Rabbit Plasma Inhibitor of the Activated Species of Blood Coagulation Factor X PURIFICATION AND SOME PROPERTIES

A naturally occurring inhibitor of activated blood coagulation Factor X has been isolated from pooled rabbit plasma and purified by the combination of Sephadex G-ZOO gel filtration, DEAE-Sephadex A-SO, and DEAE-cellulose chromatography. On microzone electrophoresis the purified product traveled as an az-globulin, and on 7.5% polyacrylamide gel disc electrophoresis at pH 9.5 it traveled as a single component close to transferrin. With exclusion chromatography on Sephadex G-75, G-100, and G-200 it emerged in the same elution volume as crystalline bovine serum albumin. On analytical disc electrophoresis, the inhibitor stained positive for glycoprotein with the periodic acid-Schiff technique. It contained 4.1% hexose, 4.6% sialic acid, and was soluble in 2.5% but not in 5% trichloroacetic acid. Greater than 80% of its original activity persisted at 56” in the 1st hour and gradually diminished to 20% by the 6th hour. The pH optimum of the inhibitor activity was between 7 and 9, and the activity was most stable at pH 6 to 8. Optimum inhibitor activity was at 37” and nondetectable at 1”. Preparative disc electrophoresis of the purified inhibitor on 15% polyacrylamide gel at pH 8.1 caused extensive aggregation of the eluted inhibitor protein with a concomitant total loss of biological activity within 48 hours of storage, even at -60”.

From the Department of Medicine, The Jewish Hospital of St. Louis, and Washington University &hod (!f Medicine, St. Louis, Missou,ri 63110 SUMMARY A naturally occurring inhibitor of activated blood coagulation Factor X has been isolated from pooled rabbit plasma and purified by the combination of Sephadex G-ZOO gel filtration, DEAE-Sephadex A-SO, and DEAE-cellulose chromatography.
On microzone electrophoresis the purified product traveled as an az-globulin, and on 7.5% polyacrylamide gel disc electrophoresis at pH 9.5 it traveled as a single component close to transferrin.
With exclusion chromatography on Sephadex G-75, G-100, and G-200 it emerged in the same elution volume as crystalline bovine serum albumin. On analytical disc electrophoresis, the inhibitor stained positive for glycoprotein with the periodic acid-Schiff technique. It contained 4.1% hexose, 4.6% sialic acid, and was soluble in 2.5% but not in 5% trichloroacetic acid. Greater than 80% of its original activity persisted at 56" in the 1st hour and gradually diminished to 20% by the 6th hour. The pH optimum of the inhibitor activity was between 7 and 9, and the activity was most stable at pH 6 to 8. Optimum inhibitor activity was at 37" and nondetectable at 1". Preparative disc electrophoresis of the purified inhibitor on 15% polyacrylamide gel at pH 8.1 caused extensive aggregation of the eluted inhibitor protein with a concomitant total loss of biological activity within 48 hours of storage, even at -60".
During the coagulation of mammalian blood in vitro some of the clotting factors essential for the formation of fibrin become activated.
Two of these activated clotting factors, activated Factor X1 (autoprothrombin C) and thrombin, are absent from serum, while others, such as Factors XII and XI, persist in the serum in their activated form. Ass., 170, 325 (1959)).
Fibrinogen is the natural substrate for thrombin, and their reaction produ&, fibrin, is a potent antithrombin (1,2). Bctween 85 and 90% of the thrombin formed during blood coagulation is immediately adsorbed by fibrin (3), whereas the remaining thrombin activity is more slowly neutralized by a substance in blood known as antithrombin III (4). Moreover, this plasma substance can also neutralize a large quantity of preformed thrombin in a slow progressive manner in the absence of fibrin (5). Activated Factor X (the enzyme responsible for the activation of prothrombin to thrombin (6)) does not clot fibrinogen, has not been shown to be adsorbed by fibrin, and does not lose its activity in the presence of thrombin.
In fact, abundant activated Factor X is found in crude preparations of thrombin, and these two activities have been separated from each other (7,8). Therefore, the rapid disappearance of activated Factor X, within a few minutes after plasma clots, may be attributed to the presence in blood of a naturally occurring blood clotting antagonist.
The disappearance of activated Factor X upon its addition to bovine serum has been previously reported by ot'hers (9). This observation has been confirmed in our laboratory and the plasma fraction responsible for the action (activated Factor X inhibitor) has been isolated and partially purified from human blood (10).
In the present communication we describe a technique for the large scale isolation and purification from rabbit plasma of t,he inhibitor to activated Factor X together with data concerning some of its physicochemical properties.

Methods
A unit of activated Factor X is defined as the activity that would evolve from 1 ml of normal human plasma, when Factor X is fully activated by an optimum amount of Russell's viper venom.
Accordingly, it was established that 1 unit of activated Faclor X gave the same clotting time (in the absence of the venom) as 1 unit of plasma Factor X, when the latter was measured in the system containing venom.
In the present study act'ivated Factor X was determined by means of a system identical with the Bachmann assay for plasma Factor X (11) with the same cephalin concentration except that the Russell's viper venom was omitted.
In constructing the calibration curve for activated Factor X, however, the Bachmann Factor X assay method using the venom was employed.
Normal human plasma served as a control.
The 280-rnp absorbing material was monitored either with a Beckman DB-G spectrophotometer or a LKB Uvicord II unit with a l-mm flow cell manufactured by LKB-Produkter AB, Bromma, Sweden, Protein was determined by the Lowry method (13) using bovine serum albumin as a standard.
Sialic acid was measured by the method of Warren (14), and the total hexose (in terms of galactose or mannose) was determined as described by Spiro (15). The Srphadex gels, 100 to 200 mesh size, were permitted to swell in the appropriate buffers for at least 1 week at either room temperature or at 4" before packing of the columns. with the Canalco unit and all reagents used were from the same supplier.
The following modified formulae for the preparation of the various reagents were kindly supplied by them.
The nomenclature designated for each solution corresponds to that used in their operation manual for preparative disc electrophoresis.
Reagent A (pH 7.6) was comprised of 11.7 g of imidazole, 48 ml of 1 N HCl, and 0.24 ml of N, N, N', N'tetramethylethylenediamine.
The mixture was dissolved and the volume brought to 100 ml with deionized water. Reagent B (pH 5.9) was comprised of 2.93 g of imidazole, 0.46 ml of N, N, N', N'-tetran~etl~ylethylenediamine, 48 ml of 1 N HCl.
The mixture was dissolved and the volume brought to 100 ml with deionized water. Reagent CN was comprised of 20 g of Prep/Cry1 (preparative grade acrylamide), 0.06 g of N ,N'methylenebisacrylamide dissolved and brought to 50 ml with deionized water.
Reagent DN was comprised of 7 g of Prep/ Cryl, 0.125 g of N, N'-methylenebisacrylamide, dissolved and brought to 50 ml with deionized water.
Reagent E was comprised of 0.004 g of riboflavin dissolved in 100 ml of deionized water.
Imidazole buffer 10 X (pH 7.6), was comprised of 2 g of imidazole, 8.5 g of glycine dissolved and brought to 1000 ml with deionized water.
The addition of /I-mercaptoethanol is optional. Elution buffer was I:8 dilution of Reagent A in deionized water. The preparative electrophoresis was carried out at 2" using a PD2/320 upper column with a 4-cm long 15'% spacer gel and a 2-cm 15yo separating gel. The protein samples for electrophoresis were loaded in 40% sucrose.
Measurement of Activated Factor X Inhibitor Activity-Unless otherwise specified, t,he activity of activated Factor X inhibitor was determined as follows.
A test fraction of 0.2 ml (appropriate dilution) and 0.2 ml of buffer (0.1 M NaCl in 0.01 M Tris-HCl, pH 8) were warmed in a 37" water bath for 30 s. To this mixture was added 0. NaCl-0.01 M Tris-HCl, pH 8, before assaying. The concentration of activated Factor X used for the detection of the inhibitor activity in the chromatographic and preparative electrophoretic fractions was approximately 3 units per ml of bovine serum albumin solution (10 g/100 ml of 0.14 M NaCl).
Puri$cation of Inhibitor Step I: Preparation of Plasma Concentrate One-liter batches of freshly frozen rabbit plasma in polyethylene bottles were allowed to thaw overnight at 4". The plasma was clarified by centrifugation at 2000 x g for 15 min at 4". It was then treated with barium sulfate (100 mg of BaS04 per ml of plasma) at room temperat,ure with continuous mechanical stirring for 30 min.
The BaS04 was removed by centrifugation at 2000 x g for 10 min at 4', the plasma decanted, and the BaS04 treatment repeated once more. The plasma was then rapidly passed through an asbestos filter pad in a Seitz filtration unit under NQ at room temperature and then lyophilized.
Step II: Gel Filtration on Sephadex G-200 Chromatographic Procedure-The plasma concentrate equivalent to 1 liter of the original material was reconstituted to 250 ml with deionized water and dialyzed against two changes of 4 liters of 0.04 M NaCl at 4" over a period of 24 hours.
The dialyzed plasma concentrate was clarified by centrifugation at 20,000 x g for 45 min at 4". Usually a small fatty layer floated on top of the centrifuged plasma. This layer was discarded by carefully decanting the plasma and filtering it through Whatman No. 1 filter paper.
The filtered plasma, containing about 60 g of protein, was brought up to 400 ml with 0.145 M NaCl and after having warmed to room temperature it was then placed on a column (10 x 100 cm) of Sephadex (Pharmacia K lOO/lOO) previously equilibrated with the eluting solution of 0.145 M NaCl at room temperature (26"). The flow rate was controlled by a hydrostatic head at 150 ml per hour and fractions of 25 ml each were collected.
Concentration of Crude Inhibitor-The inhibitor fractions were pooled, then dialyzed against three changes of 4 liters of 0.04 M NaCl at 4" for 24 hours.
The dialyzed fraction was then concentrated by lyophilization.
Step III: Chromatography on DEAE-Sephadex A-50 Preparation of Column-The DEAE-Sephadex A-50 was allowed to swell for 48 hours at room temperature in 0.05 M NaCl-0.1 M Tris-HCl, pH 8.32, and then was packed in a Lucite column with a polyethylene disc base. A layer of preswollen Sephadex G-25 about 1 cm thick was first poured into the column, allowed to settle, and then followed by the DEAE-Sephadex.
Another l-cm layer of Sephadex G-25 was stacked on top of the DEAE-Sephadex bed.
The Sephadex G-25 at the bottom of the bed facilitated a good flow rate and the top layer prevented disturbance of the DEAE-Sephadex bed when the protein fraction was being applied.
Fresh DEAE-Sephadex was used for each chromatographic procedure in the present study. Chromatographic Procedure-For each DEAE-Sephadex chromatograph, lyophilized material from two separate gel filtration runs as in Step II were pooled and reconstituted to 80 ml with 0.05 M NaCl in 0.1 M Tris-HCl, pH 8.32. This material was then dialyzed against three changes of 4 liters of the same buffer at 4' for 48 hours.
The entire fraction, containing approximately 4 to 6 g of total protein in less than 100 ml, was chromatographcd on a DEAE-Sephadex A-50 column (4 x 50 cm) previously equilibrated with 0.05 M NaCl in 0.1 M Tris-HCI, pH 8.32, at room temperature.
The flow rate was 60 ml per hour under hydrostatic pressure and fractions of 12 ml were collected. Chromatography was performed by stepwise increase in ionic strengt,h of the buffer.
Concentration of Partially Purified Inhibitor-The fractions with inhibitor activity were pooled, dialyzed against six changes of 4 liters of deionized water during a 3-hour period at room temperature, and lyophilized.
Step IV: Chromatography on DEAE-cellulose Preparation of Column-The dry cellulose was sieved to obtain the 60 to 100 mesh for packing the column.
It was pretreated as described (8) and then equilibrated with the initial buffer to be used for the chromatographic procedure.
Packing of the cellulose in a glass column was performed at room temperature.
Chromatographic Procedure-The entire lyophilized, partially purified inhibitor fraction from Step III was dissolved in less than 5 ml of deionized water and dialyzed for 24 hours against two changes of 4 liters of the buffer used for conditioning t'he DEAEcellulose.
The dialyzed fraction was centrifuged at 35,000 x g at 4" for 1 hour.
For each chromatograph 2 to 3 ml of the fraction, containing 35 to 70 mg of protein, were applied on a DEAE-cellulose column, 1 x 35 cm, previously equilibrated at 4" with 0.02 M NaCl in 0.1 M Tris-HCI, pH 8.32. A linear salt gradient elution technique was adopted with the reservoir con taining 150 ml of 0.10 M NaCl in 0.1 M Tris-HCl, pH 8.32, ant1 the mixing chamber containing 150 ml of 0.02 M NaCl in 0.1 11 Tris-HCl, pH 8.32. The flow rate was maintained at 7 to 10 ml per hour with a hydrostatic head, and fractions of 3 ml each were collected.
Concentration of Purified Inhibitor-The inhibitor fractions were pooled and concentrated at 4" by dialysis against 40% polyvinylpyrrolidone, pH 7.0.
Step V: Gel Filtration on Sephadex G-900 Chromatographic Procedure-Of the concentrated fraction (3 ml) were chromat.ographed on a Sephadex G-200 (100 to 200 mesh size) column, 2.5 x 50 cm, at 4" previously equilibrated with the eluting buffer, 0.10 M NaCl in 0.05 M Tris-HCI, pH 8.32. The elution flow rate was controled by a hydrostatic head at 0 ml per hour, and fractions of 3 ml each were collected. Two peaks of protein preceded the elution of the inhibitor.

In
The first protein peak was found to be a y-globulin and the second protein peak was mainly P-globulin as examined by microzone electrophoresis.
No inhibitor activity was detected in either protein peak. A fourth protein peak (not shown in this figure), representing greater than 75% of the total protein chromatographed, could be eluted after the inhibitor peak with 0.2 M NaCl in 0.1 M Tris-HCI, pH 8.32. This fourth peak contained predominantly albumin without inhibitor activity (10). When the inhibitor fractions were pooled and examined by microzone electrophoresis, only a single protein band migrating as an az-globulin was observed. When analyzed at a protein concentration of 200 pg per load on polyacrylamide gel disc electrophoresis, the inhibitor fraction showed a major component which migrated close to transferrin. Three to four additional slow migrating minor components could also be detected on the same gel (Fig. 7B).
These bands, however, were diffuse and stained poorly.
After lyophilization, another 10 t,o 15% loss of inhibitor activity was encountered.

DEAE-cellulose
Chromatography-As shown in Fig. 3A, when 35 mg of the inhibitor from Step III were chromatographed, the activity peaked at Fraction 32. When 35 mg of another preparation from Step III were similarly chromatographed on the same column, the activity peaked at Fraction 26, Fig. 3B, where a definite shouldering of the descending slope could be seen. If 70 mg of the same preparation were then chromatographed on the same column, Fig. 3C, t.hree distinct activity peaks were obtained.
The activity peaked at Fractions 18, 24, and 30. That t'he difference in the eluting positions of the inhibitor depicted in Fig. 3 was not an artifact was supported by the observation that in all three chromatograms, the first protein peak, without in-  Fig. 1  Fractions 32 and 34 were in a position between (pi-and ar&obulin.
Control experiments with crystalline bovine serum albumin alone, and bovine serum albumin mixed with the individual fractions, were also performed to rule out possible artifacts. When the inhibitor fractions were pooled and then subjected to electrophoresis, the resulting single band migrated as an 012globulin.
On analytical disc electrophoresis those fractions with activity displayed a single protein band when examined at 100 pg of protein per load.
When tested at a protein concentration of 200 pg, one to two additional minor slow migrating components were detected which were diffuse and stained poorly, Fig.  7C. On a few occasions, the fractions displayed a single component even at protein concentration of 300 pg when tested immediately upon elution from the column. This protein band stained positive for glycoprotein with the periodic acid-Schiff reagent'.
On storage at -20" for only 24 hours these electrophoretically homogeneous fractions now showed one to two additional minor components. This was usually accompanied by some loss of the original activity.
Attempts to rechromatograph these fractions on a DEAE-cellulose column resulted in much loss of the activity with fractions exhibiting an increasing number of multiple minor components that also stained positive for glycoprotein on disc electrophoresis.
The removal of ammonium persulfate from the polymerized gel produced no difference between the electrophoretic patterns of the inhibitor in the presence or absence of the catalyst which is a strong oxidizing agent.
Chromatography of the inhibitor on DEAE-cellulose columns two to three times the size used in the experiments in Fig. 3 did not improve the elution profile.
In fact, recovery of the inhibitor on these large columns was very poor as compared with recovery on a smaller column.
The disc elect'rophoresis pattern of the fractions from the larger columns exhibited multiple minor components, at times more than the starting material that was chromatographed.
The purified inhibitor withstood further concentration by dialysis against polyvinylpyrrolidone at pH 7 better than by either lyophilization or ultrafiltration. Xephades Gel Filtrations-Filtrations of the inhibitor fraction after DEAE-cellulose chromatography on columns of Sephatlcr G-75, G-100, and G-200 gave elution profiles similar to t'hat dc-Dieted in Fig. 4. The distribution coefficient of the inhibitor was the same as that of crystalline bovine serum albumin in a11 instances.
Passage of the purified inhibitor through Sephadex G-100 or G-200 usually resulted in eliminating the other minor slow migrating components, Fig. 70, that were previously present in Fig. 7C. In general, a lowering of the specific activity of the inhibitor was observed. Comparisons between Effect of Ultra$ltration and Polyvinylpyr-rolidont~ and Dialysis on Purified Inhibitor-Although on nnalytical disc electrophoresis the inhibitor obtained in Step V frequently showed a single protein band at a protein concentration of less than 200 pg in 100 ~1 per load, it was decided to examine the inhibitor fraction at a much higher protein concentration, without increasing the volume, to detect trace contaminants.
To do this, the inhibitor fractions had to be concentrated. Concentration by ultrafiltration through a Dia-Flo membrane was first selected because it is a simple, fast, and widely accepted method for concentrating dilute protein solutions. Approximately 40 to 50 ml of a 0.1% solution of the inhibitor from Step V in 0.10 M NaCl-0.05 M Tris-HCI, pH 8.32, was clarified by centrifugation at 30,000 x g for 60 min at 4" and then concentrated in an Amicon ultrafiltration cell, either model 52 or 202, at 2". The appropriate size Dia-Flo ultrafiltration membrane P&l-10 was used per the accompanying instructions.
Mechanical stirring of the solution filtration under Nz (less than 10 p.s.i.) was maintained at the fastest revolution rate that did not create a vortex and thus cause the protein solution to foam.
The fraction was usually reduced to 10% of its original volume.
Approximately half way through the filtration, the previously clear solution in the cell became faintly cloudy and progressively more dense as the volume was reduced. When the desired concentration was achieved, the inhibitor solution inside the cell was observed to contain fine precipitates that could be removed by centrifugation at 3000 x g for 15 min.
The clear supernatant was tested for inhibitor activity and found to contain less than 60% of the total activity initially placed in the cell. Examination of the filtrat'e indicated total absence of inhibitor activity. Therefore, the decrease of activity in the cell retentate was not caused by loss of the inhibitor in its passage through the membrane. To eliminate the possibility that the loss of activity and the appearance of fine precipitates in the purified inhibitor fraction during ultrafiltration were not caused by excessive stirring during the operation, a different batch of inhibitor was subjected to ultrafiltration without stirring.
The flow rate progressively slowed and a higher pressure was required to maintain the previous flow rate. At the end of the filtration the retentate appeared relatively clear until the stirrer was turned on slightly for 1 min, after which the fraction appeared cloudy.
Following centrifugation of the concentrate the clear supernatant was tested for inhibitor activity and was found to contain less than 7Oc/, of the original activity.
On the other hand, when the purified inhibitor was concentrated by dialysis against 40% polyvinylpyrrolidone, at 4', a process requiring 8 to 12 hours, the inhibitor retained greater than 90% of its original activity without any visible precipitation. When both inhibitor concentrates prepared by ultrafiltration and by polyvinylpyrrolidone dialysis were initially examined at 500 pg of protein per load on analytical disc electrophoresis, multiple protein bands were found in fractions concentrated by both methods; and, in addition, a striking difference in the electrophoretic pattern between these two fractions was evident. These two samples were then diluted to their preconcentration protein levels and re-examined on disc electrophoresis. The fraction concentrated by ultrafiltration (Fig. 7E) demonstrated much smearing with mult,iple, heavily stained protein bands. On microzone electrophoresis two protein bands corresponding to LYE-and a2-globulin were seen. On the other hand, the concentrate prepared by polyvinylpyrrolidone dialysis (Fig. 7F) revealed one major component with two to three additional fine bands close to it. No extremely slow migrating components similar to those in Fig. 7E were seen. On microzone electrophoresis the polyvinylpyrrolidone-concentrated inhibitor revealed only one protein band.
The preconcentration electrophoretic lbatterns of these two fractions were identical with that of Fig. 70.
Preparative Disc Electrophoresis-The purified inhibitor obtained in Step V was concentrated by dialysis against polyvinylpyrrolidone for preparative disc clectrophoresis. Because the inhibitor was most stable between pH 6 and 8 (see below), it was necessary to perform the preparative electrophoresis as close to these values as possible without affecting either the activity of the inhibitor or the resolving power of the technique. Since the inhibitor was relatively heat stable even at 56" (see below), electrophoresis at 2" was believed to pose no hazard to the protein.
The imidazole-HCl system with a running pH of 8.1 was the most reasonable one encountered.
Figs. 5 and 6 show the behavior of two different preparations of concentrated purified inhibitor on polyacrylamide gel preparative disc electrophoresis at 2". In Fig. 5, the distribution of the multiple activity peaks was narrower than that shown in Fig. 6. It is, however, significant to note that in the latter figure, three out of the four minor protein peaks still had some detectable inhibitory activity against activated Factor X. The inhibitor activity eluted from the main peak was found to be very unstable.
The specific activity across the peak was fairly constant.
At 4" it retained less than 50% of its original activity after 12 hours. It lost greater than 9076 of its activity after 24 hours especially at -25" and -60". Fig. 7 shows the various disc electrophoretic patterns of the inhibitor fraction from the main peak of Fig. 6 before and after storage at -60".
The sample, prior to freezing, appeared as a single, broad and diffuse band, Fig. 7H.
Forty-eight hours after storage at -6O", t,he fraction now showed multiple, heavily stained components, Fig. 71. At this stage, it had completely lost its biological activity to inhibit activated Factor X. Inhibitor fractions similarly obtained were also labile on storage at -2O", but appeared slightly more stable at 4". Some of the more concentrated minor components in the disc gel stained lightly with periodic acid-Schiff.
Control studies of the stability of the inhibitor from Step V incubated in the same imidazole buffer at 4" and -60" ruled out the loss of activity of the preparative disc electrophoresis fraction as caused by the particular buffer employed.
The addition of 0.14 mM P-mercaptoethanol to the preparative disc electrophoresis system did not improve the stability of the inhibitor, nor did the prior removal of the ammonium persulfate from the separating gel by electrophoresis (19). The underlying causes leading to the extensive loss of the inhibitor activity after electrophoresis on polyacrylamide gel remain to be determined.
The specific activity and the recovery of an average preparation of activated Factor X inhibitor at various purification steps are listed in Table I to electrophoresis on 15% polyacrylamide gel, pH 8.1, at 4". Five milliliters of the inhibitor sample containing approximately 30 mg of protein were subjected to electrophoresis. Electrophoresis was conducted with constant current, initially at 4 ma until the tracking dye was midway in the stacking gel, then the current was increased to 6 ma for the remainder of the experiment.
The elution flow rate was maintained at approximately 1 ml per min by adjustment of the hydrostatic head and the eluting slit in the column.  6. The elution pattern of a preparation of activated Factor X inhibitor similar to that used in Fig. 5, subjected to electrophoresis on 15% polyacrylamide gel. Of the sample containing approximately 28 mg protein, 5 ml were subjected to electrophoresis. The experimental conditions used in Fig. 5 were applied in this experiment,.
performed 16 to 24 times, and the results as a rule did not deviate markedly from those presented. Ultraviolet Spectrum-The absorption spectrum of the inhibitor in 0.1 M NaCI, with a path length of 1 cm, measured in a Beckman DB-G spectrophotometer, exhibited an ultraviolet spectrum with a maximum at 278 mp. Trichloroacetic Acid Solubility-The solubility of the inhibitor in trichloroacetic acid was determined (Table II). It was 100% soluble in 2.5% trichloroacetic acid, but totally insoluble in 5% trichloroacetic acid. Under the same experimental conditions, bovine serum albumin was 13.6y0 soluble, whereas trypsin was totally insoluble in 2.5% trichloroacetic acid. Carbohydrate Content3-The activated Factor X inhibitor contained 4.1yo hexose and 4.6% sialic acid.
Other carbohydrates were not determined.  (200 pg). Samples A to F were run on 4 cm of 7.5yo separating gels, and G to I were on l-cm spacer gels with 4 cm of 12.5% separating gels. Sample gel was not used in any tests. The sample volume was 100 ~1 and electrophoresis was conducted from 60 to 90 min at 3 ma per gel. The stained samples were stored in 7% acetic acid for photography. The 280 rnp absorbance was measured before the addition of trichloroacetic acid, and in the supernatant solution after centrifugation at 10,000 X g for 30 min.
tively defined as the amount capable of neutralizing 1 unit of activated Factor X in 10 min, at 37" (see "Methods"). % 0 100 0 13.6 Effect of pH on Inhibitor Activity-The pH dependency of the activated Factor X inhibitor activity was performed by incubating at 37", 0.1 ml of inhibitor (30 pg), 0.4 ml of buffer of indicated pH, and 0.1 ml of activated Factor X (9 units in 10% bovine serum albumin in 0.145 M NaCI) for 20 min, followed by the addition of 0.4 ml of neutralizing buffer (e.g. to neutralize reaction mixture of pH 5, buffer of pH 9 was used, and vice versa). An aliquot of the final mixture was removed and checked for residual activated Factor X activity. Buffer was Tris-maleate, 0.025 M. Control experiments in the absence of the inhibitor indicated complete stability of activated Factor X activity over the pH range studied. The results indicated maximum inhibitor activity to be at pH 7 to 9. A sharp drop in activity was noted at pH belox 7 and above 9. Effect of pH on Xtability of Inhibitor at S7"---The pH stability of the activated Factor X inhibitor activity was determined by incubating 0.1 ml of the inhibitor (300 pg) with 0.9 ml of buffer at 37' for 24 hours in a stoppered tube. At the end of the incuba-tion period, the fraction was titrated to near neutrality by adding 1.0 ml of "neutralizing" buffer (e.g. to neutralize an incubation mixture of pH 2, a buffer of pH 12 was added, and vice versa).
From this, 0.2 ml was removed and added to a tube containing 0.7 ml of 1% bovine serum albumin in 0.025 M Tris-maleate, pH 7.3, and 0.1 ml of activated Factor X (9 units), then incubated at 37". At 15 s after the addition of the inhibitor fraction, an initial activated Factor X value was assayed, and another sample tested 30 min later. The percentage of stability was calculated based on a fresh inhibitor control (no prior incubation) taken as 100%. The buffer at pH 2 to 3 was 0.025 M glycine-HCl; at pH 4 to 9, 0.025 M Tris-maleate; and at pH 10 to 12, 0.025 M glycine-NaOH. Control experiments indicated that the final ionic strength of the buffers did not influence the coagulation assay system, nor E. T. Yin, X. Wessler, and P. J. Stall did it influence the stability of activated Factor X in the absence of the inhibitor.
The results indicated that at pH 6 to 8, the inhibitor activity was completely stable under the experimental conditions. At pH 2, it lost 60% of its activity after 24 hours; whereas, at pH 12, only 30% was lost at this time interval.
The fraction remained liquid and clear at both pH extremes.
InJluence of Temperature on Inhibitor Activity-Temperature dependency of the activated Factor X inhibitor activity was performed by incubating at the indicated temperatures, 0.1 ml of inhibitor (30 pg), 0.8 ml of 1% bovine serum albumin in 0.025 M Tris-maleate, pH 7.3, and 0.1 ml of activated Factor X (9 units per ml).
At incubation times of 10, 20, and 30 min, an aliquot (0.1 ml) was removed from the incubation mixture and assayed for residual activated Factor X activity as described under "Experimental Procedure." At l", there was virtually no detectable activity during the 30.min period.
The inhibitor activity at 37" was the best of the three different temperatures studied at pH at 7.30. Stability of Inhibitor at 56"-Three milliliters of inhibitor, 50 pg per ml of 0.01 M Tris-maleate, pH 7.5, were incubated in a stoppered tube in a 56" water bath for 6 hours.
At hourly intervals, 0.2 ml of the incubation mixture was removed and tested for inhibitor activity towards activated Factor X, as described under "Experimental Procedure." The results indicated a greater than 80% stability of the inhibitor at 56" in the 1st hour and at the end of the 6th hour only 20% remained.
The fraction remained clear throughout the 6-hour incubation period. At lower inhibitor concentration (15 pg per ml), the stability of the inhibitor at 56" was insignificantly altered.

DISCUSSION
Although the inhibitor to activated Factor X is present in both plasma and serum, we have chosen to work with the former for several reasons.
By definition, serum is the fluid remaining after whole blood has coagulated.
During blood clotting the plasma coagulation factors interact and intermediates and by-products are formed.
One of the intermediate products formed is activated Factor X which is rapidly neutralized by its inhibitor. There is a significant difference between the level of activated Factor X inhibitor in plasma and serum. In addition, it is not known whether the inhibitor in serum, even though active, may not have been modified chemically or physically as a result of interactions between it and other proteinases such as plasmin and thrombin.
An analogous example is the action of thrombin and plasmin on Factor V (20). Because of these considerations it seemed preferable that activated Factor X inhibitor be isolated from plasma as close as possible to its native form.
Occasionally, the fibrinogen fraction of the pooled rabbit plasma precipitated out during Sephadex gel filtration and clogged the column.
This was especially true when the plasma had been previously frozen. Slow thawing of the frozen material at 4" enhanced the initial precipitation of a large portion of the fibrinogen and facilitated the subsequent gel filtration step. The common practice of defibrinating plasma by heating at 56", or the prior removal of the fibrinogen from the plasma by saltingout with ammonium sulfate was avoided.
In the latter technique there was a substantial loss of the inhibitor activity, unless the salt was immediately removed from the plasma. Gel filtration of the plasma concentrate was best carried out at room temperature (22-26"). At lower temperatures there was a strong possibility of fibrinogen precipitation. Chromatography of the inhibitor peak isolated from Step II on DEAE-Sephadex at room temperature afforded a faster flow rate than at 4", without any difference in the elution pattern.
The elution flow rate was not a determining factor in the effective separation of the activated Factor X inhibitor from the bulk of other proteins, especially the albumin.
However, unduly slow flow rates, even at low temperature, often resulted in low recovery of the inhibitor, and, therefore, we employed the highest flow rate possible without causing the gel bed to collapse.
Although greater than 90% of the fractions containing inhibitor activity always demonstrated a single protein band on microzone electrophoresis migrating as an at-globulin, they invariably showed other minor components on disc electrophoresis. In contrast to the activated Factor X inhibitor isolated from human plasma reported in our preliminary communication (lo), the rabbit inhibitor had less affinity for DEAE-cellulose. The latter could be eluted between 0.03 and 0.045 RI NaCl, whereas the human inhibitor required 0.075 to 0.09 M NaCl (10). Furthermore, the rabbit inhibitor was eluted over a broad peak on the DEAE-cellulose column, whereas the human inhibitor was eluted in a narrow peak.
The electrophoretic pattern on disc gels of the DEAE-cellulose chromatographed rabbit inhibitor fractions varied considerably.
Homogeneity of the individual fraction containing high activity could be obtained, provided that the starting material for this step was chromatographed with a minimum of delay from the time the pool was obtained from the DEAE-Sephadex run. However, even if minor components were present in the peak fractions, the total contamination was minute. These minor components could be produced even in fractions that were previously homogeneous, but had been stored frozen for a short period.
This suggested protein denaturation. Rechromatography of these fractions on DEAE-cellulose invariably created a greater number of minor bands on subsequent disc electrophoresis.
That failure to obtain a homogeneous species on rechromatography was not related to oxidation of the protein by the ammonium persulfate present in the polymerized gel was established by removing the catalyst (19) and adding fi-mercaptoethanol to the electrophoresis system. Thus, if the minor components were indeed denatured protein from the inhibitor, denaturation must have occurred during or after ion exchange chromatography.
Rechromatography on the DEAE-cellulose column also caused considerable lowering of the specific activity of the inhibitor.
Removal of the minor components could be achieved effectively on either a Sephadex G-100 or G-200 column. Again, the specific activity was often lowered, but not to the extent lost by rechromatography on the DEAE-cellulose column. Perhaps the greatest impact on the inhibitor, when it was exposed to purification procedures based on charge effect, was the preparative electrophoresis on polyacrylamide gel. The fresh product traveled as a single but diffuse band on the analytical disc electrophoresis (Fig. 7H).
The diffuse pattern suggested that the protein was already undergoing a certain amount of denaturation. Within a short period during storage the fraction underwent extensive denaturation of the protein molecule with a concomitant loss of activity and the appearance of multiple, heavily stained, slowly migrating bands on disc electrophoresis (Fig. 71). The failure to detect any caseinolytic or tosylarginine methyl ester esterase activity in these purified fraction+ tends to argue