Monoclonal anti-human factor VII antibodies. Detection in plasma of a second protein antigenically and genetically related to factor VII

A single-domain antibody that blocks factor VIIa activity in the absence but not presence of tissue factor

BACKGROUND

Activated factor VII (FVIIa) is pertinent to the initiation of blood coagulation. Proteolytic and amidolytic activity of FVIIa are greatly enhanced by its cofactor, tissue factor (TF).

OBJECTIVE

We aimed to generate a single-domain antibody (sdAb) that recognizes free FVIIa rather than TF-bound FVIIa.

METHODS

A llama-derived phage library was used to screen for anti-FVIIa sdAbs.

RESULTS

One sdAb, KB-FVIIa-004, bound to FVIIa, but not to its precursor FVII or to homologous proteins (prothrombin, factor X, or their activated derivatives). FVIIa amidolytic activity was inhibited by KB-FVIIa-004 (Ki  = 28-45 nM) in a competitive manner. KB-FVIIa-004 also inhibited FVIIa-mediated FX activation (Ki  = 26 nM). In contrast, KB-FVIIa-004 was inefficient in prolonging the clotting time of the prothrombin time-test, which was prolonged by a maximum of 10 s at high sdAb concentrations (10 μM). Furthermore, FVIIa/TF amidolytic activity or FVIIa/TF-mediated FX activation remained unaffected up to a 50-fold to 1000-fold molar excess of KB-FVIIa-004. These data suggest that KB-FVIIa-004 loses its inhibitory activity in the presence of TF. A KB-FVIIa-004/albumin fusion-protein (004-HSA) was generated for in vivo testing. By using 004-HSA, we observed that this sdAb blocked the therapeutic capacity of FVIIa to correct bleeding in FVIII-deficient mice.

DISCUSSION

This observation is compatible with the view that FVIIa functions independently of TF under these conditions. In conclusion, we have generated a sdAb that specifically blocks TF-independent activity of FVIIa. This antibody can be used to gain insight into the roles of TF-bound and TF-free FVIIa.

The Ratio of Factor VIIa:Tissue Factor Content within Microvesicles Determines the Differential Influence on Endothelial Cells

Tissue factor (TF)-positive microvesicles from various sources can promote cellular proliferation or alternatively induce apoptosis, but the determining factors are unknown. In this study the hypothesis that the ratio of fVIIa:TF within microvesicles determines this outcome was examined. Microvesicles were isolated from HepG2, BxPC-3, 786-O, MDA-MB-231, and MCF-7 cell lines and microvesicle-associated fVIIa and TF antigen and activity levels were measured. Human coronary artery endothelial cells (HCAECs) were incubated with these purified microvesicles, or with combinations of fVIIa-recombinant TF, and cell proliferation/apoptosis was measured. Additionally, by expressing mCherry-PAR2 on HCAEC surface, PAR2 activation was quantified. Finally, the activation of PAR2 on HCAEC or the activities of TF and fVIIa in microvesicles were blocked prior to addition of microvesicles to cells. The purified microvesicles exhibited a range of fVIIa:TF ratios with HepG2 and 786-O cells having the highest (54:1) and lowest (10:1) ratios, respectively. The reversal from proapoptotic to proliferative was estimated to occur at a fVIIa:TF molar ratio of 15:1, but HCAEC could not be rescued at higher TF concentrations. The purified microvesicles induced HCAEC proliferation or apoptosis according to this ruling. Blocking PAR2 activation on HCAEC, or inhibiting fVIIa or TF-procoagulant function on microvesicles prevented the influence on HCAEC. Finally, incubation of HCAEC with recombinant TF resulted in increased surface exposure of fVII. The induction of cell proliferation or apoptosis by TF-positive microvesicles is dependent on the ratio of fVIIa:TF and involves the activation of PAR2. At lower TF concentrations, fVIIa can counteract the proapoptotic stimulus and induce proliferation.

Activation of cancer cell migration and invasion by ectopic synthesis of coagulation factor VII

Blood coagulation factor VII (fVII) is physiologically synthesized in the liver and released into the blood. Binding of fVII to tissue factor (TF) at sites of vascular injury triggers coagulation and hemostasis. TF/fVIIa complex formation on the surface of cancer cells plays important roles in cancer biology. Although fVII is synthesized by hepatocellular carcinoma, it remained unclear how TF/fVIIa complex formation and promigratory signaling can occur for most other cancers in extravascular locations. Here, we show by reverse transcription-PCR analysis that nonhepatic cancer cell lines constitutively express fVII mRNA and that endogenously synthesized fVIIa triggers coagulation activation on these cells. fVIIa expression in cancer cells is inducible under hypoxic conditions and hypoxia-inducible factor-2 alpha bound the promoter region of the FVII gene in chromatin immunoprecipitation analyses. Constitutive fVII expression in an ovarian cancer cell line enhanced both migration and invasion. Enhanced motility was blocked by anti-TF antibodies, factor Xa inhibition, and anti-protease-activated receptor-1 antibody treatment, confirming that TF/fVIIa stimulated migration by triggering cell signaling. This study shows that ectopic synthesis of fVII by cancer cells is sufficient to support proinvasive factor Xa-mediated protease-activated receptor-1 signaling and that this pathway is inducible under hypoxia.

The P303T mutation in the human factor VII (FVII) gene alters the conformational state of the enzyme and causes a severe functional deficiency

We report the results of in vitro expression and biochemical characterization of the naturally occurring type II mutation Pro303Thr (P303T) in the factor VII (FVII) gene. Recombinant activated mutated FVII (FVIIa303T), compared with the activated wild-type FVII (FVIIaWT), showed reduced amidase activity toward synthetic substrates, especially when the observed reduced binding affinity for human soluble tissue factor (TF) (K(d) from 4.4 nmol/l for FVIIaWT to 17.3 nmol/l for FVIIa303T) was overcome by a fully saturating TF concentration. Likewise, factor X (FX) hydrolysis by FVIIa303T showed a reduced activity in the absence (and more severely in the presence) of TF (k(cat)/K(m) from 2.3 x 10(7)/mol/l s for FVIIaWT to 8.7 x 10(5)/mol/l s for FVIIa303T). These results showed that the mutant FVIIa is more shifted toward a zymogen-like form compared to FVIIaWT, suggesting that P303 facilitates the conformational transitions that stabilize the active form of FVIIa. The alteration of these allosteric equilibria is especially evident in the presence of TF, which was unable to shift the equilibrium toward a fully active FVIIa form. Additional experiments showed that both TF-catalysed FVII303T autoactivation and FVII303T activation by activated FX in the presence of TF were severely impaired, mainly because of an increase of the K(m) value. Altogether, these defects may explain the severe bleeding symptoms in a patient carrying the FVIIP303T mutation.

Characterization of the protein Z–dependent protease inhibitor

Protein Z-dependent protease inhibitor (ZPI) is a 72-kd member of the serpin superfamily of proteinase inhibitors that produces rapid inhibition of factor Xa in the presence of protein Z (PZ), procoagulant phospholipids, and Ca++ (t1/2 less than 10 seconds). The rate of factor Xa inhibition by ZPI is reduced more than 1000-fold in the absence of PZ. The factor Xa–ZPI complex is not stable to sodium dodecyl sulfate–polyacrylamide gel electrophoresis, but is detectable by alkaline–polyacrylamide gel electrophoresis. The combination of PZ and ZPI dramatically delays the initiation and reduces the ultimate rate of thrombin generation in mixtures containing prothrombin, factor V, phospholipids, and Ca++. In similar mixtures containing factor Va, however, PZ and ZPI do not inhibit thrombin generation. Thus, the major effect of PZ and ZPI is to dampen the coagulation response prior to the formation of the prothrombinase complex. Besides factor Xa, ZPI also inhibits factor XIa in the absence of PZ, phospholipids, and Ca++. Heparin (0.2 U/mL) enhances the rate (t1/2 = 25 seconds vs 50 seconds) and the extent (99% vs 93% at 30 minutes) of factor XIa inhibition by ZPI. During its inhibitory interaction with factor Xa and factor XIa, ZPI is proteolytically cleaved with the release of a 4.2-kd peptide. The N-terminal amino acid sequence of this peptide (SMPPVIKVDRPF) establishes Y387 as the P1 residue at the reactive center of ZPI. ZPI activity is consumed during the in vitro coagulation of plasma through a proteolytic process that involves the actions of factor Xa with PZ and factor XIa.

Characterization of the protein Z–dependent protease inhibitor

Protein Z-dependent protease inhibitor (ZPI) is a 72-kd member of the serpin superfamily of proteinase inhibitors that produces rapid inhibition of factor Xa in the presence of protein Z (PZ), procoagulant phospholipids, and Ca++ (t1/2 less than 10 seconds). The rate of factor Xa inhibition by ZPI is reduced more than 1000-fold in the absence of PZ. The factor Xa–ZPI complex is not stable to sodium dodecyl sulfate–polyacrylamide gel electrophoresis, but is detectable by alkaline–polyacrylamide gel electrophoresis. The combination of PZ and ZPI dramatically delays the initiation and reduces the ultimate rate of thrombin generation in mixtures containing prothrombin, factor V, phospholipids, and Ca++. In similar mixtures containing factor Va, however, PZ and ZPI do not inhibit thrombin generation. Thus, the major effect of PZ and ZPI is to dampen the coagulation response prior to the formation of the prothrombinase complex. Besides factor Xa, ZPI also inhibits factor XIa in the absence of PZ, phospholipids, and Ca++. Heparin (0.2 U/mL) enhances the rate (t1/2 = 25 seconds vs 50 seconds) and the extent (99% vs 93% at 30 minutes) of factor XIa inhibition by ZPI. During its inhibitory interaction with factor Xa and factor XIa, ZPI is proteolytically cleaved with the release of a 4.2-kd peptide. The N-terminal amino acid sequence of this peptide (SMPPVIKVDRPF) establishes Y387 as the P1 residue at the reactive center of ZPI. ZPI activity is consumed during the in vitro coagulation of plasma through a proteolytic process that involves the actions of factor Xa with PZ and factor XIa.

Mechanism underlying factor VII deficiency in Jewish populations with the Ala244Val mutation

We investigated a Sephardic Jewish patient with a mild bleeding diathesis whose plasma levels of factor VII coagulant activity and factor VII antigen were 7% and 9% of normal, respectively. Sequencing demonstrated homozygosity for the Ala244Val mutation and the Arg353Gln polymorphism, which is associated with a modest decrease in factor VII levels. To elucidate the mechanism by which Ala244Val reduced factor VII levels in this patient, transient transfections were performed in COS-1 cells with wild type and mutant factor VII cDNAs and factor VII antigen levels in cell lysates and conditioned media were measured. The secretion of the mutant protein (FVII244V) into the media was 20% of wild type (FVIIwt), and intracellular levels of FVII244V were 60% of FVIIwt. A construct encoding Ala244Val along with the Arg353Gln polymorphism decreased the factor VII level in the media to that observed in the patient's plasma. Pulse-chase experiments demonstrated that FVII244V did not accumulate intracellularly and that low levels of the abnormal protein were maintained throughout the chase. To test the hypothesis that FVII244V results in an unstable molecule, amino acids with smaller (Gly) or larger (Phe) side chains were substituted for Val244 by site-directed mutagenesis. Transient transfection assays with these constructs demonstrated that the side chain of amino acid 244 is crucial in maintaining a proper conformation of the molecule. We conclude that Ala244Val results in a factor VII molecule that is unstable and is probably degraded intracellularly.

Characterization of Two Naturally Occurring Mutations in the Second Epidermal Growth Factor-Like Domain of Factor VII

We investigated the mechanisms responsible for severe factor VII (FVII) deficiency in homozygous Italian patients with either Gly97Cys or Gln100Arg mutations in the second epidermal growth factor domain of FVII. Transient expression of complementary DNA coding for the mutations in COS-1 cells showed impaired secretion of the mutant molecules. Using stably transfected Chinese hamster ovary (CHO) cells, we performed pulse-chase labeling studies, immunohistochemistry, and experiments with inhibitors of protein degradation, showing that FVII-Cys97 did not accumulate intracellularly but was degraded in a pre-Golgi, nonlysosomal compartment by a cysteine protease. In stably transfected CHO cells expressing FVII-Arg100, the level of intracellular FVII was not increased by several inhibitors of protein degradation, but FVII-Arg100 was retained in the endoplasmic reticulum for a longer period of time than wild-type FVII. FVII-Arg100 had a lower apparent molecular weight than did wild-type FVII under nondenaturing conditions, which is attributable to misfolding due to abnormal disulfide bond formation.

Characterization of Two Naturally Occurring Mutations in the Second Epidermal Growth Factor-Like Domain of Factor VII

We investigated the mechanisms responsible for severe factor VII (FVII) deficiency in homozygous Italian patients with either Gly97Cys or Gln100Arg mutations in the second epidermal growth factor domain of FVII. Transient expression of complementary DNA coding for the mutations in COS-1 cells showed impaired secretion of the mutant molecules. Using stably transfected Chinese hamster ovary (CHO) cells, we performed pulse-chase labeling studies, immunohistochemistry, and experiments with inhibitors of protein degradation, showing that FVII-Cys97 did not accumulate intracellularly but was degraded in a pre-Golgi, nonlysosomal compartment by a cysteine protease. In stably transfected CHO cells expressing FVII-Arg100, the level of intracellular FVII was not increased by several inhibitors of protein degradation, but FVII-Arg100 was retained in the endoplasmic reticulum for a longer period of time than wild-type FVII. FVII-Arg100 had a lower apparent molecular weight than did wild-type FVII under nondenaturing conditions, which is attributable to misfolding due to abnormal disulfide bond formation.

Regulation of extrinsic pathway factor Xa formation by tissue factor pathway inhibitor

Tissue factor (TF) pathway inhibitor (TFPI) regulates factor X activation through the sequential inhibition of factor Xa and the VIIa.TF complex. Factor Xa formation was studied in a purified, reconstituted system, at plasma concentrations of factor X and TFPI, saturating concentrations of factor VIIa, and increasing concentrations of TF reconstituted into phosphatidylcholine:phosphatidylserine membranes (TF/PCPS) or PC membranes (TF/PC). The initial rate of factor Xa formation was equivalent in the presence or absence of 2.4 nM TFPI. However, reaction extent was small (<20%) relative to that observed in the absence of TFPI, implying the rapid inhibition of VIIa.TF during factor X activation. Initiation of factor Xa formation using increasing concentrations of TF/PCPS or TF/PC in the presence of TFPI yielded families of progress curves where both initial rate and reaction extent were linearly proportional to the concentration of VIIa.TF. These observations were consistent with a kinetic model in which the rate-limiting step represents the initial inhibition of newly formed factor Xa. Numerical analyses of progress curves yielded a rate constant for inhibition of VIIa.TF by Xa.TFPI (>10(8) M-1.s-1) that was substantially greater than the value (7.34 +/- 0.8 x 10(6) M-1.s-1) directly measured. Thus, VIIa.TF is inhibited at near diffusion-limited rates by Xa.TFPI formed during catalysis which cannot be explained by studies of the isolated reaction. We propose that the predominant inhibitory pathway during factor X activation may involve the initial inhibition of factor Xa either bound to or in the near vicinity of VIIa.TF on the membrane surface. As a result, VIIa.TF inhibition is unexpectedly rapid, and the concentration of active factor Xa that escapes regulation is linearly dependent on the availability of TF.

Factor VII Deficiency Caused by a Structural Variant N57D of the First Epidermal Growth Factor Domain

We have previously described a kindred with factor VII (FVII) deficiency whose members exhibited reduced procoagulant activity relative to FVII antigen concentration. In this report, the molecular genetic basis of the FVII defect has been determined to be a heterozygous substitution of Asp for Asn at position 57 in the first epidermal growth factor (EGF) domain. Recombinant FVII (N57D) cDNA was created by site-directed mutagenesis and transiently expressed in human 293 cells. The transfected cells synthesized an immunoprecipitable protein with an apparent molecular weight of 50 kD. Quantitation of expression by FVII enzyme-linked immunosorbent assay indicated that mutant protein yields were consistently low, typically 10% to 30% of wild-type FVII. FVII (N57D) protein did not accumulate intracellularly, and Northern blot analysis indicated equivalent FVII mRNA levels in 293 cells expressing either wild-type FVII or FVII (N57D). Secreted FVII (N57D) protein did not bind tissue factor, exhibited no procoagulant activity, and failed to bind a conformation-dependent monoclonal antibody specific for the first EGF domain of FVII. Molecular modeling of the first EGF domain of FVII predicted that the N57D amino acid substitution would disrupt tertiary bonding structure. We conclude that the N57D mutation affects folding of the first EGF domain of FVII resulting in decreased cellular secretion of a mutant FVII molecule, which is unable to bind tissue factor and is therefore biologically inactive.

Factor VII Deficiency Caused by a Structural Variant N57D of the First Epidermal Growth Factor Domain

We have previously described a kindred with factor VII (FVII) deficiency whose members exhibited reduced procoagulant activity relative to FVII antigen concentration. In this report, the molecular genetic basis of the FVII defect has been determined to be a heterozygous substitution of Asp for Asn at position 57 in the first epidermal growth factor (EGF) domain. Recombinant FVII (N57D) cDNA was created by site-directed mutagenesis and transiently expressed in human 293 cells. The transfected cells synthesized an immunoprecipitable protein with an apparent molecular weight of 50 kD. Quantitation of expression by FVII enzyme-linked immunosorbent assay indicated that mutant protein yields were consistently low, typically 10% to 30% of wild-type FVII. FVII (N57D) protein did not accumulate intracellularly, and Northern blot analysis indicated equivalent FVII mRNA levels in 293 cells expressing either wild-type FVII or FVII (N57D). Secreted FVII (N57D) protein did not bind tissue factor, exhibited no procoagulant activity, and failed to bind a conformation-dependent monoclonal antibody specific for the first EGF domain of FVII. Molecular modeling of the first EGF domain of FVII predicted that the N57D amino acid substitution would disrupt tertiary bonding structure. We conclude that the N57D mutation affects folding of the first EGF domain of FVII resulting in decreased cellular secretion of a mutant FVII molecule, which is unable to bind tissue factor and is therefore biologically inactive.

Platelet coagulation factor XIa-inhibitor, a form of Alzheimer amyloid precursor protein

An inhibitor of coagulation factor XIa was purified from serum-free conditioned medium of HepG2 liver cells. Platelets stimulated with thrombin or calcium ionophore (A23187) secrete a protein functionally and immunologically identical to the inhibitor, implying a role for this inhibitor in hemostasis. Analysis of the amino-terminal amino acid sequence and immunologic reactivity showed the inhibitor to be a truncated form of the Alzheimer's amyloid precursor protein that contains a Kunitz-type serine protease inhibitor domain and at least a portion of the amyloid beta protein. It inhibits factor XIa and trypsin with a Ki of 450 +/- 50 pM and 20 +/- 10 pM, respectively. Heparin (1 unit/ml) did not significantly effect inhibition of trypsin, but inhibition of XIa was 15 times greater (Ki = 25 +/- 15 pM) in the presence of heparin.

Identification of the 1.4 kb and 4.0 kb messages for the lipoprotein associated coagulation inhibitor and expression of the encoded protein

Lipoprotein-Associated Coagulation Inhibitor (LACI) is a factor Xa dependent inhibitor of the factor VII(a)/Tissue Factor catalytic complex. Deduced from partial cDNA sequence, LACI's amino acid sequence has recently been reported. Northern blot analysis showed LACI cDNA hybridizes to RNAs of 1.4 and 4.0 kb in size. To complete the characterization of the LACI message(s), overlapping LACI cDNAs were isolated from a human endothelial cell library. Sequence analysis revealed the clones' inserts span 4023 bases of sequence, consisting of 381 bases of 5' untranslated sequence, an open reading frame of 912 bases, 2682 bases of 3' untranslated sequence and 48 bases of poly(A) sequence. In addition, a short 1.4 kb insert which encodes for LACI was found to contain 49 bases of 3' untranslated sequence and a 3' poly(A) tail. The 1.4 kb of sequence is contained in the 4.0 kb sequence, except for 14 bases of 5' sequence, suggesting that the LACI messages arise by the use of alternative termination and polyadenylation signals during processing. Northern blot analysis of RNA isolated from cells treated with actinomycin D showed both RNA species appear to be relatively stable. Using a bovine papilloma virus vector, LACI cDNA was transfected into mouse C127 fibroblasts. The recombinant LACI is recognized by polyclonal anti-LACI IgG, binds to factor Xa and inhibits VII(a)/Tissue Factor activity in a similar fashion as LACI purified from HepG2 cell conditioned media.

Monoclonal antibodies to human factor VII: production of immunodepleted plasma for VII:C assays

A high affinity monoclonal antibody to factor VII (RFF-VII/1), coupled to sepharose, was used to immunodeplete factor VII from normal plasma. The plasma could be used as a substrate in a one stage coagulation assay and performed as well as, or better than, commercially available factor VII deficient plasma or plasma from congenitally deficient factor VII patients. Plasma from normal donors (n = 20), patients with liver disease (n = 20), and patients receiving warfarin (n = 20), or congenitally factor VII deficient patients (n = 7) was assayed for VII:C concentration in a one stage coagulation assay. The concentration of VII:C detected with the immunodepleted plasma substrate was comparable in all cases with that seen with a commercially available substrate (r = 0.95).

Monoclonal antibodies to human factor VII: a one step immunoradiometric assay for VII:Ag

Three mouse monoclonal antibodies (RFF-VII/1, RFF-VII/2, and RFF-VII/3) which bind specifically to different epitopes on human factor VII antigen were raised. Two of the antibodies, RFF-VII/1 and RFF-VII/2, bound strongly to factor VII antigen (VII:Ag), but only RFF-VII/1 and RFF-VII/3 were potent inhibitors of factor VII coagulation activity (VII:C). RFF-VII/1 and RFF-VII/2 were used in a one step, double monoclonal immunoradiometric assay for VII:Ag. This was highly reproducible and detected as little as 0.05U/dl VII:Ag. Values for VII:Ag obtained for plasma samples from normal subjects (n = 20), patients with liver disease (n = 20), patients treated with warfarin (n = 20), and those congenitally deficient in factor VII (n = 7) correlated very well (r = 0.96) with data obtained in a radioimmunoassay using polyclonal rabbit antiserum to factor VII. This simple and sensitive monoclonal antibody based assay offers a convenient method for the detection of VII:Ag in various disease states.

Absence of thrombosis in subjects with heterozygous protein C deficiency

Protein C deficiency has been thought to be associated with an increased risk of venous thrombosis. To establish a normal range of values, we used a two-site monoclonal-antibody assay to measure protein C levels in 699 healthy adults. The distribution was log normal; 95 percent of the values ranged from 70 to 140 percent of the overall mean (4.03 micrograms per milliliter). Two subjects had levels more than 3.5 SD below the mean (34 and 50 percent), consistent with heterozygous deficiency. We also screened 4723 other blood donors and found 8 additional unrelated subjects with levels from 33 to 51 percent of normal. Autosomal inheritance of heterozygous protein C deficiency was confirmed in them by a detailed study of four families. Levels from 55 to 65 percent of normal are consistent with either heterozygous deficiency or the lower end of the distribution of normal values, and were found in 79 of 5422 subjects when the two groups were combined. None of the subjects had any history of venous thrombosis. We conclude that heterozygous deficiency of protein C has a prevalence of 1 in 200 to 300, a figure consistent with the known number of homozygous infants recently identified, and that levels consistent with heterozygous deficiency are found in 1 in 60 healthy adults but are not detectably associated with a risk of thrombosis.

Monoclonal antibodies to factor VIII: their application in immunoblotting for the visualization of factor VIII in therapeutic concentrates and plasma

Two monoclonal antibodies (McAb) termed 12A4 and 19C1 have been raised against human factor VIII. In immunoassays 12A4 bound to factor VIII antigen (VIII:Ag) in plasma but not serum whilst 19C1 bound to VIII:Ag in both plasma and serum. Both McAb were shown by immunoblotting to react with the carboxy (C) terminal polypeptide of factor VIII which appeared as a doublet with a molecular weight (Mr) of 77,000/75,000. The C terminal factor VIII polypeptide was detectable by immunoblotting in each of 12 therapeutic factor VIII concentrates, from six different manufacturers, although its level was variable. Factor VIII was visualized in plasma by immunoblotting following its immunoadsorption and elution from agarose-bound monoclonal antibodies. No Mr 77,000/75,000 bands were detectable in plasma obtained from 13 unrelated CRM- haemophiliacs whilst 11 CRM+ haemophilic plasmas from seven kindred were shown to have a C terminal factor VIII polypeptide of normal molecular size.

Isolation of the tissue factor inhibitor produced by HepG2 hepatoma cells

Progressive inhibition of tissue factor activity occurs upon its addition to human plasma (serum). This process requires the presence of factor VII(a), facto-X(a), Ca2+, and another component in plasma that we have called the tissue factor inhibitor (TFI). A TFI secreted by HepG2 cells (human hepatoma cell line) was isolated from serum-free conditioned medium in a four-step procedure including CdCl2 precipitation, diisopropylphosphoryl-factor Xa affinity chromatography, Sephadex G-75 superfine gel filtration, and Mono Q ion-exchange chromatography. The purified TFI contained a predominant band at Mr 38,000 on NaDodSO4/polyacrylamide gel electrophoresis that comigrates with inhibitory activity. Like the activity present in plasma, this TFI requires the presence of factor VII(a), factor X(a), and Ca2+ to express inhibitory activity. Its specific activity (assuming an extinction coefficient of 10 at 280 nM, for a 1-cm path length through a 1% solution) was 9800 units/mg of protein, where 1 unit of TFI activity was defined as that present in 1 ml of normal pooled serum.