyl sulfate  for 12 min at 23". iPr,PF was added to the effluent from the soybean trypsin inhibitor column to give a concentration of 1 mM. After 10 min, this solution was applied to a column (1.5 X 25 cm) of QAE-Sephadex Q-50 equilibrated with 0.02 M Tris-HCl, pH 7.5, 0.1 M NaCl. The column was eluted with a linear gradient (150 ml per gradient chamber), 0.1 M NaCl to 0.6 M NaCl in 0.02 M Tris-HCl, pH 7.5, at 40 ml per hour. Four-milliliter fractions were collected.
O-O, absorbance (ABS), 280 nm. electrophoresis equal to or slightly less than thrombin (37,000), consistent with the value predicted from the sum of the molecular weights of Fragment 1 and Fragment 2 (36,000) (5). An aliquot from the peak which contained the presumed Fragment 1.2 (Fraction 70, Fig. 1B) was incubated with thrombin and the products were examined by sodium dodecyl sulfate gel electrophoresis. Two bands were seen which underwent co-electrophoresis with Fragment 1 and Fragment 2 (data not shown, see results with purified Fragment 1.2 below).
This result is consistent with the formation of Fragment 1 a2 in a reaction mixture in which thrombin is rapidly inhibited by iPrzPF.
A careful comparison of the elution profiles from the QAE-Sephadex columns indicates two other important differences. First, the ratio of the areas of the Fragment 2 to Fragment 1 peaks is 0.7 in the absence of iPrzPF and 1.0 in the presence of iPrtPF, consistent with an increased amount of a product which behaves chromatographically like Fragment 2. As some unreacted prothrombin is present with the Fragment 1, this ratio is actually greater than 1. Second, resolution of the two Fragment 2 peaks is reduced.
The well resolved Fragment 2 doublet is a reproducible feature of complete activation mixtures (5). The difference in the thrombin peaks between the two chromatograms ( Fig. 1) does not reflect real differences in the thrombin, since chromatography of mixtures at lower ionic strength results in elution of only a single thrombin peak (5). Also, the initial ionic strength of the reaction mixture is higher than the control reaction because of the additional Tris buffer added to compensate for the 10 mM iPrzPF hydrolysis.products and probably is responsible for the greater amount of "breakthrough" thrombin seen in Fig. 1B.
Characterization of Fragment 1 .%--Fragment 1.2 could be prepared on a large scale and in high yield only when the prothrombin concentration was reduced in the activation mixture and a separate soybean trypsin inhibitor-Sepharose column was used to remove the X, completely from the reaction mixture.
Activation conditions are given in the legend of Fig. 2. After activation, reaction products were passed over a soybean trypsin inhibitor-Sepharose column (1.5 x 10 cm) equilibrated with 0.02 M Tris-HCl, pH 7.5, 0.1 M NaCl.
This effluent then was applied to a QAE-Sephadex column.
The elution profile from this preparative scale experiment, showing only the activation fragment region of the chromatogram, is shown in Fig. 2. The sodium dodecyl sulfate electrophoresis gels show the expected Fragment 1.2 and also indicate that the usual Fragment 1 peak contains primarily unreacted prothrombin. Fragment 1.2 was separated from Fragment 2 by gel filtration on Bio-Gel P-100 (Fig. 3). Pooled Fragment 1.2, as shown in Fig. 3, was examined by sodium dodecyl sulfate gel electrophoresis and electrophoresis at pH 9.5 (Fig. 3, inset). Only a single component is seen in either of the electrophoresis systems.
A sample of the isolated Fragment 1.2 (40 pg/200 ~1) was incubated with thrombin (4 pg) for 40 min at 23". Fig. 4A shows sodium dodecyl sulfate electrophoresis gels of the starting Fragment 1.2 and the products formed by thrombin. shows the same materials as seen by pH 9.5 disc electrophoresis. In both electrophoresis systems the reaction products underwent co-electrophoresis with isolated Fragment 1 and Fragment 2 from ordinary activation mixtures.
A trace of remaining Fragment 1.2 which was not detected by electrophoresis at pH 9.5 can be seen in the sodium dodecyl SO1 electrophoresis gel. Fragment 1.2 (1.35 mg) from gel filtration (Fig. 3) was incubated with 30 pg of thrombin as described above, except that Caz+ was omitted.
After 1% hours, the reaction mixture was subjected to analytical ion exchange chromatography (5) on QAE-Sephadex Q-56 (Fig. 5). Four peaks were eluted: the added thrombin (Fraction 14), Fragment 1 (Fraction 56), and the characteristic two peaks of Fragment 2 (Fractions 64 and 69). The ratio of the areas of the Fragment 2 to Fragment 1 peaks was 0.7, identical with the ratio for these two peaks in Fig. 1A. Furthermore, the elution positions for the Fragment 1 and the Fragment 2 derived from Fragment 1.2 are identical with these of the fragments derived from prothrombin on activation with [X,,V,phospholipid,Ca2+] or X, alone (Fig. 1A).
- Pooled fractions 70 to 78, from QAE-Sephadex chromatography (Fig. 2) were concentrated by ultrafiltration with an Amicon Diaflo concentrator (UM-10 membrane) and a sample, 4.7 mg in 1.1 ml was applied to a column (1.5 X 90 cm) of Bio-Gel P-100 which was equilibrated in 0.02 M Tris-HCl, pH 7.5, 0.5 M NaCl. The column was eluted at room temperature in the upward direction at 15 ml per hour. One-milliliter fractions were collected.
The Fragment 1.2 was pooled as marked by the bar and then examined by acrylamide gel electrophoresis at pH 9.5 (Gel A) and by sodium dodecyl sulfate gel electrophoresis Sodium dodecyl sulfate electrophoresis gels from t'his experiment are shown above t'he column peaks (Fig. 5). Amino acid composit,ions of Fragment 1.2 from the Uio-Gel P-100 column (Fig. 3) and Fragment 1 and Fragment 2 from t'he QAE-Sephadex column (Fig. 5) are given in Table I. The composit,ions of isolated Fragments 1 and 2 from prot'hrombin activated with [X,, V, phospholipid, Ca*+] (5) are compared with Fragments 1 and 2 derived from cleavage of Fragment 1.2 (Table II).
No significant differences were found bet'ween Fragment 1 derived from Fragment 1.2 (Table I) and Fragment 1 described previously (Table II).
Likewise, Fragment 2 derived from Fragment 1.2 is identical with Fragment 2 from prothrombin activation mixtures (Table II). Factor x', Alone in Presence of Diisopropyljuorophosphate-Fragment 1.2 was successfully isolated from prothrombin activation mixtures in which X, alone was the activator only if both the X, and thrombin were removed from the mixture prior to chromatography on QAE-Sephadex.
Prothrombin in 10 mM iPr$F was incubated with X,.
Approximately one-half the starting prothrombin was still present when the act,ivation was terminat,ed as assessed by sodium dodecyl sulfate gel electrophoresis.
The partial activation mixture was passed through a column (0.9 X 17 cm) which consisted of a 7-cm bed of sulfopropyl Sephadex W-50 layered over a lo-cm bed of soybean trypsin inhibitor-Sepharose.
Under these conditions, X, was adsorbed t,o the soybean trypsin inhibitor-Sepharose column, and the thrombin was adsorbed to the sulfopropyl Sephades (24). The effluent from t'his column was brought to 1 mM in iPrZF and then chromatographed on QAE-Sephadex, as described in Fig. 1. Two features of t'his chromatogram (Fig. 6) are particularly significant.
First, neither Fragment 1 nor Fragment 2 is seen in the sodium dodecyl sulfate electrophoresis gel of the final reaction mixture, nor in the column peaks in which they are usually found.
Second, a comparison of the mobilit'y of Fragment 1.2 in sodium dodecyl sulfate electrophoresis gels (gels above Fractions 66 to 78) with the mobility of the thrombin (gels  6. Isolation of Fragment 1.2 (F-l .2) from the products of prothrombin (P) activation by Factor X, alone in the presence of iPrnPF. Prothrombin, 6.3 mg in 4 ml of 0.03 M Tris-HCl, pH 7.5, 0.1 M NaCl, 10 mM iPrzPF was incubated with X, (70 pg) for 135 min at 23". The reaction mixture was passed through a column, 0.9 X 17 cm, equilibrated in 0.02 M Tris-HCl, pH 7.5, 0.1 M NaCl, which consisted of 7 cm of sulfopropyl Sephadex layered over 10 cm of soybean trypsin inhibitor-Sepharose.
After passage through the column, the pooled column eluate (10 ml, 0.752 Az80 per ml) was made 1 mM in iPrzPF.
above Fractions 15 to 30) suggests that in more heavily loaded gels, Fragment 1.2 may not be distinguished from thrombin or Intermediate 2 (5).
Failure of Factor X, to Form Fragments 1 and 2 from Isolated Fragment 1 .%--Fragment 1.2 (30 pg in 300 ~1) was incubated with [X,, 1 pg; V, 2 I.rg; phospholipid, 1 pg; CaC12, 10 mM] for 12 min at 23". The activation mixture was analyzed by sodium dodecyl sulfate electrophoresis. Neither Fragment 1 nor Fragment 2 was detected in the fixed, stained gel. An equimolar sample of prothrombin would have been activated completely in less than 2 min under these conditions (5).
In a second attempt to cleave Fragment 1.2 with X,, 30 pg of Fragment 1.2 in 300 ~1 was incubated with 1.5 pg of X, for 40 min at 23". No cleavage of Fragment 1.2 was detected. Since Factor X, did not cleave Fragment 1.2 either alone or in conjunction with the other activation components, and since activation can be carried out without the formation of appreciable Fragment 1 (Figs. 2,6) it appears that Fragment 1 is released from prothrombin only as a result of proteolysis by thrombin.
Nllz-term&a2 Amino Acid Analysis of Fragment 1 .2-Dansylalanine was the only residue present in sufficient quantity on the thin layer chromatogram to represent an NHAerminal amino acid. This observation, coupled with the data of Magnusson (15) which indicate that thrombin (NHz-terminal threonine and isoleucine) comes from the COOH-terminal end of prothrombin and the previously determined alanine NHz-terminus of prothrombin (25-27) (confirmed in this laboratory), indicates that Fragment 1.2 arises from the NHAerminal end of the prothrombin polypeptide chain.

I.2 Formation during Prothrombin
Activation-In view of the previous observation that Intermediate 2 and Fragment 1.2 undergo co-electrophoresis in sodium dodecyl sulfate gels, unambiguous demonstration of Fragment 1.2 during activation could not be obtained by this technique. Disc electrophoresis in urea, however, resolves Fragment 1.2 and Intermediate 2. A time course of prothrombin activation by [X,, V, phospholipid, Ca*+] in the absence of iPrzPF is shown in Fig. 7. The identity of each of the bands in the urea gels was determined by co-electrophoresis with the isolated intermediates and fragments. The bands from the top to the bottom of the gels are: 1, thrombin, which aggregates under these electrophoresis conditions (28) ; 2, Intermediates 1 and 2, which are not resolved in this system and are characteristically multiplet patterns similar to the pattern described for thrombin (28); 3, prothrombin; 4, Fragment 1.2; 5, Fragment 1 (diffuse and broad); and 6, Fragment 2 (very sharp). It is seen from the time course that Fragment 1.2 is formed during prothrombin activation in the absence of iPrzPF. However, as a consequence of proteolysis by thrombin, only Fragments 1 and 2 are found in the final reaction mixture. Fig. 8 demonstrates Fragment 1 a2 formation and subsequent degradation with Factor X, alone as the activator. The primary difference in the two systems is in the time required for activation, i.e. 8 min for complete activation with [X,, V, phospholipid, Caz+] versus 180 min for 10% activation with X, alone (determined by assay).

DISCUSSION
Prothrombin activation in the presence of iPrzPF results in the formation of a stable activation fragment (Fragment 1.2) which consists of the covalently linked Fragments 1 and 2 described previously (5). This fragment is readily cleaved by thrombin to form Fragments 1 and 2. However, it was not cleaved by X,, either alone or in the presence of V, phospholipid, and Ca*+. In view of these results, formation of the "activation intermediate" designated Intermediate 1, which can be observed during prothrombin activation, must occur primarily and probably exclusively as a result of proteolysis by thrombin and not X,. Demonstration of an NHt-terminal alanine residue in Fragment 1.2, the same NH2 terminus as prothrombin, supports a previous proposal (15) that thrombin (NHz-terminal threonine and isoleucine) is derived from the carboxyl end of the prothrombin polypeptide chain. Since Fragment 1 is removed from prothrombin to give Intermediate 1, a single polypeptide chain (5, 12), Fragment 1 must be derived from the NH*-terminal end of the prothrombin polypeptide chain.  On the basis of these observations and the results presented previously (5), Scheme 1 is proposed to describe the process of prothrombin activation.
The symbol, 0, in the schematic map, represents the position in prothrombin at which peptide bonds are cleaved during prothrombin activation.
The lengths of the line segments, which represent the activation products, are proportional to the number of amino acid residues in each ac: tivation product (5). An activation pathway is defined by the order of bond cleavages.
The existence of pathway A is supported by the following evidence: (a) the appearance and disappearance of Intermediate 1 during prothrombin activation, Fig. 3 of the preceding paper (5) ; (b) the peptide bond linking Fragment 1 with Fragment 2 is not cleaved by Factor X,; (c) isolation of Intermediate 1 and Fragment 1 after incubation of prothrombin with thrombin; (d) isolation of Intermediate 2 and the demonstration that it can be converted to thrombin (5) ; and (e) the observed appearance and disappearance of Intermediate 2 during activation of isolated Intermediate 1 (Fig. 11  obtained, the question of whether pathway U or C (or both) is