Identity of Plasma-activated Factor X Inhibitor with Antithrombin III and Heparin Cofactor*

Abstract Antithrombin III activity cochromatographed with the activity of the inhibitor to activated Factor X during the purification of the inhibitor on Sephadex G-200, DEAE-Sephadex, and DEAE-cellulose columns. Both activities also comigrated on preparative polyacrylamide gel disc electrophoresis. When a fraction of activated Factor X inhibitor was incubated at 56°, at pH 7.5, a simultaneous loss of antithrombin III and activated Factor X inhibitor activities resulted. A complete neutralization of the antithrombin III activity of the activated Factor X inhibitor fraction with thrombin rendered the inhibitor incapable of subsequently inhibiting activated Factor X and vice versa. When the inhibitor was chromatographed on a Sephadex G-200 column, antithrombin III, heparin cofactor, and activated Factor X inhibitor activities cochromatographed with the 280 mµ absorbing material in the activated Factor X inhibitor sample applied. To demonstrate further the identity of these three anticoagulant activities, a series of gel filtration experiments on a single column of Sephadex G-200 was performed in which fractions of activated Factor X inhibitor alone, heparin alone, and heparin mixed with the inhibitor were separately chromatographed in the presence of 0.01 m, 0.10 m, 0.20 m,and 0.30 m NaCl in 0.02 m Tris-maleate, pH 7.2, at 26°. The relative elution volume, Ve/V0 of activated Factor X inhibitor alone in 0.01 m NaCl was 1.83, and in 0.10 m to 0.30m NaCl was 1.77. Heparin behaved as a heterogenous substance with molecular weights ranging from 40,000 to greater than 200,000, at the salt concentrations studied. When a mixture of activated Factor X inhibitor and heparin was chromatographed, the two components eluted as a complex. A single broad heparin cofactor peak cochromatographed with the activated Factor X inhibitor protein and the heparin when high ionic strength eluent was employed. The fractions demonstrated antithrombin III activity when the heparin in the heparin cofactor peak fractions was neutralized with protamine sulfate. The plasma-activated Factor X inhibitor was shown to be unrelated to a hepatic anticoagulant that had been claimed to be a specific inhibitor of activated Factor X. It is concluded, therefore, that the biological activities variously termed activated Factor X inhibitor, antithrombin III, and heparin cofactor activities all belong to a single blood proteinase inhibitor with broad specificity. Together with other data presented, it is suggested that the key function of activated Factor X inhibitor as a natural anticoagulant may be primarily concerned with regulating hemostatic balance through neutralization of activated Factor X, a reaction profoundly enhanced by traces of heparin. It is for these reasons we propose that this low molecular weight α2-globulin inhibitor be termed activated Factor X inhibitor.


Xaterials
The reagents used in this report have already been described in the preceding two communications (1, 24).

Methods
The act,ivated Factor X inhibitor activity in the chromatographic effluents, and other investigative techniques, unless otherwise stated, are described in the preceding two reports (1, 24). The antithrombin III activity from column effluents were routinely assayed as follows. A 0.2-ml fraction (1:2 to 1: 5 dilution in 0.145 M NaCl), 0.2 ml of NaCl (0.145 M), and 0.1 ml of thrombin (2 units) were added in the indicated order to a siliconized tube at 37", and at 15 s and 5 min after the addition of the thrombin, O.l-ml aliquots were removed and added to 0.2 ml of fibrinogen solution.
The difference in clotting times between the 15-s and 5-min samples represented the potency of the antithrombin III activity.
,411 experiments involving thrombin were performed with siliconized glassware. The heparin cofactor and hepnrin assays were performed as described in the appropriate sections.

Identity
of Activated Factor X Inhibitor with Antithrombin III-It has been observed (3,25) that an heterogeneous fraction of plasma antithrombin III was also capable of inhibiting activated Factor X (autoprothrombin C). When the antithrombin III fraction was saturated with thrombin, it was found that the resulting mixture subsequently became incapable of neutralizing activated Factor X. From this it was concluded that the two activities responsible for the neutralization of thronlbin and activated Factor X belonged to a singIe molecule. As we have previously demonstrated, highly purified fractions of activated Factor X inhibitor were capable of inhibiting thrombin (1). The following experiments were performed to establish the relationship between the inhibitor and antithrombin III. Titration of Activated Factor X Znhibitor with Thrombin and Activated Factor X-When 1 ml of inhibitor (100 @Lg.) was incubated with 1 ml of t'hrombin (200 NIH units) in a siliconized tube at 37" for 180 min, less than 0.05 unit of thrombin remained in the reaction misture.
From this primary reaction mixture, 0.3 ml was removed and incubated with 0.1 ml (2 units) of thrombin for 10 min at 37". Likewise, another 0.3.ml sample from the primary incubation mixture was removed and allowed to react with 0.1 ml (2 units) of activated Factor X. Approximately 98% of the thrombin and 100% of the activated Factor X remained in the secondary incubation mixtures. Control experiments were performed in which the inhibitor was incubated for 3 hours at 37" with a volume of NaCl (0.145 MM) replacing either activated Factor X or thrombin, and then tested for its ability to inhibit either enzyme. It was found that the inhibitor activity was still fully intact.
In another series of experiments in which the inhibitor was approximately 990/, quenched with activated Factor X, t,he ability of the reaction mixture further to inhibit thrombin was completely lost. Heat Stability of Two Anticoagulant Activities in Purified Fraction of Activated Factor X Inhibitor-A purified fraction of activated Factor X inhibitor was heated in a stoppered tube at 56" and pH 7.5 for 6 hours.
Samples of the heated fract,ion were removed hourly and assayed for antithrombin III and activated Factor X inhibitor activities.
Prior to heat treatment, the fraction was serially diluted to establish a calibration curve for each activity.
The concentration of the fraction to be heated was adjusted such that 2 units of activated Factor X and 18 units of thrombin were inactivated sepasately in 10 min under the described conditions (1). The loss of both the antithrombin III and activated Factor X inhibitor activities closely followed each other (Fig. 1).
Chromatographic and Electrophoretic Behavior of Antithrombin III during Purification of Activated Factor X Inhibitor--As shown in Fig. 2, in the initial isolation of the activated Factor X inhibitor on Sephadex G-200, the antithrombin III activity cochromatographed with it. When this peak was pooled and further chromatographed on a DEAE-Sephadex column, Fig. 3, the same pattern as in Fig. 2 was obtained. During the subsequent steps in the purification procedure (24)) antithrombin III activity consistent,ly cochromatographed with the activated Factor X inhibitor activity on the DEAE-cellulose column. When the DEAE-cellulose chromatographed, pooled fraction containing high activated Factor X inhibitor activity was filtered either on a Sephadex G-100 or Sephades G-200 column, the antithrombin III also cochromatographed with the activated Factor X inhibitor (not shown). These two activities were also eluted together at pH 9 on the ion exchange cellulose columns.
A pooled fraction of activated Factor X inhibitor from three different DEAE-cellulose runs was fractionated by preparative polyacrylamide gel disc electrophoresis at pH 8.  with the protein in a purified fraction of activated Factor X inhibitor further suggested that all three activities might belong to a single protein molecule.
The following experiments mere performed to demonstrate further the relation of heparin cofactor to activat,ed Factor X inhibitor and, therefore, to antithrombin III. Sephadex G-200 packed in a single Lucit,e column, 0.95 x 40 cm, was used for all the filtration experiments carried out at 26" with a flow rate of 3 ml per hour controlled with a hydrostatic head.
Fractions of 1 ml each were collected in a LKB fraction collector.
For each set of filtrations and using the same salt concentration, 50 ml of the buffer were allowed to pass through the column between each run, and when a buffer of a different ionic strength was used, the column was re-equilibrated with 200 ml of the desired buffer before use. The experiments presented in Fig. 6 were chromatographed in the following order: First, System A; second, System D; third, System B; and last, System C. Dextran blue was used as a void volume (V,) marker for each set of filtrations using different buffers. The total volume of the samples loaded was always 1 ml for each experiment and the materials to be chromatographed were contained in the same buffer as that for elution.
The various concentrations of NaCl were dissolved in 0.02 M Tris-HCl, pH 7.2.
When 1.25 mg of the inhibitor alone was chromatographed on the Sephadex G-200 column, it emerged as a symmetrical peak irrespective of the ionic strength of the buffers used. The relative elution volume, V,:Vo (where V, is the elution volume of the inhibitor and Vo is the void volume), of the inhibitor with 0.01 M NaCl was 1.83 and that with 0.10 M, 0.20 M, and 0.30 M NaCl was 1.77. When 500 USP units of heparin alone were chromatographed, the elution profiles were typical of those for macroheterogeneous substances. At a low salt concentration, such as 0.01 M, part of the heparin fraction behaved like a compound with a molecular weight approaching 200,000 (the exclusion limit of Sephadex G-200) with V,: VO equal to 1. The remainder of the fraction was eluted over a broad peak with a V,:VO equal to 1.83, the same as that for the inhibitor alone. However, when the heparin was eluted with 0.10 M NaCl, the V,:VO was 1.38. The activity spread over the entire gel bed with a slight shouldering on the descending slope of the peak. At 0.20 M NaCI, the heparin was eluted in a slightly narrower peak with a 8,: VO equal to 1.53. Finally, when the ionic strength was increased to 0.30, there was less activity near the VO region and the V,: Vo of the heparin was 1.77, the same as that of the inhibitor alone. In all instances, the heparin behaved like a heterogeneous compound with a molecular weight ranging from approximately 40,000 to greater than 200,000.
When a fraction of activated Factor X inhibitor was chromatographed together with 500 units The activated Fa.ctor X inhibit,or activity was measured as in the preceding paper (24) and the antithrombin III was measured as described under "Methods," but undiluted.
The following method was used in the determination of heparin cofactor activity.
A 0.1~ml sample of the column effluent diluted 1:4 in 0.145 M NaCl was incubated in a siliconixed tube with 0.1 ml of heparin (1 unit) and 0.2 ml of fibrinogen (Fraction I-4) at 37" for exactly 10 s and the mixture clotted with 0.1 ml of thrombin (3 units).
of heparin, the elution profile for each substance was altered for each salt concentration used. The inhibitor protein (measured by 280 rnp absorbance), when eluted at an ionic strength of less than 0.3, behaved as a heterogeneous substance with a wide range of molecular weights. It cochromatographed with the heparin.
At 0.30 M NaCI, the inhibitor was eluted near symmetry.
In all four experiments, both the heparin activity and the inhibitor peaked together, and had the same V,: V, as that of the inhibitor when chromatographed alone under the same conditions. Therefore, the relative elution volume of the heparin, when chromatographed together with the inhibitor was shifted to the right, indicative of complex formation with the inhibitor protein. The fractions were furthermore tested for heparin cofactor activity, that is, the "immediate" antithrombin effect. a major heparin cofactor activity peak was found in the void volume region of the Sephadex column using 0.01 M NaCl as eluent. The rest of the activity spread over the entire column. The activity peak did not cochromatograph with either the inhibitor protein or the heparin activity.
When the heparin plus inhibitor mixture was chromatographed and eluted with either 0.10 M or 0.20 M NaCl, two distinct heparin cofactor activity peaks separated from both the inhibitor protein and heparin peaks were seen for each run.
The inhibitor protein and heparin peaks showed less skewing with increasing ionic strength.
Bt an ionic strength of 0.3, the inhibitor protein peak was nearly symmetrical.
There The residual activity of each enzyme was then determined at the stated time periods.
As shown in Table I  activated Factor X remained. In comparison, when 5 pg of inhibitor replaced the hepatic inhibitor in the test system greater than 95% of the activated Factor X was neutralized in 60 min.
The amount of hepatic inhibitor employed in the above experiment was approximately 12 times greater (per ml of reaction mixture) than that used by Deykin et al. (23). As anticipated, however, 350 pg of the hepatic fraction that failed to inhibit activated Factor X in the specific assay system was still capable of prolonging the partial thromboplastin time from 82 to greater than 600 s. This anticoagulant effect was partially corrected by protamine sulfate.
When 350 pg of the material in tube 36 were incubated with 5 pg of activated Factor X inhibitor and 10 units of activated Factor X in a total volume of 1 ml, 9 units of activated Factor X were neutralized in 20 min instead of 60 min when the hepatic fraction was omitted.
This indicated an acceleration effect on the activated Factor X inhibitor inhibition of activated Factor X equivalent to 0.01 unit of heparin.
However, this trace amount of heparin-like activity in tube 36 could not account for the heat-labile anticoagulant activity observed in the partial thromboplastin time test. These data strongly suggest that the liver inhibitor is not identical with our plasma inhibitor and is not a specific inhibitor of activat.ed Factor X.

DISCUSSION
As one reviews the literature on this subject, one becomes aware that the issue of the identity of antithrombin III with heparin cofactor has remained unresolved. Seegers et al. (5) in 1942, for example, claimed identity for antithrombin III and heparin cofactor. Yet, in 1954, Fell et al. (17) reported that these two activities belonged to different substances without referring to Seeger's earlier findings.
More recently, Abildgaard (13) prepared a fraction of antithrombin III by disc electrophoresis and found it to possess heparin cofactor activity. From this single observation he concluded that both antithrombin III and heparin cofactor were identical with one another.
Porter, Porter, and Shanberge (21), using Sephadex gel filtration and electrophoretic techniques on heparinized plasma, obtained certain elution patterns that they interpreted as proof that they had partially separated heparin cofactor from antithrombin III.
Based on their observations they concluded that these two activities were different plasma components.
Ganrot (22), on the basis of his work, came to a similar conclusion.
The data presented in Figs In general, gel filtration permits the fractionation of compounds with different molecular weights or sizes. In rare instances, difficulties are encountered because of the nature of the dextran gel or the molecular structure of the materials fractionated.
Changes occur, moreover, with the varying ionic strength of the solvent employed.
Often, departure from ideal molecular sieve behavior is observed. Dextran gel contains a small number of free hydroxyl groups and behaves as a weak polyanion.
Unless these charges are suppressed, highly acidic compounds will be excluded from the gel matrix regardless of their molecular weights (26-36).
Heparin is one of these compounds, as demonstrated by Skalka (32), whose findings were confirmed here when heparin alone was chromatographed on Sephadex G-200 and eluted with different ionic strength buffers (Fig.  6). In all instances heparin was eluted as a heterogeneous compound with apparent molecular weights ranging from approximately 40,000 to greater than 200,000.
The elution profiles of heparin alone on the Sephadex column have served as controls for the behavior of this compound under the experimental conditions chosen, in this study, to demonstrate the identity of activated Factor X inhibitor with heparin cofactor activity. Porter et al. (20,21) did not indicate that they had performed such controls, but had simply assumed that since heparin is a low molecular weight substance (less than 16,000) it should completely penetrate the gel matrix and thus be eluted much later than albumin.
Such an assumption, however, is valid only for globular substances that do not interact with the gel.
The experiments in which heparin was chromatographed together with the inhibitor to activated Factor X (Fig. 6,  with those of heparin cofactor activity were obtained. The following explanations are offered for this phenomenon. In low ionic strength solvents there is much electrostatic interaction between the heparin molecule of the heparin-inhibitor complex and the dextrdn gel. This causes the electroheterogeneous population of heparin molecules to rearrange into groups carrying a similar number of electrical charges.
The group capable of causing higher electrostatic repulsion will, therefore, be excluded faster from the gel. These highly charged groups possess greater anticoagulant activity and may reflect the presence of larger amounts of sulfate ester groups at all primary hydroxyl positions.
It is the 6-O sulfate group that confers the anticoagulant activity on the heparin molecule rather than the actual molecular weight (37). Using heparinized whole plasma, Porter et al. (21) interpreted a similar, but less defined, phenomenon as a separation of the heparin cofactor from the antithrombin III. This separation occurred, in their view, by virtue of the heparin complexing with its plasma cofactor thereby resulting in the formation of a larger molecular weight complex that was eluted earlier than antithrombin III. In our study, when ionization of the free charged groups is suppressed by high salt concentration, e.g. in the experiment employing 0.3 M NaCl (Fig. 6), electrostatic interactions are minimized, and, therefore, the heparin cofactor cochromatographed with the activated Factor X inhibitor.
Porter et al. (21) attempted further to support their claim that heparin cofactor and antithrombin III are two different substances by their failure to detect the latter activity in a fraction containing heparin plus cofactor, after neutralization of the heparin.
They reasoned that, if heparin cofactor and antithrombin III belonged to a single substance, then the neutralization of the anticoagulant in their isolated fraction containing heparin cofactor activity should demonstrate progressive antithrombin activity. Although this is a valid argument, they did not determine the amount of heparin cofactor protein in the fractions they examined. When the heparin in the fractions containing heparin cofactor activity (Fig. 6) is neutralized with protamine sulfate, these fractions demonstrate antithrombin III activity.
The amount of antithrombin III activity detected was dependent upon the quantity of activated Factor X inhibitor protein present in each fraction.
The data presented in Table I  further substantiate the thesis that heparin cofactor, antithrombin III, and activated Factor X inhibitor activities are identical.
The observed increase in the inhibition of either enzyme in the presence of heparin over those values in the absence of heparin is related to the fact that the control experiments (no added heparin) were not allowed to proceed to equilibrium with the inhibitor. It is evident that, in the presence of heparin, the inhibitory action of minute quantities of activated Factor X inhibitor on either activated Factor X or thrombin can be maximally achieved instantaneously, whereas, in its absence, the time period required to achieve the same degree of inhibition would be so great that it might even exceed the half-life of the inhibitor.
Great precautions should also be exercised, in all these studies, to rule out any contamination of activated  (23) reported that they had isolated a specific inhibitor of activated Factor X from bovine liver.
This inhibitor prolonged both the intrinsic and extrinsic coagulation assay systems. No specific assay for activated Factor X inhibition was employed. The inhibitor was completely inactivated at 56" within 10 min, and prolonged slightly the thrombin-fibrinogen reaction.
The addition of the crude liver inhibitor fraction to a thrombin generation test system caused a reduction of the final amount of thrombin formed from 0.65 to 0.35 unit.
From this experiment the! inferred that the inhibitor acted specifically against activated Factor X, despite the fact that they were working with a multicomponent reaction system, in which an adverse effect on an! one of the reactants could influence the over-all kinetics.
W'c were interested in this hepatic inhibitor primarily to determine its relationship to our plasma inhibitor. Having repeatedly failed to detect any specific inhibition on purified actiratctl Factor X with hepatic inhibitor fractions that prolonged the Quick prothrombin time and the partial thromboplastin time, prepared according to their published method (23), we obtained from Dr. Deykin for comparative studies the original fractions used in his reported experiments.
No inhibition 01 activated Factor X was observed when it was incubated with the hepatic inhibitor fractions supplied by Dr. Deykin. This was true even though the concentration of this inhibitor fraction employed in our test system was increased up to 12-fold the amount used in their published report without any loss of activated Factor X activity. The behavior of this hepatic in hibitor was, in fact, rather reminiscent of the one observed by us previously in liver perfusates (42). In conclusion, based on the data presented both in this and the preceding reports (1,43), we propose that the key function of activated Factor X inhibitor as a natural anticoagulant may be primarily concerned with regulating hemostatic balance through neutralization of activated Factor X; a reaction profoundly enhanced by traces of heparin.
Whether we are dealing with a polyvalent inhibitor or one with a single substrate site for the enzymes studied is not conclusively demonstrated. However, the fact that both activated Factor X and thrombin (as well as plasmin and trypsn?) are all inhibited by the activated Factor 9 2 E. T. Yin and S. Wessler, unpublished data. by guest on March 17, 2020 http://www.jbc.org/ inhibitor, and that these activities belong to the serine group of enzymes, tends to favor the view that these anticoagulant activities are functions of a single site.
We further propose that this wglobulin inhibitor previously termed antithrombin III or helxwin cofactor, be now appropriately designated activated Factor X inhibitor.
Finally, there are several implications of this investigation that are of clinical interest.
One of these, perhaps, deserves mention here: namely, the possibility that an altered level of plasma-activated Factor X inhibitor may serve as au indicator that intravascular coagulation may have occurred. For, if systemic hypercoagulability is reflected in the level of circulating activated Fact,or X inhibitor activity, the methodology herein presented is capable of determining the validity of such