isolation and characterization of human factor viii ... journal of b~ologicli. chemistry vol. 218,...

THE JOURNAL OF B~OLOGICLI. CHEMISTRY Vol. 218, No. 11, Issue of June 10 pp. 3946-3955, 1973

Printed in U.S.A.

Isolation and Characterization of Human Factor VIII (Antihemophilic Factor) *

(Received for publication, December 21, 1972)

n/IARK E. LEGAZ, GOTTFRIED SCHMER, RICHARD B. COUNTS, AND EARL W. DAVIE

From the Departments of Biochemistry and Laboratory Medicine, University of Washington, School of Medicine, Seattle, Washington 98195

SUMMARY

Factor VIII (antihemophilic factor) has been purified approximately 500.fold from the cryoprecipitate fraction of human plasma. The isolation procedure involves adsorption of contaminants with Al(OH),, column chromatography on tricalcium citrate-cellulose, precipitation with concanavalin A, and an agarose gel filtration step. The linal product is homogeneous when examined by zone electrophoresis, sedimentation equilibrium, and immunoelectrophoresis. The molecular weight determined by sedimentation equilibrium is 1.12 million + 98,000. After reduction of the protein with Z-mercaptoethanol or dithiothreitol, subunits are formed which migrate as one band in polyacrylamide gel electrophoresis and zone electrophoresis. The subunits are heterogeneous, however, in the ultracentrifuge, apparently due to substantial aggregation. The smallest species which could be detected has a molecular weight of 105,000 f 5,000. The molecular weight of the subunit determined by sodium dodecyl sulfate (SDS) gel electrophoresis was 240,000. The latter value may be high, however, due to the fact that human Factor VIII contains approximately 6% carbohydrate (hexose, hexosamine, and neuraminic acid) and the molecular weights of glycoproteins determined by SDS gel electrophoresis tend to be high. Antibodies prepared in rabbits against human Factor VIII inhibit both human and bovine Factor VIII activity. Antibodies to the highly purified human Factor VIII also form a precipitin line in immunoelectrophoresis experiments with the cryoprecipitate fraction prepared from hemophilic plasma, indicating that an abnormal Factor VIII molecule is present in the plasma of individuals with classic hemophilia. Other general properties of human Factor VIII, including its amino acid composition, thrombin modification, and turnover in hemophilic dogs, a,re also reported.

Factor VIII (antihemophilic factor) is a plasma protein which participates in the middle phase of intrinsic blood coagulation

* Preliminary reports of this work have appeared elsewhere (1, 2). This work was supported in part by Research Grants GM 10793 and HIi: 11857 from the National Institutes of Health. M. E. L. and R. B. C. were supported by National Institutes of Health Training Grants GM 00052 and AM 05130, respectively. The nomenclature for various coagulation factors is that recom- mended by an international nomenclatlwe committee (3).

(4). It is inactive in individuals with classic hemophilia, a dis- ease which is characterized by a sex-linked coagulation disorder. Factor VIII isolated from bovine plasma is a large glycoprotein with a molecular weight of approximately 1.2 million (5). It is composed of a number of small subunits which are similar or identical, and these are held together by disulfide bonds. Factor VIII has also been highly purified from human plasma. Hersh- gold et al. (6) purified human Factor VIII approximately lO,OOO- fold by a procedure which included polyethylene glycol fractiona- tion and a gel filtration step. More recently, Marchesi et al. (7) have reported the isolation of human Factor VIII by a method which includes a limited digestion step with chymotryp- sin for the removal of fibrinogen. The purification of bovine Factor VIII, as developed in our laboratory, involved a key step employing calcium citrate-cellulose column chromatography (5). This procedure was very effective in separating bovine Factor VIII from fibrinogen. In the present manuscript, we report the isolation and characterization of human Factor VIII from Blood Bank cryoprecipitat,e by a procedure which also includes chromatography on a calcium citrate-cellulose column. With this procedure, a relatively stable product of high purity can be made in milligram quantities. This preparation is suitable for various physicochemical and biological studies, many of which are also reported in this paper.

ESPERIivlENTAL PROCEDURE

Materials

Aluminum hydroxide gel (Amphojel without flavor) was purchased from Wyeth Laboratories and rabbit brain cephalin, e-amino-n-caproic acid, Tris (Trizma base), N-acetylneuraminic acid, and p-dimethylaminobenzaldehyde were purchased from Sigma. Barium sulfate (x-ray grade) was purchased from Merck (U.S.), and kaolin (acid wash) was obtained from Fischer Scien- tific Co. Sodium barbital and 5-sulfosalicylic acid were purchased from Mallinckrodt Chemical Works. Sodium azide N , N’methylenebisacrylamide, polyacrylamide, and N , N, N’, N’-tetramethylethylenediamine (TEMED) were products of Eastman Organic Chemicals. Polyethylene glycol (Carbowax 6000) was obtained from Union Carbide Corp., and imidazole, (ethylenedinitrilo)-tetraacetic acid tetrasodium salt (EDTA), ammonium persulfate, and calcium citrate [Ca3(C&H50,)2.4 HtO] were obtained through Matheson, Coleman and Bell.

Whatman CFll fibrous powder was obtained from Reeve Angel. The final gel filtration step was carried out with agarose

3946

by guest on June 27, 2018http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

3947

A-15m supplied by Bio-Rad. Agarose, purchased from Marine Colloids, Inc., Springfield, N. J., was employed for zone electrophoresis and immunoelectrophoresis. Methyl-a!-n-glucopyranoside (B grade) and dithiothreitol were obtained from Calbiochem. Centrolex-P, purchased from Central Soya, Chicago, Ill., was used as a platelet substitute. Dri Film silicone (SC 87) was a product from General Electric Co., and all glassware was coated with this material.

The 5,5’-dithiobis@nitrobenzoic acid) was obtained from Aldrich Chemical Co., Inc. Bovine thrombin was purchased from Parke Davis and Co. and further purified by the method of Thompson and Davie (8). Concanavalin A was prepared according to the method of Agrawal and Goldstein (9). Some of these preparations apparently contained protease activity which readily destroyed Factor VIII activity, and these preparations were heated at 56” for 30 min to overcome this problem. Bovine Factor VIII was prepared by the method of Schmer et al. (5). All other chemicals were commercial preparations of the highest quality available.

Methods

Protein concentration was measured by the ninhydrin reaction after alkaline hydrolysis according to the method of Moore (lo), absorption at 280 nm, or by the method of Lowry et al. (11). The hesose content of Factor VIII was determined by the phenol- sulfuric acid method of Dubois et al. (12), employing a 1:l mixture of glucose and galactose as a reference standard. Hexos- amine was determined by the Elson and Morgan reaction as described by Garde11 (13), with galactosamine as a standard. For the quantit,ative determination of sialic acid, the thiobar- bituric acid assay of Warren (14) was used, with N-acetylneuraminic acid as a reference standard.

Factor VIII was reduced and alkylated by dissolving 5 mg of protein in a total volume of 1.5 ml containing 720 mg of recrystallized urea, 0.1 ml of a 0.1 M EDTA solution, pH 8.6, 0.015 ml of 14 M 2-mercaptoethanol, and 0.6 ml of a 0.1 M Tris- HCl solution, pH 8.6. The above reaction mixture was allowed to stand for 4 hours at room temperature under nitrogen, followed by t,he addition of 0.2 ml of 1 N NaOH containing 56 mg of recrystallized monoiodoacetic acid. After 15 min, the solution was dialyzed at room temperature overnight against a solution containing 0.2 M 2-mercaptoethanol, 8.0 M urea, and 0.5 M

sodium barbital buffer, pH 8.6. This preparation was then employed directly for zone electrophoresis experiments. For polyacrylamide gel electrophoresis, the protein was reduced with 0.065 M dithiothreitol or 0.2 M 2-mercaptoethanol in 0.1 M phosphate buffer, pH ‘7.0, containing 2y0 SD&r 0.002 M EDTA, and 4.0 M urea.

Antibody Preparation-Rabbits were immunized as previously described (5) with either the highly purified human or bovine Factor VIII or material which had only been purified through the early stages of the procedure (up through the adsorption and elution from the tricalcium citrate-cellulose column).

Electrophoresis-Polyacrylamide gel electrophoresis of the reduced Factor VIII was carried out by a modification of the general met,hod of Weber and Osborn (15). In the present experiments, electrophoresis was performed in 0.2 M phosphate buffer, pH 7.0, containing 0.270 SDS. A 5% gel was employed and electrophoresis was carried out for 5 to 7 hours at room temperature. The gels were stained for protein with Coomassie blue. When staining for carbohydrate, the SDS was removed from the gels by dialysis for 24 hours against a solution containing 5% tri-

The abbreviation used is: SDS, sodium dodecyl sulfate.

chloroacetic acid and 56/, sulfosalicylic acid. The gels were then stained by the method of Zacharius et al. (16).

Zone electrophoresis was carried out on microscope slides precoated with 1 ml of 0.5% agarose in water and allowed to dry at 37”. The slides were then layered with 1.6 ml of 0.5% agarose solution containing 2% unpolymerized acrylamide and 0.05 M

sodium barbital buffer, pH 8.6, as described by Williams and Chase (17). Samples (10 ~1 containing 10 to 20 pg of protein) were placed in a small well and electrophoresis was carried out at room temperature for 45 min with 150 volts and 5 ma per slide. The slides were stained for protein with Amido black.

Immunoelectrophoresis was performed on microscope slides (25 x 75 mm) coated with 0.5% agarose following the method of Scheidegger (18). Essentially the same conditions were employed as those for zone electrophoresis. Samples were diluted in 0.05 M sodium barbital buffer, pH 8.6, and run for 30 to 45 min. Antibody was added to the center trough and allowed to diffuse for at least 48 hours. The slides were then soaked in 0.15 RZ NaCl for at least 48 hours and then dried at 37”. They were subsequently stained for protein with Coomassie blue in meth- anol-acetic acid-water (5: 1:5) for 5 min and destained in i% acetic acid followed by a 95% ethanol wash.

Amino Acid Analysis-Samples for amino acid analysis were prepared by the method of Moore and Stein (19). These samples (2 to 4 mg) were placed in 1 ml of 6 x constant boiling HCl, deoxygenated, and sealed in glass tubes. Hydrolysis was carried out for 24, 48, 72, and 96 hours at 110’. Amino acid analyses were carried out with a Spinco model 120 amino acid analyzer according to the method of Spackman et al. (20). Samples from two different preparations of Factor VIII were employed and gave essentially identical results. The reported values for serine and threonine were extrapolations to zero time hydrolysis, whereas isoleucine, leucine, and valine values were the average of the 96-hour hydrolyses. Half-cystine and methionine were determined after performic acid oxidation according to Hirs (21). Tryptophan was determined by the method of Bencze and Schmid (22). Free sulfhydryl groups were determined by the method of Ellman (23) after denaturation of the protein in 8 M

urea. Ultracentrifuge Analysis-Sedimentation equilibrium experi-

ments were performed as described by Harris et al. (24) and Seery et al. (25). Samples of human Factor VIII were prepared by dialyzing 2 to 4 mg of protein against 0.04 M Tris-HCl, pH 7.4, and 6 M guanidine at 4” for 12 to 14 days prior to each experiment. For subunit determinat,ions, 0.04 M dithiothreitol was added to the above buffer and the flask flushed with nitrogen and sealed with a rubber stopper. This procedure avoided oxidation of the dithiothreitol and insured total reduction of the protein. In all experiments, the concentration in each of the three cells was 1.0, 0.7, and 0.5 mg per ml, respectively. Calculations were made by the methods discussed by Teller et al. (26) and Teller (27). The sedimentation equilibrium data were obtained using the short column, high speed technique of Yphantis (28). A six- channel Kel-F centerpiece was employed, and the interference patterns were recorded on Kodak II-G photographic plates. Rayleigh interference fringes were measured by a modified Nikon microcomparator automated as described by DeRosier et al. (29). This method yields far more precise data than those collected manually. All experiments were conducted at 20”. The partial specific volume of the protein moiety was calculated using the amino acid composition according to the method of McMeekin et al. (30). To correct this value for 6% carbohydrate content (see Table IV), a value of 0.62 ml per g was em-


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

394s

ployed (31). Thus, the actual value of 0 = (0.74 ml per g X 0.94) + (0.62 ml per g x 0.06) = 0.73 ml per g. Other pro- cedures and computations were performed as described (32). The SDS gel electrophoresis method of Weber and Osborn (15) was employed for estimating the molecular weight of Factor VIII subunits.

Clotting Assays-Platelet-deficient titrated plasma was prepared from normal human blood and patients with various coagulative abnormalities as previously described (32). In this procedure, plasma was not permitted to come in contact with glass and was frozen and stored at -80” until employed. Pooled platelet-deficient titrated human plasma was employed as a reference standard in the Factor VIII assay and was stored at -80”.

Factor VIII activity was measured by a modification of the kaolin-activated partial thromboplastin time employing the methods of Langdell et al. (33) and Proctor and Rapaport (34). In this assay, 0.1 ml of Factor VIII-deficient plasma (Factor VIII level less than 1%) was incubated for 10 min with 0.1 ml of an appropriate dilution of Factor VIII and 0.1 ml of a solution containing 0.05% Centrolex-P and 0.2% kaolin in 0.01 M Tris- HCl, pH 7.4, and 0.15 M NaCl. The Centrolex-P-kaolin solution was preincubated for 10 min at 37” prior to use. The assay mixture was then recalcified by the addition of 0.1 ml of 0.033 M

CaC12. The time required to form a visible clot after the addition of Ca2+ was recorded with a stop watch. Further details of this assay are described elsewhere (5). One unit of Factor VIII was arbitrarily chosen as that amount of activity present in 1 ml of fresh human plasma.

Fibrinogen was assayed employing the methods of Jacobsson (35) and Clauss (36). Various other coagulation factors (Factors VII, IX, X, and XII) and plasminogen were determined essentially as described by Biggs and Macfarlane (37).

Turnover Studies in Hemophilic Dogs-Studies of bovine and human Factor VIII turnover were performed in hemophilic beagles obtained from Dr. R. G. Buckner of Oklahoma State University. These animals weighed about 8 kg each. Ten minutes before the turnover experiments began, the dogs were given 20 mg of diphenhydramine hydrochloride intravenously. Ten milliliters of the hemophilic dog blood were then drawn and placed in a tube containing 1 ml of 4% sodium citrate. Factor VIII was then dissolved in 10 ml of sterile NaCl solution and slowly injected into a cephalic vein. Two-milliliter samples were drawn into sodium citrate at 15 and 30 min after injection, then at 2- to a-hour intervals for the first 12 hours. The final sample was removed at approximately 18 hours after infusion of Factor VIII. Samples were centrifuged at 2,000 x g for 10 min and the platelet-poor plasma quick-frozen in Dry Ice and ethanol and stored at -80” until assayed. All samples from a single infusion were thawed and assayed at the same time.

serum diluted l/20 in Verona1 buffer, and 0.1 ml of 0.035 M

CaClz containing rabbit brain cephalin. The sample was then incubated for 8 min. At this time, 0.1 ml of the incubation mixture was added to 0.1 ml of 0.035 M CaCl2, followed by 0.1 ml of normal human plasma. The clotting time was then measured.

For assay of the dogs’ intrinsic Factor VIII levels, standard curves were made using dilutions of normal dog plasma. The hemophilic dogs all had basal levels of 3.5 to 5.0% Factor VIII by this assay. For survival assays of bovine or human purified Factor VIII, standard curves using dilutions of normal bovine plasma or normal human plasma were used to determiue the activity of these substances in the dogs’ plasma.

Purijkation of Human Factor VIII-The preparation of purified human Factor VIII was performed with Blood Bank cryoprecipitate as the starting material (42). In a typical experiment, the contents of each of 10 plastic bags of Factor VIII cryoprecipitate were dissolved individually in 5 ml of 0.05 M

citrate and 0.5% e-amino-n-caproic acid, pH 6.8, at 37” and transferred to a 40.ml polystyrene centrifuge tube containing 1 ml of a 307. Al(OH)z solution. Each bag was rinsed with an additional 5 ml of citrate buffer and the wash solution transferred to the corresponding centrifuge tube. After an initial mixing with the Al(OH)s, the contents of each centrifuge tube were combined in a polypropylene beaker and stirred for an additional 15 min at room temperature. The Al(OH)3 was subsequently removed by centrifugation at 16,000 x g for 10 min at 22”. The supernatant (approximately 140 ml) was then subjected to chromat’ography at room temperature on a calcium citrate-cellulose column. This column was prepared by pouring a slurry of 150 g of calcium citrate prewashed with 1 liter of H20 and 150 g of cellulose powder suspended in 1 liter of 0.05 1~ citrate and 0.5% e-amino-n-caproic acid, pH 6.8, into a polypropylene column (7.5 X 30 cm). The cryoprecipitate adsorbed with Al(OH)x was then allowed to enter the column, followed by extensive washing with citrate buffer until the effluent had an absorbance of less than 0.05 at 280 nm. This is accomplished with approximately 1.8 to 2.0 liters.

After the initial citrate wash, the column was transferred to the cold room at 4” and Factor VIII activity was eluted with a solution containing 0.1 112 ethylenedinitrilo-tetraacetic acid (tetrasodium salt), 0.2 M Tris-HCl, and 0.5% e-amino-n-caproic acid, pH 8.5, at a flow rate of 8 to 10 ml per min. The flow rate was reduced to 2.5 ml per min after 400 ml of effluent had passed through the column, and 25.ml fractions were then collected.

Factor VIII from the hemophilic dogs was assayed by a two- stage method which does not require Factor VIII-deficient plasma as a substrate. This procedure is a modification of that previously described by Pitney (38) and Biggs el al. (39). In this method, 0.1 ml of a l/20 dilution of bovine serum was used in place of hemophilic plasma adsorbed with Al(OH) 3 as a source of Factor V. Rabbit brain cephalin prepared by the method of Bell and Alton (40) was substituted for plat,elets as a phospho- lipid source. Dilutions of plasma were made in 0.03 XI Verona1 buffer, pH 7.35 (41). The assay mixture consisted of 0.1 ml each of the samples which had been adsorbed for 10 min at 37” with l/10 volume of Al( suspension (Amphojel), 0.1 ml of

Fractions containing the majority of the Factor VIII activity (-220 ml) were mixed with ?{a volume of 40% polyethylene glycol in 0.25 M NaCl and 0.2 M imidazole, pH 6.5. This solution was then placed in an ice-water bath for 10 min after which the precipitate was centrifuged at 4” and 7000 x g for 10 min. The supernatant was then discarded and the precipitat,e dissolved in 15 to 20 ml of 0.2 M KC1 and 0.02 M Tris-HCl, pH 7.4, at 37”. Further concentration of the Factor VIII was accomplished by the addition of 1.8 ml of concanavalin A (25 mg per ml dissolved in 1.0 M NaCl). The solution was mixed and then allowed to stand at room temperature for 10 min. The flocculent precipitate was centrifuged at room temperature for 5 min at 2000 x g and then dissolved at 37” in 2 ml of 0.25 M NaCl-0.02 M imidazole (pH 6.5) containing lO70 methyl-Lu-u-glucopyranoside (B grade) and 0.03% sodium azide and allowed to stand overnight at 4”.

The final step in the purification was passage of the concanavalin A-glycoprotein solution (2.5 to 3 ml) through a column (1.5 x 100 cm) of A15m agarose pre-equilibrated with the same

normal dog serum diluted l/10 in Verona1 buffer. 0.1 ml of bovine methyl-ol-n-glucopyranoside buffer solution as above. Elution


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

3949

of purified Factor VIII was performed at a rate of 15 ml per hour, and 3-ml samples were collected after 40 ml had passed through the column. The contents of the first four tubes, containing approximately 75F0 of the Factor VIII activity, were combined and the Factor VIII was precipitated with $io volume of 40% polyethylene glycol (15 min at 0”) followed by centrifugation at 16,000 x g for 10 min at 4’. The Factor VIII precipitate was dissolved in 2.0 ml of 0.01 M Tris-HCl, pH 7.4, 0.15 M NaCI, and 0.03 M sodium azide and stored at 4”. Under these conditions, human Factor VIII activity was stable for about 2 weeks; after 4 weeks, the activity drops approximately 50%. When quick- frozen and stored at -8O”, the stability was greatly enhanced.

The yield from the agarose gel filtration step improves mark- edly after the column has been used three or four times. The column call then be used another 25 to 30 times if it is washed with 600 to 800 ml of the imidazole buffer containing methyl- oc-n-glucopyranoside prior to reutilization. The yield for this step can also be increased by including those tubes on the ascend- ing portion of the first peak (see lower panel, Fig. 1). These preparations, however, contain trace amounts of fibrinogen which can be detected by disc gel electrophoresis. In some Factor VIII preparations, all of the tubes in the first peak which contained Factor VIII activit,y were combined and the fibrinogen in these preparations was removed by passing the solution through an agarose column to which human fibrinogen antibody was co- valently bound (43, 44).

In more recent experiments, the Al(OH)z adsorbed cryoprecipitate solution was stirred for 15 min at 20” with 150 g of prewashed calcium citrate in 250 ml of 0.05 M citrate buffer. The slurry was then centrifuged for 5 min at 7000 x g at 16” and the calcium citrate then washed two more times with 500 ml of citrate buffer and centrifuged at 4”. The calcium citrate precipitate was then mixed with 1 liter of citrate buffer containing 150 g of cellulose powder and the mixture placed in a polypropylene column (7.5 x 30 cm) at 4”. The column was allowed to set,tle and the Factor VIII was eluted as described above. This modification speeds up the purification procedure by 1 to 2 hours.

RESULTS

Preparation of Human Factor VIII-The purification steps for

human Factor VIII and the yield for a typical preparation are shown iu Table I. The cryoprecipitate employed contained Factor VIII purified approximately lo- to 18fold with a recovery

of about 30%. The starting material varied, however, from 5- to 20.fold in purification and 25 to 45% in yield from plasma. Residual amounts of Factors VII, IX, and X and prothrombin activity were removed from the cryoprecipitate by an AI

TABLE I

Purf,ficalion o.f human Factor VIII from cryoprecipitate

The st,arting material employed was the cryoprecipitate prepared at’ t’he King County Blood Bank from 10 individual donors. This is equivalent to approximately 2600 ml of plasma. -.

Purification step Total protein

Cryoprecipitnte 11,600 Al(OH), 9,990 Calcium citrate-cellulose. 96 Concanavalin A-agarose

filtration. 5

- ) _-

Total units Yield

___~ unrls/mg

750 0.065 750 0.075 100 260 2.5 34.7

170 134 22.6

-

-

Purifi- cation

1 1.1

38.5

523

adsorption step, and the protein was then subjected to calcium citrate-cellulose column chromatography (top panel, Fig. 1). A major portion of the protein applied to the column appeared in the solvent front and contained approximately 10% of the Factor VIII activity, while the adsorbed Factor VIII was eluted in the second protein peak. The final step involved precipitation of Factor VIII by concanavalin A and subsequent gel filtration on agarose A-15m (lower panel, Fig. 1). Factor VIII activity appeared in the breakthrough peak from the agarose column which separated it from concanavalin A and the remain- ing traces of fibrinogen as well as other contaminating proteins. Only a portion of this peak, as indicated by the bar, was pooled and concentrated by polyethylene glycol precipitation. The purification of human Factor VIII from cryoprecipitate was approximately 500.fold. This is equivalent to an over-all purification of approximately 3000. to 5000-fold from plasma. Human Factor VIII prepared by this method was free of fibrinogen, prothrombin, Factors V, VII, IX, X, and XIII.

Zone Electrophoresis-Human Factor VIII was subjected to electrophoresis at pH 8.6 employing microscope slides layered with agarose and nonpolymerized acrylamide (Fig. 2). After electrophoresis for 45 min, a single spot was obtained for Factor VIII (Sample 1). Following reduction with 2-mercaptoethanol in 8 M urea (Sample 2) or reduction followed by alkylation with iodoacetate (Sample 3), single spots were also obtained. As expected, the addition of carboxymethyl groups by alkylation

4.0

3.6

3.0 2.8

0.6

4

0.2

55 75 95 II5 135 155 Column Effluent lmN

FIG. 1. Calcium citrate-cellulose column chromatography of an Al(OH), adsorbed cryoprecipitate (top panel) and gel filtration of partially purified human Factor VIII on agarose A-15m (bottonz panel). The adsorbed cryoprecipitate (140 ml in 0.05 M sodium citrate, 0.59;;: t-amino-n-caproic acid, pH 6.5) was applied to the calcium citrate-cellulose column and eluted as described under “Methods.” Similarly, 2 to 3 ml of the concanavalin A precipitated Factor VIII were eluted from the agarose A-15m column with 0.25 ?M NaCl-0.02 M imidazole buffer, pH 6.5, containing 10yG methyl-a-n-glucopyranoside and 0.03 M sodium azide. O-O, protein concentration as determined by absorbance at 280 nm; l - - -0, Factor VIII activity.


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

3950

1 \

anode 2 cathode

3

T starting well

FIG. 2. Zone electrophoresis of human Factor VIII, reduced Factor VIII, and reduced and alkylated Factor VIII. Factor VIII (10 ~1, containing 10 to 20 fig of protein) was added to each of the wells and electrophoresis carried out as described under “Methods.” Sample 1 is native Factor VIII, Sample 2 is reduced Factor VIII, and Sample 3 is reduced and alkylated Factor VIII.

I

2

6

8

FIG. 3. Immunoelectrophoresis of highly purified human Factor VIII, partially purified human Factor VIII, and highly purified human fibrinogen. Immunoelectrophoresis was carried out for 75 min as described under “Methods.” The antibodies employed were also prepared as described under “Methods.” Various protein solutions were placed into the appropriate well, and electrophoresis performed as described under “Methods.” Upo,n completion of the electrophoresis, 50 ~1 of the appropriate rabbit antibody solution were applied to the center trough. Wells 1 and 2 represent two different Factor VIII preparations (10 ~1 containing 10 to 15 pg of protein) which were precipitated by a rabbit antibody to highly purified Factor VIII and stained for protein with Coomassie blue. Well S contained 15 pg of highly purified Factor VIII, while Well 4 contained 25 fig of protein from the calcium citrate column, and the center trough contained the same antibody as employed in the top slide. Well 5 contained 40 pg of protein from the calcium citrate column, while Well 6 contained 20 pg of highly purified Factor VIII and the center trough contained rabbit antibody against those proteins present in the calcium citrate eluate. Well 7 contained 45 pg of 99% clottable human fibrinogen, and Well 8 contained 20 rg of highly purified human Factor VIII. The center trough contained an antibody prepared against partially purified Factor VIII. The anode was at the right in all slides and protein migration was from left to right.

increases the mobility of the Factor VIII relative to the reduced or native Factor VIII.

Immunoelectrophoresis-As another criterion for purity, human Factor VIII was subjected to immunoelectrophoresis on agarose slides (Fig. 3). The antibody employed in these experiments was prepared in rabbits against highly purified Factor VIII and a partially purified Factor VIII preparation which contained fibrinogen. These antibodies, as shown later, readily inhibit Factor VIII activity. In Samples 1 and 2, the highly purified Factor VIII was subjected to electrophoresis followed by im- munodiffusion against rabbit antibody to highly purified Factor VIII. After staining with Coomassie blue, single sharp precipitin lines were obtained.

The antibody prepared against highly purified Factor VIII also yields a single precipitin line with partially purified Factor VIII. The pattern shown in Sample 3 contains highly purified Factor VIII, while Sample 4 contains partially purified Factor VIII. In Sample 5, a partially purified Factor VIII preparation as well as an antibody against partially purified Factor VIII were employed. Many precipitin lines are obtained with the partially purified preparation of Factor VIII; however, the antibody yields only a single precipitin line with the highly purified Factor VIII (Sample 6), indicating removal of many contaminating proteins during the final purification. Sample 7 shows the precipitin line for fibrinogen with an antibody to partially purified Factor VIII, while Sample 8 contains only the highly purified Factor VIII. The precipitin lines shown in these experiments were identical with those observed by indirect lighting, indicating that the staining procedure was detecting all of the precipitin lines present.

Polyacrylamide Gel Electrophoresis of Human Factor VIII- Human Factor VIII, like bovine Factor VIII (5), did not enter a 3.25% polyacrylamide gel even in the presence of 8 M urea. After reduction, a single protein band was obtained in 5% polyacrylamide gels in the presence of SDS. These experiments, shown in Fig. 4, were performed with 20, 30, and 50 pg samples of protein (Samples 1, 2, and 3, respectively). Sample 4 contained an equal amount of reduced human Factor VIII and a highly purified sample of reduced bovine Factor VIII. Sample 5 contained only reduced bovine Factor VIII. The results of these experiments support the idea that human Factor VIII is made up of subunits which are similar or perhaps identical, and that these subunits are held together by disullide bonds. Fur- thermore, the subunits in human as well as bovine Factor VIII appear to be essentially the same size.

The subunits in human Factor VIII contain carbohydrate, as shown in the SDS polyacrylamide gel electrophoresis patterns in Fig. 5. In these experiments, the protein was reduced with 0.1 M 2-mercaptoethanol. The gel on the right was stained for protein with Coomassie blue, and the gel on the left for carbohydrate with the periodic acid-Schiff reagent. A single protein and a single carbohydrate band running in the same position was obtained for each sample. These data indicate that human Factor VIII, like bovine Factor VIII, is composed of glycoprotein subunits.

The subunits of human Factor VIII migrate slower in SDS gel electrophoresis than the (Y, fl, and y subunits of fibrinogen, which have molecular weights of 47,000, 56,000, and 70,900, respectively (45). The molecular weight estimated for reduced Factor VIII by the SDS method was 240,000.

Ultracentrijugation of Human Factor VIII-Sedimentation equilibrium studies of human Factor VIII indicated that the protein was homogeneous at all protein concentrations tested. The molecular weight was calculated to be 1.12 million i 98,000


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

3951

methionine, tyrosine, and tryptophan content. Neither human nor bovine Factor VIII was found to contain free sulfhydryl groups.

The carbohydrate compositions for human Factor VIII and that previously determined for bovine Factor VIII (5) are pre- sented in Table IV. The human preparation contains approximately 6% carbohydrate, including 2.2% hexose, 2.7% hexosamine, and 0.9% neuraminic acid. In contrast, bovine Factor VIII contains approximately 20% carbohydrate. Also, there are approximately equal amounts of hexose and hexosamine in human Factor VIII, while bovine Factor VIII contains approximately twice as much hexose as hexosamine.

Inactivation of Human and Bovine Factor VIII by Antibodies Prepared in Rabbits against These Two Proteins-In previous experiments, we have shown the inhibition of bovine Factor VIII by antibodies prepared in rabbits against this protein (5). Similar results were obtained with antibodies prepared in rabbits against human Factor VIII (Fig. 6). In these experiments, the human or bovine Factor VIII was preincubated for 30 min at 37” with various concentrations of the antibody. The mixture was

(Table II). When the protein was reduced with 0.04 M dithiothreitol, the preparation was heterogeneous. The molecular weight of the smallest species was 105,000 f 5,000, a value similar to that obtained for bovine Factor VIII (5). Considera- ble aggregation of the Factor VIII subunits was probably occur- ring since this preparation showed only a single band on disc gel electrophoresis and zone electrophoresis.

Chemical Composition of Human Factor VIII-The amino acid composition for human Factor VIII is shown in Table III along with the amino acid composition for bovine Factor VIII (5). It is clear that the amino acid contents of the two proteins are considerably different. Both, however, have a relatively low

FIG. 4. SDS polyacrylamide gel electrophoresis of highly purified reduced human and reduced bovine Factor VIII. From left to right, Gels 1,2, and S contained 20,30, and 50 pg of purified human Factor VIII, respectively. Gel Q represents 50 pg of a mixture of equal amounts of human and bovine Factor VIII, while Gel 6 contained 50 pg of bovine Factor VIII alone. The protein on these gels was stained with Coomassie blue. The anode was at the bottom of the gel.

FIG. 5. SDS polyacrylamide gel electrophoresis of reduced human Factor VIII stained for protein or carbohydrate. Both gels contain 75 pg of Factor VIII. The gel on Ihe left was stained for carbohydrate, while the gel on the right was stained for protein as described under “Methods.” The anode was at the bottom of the gel.

TABLE II

Molecular weight of human Factor VIZI by sedimentation equilibrium

rPm

Denaturedc . . . . Denatured and reduced in

dithiothreitold . .

5,090

14,000

i

0.733

0.733

Molecular weights”

5 These values represent an average of the values obtained for the whole cell average molecular weights at the three different protein concentrations employed as described under “Methods.”

b Mi refers to the smallest species which could be calculated by the methods described by Teller et al. (26) and Teller (27). The uncertainties are weighted root mean square values in this column, but standard deviations for all other values.

c Buffer: 0.04 M Tris-HCl, 6.0 M guanidine HCI, pH 7.5. d Buffer: same as c with the addition of 0.04 M dithiothreitol.


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

3952

TABLE III Axino acid composition of human and bovine Factor VIII

Amino acid

Lysine ........................

Histidine ..................... Arginine ..................... Aspartic acid. ................ Threonine ................... Serine ........................ Glutamic acid. ............... Proline ....................... Glycine ..................... Alanine ...................... Half-cystineb ................ Valine ........................ Methionine. .................. Isoleucine .................... Leucine ...................... Tyi osine ..................... Phenylslanine ................ Tryptophanc ..................

Ii Hu---n Factor VIII Bovine Factor VIII

39.9 53.8 22.5 22.3 39.9 45.5 85.4 85.1 60.1 43.8 70.8 57.5 89.9 84.4 61.2 50.2 62.9 68.5 43.8 49.0 65.7 57.6 75.8 99.3

8.4 11.3 33.7 36.4 65.2 90.3 20.2 16.0 27.0 26.9 24.7 17.5

Q The amino acid composition is expressed in residues per 165 g of protein (not glycoprotein).

b Iletermined as cpsteic acid. c I>etermined by the spectrophotometric assay.

TABLE IV Carbohydrate composition of human and bovine Factor VIII

Carbohydrate Human Factor VIIT Bovine Factor VIII

midues/ 100,000 ga 70 residues/

100,000 g %

Hexose 12.2 2.2 61.6 11.1 Hexosamine 15.0 2.7 32.2 6.8 Neruaminic acid. 2.7 0.9 5.7 1.7

Total. 5.8 19.6

a The carbohydrate content is expressed in residues per 165 g of glycoprotein (not protein).

then diluted 300. to 500-fold and an aliquot assayed in Factor VIII-deficient plasma. The solid curue on the left is the neutralization of human Factor VIII by antibody prepared against highly purified human Factor VIII, and the solid curve on the right is the neutralization of human Factor VIII by antibody prepared against bovine Factor VIII. These experiments indicate that the antibodies prepared against either human or bovine Factor VIII inactivate human Factor VIII. Approximately 20 times more antibody to bovine Factor VIII, however, was required to cause 50y0 inhibition of human Factor VIII than that required to cause the same inhibition with bovine Factor VIII (dashed curve on left). Similarly, antibodies to either human or bovine Factor VIII neutralize bovine Factor VIII. These experiments are represented by the dashed curves. The dashed curve on the left represents bovine Factor VIII versus antibody to bovine Factor VIII, and the dashed curve on the right is bovine Factor VIII versus antibody to human Factor VIII. From these experiments, one can conclude that bovine Factor VIII was in- hibited by antibodies to both human and bovine Factor VIII. Furthermore, approximately 20 times more antibody to human Factor VIII was required to neutralize bovine Factor VIII than that required for human Factor VIII. Thus, the antibodies pre-

80-

70-

2 I$ 60-

E “, 50- .2 -? $j 40-

k 30 -

20 -

IO-

I -=h., --a.- I I I I -0.8 -0.4 0 0.4 0.8 1.2 1.6 2.0

Log Anfibody Cont.

FIG. 6. Neutralization of human and bovine Factor VIII activity with either rabbit anti-human or anti-bovine Factor VIII antibodies. The two curve.s on the left represent the homologous system; i.e. human Factor VIII versus antibody to human Factor VIII (O--O ), and bovine Factor VIII versus antibody to bovine Factor VIII (O- - -0). Conversely, the two curves 012 the right represent the heterologous system; i.e. human Factor VIII versus antibody to bovine Factor VIII (O-O ), and bovine Factor VIII UeTsus antibody to human Factor VIII (O- - -0 ). Samples were incubated and assayed as described in the text and under “Methods.”

pared against Factor VIII are approximately 20 times more potent as inhibitors toward the homologous protein as compared to the heterologous molecule.

Immunoelectrophoresis of Cryoprecipitate from Normal and Herrwphilic Plasma-With an antibody prepared in rabbits from a highly purified human Factor VIII, it was possible to test for the presence of cross-reactive material in the cryoprecipitate of an individual with classic hemophilia. Immunoelectrophoresis experiments employing highly purified Factor VIII and a cryoprecipitate from pooled normal plasma and from Factor VIII- deficient plasma are shown in Fig. 7. Single sharp precipitin lines were obtained with pure Factor VIII (Samples 1 and 2), normal cryoprecipitate (Sample 3), and cryoprecipitate from a patient with classic hemophilia (Sample 4).

Turnover of Factor VIII in Hemophilic Dogs-The final proof of identity of a Factor VIII preparation is evidence that it will elevate the plasma level of Factor VIII in an individual with classic hemophilia. For our initial studies, hemophilic dogs were employed since human or bovine Factor VIII readily corrects hemophilic dog plasma in vi&o. Highly purified human Factor VIII (280 units) was given by intravenous injection, and samples were removed from the hemophilic dogs during the next 18 hours. The initial recovery of Factor VIII was approximately 56% of the infused Factor VIII assuming a blood volume of 90 ml per kg of body weight (46). The half-life in viva for human Factor VIII was approximately 4 hours. In similar experiments with highly purified bovine Factor VIII (360 units), the initial recovery was approximately 557, and the half-life in vivo was 9 hours. It is


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

4

FIG. 7. Immunoelectrophoresis pattern of highly purified Factor VIII and the cryoprecipitate prepared from normal and hemophilic plasma. Immunoelectrophoresis was carried out for 45 min and antibodies to the highly purified Factor VIII were employed. Wells 1 and B contained highly purified Factor VIII. Well 3 contained approximately 500 pg of cryoprecipitate prepared from normal plasma, and Well .J contained approximately 500 erg of cryoprecipitate prepared from hemophilic plasma. The anode was at the right and protein migration was from left to right.

1 160

140 t

I 1 I

IO 20 30 40 50 60 Time fminl

FIG. 8. Thrombin modification of human Factor VIII. Factor VIII (25 units per ml) was incubated at 4” in 0.15 M NaCl-0.01 M Tris-HCl, pH 7.4, in the presence (A--A) and absence (O-----O ) of bovine thrombin. The thrombin concentration employed in this experiment was 0.1 unit per ml. At various times, aliquots were removed, diluted, and assayed for Factor VIII activity as described under “Methods.”

clear that either human or bovine Factor VIII corrects the clotting defect in the plasma of hemophilic dogs, and the isolation of these proteins has not resulted in their partial degradation leading to an abnormally rapid clearance. A half-life of 4 to 9 hours of human or bovine Factor VIII in hemophilic dogs is similar to that obtained for human Factor VIII in patients with classic hemophilia (47, 48).

3953

Thrombin Modi$cation of Human Factor VIII-Numerous investigators have observed a large increase in the apparent activity of Factor VIII in the presence of trace amounts of thrombin (49-52). Similar results were obtained in the present experiments with highly purified bovine thrombin and human Factor VIII. In the presence of thrombin (0.1 unit per ml), the apparent Factor VIII activity increased approximately 16.fold after incubation for 20 min at 4” (upper curse, Fig. 8). In these experiments, the weight ratio of human Factor VIII to thrombin was 100: 1. Modification of human Factor VIII was also obtained with lower levels of thrombin (0.01 to 0.05 unit per ml), although the increase in specific activity of Factor VIII was substantially lower under these conditions.

The thrombin-modified human Factor VIII was unstable, however, and its biological activity fell during prolonged incubation at 4’. This is in contrast to bovine Factor VIII which is stable for many hours after thrombin modification.a

The molecular basis for the thrombin modification has not been established. No difference in the electrophoretic mobility of native and thrombin-modified Factor VIII was observed by immunoelectrophoresis experiments. Also, in SDS polyacrylamide gel electrophoresis, no change was observed in the protein pattern for the Factor VIII subunits before and after thrombin activation. These results suggest that only very minor proteol- ysis of the Factor VIII molecule has occurred during the thrombin modification reaction.

DISCUSSION

Human Factor VIII, like bovine Factor VIII, is a high molecular weight glycoprotein. The carbohydrate level in human Factor VIII is approximately 6%, which is substantially lower than that found in bovine Factor VIII (Table IV). The carbohydrate content of human Factor VIII found in the present experiments is very similar to that recently reported by Marchesi et al. (7) and by McKee (53), but it is lower than the value of 10% reported by Hershgold et al. (6). In preliminary experiments, it was found that human Factor VIII prepared by the present method contains approximately 5% lipid. Other investigators have also noted the presence of lipid in human Factor VIII (6, 7). The lipid may be a contaminant, however, since Ratnoff et al. (54) found that the lipid content of human Factor VIII varied depending on the chylomicron content of the starting plasma. No lipid was found in the highly purified bovine Factor VIII (5).

The amino acid compositions of bovine and human Factor VIII are different as shown in Table III. Some immunological cross- reactivity, however, exists between these two proteins. Anti- bodies prepared against each of these two highly purified proteins will inactivate Factor VIII from either human or bovine sources. The potency of the respective antibodies against the homologous protein, however, is much greater than that toward the heterologous protein. These experiments suggest that there may be large regions in the human and bovine Factor VIII molecule which differ, but active site regions important for biological activity may be very similar or perhaps identical in both proteins. Cross-reactivity of human or bovine Factor VIII with anti- serum to crude bovine or human Factor VIII has also been studied by Denson (55) who found similar results.

With an antibody prepared against highly purified human Factor VIII, it was also possible to demonstrate the presence of an inactive cross-reactive Factor VIII molecule present in the

* A. R. Thompson, E. P. Kirby, K. Fujikawa, and E. W. Davie, unpublished results.


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

3954

cryoprecipitatc of an individual with classic hemophilia. These experiments, which have been carried out with only one individual with classic hemophilia, confirm the results published by other investigators (5660) who have reported t,he presence of an abnormal Factor VIII protein in individuals with less than 1% Factor VIII activity. The present data, however, are the first to employ an antibody prepared against highly purified Factor VIII with well defined physicochemical properties. These data are consistent with the hypothesis that an abnormal Factor VIII molecule is synthesized in individuals with classic hemophilia. Furthermore, this prot.ein is probably inactive due to an amino acid replacement in some essential portion of the molecule. It also appears probable that several classes of abnormal Factor VIII molecules will be found in individuals with classic hemophilia, but, further investigations will be required to test this possibility.

The molecular weight of human Factor VIII, determined by sedimentation equilibrium experiments, was 1.12 million f 98,000. In these experiments, the intact protein was homogeneous at all concentrations tested. This value is very similar to that found for the bovine Factor VIII, although this protein was found to be heterogeneous in sedimentation equilibrium experiments (5). In sedimentation velocity experiments, it was found that human Factor VIII has an szo,w value of 23.7 f 0.5 when examined in 0.01 M Tris-HCl, pH 7.4, 0.15 M NaCl, and 0.05 M e-amino-n-caproic acid. This value is consistent with that expected for a large macromolecule. In the present studies, there were no indications of active Factor VIII molecules of smaller molecular weight. Similar results were previously obtained for bovine Factor VIII (5).

The molecular weight of the subunit of human Factor VIII remains unclear. Sedimentation equilibrium studies of the subunits formed in the presence of dithiothreitol indicate that the preparation was heterogeneous. The molecular weight of the smallest species was calculated as 105,000 + 5,000. It appears likely that considerable aggregation of similar or identical subunits was taking place during these studies, however, since the reduced Factor VIII gave only one band in polyacrylamide gel elcctrophoresis and zone electrophoresis. The molecular weight determined by SDS gel electrophoresis is very large compared to that calculated from sedimentation equilibrium experiments. Similar results have been obtained by others (7, 53). The molecular weight determined by SDS gel electrophoresis may be high, however, due to t.he presence of carbohydrate in Factor VIII. High molecular weights of glycoproteins have been noted by others (61, 62) who have found an anomalous electrophoretic behavior of the glycoproteins isolated from erythrocyte mem- branes.

Highly purified human Factor VIII, like bovine Factor VIII,2 readily undergoes thrombin activation. Whether thrombin modification of Factor VIII and perhaps Factor V is related to a hypercoagulable state is unclear at the present time. Thrombin modification of Factor VIII in viva is most likely under strict regulatory control which may be abnormal in patients suffering from hypercoagulability. The molecular events associated with the thrombin modification of Factor VIII are presently under investigation in our laboratory.

Acknowledgments-We wish to express our sincere thanks to Drs. Robert Meyer and Kazuo Fujikawa for valuable discus- sions, to Dr. David Teller for his assistance in the sedimentation

equilibrium experiments, and to Charles Nicholas and Richard Granberg for excellent technical assistance. We are also in- debted to the King County Blood Bank and Dr. Dennis Donohue for the human cryoprecipitate which made these studies possible.

REFERENCES

1. SCHMER, G., LEGAZ, M. E., AND DAVIE, E. W. (1972) in Inter- national Society of Thrombosis and Haemostasis Abstracts, p. 24

2. LI.:GAZ, M. E., SCHMER, G., AND DAVIE, E. W. (1972) Circula- tion (Suppl. 2) 46, 53

3. WRIGHT, I. (1959) J. Amer. Med. Ass. 170, 325 4. DAVIE, E. W., HOUGIE, C., AND LUNDBLAD, R. L. (1969) in

Recent Advances in Blood Coaaulation (POLLER. L., ed) pp. 13-28, J. & A. Churchill, Lth., London ’

5. SCHMER, G., KIRBY, E. P., TF,LLICR, D. C., AND DAVIE, E. W. (1972) J. Biol. Chem. 247, 2512-2521

6. HISRSHGOLD, E. J., DAVISON, A.M., AND JANSZF,N, M. E. (1971) J. Lab. Clin. Med. 77, 185-205

7. MARCHESI, S. L., SHULMAN, N. R., AND GRALNICI~, H. 1~. (1972) J. Clin. Invest. 51, 2151-2161

8. THOMPSON, A. R., AND DAVIE, E. W. (1971) Biochim. Biophys. Acta 250, 210-215

9.

10. 11.

12.

13.

AGRAWAL, B. B. L., AND GOLDSTEIN, I. J. (1965) Biochem. J. 96, 23C-25C

MOORE, S. (19G8) J. Biol. Chem. 243, 6281-6283 LOWRY, 0. H., ROSEBROUGH, N. J., FARR, A. L., AND RANDALL,

R. J. (1951) J. BioZ. Chem. 193, 265-275 DUBOIS, M., GILLES, K. A., HAMILTON, J. K., RI~:I~J<RS, P. A.,

AND SMITH, F. (1956) Anal. Chem. 28, 350-356 GAIZDELL, S. (1957) in Methods of Biochemical Analysis (GLICK,

D., ed) Vol. VI, pp. 289-317, Interscience Pubhshers, Inc., New York

14. 15.

16.

17.

WARR~:N, L. (1959) J. BioZ. Chem. 234, 1971-1975 WEBER, K., AND OSBORN, M. (1969) J. BioZ. Chem. 244, 4406-

4412 ZACHARIUS, Ii. M., ZELL, T. E., MORRISON, J. H., AND WOOD-

LOCK, J. J. (1909) Anal. Biochem. 30, 148-152 WILLIAMS, C. A., A4~~ Chasms, ;I/. W. (1971) Methods in Zm-

munology and Immunochemistry, Vol. III, p. 103, Academic Press, New York

18. SCHEIDEGGI:R, J. J. (1955) Znt. Arch. Allergy Appl. Zmmunol. 7, 103~110

19. MOORE, S., AXD STTXN, W. H. (1963) Methods Enzymol. 6, 819-831

20. SPXKMAN, I>. H., STEIN, W. H., AND MOORE, 8. (1958) Anal. Chem. 30, 1190-1206

21. 22.

HIRS, C. H. W. (1967) Methods Enzymol. 11, 59-62 BENCZE, W. L., AXD SCHMID, K. (1957) Anal. Chem. 29, 1193-

1196 23. 24.

25.

ELLMAN, G. L. (1959) Arch. Biochem. Biophys. 82, 70-77 HARRIS, C. I+:., KOBES, 1~. D., TIGLLER, D. C., AND RUTTIER,

W. J. (1969) Biochemistry 8, 2442-2454 S~ERY, V. L., FISCHER, 71:. H., AND TELLER, D. C. (1970)

Biochemistry 9, 3591-35!)8 26.

27. 28. 29.

TI~LLXR, D. C., HORBETT, T. A., RICHARDS, E. G., AND SCI-IACH- MAN, H. K. (1969) Ann. N. Y. Acad. Sci. 164, G6-101

TELLER, D. C. (1972) Methods Enzymol. 26C, in press YPHANTIS, D. A. (1964) Biochemistry 3, 297-317 DBROSIER, D. J., MUNK, P., AND Cox, D. J. (1972) Anal.

Biochem. 60, 139-153 30.

31. 32.

33.

34.

MCMEEKIN, T. L., GROVES, M. L., AND HIPP, N. J. (1949) J. Amer. Chem. Sot. 71, 3298-3306

LONGSWORTH, L. G. (1953) J. Amer. Chem. Sot. 76, 5705-5709 FUJIKAWA, K., LEGAZ, M. E., AND DAVIE, E. W. (1972) Bio-

chemistry 11, 48824891 LANGDELL, R. D., WAGNXR, IL. H., AND BRINI~HOUS, K. M.

(1953) J. Lab. CZin. Med. 41, 637-647 PROCTOR, R. R., AND RAPaPORT, S. I. (1961) Amer. J. CZin.

Pathol. 36, 212-219 35. JACOBSSON, K. (1955) &and. J. CZin. Lab. Invest. (Suppl. 14)

7, 7-102 36. CLAUSS, V. A. (1957) Acta Haematol. 17, 237-246


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

37. BIGGS, R., AND MACFARLANE, R. G. (19G2) Human Blood Coagulation and Its Disorders, 3rd Ed., pp. 380-421, F. A. Davis Co., Philadelphia

38. PITNEY, W. R. (1956) Brit. J. Haematol. 2, 250-264 39. BIGGS, R., EVELING, J., AND RICHARDS, G. (1955) Brit. J.

Haematol. 1, 20-34 40. BELL, W. N., AND ALTON, H. G. (1954) Nature 174, 880-881 41. OWREN, P. A. (1949) &and. J. Clin. Lab. Invest. 1, 81-83 42. POOL, J. G., AND SHANNON, A. E. (1965) Xew Engl. J. Med. 273,

1443-1447 43. PORATH, J., AXEN, R., AND ERNBACK, S. (1967) Nature 216,

1491-1492 44. CUATRECASAS, P. (1970) J. Biol. Chem. 245, 3059-3065 45. MCKEE, P. A., ROGERS, L. A., MARLER, E., AND HILL, R. L.

(1966) Arch. Biochem. Biophys. 116, 271-279 46. ALTMAN, P. L. (1961) in Blood and Other Body Fluids (DITTMER,

D. S., ed) p. 4, Federation of the American Society of Ex- perimental Biology, Washington, D. C.

47. BIGGS, R., AND DENSON, K. W. E. (1963) Brit. J. Haematol. 9, 532-547

48. POOL, J. G., COHEN, T., AND CREGER, W. P. (1967) Brit. J. Haematol. 13, 822-828

49. RAPAPORT, S. I., SCHIFFMAN, S., P.~TcH, M. J., AND AMES, S. B. (1963) Blood 21, 221-236

50. RAPAPORT, S. I., HJORT, P. F., AND P-ITCH, M. J. (1965) Sand. J. Clin. Lab. Invest. (Suppl. 84) 17, 88-100

51. ~ZGE-ANWAR, A. H., CONNELL, G. E., AND MUSTARD, J. F. (1965) Blood 26, 500-509

52. BIGGS, R., MACFARLBNE, R. G., DENSON, K. W. E., AND ASH, B. J. (1965) Brit. J. Haematol. 11, 276

53. MCKEE, P. A. (1970) Fed. Proc. 29, 647Abs 54. RATNOFF, 0. D., KASS, L., AND LANG, P. D. (1969) J. Clin.

Invest. 43, 957-962 55. DENSON, K. W. E. (1967) The Use of Antibodies in the Study

of Blood Coagulation, F. A. Davis Co., Philadelphia 56. ZIMMERMAN, T. S., RATNOFF, 0. D., AND POWELL, A. E. (1971)

J. C&in. Invest. 60, 244-254 57. BENNETT, E., AND HUEHNS, E. R. (1970) Lancet 2, 956-958 58. SHAPIRO, G. A., AND McKm:, P. A. (1970) Clin. Res. 18, 615 59. STITES, D. P., HERSHGOLD, E. J., PERLMAN, J. D., AND FUDEN-

BERG, H. H. (1971) Science 171, 196-197 60. BOUMA, B. N., WIEGERINCK, Y., SIXMA, J. J., VAN MOURII~,

J. A., AND MOCHTAR, I. A. (1972) Nature New Biol. 236, 104- 106

61. SEGREST, J. P., JACKSON, R. L., ANDRIGWS, E. P., AND MAR- CHESI, V. T. (1971) Biochem. Biophys. Res. Commun. 44, 390-395

62. BRETSCHER, M. S. (1971) Nature New Biol. 231, 229-232


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Mark E. Legaz, Gottfried Schmer, Richard B. Counts and Earl W. DavieIsolation and Characterization of Human Factor VIII (Antihemophilic Factor)

1973, 248:3946-3955.J. Biol. Chem.

http://www.jbc.org/content/248/11/3946Access the most updated version of this article at

Alerts:

When a correction for this article is posted•

When this article is cited•

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

http://www.jbc.org/content/248/11/3946.full.html#ref-list-1

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


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/content/248/11/3946

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;248/11/3946&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/248/11/3946

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=248/11/3946&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/248/11/3946

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/248/11/3946.full.html#ref-list-1

http://www.jbc.org/

isolation and characterization of human factor viii ... journal of b~ologicli. chemistry vol. 218,...

Documents