PUREFICATIGN OF BOWNE BECOD CLONING FACTQR VHI iAM‘EHfiMGFHE‘LEC GLOEUiLIN) Thurs €09 Hm Dogma a? {352. D; MECBKGAN S'FATE UNWERSETY John Edward Mercer 1965 LIBRARY S [th15 - . Michigan 8 as: Univcrxit y WW... ~ This is to certify that the thesis entitled PURIFICATION OF BOVINE BLOOD CLOTTING FACTOR VIII (ANTIHEMOPHILIC GLOBULIN) presented by John Edward Merc er has been accepted towards fulfillment of the requirements for PhoDo degree in Ph281010gy 593/) " , ) _, "I 7, t/' - [JIJAiPA/L/f {I ,. Major professor Date February 15, 1966 0-169 i i ABSTRACT PURIFICATION OF BOVINE BLOOD CLOTTING FACTOR VIII (ANTIHEMOPHILIC GLOBULIN) by John Edward Mercer The purpose of this project was to study methods for the purification of bovine blood clotting Factor VIII that could be applied to purification of this factor for clinical use in man. The purification procedure required that conditions be estab- lished for the separation of Factor VIII from fibrinogen, resulting in minimal losses of Factor VIII activity. The Blomback procedure1 was modified, based on the original observations of Simonettiz, by the elimination of sodium citrate from the extraction buffers and remaining purification procedure and replacing it with sodium chloride of equal ionic strength to permit heat denaturation of fibrinogen at 56°C with minimal loss of Factor VIII activity. The modification of the procedure resulted not only in reduced heating losses of Factor VIII but also in changes in the protein distribu- tion, causing further increases in purification. Antisera prepared against partially purified Factor VIII showed that the fibrinogen-free Factor VIII preparations contained three immunologic components. Electrophoretic analysis on 7M urea-starch gel showed that three components were present in partially purified Factor VIII. Amino acid composition of several preparations showed John Edward Mercer 2 that preparations were quite uniform. Removal of a cryoglobulin fraction from the partially purified Factor VIII preparation resulted in minor losses of Factor VIII activity. Gel filtration of the preparation through G-ZOO Sephadex resulted in partial separ- ation of two components, with the peak of Factor VIII activity coincident with the major (slow) component. Immunodiffusion studies confirmed the presence of single components in the isolated fractions representing the individual peaks and the presence of two components in the fractions representing the mixed peaks. 1Blomback, M. 1958. Purification of Antihemophilic Globulin. I. Some Studies on the Stability of Antihemophilic Globulin Acti~ vity in Fraction 1-0 and a Method for its Separation from Fibrinogen. Arkiv. f. kemi, lg, 387. 2Simonetti, C., G. Casillas, and A. Pavlovsky. 1961. Purification du Facteur VIII Antihemophilique (FAH). Hemostase, l, 57. PURIFICATION OF BOVINE BLOOD CLOTTING FACTOR VIII (ANTIHEMOPHILIC GLOBULIN) By John Edward Mercer A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Physiology 1966 to my wife, Marilyn, for constant faith and encouragement ii ACKNOWLEDGEMENTS The author wishes to express sincere appreciation to Dr. E. P. Reineke of the Department of Physiology and Dr. H. A. Lillevik of the Department of Biochemistry, co-chairmen of his guidance committee for their advice and guidance in research and preparation of this thesis. Sincere appreciation is also due Dr. L. F. Wolterink, Dr. W. D. Collings and Dr. P. O. Fromm for their guidance and encouragement throughout the course of his studies. Acknowledgement is due the Michigan Department of Public Health for providing facilities, materials and technical assis- tance, and to Dr. L. A. Hyndman, Dr. H. Gallick, and other associates in the Division of Biologic Products for their helpful advice and encouragement. Special thanks is due Dr. B. H. Olson and Dr. G. L. Goerner of the Division of Antibiotic and Fermentation for performing the amino acid analysis. This investigation was financially supported by funds from the Michigan Department of Public Health, and the American National Red Cross. iii TABLE OF CONTENTS I. INTRODUCTION . II. HISTORICAL . III. MATERIALS AND METHODS A. Apparatus . B. Materials and Reagents 1. Chemical Reagents 2. Clotting Reagents C. Methods . . . . . . . . . 1. Collection of Bovine Blood Fractionation of Clarified Plasma Assay of Factor VIII Activity Preparation of Antisera Against Human Plasma Bovine Plasma and Bovine Factor VIII . Immunoelectrophoretic Analysis Immunodiffusion Techniques . Starch Gel Electrophoresis Urea Gel Electrophoresis Tyrosine-Tryptophane Ratios Amino Acid Analysis J-‘LJJN OKDGDVO‘UI 1 IV. EXPERIMENTAL . A. Modification of Blomback Procedure 1. First Modification, A Comparative Study . . 2. Complete Modification . . 3. Second Comparative Study of Mercer Modification B. Antisera . . . . 1. Preparation of Antigens 2. Immunization and Testing 3. Preparation of Anti- Bovine Factor VIII by Repository Immunization C. Immunoelectrophoretic Analysis of Bovine Factor VIII . . . . . . . . D. Gel Electrophoretic Studies . . . E. Gel Filtration of Bovine Factor VIII F. Tyrosine- Tryptophane Ratios iv Page 18 18 19 19 22 25 25 26 3O 31 34 36 39 39 4O 42 44 44 44 SO 50 55 55 S9 60 63 64 67 73 TABLE OF CONTENTS - Continued Page G. Amino Acid Analysis . . . . . . . . . 73 1. Glycine Removal . . . . . . . . 73 2. Amino Acid Analysis . . . . . . 76 V. DISCUSSION . . . . . . . . . . . . . . 83 A. Modification of the Blomback Procedure . . 83 B. Preparation of Antisera . . . . . . . 87 C. Immunoelectrophoretic Analysis . . . . . 88 D. Gel Electrophoresis . . . . . . . . . 89 E. Gel Filtration Studies . . . . . . . . 90 F. Cryoglobulin Studies . . . . . . . . 91 G. Amino Acid Analysis . . . . . . . 93 H. Tyrosine-Tryptophane Ratio and Content . . 94 I. Significance . . . . . . . . . . . 95 VI. SUMMARY . . . . . . . . . . . . . . . 97 REFERENCES . . . . . . . . . . . . . . . . 100 TABLE II. III. IV. VI. VII. VIII. XI. XII. XIII. LIST OF TABLES Plasma Clotting Factors Discovered Between the Years 1947 and 1955 Summary of Data Obtained from Four Fractionation Runs with the Modified Procedure (Mercer) Comparison Between Two Comparative Studies (Runs 1 and 6) . Comparison of the Blomback Procedure and the Modified Procedure (Mercer) Antibody Titers from Injection of Adjuvant Enriched Bovine Factor VIII Six Weeks After Injection Results of Immunoelectrophoretic Analysis of Plasma and Factor VIII Preparations at Various Stages of Purification . Comparison of Factor VIII Activity of Fraction I-lA Before and After Cryoglobulin Removal Results of Immunodiffusion Studies on Factor VIII Preparations After Heat Treatment, Cryoglobulin Removal, and Gel Filtration . . . Results of Tyrosine-Tryptophane Ratio and Content Determination by the Beavan and Holiday Method Results of Amino Acid Analysis of Three Lots of Bovine Factor VIII Nitrogen Content of Three Lots of Factor VIII Calculated from Amino Acid Analysis Data . Comparison of Tyrosine and Tryptophane Content from Amino Acid Analysis and the Beavan and Holiday Technique . . . . . . . . . . Calculation of Tyrosine-Tryptophane Molar Ratios and Concentrations vi Page 52 57 58 62 63 70 72 74 77 78 78 82 LIST OF TABLES - Continued TABLE XIV. XV. Page Comparison of the Effects of Heat Treatment for Two Minutes at 56°C of Fraction I-0 and Fraction I-1A, Prepared by the Blomback Procedure and the Initial Modified Procedure (Mercer) . . . . . . . . 84 Comparison of the Effects of Heat Treatment for Five Minutes at 56°C of Fraction I-0 and Fraction I-lA Prepared by the Blomback Procedure and the Modified Procedure (Mercer) . . . . . . . . 85 vii FIGURE 1. 10. 11. LIST OF FIGURES A tentative mechanism for the intrinsic conversion of prothrombin to thrombin in human plasma. Davie et a1. (1964) The scheme for the preparation of Factor VIII by the Blomback procedure (Blomback, 1958) Comparison of Blomback procedure and modified proce- dure (Mercer) on the basis of Factor VIII activity units/ml. (First comparative study) . Comparison of Blomback procedure and modified proce- dure (Mercer) on the basis of Factor VIII activity units/mg. protein. (First comparative study) Comparison of the Blomback and the modified procedure (Mercer) on the basis of protein mg/ml. (Fig. 5-A), Factor VIII activity units/ml. (Fig. 5-B), and Factor VIII activity units/mg protein. (Fig. 5-C) The scheme for the preparation of Factor VIII by the modified procedure . . Comparison of Blomback procedure and modified proce- dure (Mercer) on the basis of Factor VIII activity units/ml. (Second comparative study) . Comparison of Blomback procedure and modified proce- dure (Mercer) on the basis of Factor VIII activity units/mg protein. (Second comparative study) Comparison of the Blomback and the modified procedure (Mercer) on the basis of protein mg/ml. (Fig. 9-A), Factor VIII activity units/ml. (Fig. 9-B), and Factor VIII activity units/mg protein (Fig. 9-C) . Variations in in vivo antibody titers over a fifteen- week period . . . . Starch gel electrophoresis patterns of bovine Factor VIII at various stages of purification viii Page 29 46 48 49 51 53 54 56 61 65 LIST OF FIGURES - Continued FIGURE 12. 13. 14. 15. Urea starch gel electrophoresis patterns of bovine Factor VIII at various stages of purification . Elution patterns of protein and Factor VIII activity form gel filtration of Factor VIII (Fraction I-lAA 56°) on G-200 Sephadex. A typical elution pattern obtained from the separa- tion of Factor VIII and free glycine by gel filtra- tion through G-25 Sephadex. A portion of a typical absorption spectrum as used to calculate tyrosine-tryptophane ratios by the method of Bencze and Schmid (1957) . ix Page 66 71 75 8O I. INTRODUCTION Blood coagulation mechanisms and components have been under intensive investigation for the past two or three decades. These studies have, for the most part, involved the use of the various com- ponents in crude form in systems designed to determine (sometimes inaccurately) the action or activity of the relatively impure component. In coagulation research, as in most areas of scientific investigation, the research must be classified as either basic research, i.e., the study of coagulation mechanisms, or applied research, the preparation of products for clinical control of hemorrhage. It is in the second category that much of the study on classical hemophilia (Factor VIII deficiency) must be placed. Classical hemophilia (Factor VIII deficiency) is inherited as a sex-linked recessive trait, and expresses itself as a mild to severe disorder. Clinically, bleeding episodes are treated by the adminis- tration of fresh whole blood, fresh frozen plasma, or one of the several types of Factor VIII concentrates available. The extent of the therapy is dependent on the severity of the episode. Plasma or whole blood therapy is necessarily limited to minor episodes, as the circulatory overloading attendant with administration of large volumes of these products could tend to increase, rather than improve the bleeding disorder. The increasing demands for fresh banked blood for other uses has limited the supply available for preparation of clinically effective Factor VIII concentrates. Factor VIII concentrates of animal origin have been proven clinically effective, but have been limited in use due to their antigenicity. Purification studies on Factor VIII of animal origin can therefore have several advantages. Knowledge gained from studies of animal preparations, hopefully, can be applied to work with concentrates of human origin. Purified animal Factor VIII can also be used to study coagulation mechanisms. Purification of Factor VIII preparations of animal origin may lead to more extensive clinical application. Reducing the number and quantity of potential antigens present in highly purified preparations could tend to decrease the antigenicity of these preparations thus reducing the limitations of their clinical use. The purpose of this project was to study the purification of bovine Factor VIII. It is hoped to apply the findings to the purifi- cation of Factor VIII of human origin, and possibly to the preparation of highly purified Factor VIII of animal origin for clinical use in man. II. HISTORICAL A. General Coagulation From the publication of De Motu Cordis by Harvey in 1628 on the function of the heart and circulatory system, followed by observations of Malphigi in 1666 of fibrin strands from the washed blood clot, there had been sporadic, but continued interest in the phenomena of hemo- stasis. It remained for Petit in 1731 to make the first scientific approach to the physiology of hemostasis. He observed the vascular plugging effect of clots, and their adherence to the intima of the vessel. Morand (1736) postulated that arteries underwent longitudinal contraction and this decreased the lumen to effect hemostasis. This later concept was more widely accepted than the ideas of Petit by clinicians of that era. It was not until 1805 that Jones presented a concept of hemo- stasis that combined the two previous theories of clotting and vascular contraction. It was about this time that Otto (1803) first clinically recognized hemophilia. The classical theories of Morawitz (1904) and Fuld and Spiro (1904) were the first significant attempts to correlate existing data and condense it into two definitive equations. Ca++ a; Thrombin Thrombokinase Prothrombin Fibrinogen +~Thrombin._——————) Fibrin The fundamental bases of these equations were that thrombokinase released from injured tissue cells in the presence of calcium and prothrombin-———; thrombin. This theory immediately met with objections in that platelets were not considered as part of the theory. Bizzozero (1882) had previously shown that platelets were somehow involved in coagulation. It was later shown by Collingwood and McMahon (1912) that a tissue source of thromboplastin (thrombokinase) was not necessary for coagu- lation. These findings caused a revision of thinking about the proposed theory and gave rise to the distinction of two mechanisms that triggered coagulation. The extrinsic mechanism was based on release of tissue thromboplastin, and the intrinsic mechanism based on the formation of a thromboplastin from plasma components. Brinkhous (1947) proposed that the source of intrinsic thrombo- plastin was platelets, and that its release was dependent on a 1ytic factor in plasma that he called thrombocytolysin. Quick (1947) showed that platelets pg; g; contain only small amounts of thromboplastin, and proposed that intrinsic thromboplastin was formed by the reaction of a platelet factor and a plasma factor, which he called thrombo- plastinogen. The thromboplastinogen of Quick and thrombocytolysin of Brinkhous have since been shown to be the same factor, namely Factor VIII (antihemophilic globulin). The literature on general coagulation is too vast for complete coverage in this writing. In synopsis, the most noteworthy was the discovery, from 1947 to 1955, of six new plasma factors involved in the extrinsic and intrinsic coagulation mechanisms. Table I gives a list of these six factors, with reference to the original publication. There have been many excellent reviews of the clotting literature recently published. For a concise review of the pertinent literature leading to latest theories of the intrinsic coagulation mechanisms proposed by MacFarlane et a1. (1964) as an enzyme cascade, or by Davie et a1. (1964) as a waterfall sequence shown in Figure l, the reader is referred to the work of Gaston (1964) or Gallick (1965). B. Hemophilia In 1905, Sahli suggested that hemophilia was due to a deficiency of thrombokinase. This suggestion and the earlier suggestions of Bizzozero (1882) and Hayem (1898) that platelets were the source of thrombokinase, led to the conclusion that platelets were abnormal in hemophilia. The first suggestion that the defects in hemophilia were in the plasma, not the platelets, was from the work of Addis (1911). Addis concluded that prothrombin was the abnormality in hemophilia. Frank and Hartmann (1927) showed that small amounts of prothrombin- free normal plasma when added to hemophilic plasma corrected the deficiency. This finding was most significant in that it suggested that the deficiency was not due, as previously concluded, to platelets or prothrombin abnormalities, and thus marked the beginning of a new approach to the study of hemophilia. A globulin fraction was later isolated from normal plasma by Patek g£_gl (1936 and 1937) that, when injected intravenously into a hemophiliac, normalized the clotting time of the subject. These _...: \.\.~._ 3.......f ..._~ ......)a..u: ~...-...>....J..‘A~ au~.«..v...~ r...‘ -..\..v ~223~s~afi .N mcowumuwanam umuwm mEmz HmMUmeo mmmfi .Hm no “monumm HHx ecuomm neuumw :mEowwm mmma : : Acoo wood .Hm um coBO HH> pouumm cHnEOHLDOHQ Esuom mmmfi wHoz ,33 33.3 89$ Nema amuse > neuomm cflasnon Haumuoaooo< Meow Mogus< mamz :oEEoo umoz mmma one nemfi mummw mfiu cooBumm wouo>oumfln mHODUmm wcHuuoHU mEmmHm .H manna Foreign Surface A XI (PTA) Activated XI m IX (PTC) Activated IX Ca'l‘4k phospholipid VIII (AHG) ctivated VIII X (Stuart factor) Activated X Ca++ phospholipid V (AcG) Activated V m 11 (Prothrombin) Thrombin Fig. l. A tentative mechanism for the intrinsic conversion of prothrombin to thrombin in human plasma. Davie et a1. (1964). results were confirmed later by Bendien et a1. (1939) and Taylor et a1. (1945). It was thus established that hemophilia was due to an abnor- mality or lack of a plasma constituent. With the discovery of other plasma components necessary for coagulation, the term hemophilia required further classification to avoid confusion with these related disorders. With the varied approaches to coagulation research during the ensuing years, there developed a multiplicity of terms to describe the clotting factors. An international committee on nomenclature of blood clotting factors was formed, and has assigned Roman numeral designations to the accepted blood clotting factors in an effort to alleviate this situation. C. Purification of Factor VIII In the past, several approaches have been used to isolate and further purify Factor VIII from human and various animal sources, including bovine, porcine, canine, ovine, and equine plasmas. The isolation techniques are based on precipitation of the factor from plasma by ethanol (Cohn g£_§1., 1946), ether (Keckwick and Mackay, 1946, 1954), inorganic salts (Bidwell, 1955a, 1955b), and amino acids (Wagner g£_gl., 1964), or by selective adsorption of Factor VIII from plasma or plasma fractions on kaolin (Seegers g£_a1., 1957), trical- cium phosphate (Niemitz §£_§1., 1961), and tricalcium citrate (Blomback 3.531., 1961). Alexander and Landwehr (1948) demonstrated, that from their experience with Cohn Fraction I, between 10% and 35% of the plasma Factor VIII activity was recovered in this fraction. These values cannot be taken as being absolute, in that the great variation in assay systems that have been used in the past and present are not always specific for the factor being assayed. Cohn's Fraction I has been the starting material for many and varied studies on the further purification of Factor VIII. The isolation of Factor VIII from Fraction I first requires its separation from fibrinogen which is the major component of the fraction. Spaet and Kinsel (1953) accomplished the separation by heat denaturation of fibrinogen, followed by absorption of Factor VIII and reprecipitation by pH adjustment. Paper electrophoretic analysis of their purified Factor VIII preparation revealed that it contained a single component. However, it is possible that more refined techniques would have revealed the presence of additional components. Blomback (1958) developed methods for further purifi- cation of Fraction I, utilizing glycine ethanol buffers to extract the trace contaminants of the fraction, and a procedure for separa- tion of Factor VIII activity and fibrinogen. The separation was only partial with 13% of the fibrinogen remaining in the Factor VIII. Van Creveld g£_§1. (1959, 1960, 1961), using ECTEOLA - cellulose chromatography were able to quantitatively separate fibrinogen and Factor VIII from Fraction I. A procedure using tricalcium phosphate to selectively adsorb Factor VIII from Fraction I was described by Niemetz §£_§1. (1961). Complete separation was not obtained, as 20% of the fibrinogen remained with the Factor VIII. Janiak and Soulier (1962), co-workers of Niemetz, reported on an extension of this pro- cedure involving subsequent washings of the tricalcium phosphate - 10 Factor VIII complex, followed by elution and stepwise ethanol fractionation of the eluate. The final product had 3 to 10 times the activity of an equal volume of plasma, representing a 30 to 50 fold purification. Tannic acid was used by Pavlovsky e£_§1. (1961) to selectively precipitate fibrinogen from Fraction I-O prepared by the Blomback procedure. Their data showed that complete separation was obtained with apparent full recovery of Factor VIII activity. Blomback et a1. (1961) used tricalcium citrate in an attempt to further purify Factor VIII from Fraction I-lA. They showed that Factor VIII was apparently adsorbed on tricalcium citrate. When the tricalcium citrate was eluted (dissolved) in 0.1M EDTA, a precipitate containing protein and lipid materials remained. When this precipi- tate was extracted with organic solvents, the lipid and the Factor VIII activity were found in the organic phase. This lipid extract contained low levels of Factor VIII activity as did the original supernatant from the adsorption step. When the supernatant and lipid extract were recombined, the resulting activity was greater than the sum of the separate activities. They also found that when supernate from hemophilic Fraction I-1A combined with normal lipid extract, there was little increase in activity. Lipid extracts from hemophilic Fractions I-lA were found to be inactive and devoid of phosphorous. These results suggested that Factor VIII is possibly a protein phos- pholipid complex. Procedures not involving the initial precipitation of Factor VIII by the procedure of Cohn g£_al. (1946) have been described by several groups. 11 Keckwick and Mackay (1946, 1965) developed a procedure for the fractionation of plasma proteins using ether (1946), and devised methods for aseptic fractionation of plasma proteins with ether (1954). The original process of 1946 was developed using ether, because of the severe shortage of ethanol in England at that time. Keckwick and Wolf (1957) developed a method of further purification of Factor VIII based on ether fractionation, very similar to the Blomback procedure, involving two washings of the crude fraction before lyophilization. Potassium phosphate and sodium citrate were used for isolation and purification of Factor VIII from animal plasma by Bidwell (1955a and b) resulting in 100 to 400 fold purification. The products were contaminated with fibrinogen that could not be removed by heating without complete loss of Factor VIII activity. Michael and Tunnah (1963) studied the further purification of crude porcine Factor VIII, isolated by the Bidwell procedure, using ion exchange resins. The purified preparations obtained were at least 50% fibrinogen, and required stabilization. Factor VIII has been isolated from plasma by adsorption on kaolin (Lorand and Laki, 1954). Seegers and his co-workers (1957, 1959) have utilized this technique followed by alcohol fractionation to extensively purify Factor VIII (platelet cofactor 1). Ultra- centrifugal analysis of their preparations revealed two to three components. With preparative ultracentrifugation, they were able to reduce the preparation to one component (Sw20 = 7.1), but when sub- jected to Ouchterlony analysis, two components were revealed. 12 Amino acids -- glycine, p alanine, and )’mninobutyric acid were used by Wagner g£_al. (1964) as protein precipitants to purify Factor VIII from canine plasma. Their best preparations were purified 2000 times in terms of protein nitrogen. With all of the varied methods used to purify Factor VIII from human and animal plasmas, there has as yet been no report of purifi- cation of this factor to a single component as measured by several parameters of purity. D. Hemophilia Therapy The first treatment of hemophilia reported by Lane in 1840 involved a direct transfusion to a bleeder with good clinical results. The use of crude Factor VIII concentrates has been on the increase in the past decade. In Michigan, however, human Factor VIII concen- trates have been produced for clinical trial by the Bureau of Laboratories, Michigan Department of Public Health since 1948 (Anderson, 1964), and until the mid 1950's, the Michigan Department of Public Health was the only producer of such concentrates in the United States. Many of the treatment failures obtained with these products have been due to improper use for treatment of hemorrhage not due to Factor VIII deficiency, or due to insufficient dosage. Distribution studies show that the extravascular Factor VIII pool may be at least 1.5 times as great as the vascular pool (Adelson .g£_§1., 1963).. It was further shown that the survival of injected Factor VIII was biphasic. The first phase, the intravascular - 13 extravascular equilibration phase, has a half life of 3.8 hours, and the utilization and destruction phase has a half life of 2.9 days in normal humans. In a hemophiliac during hemorrhage, the utilization would be expected to be much more rapid, causing an apparent shortening of the equilibration phase. This has been confirmed by Brinkhous (1964) in studies with Irish setter dogs suffering from hemophilia. After an initial priming dose, the equilibration phase was rapid and thought to be a combination of equilibration and utilization. Sub- sequent administration showed that there was slower utilization. With the cessation of bleeding, utilization approached a normal rate with an apparent half life of 18-24 hours. It was also noted that Factor VIII, when administered as concentrates, seemed to have a longer circulating half life than when administered as plasma (Brinkhous, 1964). Douglas (1958) studied the effects of massive plasma infusion. He found that one liter of plasma raised the circulating Factor VIII level to 14% of normal, and similar infusions were required every 6 to 12 hours to maintain this level. He determined that the circulating half life of Factor VIII from plasma infusion was nine hours. Replacement therapy should be based on replenishing both the vascular and extravascular pools at the onset of therapy in order to maintain theraPEutic levels of Factor VIII in the vascular pool. McMillan g£_gl.(1961) found that preparation of patients for surgery required raising plasma Factor VIII levels 30% to 60% of 14 normal initially, and maintaining plasma levels at 15% to 30% for a minimum of 10 days. There have been several reports on the use of human Factor VIII concentrates in the treatment of hemorrhage and preparation of hemo- philiacs for elective surgery. (Gugler, 1961; Pavlovskng£_al., 1961; McMillan g£_al., 1961; Maycock §£_§1., 1963; and Field §£_gl., 1963). These five reports entail 104 patients, many of whom were treated for more than one hemorrhagic episode. In the report by Maycock g£_§1. (1963), six individuals were treated for a total of 84 individual bleeding episodes over a three-year period. In all cases reported, there was only one case that developed refractoriness to replacement therapy with Factor VIII concentrates. Many and varied reactions were noted during infusion of Factor VIII preparations that were readily controlled by administration of antihistamines or corti- costeroids, or by reduction of the infusion rate. Recently, Pool g£_al. (1964) have developed a simple new approach to the concentration of Factor VIII from plasma, involving a cryo- globulin precipitate obtained from fresh plasma. The procedure achieves a 50 fold concentration of Factor VIII. A single dose of this preparation containing 750 mg protein raised the circulating Factor VIII level in a patient with severe hemophilia to 40% of normal. There has been too limited clinical experience with this type of preparation to evaluate its potential worth in prolonged therapy. There have been many reports by investigators in England con- cerning the use of Factor VIII concentrates of animal origin in major 15 surgery, treatment of extensive injury, and bleeding episodes of hemophiliacs (MacFarlane g£_§1., 1954; Frankel and Honey, 1955; Biggs, 1960; Handly_g£_al., 1961; Ingram, 1962; and Smith, 1965). Treatment of patients (with Factor VIII of animal origin) has been limited to a single regimen not exceeding 14 days. This limi- tation was imposed because of the possible sensitization to the animal source material, because of decreasing response to Factor VIII present, and because of evidence of platelet agglutinins in the animal products. The concentrates of animal Factor VIII used to date have been crude, thus increasing the chances of sensitization to the several antigenic (protein) components. Follow-up studies on several patients treated with animal Factor VIII preparations over a period of weeks or months, failed to show demonstrable sensitization to animal Factor VIII. In most cases treated, the hemostatic effect of the animal Factor VIII preparations was excellent. Reactions encountered during transfusion were manage- able by reduction of the infusion rate or administration of antihista- minic drugs. Boudreaux and Frampton (1960) reported on an apparently peroral acting peanut factor that produced hemostasis in hemophilia. Boudreaux, himself a hemophiliac, noticed that ingestion of peanuts seemed to improve his condition. They suggested that the genetic block to Factor VIII synthesis could have been partially overcome by the peanut factor in the diet. Astrup g£_a1. (1960) suggested that the explana- tion of the phenomenon was based on an upset of the dynamic equilibrium of fibrin formation and fibrin resolution by a factor in peanuts that l6 delayed fibrin resolution. Brakman gt_§1. (1962) showed that ingestion of peanuts was regularly followed by a decrease in the fibrinolytic activity of the blood, and suggested that the hemostatic effect was an indirect effect of an antiprotease present in peanuts. Astrup_g£_al. (1962) isolated a protease inhibitor from peanut flour that contained nearly all of the antiprotease activity of the original flour as assayed by the fibrin plate method against a plasminogen activator. When the extract as administered to a patient, the results were similar to those obtained with the crude flour. Strauss §£_§1.(l965) studied the effect of epsilon amino caproic acid (EACA), itself a potent antifibrinolytic agent, on severe hemophilia, and found that it had no effect. E. Inhibitors of Factor VIII The existence of inhibitors or anticoagulants directed against Factor VIII has been recognized for several decades. (Joules _g£_a1., 1938; Lozner g£_§1., 1940; Munro §£_al., 1943; Verstraete, 1957; Breckenridge_g£_al., 1962; Leitner §£_§l., 1963). The presence of these anticoagulants against Factor VIII in hemophiliacs has raised the question of whether hemophilia is due to a deficiency of Factor VIII synthesis, or whether it is primarily due to the presence of abnormally high concentrations of inhibitors speci- fic for Factor VIII. In the past, several investigators have supported the latter hypothesis. Verstraete (1957) studied an anticoagulant from a hemophiliac subject that was found to be identical with one isolated from a 17 normal subject. He also studied hemophiliacs in which anticoagulants could not be demonstrated. Leitner g£_§1. (1963) studied a purified inhibitor of Factor VIII, and found that dissociation of the Factor VIII inhibitor complex could not be demonstrated. Extensive studies on the inhibitor hypothesis of hemophilia have been made by Mammen (1963). He has been able to demonstrate that an inhibitor present in hemophilic plasma can be denatured by freeze drying, or ether extraction, that is labile to acid and to storage at room temperature. Denaturation of the inhibitor in hemo- philia plasma with ether enabled the isolation of the normal complement of Factor VIII from the plasma. He concluded that hemo- philiacs have normal amounts of Factor VIII. The inhibitor described by Mammen is evidently not similar to that described by Leitner g£_§1.(1963) in that their purified inhib- itor was very stable to changes in temperature and pH, and the Factor VIII inhibition with their inhibitor was irreversible. Mammen (1964) has also reported the isolation of Factor VIII from bovine serum. McLester §£_§1. (1965) have presented data that they feel further substantiate the hypothesis that hemophilia is due to a deficiency of Factor VIII. Regardless of the mechanisms of the pathologic origin of hemo- philia, the fact remains that the most successful therapy of the disorder has been the administration of high potency Factor VIII concentrates . III. MATERIALS AND METHODS The methods herein described are standard laboratory procedures. Any modification of their effectiveness in the purification of bovine Factor VIII preparations will be described and documented in the Experimental section of this thesis. A. Apparatus Following is a list of the standard laboratory equipment and apparatus used in this project, listed according to function. Temperature Control--Two Precision-Freas Model 160 constant temperature water baths, one maintained at 37°C + 0.1°C, and one maintained at 56°C t-O.5°C. One Aminco Model 4-8615 50 gallon con- stant temperature bath filled with 40% (v/v) ethanol, maintained at -4°C }-0.1°C. A constant temperature cold room was maintained at 0 to +4°C. Centrifugation--International refrigerated centrifuges, one Model REF and one Model PR-2. 42H Measurement--One Beckman Model G pH meter with glass electrode. .Absorbance Measurement--One Beckman Model DU spectrophotometer with quartz cells. One Coleman Jr. spectrophotometer with pyrex CUVECCES . 18 19 Fraction Collection-éOne Research Specialties Co. Model 1205 fraction collector with volumetric measuring siphons of 1 m1 and 10 m1 capacity. Electrophoretic Equipment--One Beckman Spinco Durrum Cell used for immunoelectrophoresis on 25 x 75 mm microscope slides. Buchler Model 3-1072 vertical gel mold used for starch gel and urea starch gel electrophoresis. Power supplies for electrophoresis, Heathkit Models PS-3 and IP 32 variable voltage power supplies. Time Measurement--Ga11et one-fifth second stop watch. General Electric 20-second interval timer. Mechrolab Model 202 clot timer. Freeze Drying Eguipment--One Virtis lucite drying chamber with the following accessory equipment: Virtis temperature controller, Virtis cold finger condenser, Virtis McLeod gauge, and Welch Duoseal vacuum pump. B. Materials and Reagents 1. Chemical Reagents Chemicals-~All inorganic and organic chemicals were of reagent grade unless otherwise specified. ,pH 6.8 Glycine-Citrate-Ethanol Buffer (Ionic Strength 0.33): 75 gm (1M) of glycine and 16.2 gm (0055M) trisodium citrate were dissolved in 800 m1 of distilled water. The pH was adjusted to 6.8 with concentrated hydrochloric acid. 68.5 ml of 95% ethanol was added and the total volume adjusted to 1 liter with distilled water. 20 pH 6.8 Glycine-Citrate-Saline-Ethanol Buffergjlonic Strength Q;§3): 75 gm (1M) of glycine, 1.62 gm (.0055M) trisodium citrate, 17.55 gm (.3M) of sodium chloride were dissolved in 800 m1 of distilled water. The pH was adjusted to 6.8 with concentrated hydrochloric acid, 68.5 ml of 95% ethanol was added and the total volume adjusted to 1 liter with distilled water. ,pH 6.8 Glycine-Saline-Ethanol Buffer (Ionic Strength 0.33): 75 gm (1M) of glycine and 19.3 gm (.33M) sodium chloride were dissolved in 800 m1 of distilled water. The pH was adjusted to 6.8 with dilute (approximately 1M) hydrochloric acid, 68.5 ml of 95% ethanol was added, and the total volume adjusted to 1 liter with distilled water. pH 6.8 Citrate Buffer 0.055M (Ionic Strength 0.33): 16.2 gm of trisodium citrate were dissolved in 900 m1 of distilled water. The pH was adjusted to 6.8 with concentrated hydrochloric acid and the total volume adjusted to 1 liter with distilled water. pH 6.8 Sodium Chloride 0.33M (Ionic Strength 0.33): 19.3 gm of sodium chloride were dissolved in 900 ml of distilled water. The pH was adjusted to 6.8 and the total volume was adjusted to 1 liter with distilled water. pH 6.8 Chloride-Citrate-Dextrose Buffer (CCD),(Ionic Strength 0.175): 1.46 gm sodium chloride, 7.35 gm trisodium citrate and 30 gm dextrose were dissolved in 900 m1 of distilled water. The pH was adjusted to 6.8 and the total volume adjusted to 1 liter with distilled water. 21 pH 6.8 Chloride-Dextrose Buffer (Ionic Strength 0.175):, 10.22 gm of sodium chloride and 30 gm of dextrose were dissolved in 900 ml of distilled water. The pH was adjusted to 6.8 and the total volume adjusted to 1 liter with distilled water. ,pH 8.2 Barbital Buffer (Ionic Strength 0.1): 15.85 gm of sodium diethylbarbiturate were dissolved in 600 ml of distilled water 0.1N HCl is added until the desired pH is reached (£3230 ml) and the total volume is adjusted to 1 liter with distilled water. This stock buffer solution is diluted 1:2 with distilled water for use. pH 7.3 Imidazole-Saline Buffer (Ionic Strength 0.8): 3.4 gm of imidazole and 3.85 gm of sodium chloride were dissolved in distilled water. 0.1N HCl was added until the desired pH was reached (C3183 m1) and the total volume adjusted to 1 liter with distilled water. pH 6.8 Imidazole-Saline Buffer (Ionic Strength 0.33): 3.4 gm of imidazole and 17.54 gm of sodium chloride were dissolved in 500 m1 of distilled water. The pH was adjusted by adding 0.1N hydrochloric acid, requiring approximately 300 ml. The volume was then adjusted to 1 liter with distilled water. ,pH 8.6 EDTA-Borate-Tris Buffer--This was prepared according to the procedure of Boyer'e£_al. (1963) with one modification. Disodium EDTA was substituted for EDTA free acid on an equimolar basis and the final pH was adjusted to 8.6 with concentrated hydrochloric acid. The buffer was prepared by dissolving 7.44 gm of disodium-ethylene diamine tetraacetic acid, 30.92 gm of boric acid and 109 gm of tris- hydroxymethyl-aminomethane in 800 ml of distilled water, the pH 22 adjusted to 8.6, and the volume adjusted to 1 liter. This buffer is a concentrated stock and is diluted for use as specified elsewhere in this thesis. 53.37. (v/v) Acetate Buffered Ethanol--560 m1 of 957. ethanol and 5.62 ml 80 x concentrated acetate buffer was diluted to 1 liter with distilled water. pH 8.5 Borate:§aline Buffer--6.184 gm of boric acid, 9.536 gm of sodium tetraborate .10H20 and 4.384 gm of sodium chloride were dissolved in 800 ml of distilled water. The pH was checked and adjusted if necessary to 8.5 with dilute HCl or NaOH. The volume was then adjusted to 1 liter with distilled water. Phosphate-Saline Buffer (Immunodiffusion)--Phosphate-saline buffer was prepared as two solutions A and B. Solution A was pre- pared by dissolving 4.5 gm of sodium chloride in 250 ml of distilled water. Solution B was prepared by dissolving 4.1 gm of disodium monohydrogen phosphate and 1.18 gm of potassium dihydrogen phos- phate in 250 m1 of distilled water. Equal volumes of solutions A and B were mixed for use in immunodiffusion studies on cellulose acetate membranes as described elsewhere in this thesis. 2. Clotting Reagents a. Calcium Chloride A 1M stock solution of calcium chloride was prepared by dissolving 110.99 gm of anhydrous Ca012 in 1 liter of double dis- tilled water. This stock solution was used to prepare 0.025M, 0.03M and 0.06M solutions of calcium chloride by pipetting the appropriate 23 amount of 1M stock solution into volumetric flasks and diluting to the mark with distilled water. b. Hemophilic Substrate Plasma , Blood 500 ml was drawn from a subject (Ritchie) with a mild deficiency of antihemophilic globulin into a double pak plastic blood donor bag (Fenwal), containing 75 m1 of NIH solution A anti- coagulant. The blood was centrifuged, the plasma expressed into the satelfite bag, and the cellular components returned to the donor. The plasma was clarified by centrifugation at 4000 x G to remove cells and dispensed into small aliquots in silicone-coated vials, quick- frozen in a dry ice-ethanol bath and stored frozen at -30°C. c. Kaolin Suspension 5 mg/ml (Mallinckrodt N.F. colloidal) was prepared by sus- pending 1 gm of kaolin in a 23200 m1 of 0.9% saline. The suspension was centrifuged for 20 minutes at 2000 RPM and the supernatant decanted and discarded. This washing procedure was repeated three additional times. The washed kaolin was suspended volumetrically to a total volume of 200 ml in 0.9% saline and stored at room temperature. d. Inosithinngoya Bean Phospholipid) (Associated Concen- trates) 0.2 mg/ml 1 gm of dried inosithin was suspended in 400 ml of 0.9% saline. The suspension was allowed to stand overnight at room tem- perature to insure complete solution. The volume was adjusted to 500 ml with 0.9% saline mixed well and dispensed into a 2 ml aliquot in small (3 m1) serum bottles, stoppered and stored at -20°C. 24 e. Saline Solution 0.9% CW/V) 9 gm of NaCl was dissolved volumetrically in 1 liter of distilled water. f. (3.8%"(ELV) Trisodium Citratengakers Analyged Reagent slide.) 38 gm of trisodium citrate was dissolved volumetrically in 1 liter of distilled water. g.. Standard Factor VIII The standard Factor VIII preparation used in this study was a lyophilized Cohn Fraction I with an established potency of 1.52 Surgenor units/mg protein. The working standard was prepared by dissolving 100 mg of the dried Fraction I in 10 m1 of 0.9% saline. The potency of the redissolved standard was determined by running micro Kjeldahl protein determination on the standard and multiplying the protein content (mg/m1) by the potency (Surgenor units/mg protein) to deter- mine the potency on a unit/m1 basis. The potency of this standard was originally established by many comparisons with a preparation that was originally standard- ized by Dr. D. M. Surgenor and found to contain 11.25 Surgenor units/10 mg dry weight. Definition of Unitage - The Surgenor Unit The Surgenor unit was established by assigning a potency of 1 unit of Factor VIII activity to a Specific Cohn Fraction I then in use in Dr. D. M. Surgenor's laboratory. Based on this assigned 25 potency, Dr. Surgenor established that average normal plasma contained 46 Surgenor units per milliliter. The "D.B.S." Unit_ Dr. D. Aronson of the Division of Biologics Standards, National Institutes of Health, has defined a unit of Factor VIII activity as that amount of activity (Factor VIII) found in one milliliter of average normal plasma. Enzyme activity is normally measured by the amount of a specific substrate converted in a given time. In the case of Factor VIII this is difficult in that its specific substrate has not as yet been definitely established. C..- We 1. Collection of Bovine Blood Nine liters of bovine blood were collected during the bleed- out procedure attendant to slaughter of one cow1 into a large stainless steel vessel containing 1 liter of 3.8% sodium citrate or 10% EDTA. The blood was mixed rapidly and thoroughly with the anti- coagulant and strained through several layers of surgical gauze to remove any foreign materials that might have fallen in during the bleeding procedure. The foam was also removed by the gauze. The blood was transported to the laboratory and cooled to 0 to +4°C during the centrifugation procedure. Plasma was separated by 1Due to the nature of blood source (slaughter house) it was not possible to control the breed of animal used in this study. 26 centrifugation in 1 liter plastic bottles at 1340 x G for 60 minutes followed by aspiration of the plasma from above the cell pack. Pooled plasma was clarified by centrifugation at 4000 x G for 30 minutes to remove platelets and extraneous cellular components carried over during the pooling procedure, and the clarified plasma removed by decantation. The clarified plasma was either fraction- ated immediately, or quick-frozen in a dry ice-ethanol bath and held at -40°C until fractionated. 2. Fractionation of Clarified Plasma at; Precipitation of Fraction I (containing AHG, fibrinogen and traces of other plasma proteins) Plasma Fraction I was isolated from fresh bovine plasma according to the method of Cohn e£_§1. (1946). The plasma was cooled to 0°C. 177 m1 of 53.3% ethanol (-10°C), mixed with 1 ml of 80 x concentrated acetate buffer pH 4.0, was slowly added to each liter of plasma with constant stirring. During the ethanol addition, the plasma was further cooled to -3°C and stirred for 30 minutes after the completion of the ethanol addition step, followed by a standing time of 45 minutes at -3°C to -4°C. The precipitate was sedimented by centrifugation at 1340 x G for 15 minutes. Following centrifugation, the supernatant plasma-ethanol mixture termed Super I was decanted, measured, and sampled for protein and potency assay, and the bulk of the supernatant was discarded. The precipitated wet paste, termed Fraction I was further purified by the procedures described in the Experimental section of this thesis. 27 b.‘ Preparation of Fraction I-0_H Fraction I-0 was prepared according to the procedure of Blomback (1958) as follows: Fraction I paste was resuspended to one-fourth of the original plasma volume in glycine-citrate-ethanol buffer pH 6.8 at -3°C, and stirred for 30 minutes. The suspended paste was sedimented by centrifugation at 1340 x G at -3°C for 15 minutes, the supernatant liquid termed Extract]. (E-l) was decanted, measured, sampled for protein determination and the bulk of the E-l discarded. A second extraction of the Fraction I paste was carried out by the same procedure. The supernatant termed Extract 2 (E-2) was measured, sampled for protein determination and the bulk of the E-2 discarded. The wet paste was redissolved in 0.055M citrate buffer pH 6.8 at 30°C to one-eighth of the original plasma volume. The resulting solution termed Fraction I-0 was sampled for protein and potency determination and submitted to further purification to Fraction I-lA as described below. c. Preparation of Fraction I-lA During this purification procedure there was a partial separa- tion of antihemophilic globulin (Factor VIII) and fibrinogen (Factor I) Fraction I-0 solution was diluted to a protein concentration of 0.75% (7.5 mg/ml) with .055M citrate buffer, based on biuret pro- tein determination. The dilute Fraction I-0 was further diluted with two volumes of cold 0.45M glycine and Fraction I-lA was precipitated by dropwise addition of 10% (v/v) ethanol to a final concentration 28 of 0.5% (v/v) with constant stirring at 0°C. The suspension was stirred for 15 minutes after completion of the ethanol addition and Fraction I-lA was sedimented by centrifugation at 1340 x G for 15 minutes at 0°C. The supernatant (Super I-lA) was decanted, sampled, and the bulk of the supernatant discarded. The precipitate was redissolved in one-eighth of the original plasma volume of CCD buffer at 30°C. The fraction was sampled for potency and protein determin- ations, and the bulk of the fraction was lyophilized pending further purification. Figure 2 gives a schematic presentation of the Blomback procedure. d.) Heat Denaturation of Fibrinogen Fibrinogen was removed from partially purified Factor VIII by heat denaturation at 56°C for a period of 2 to 5 minutes as follows: The fraction to be heat treated was placed in an erlen- meyer flask with a capacity of at least 2% times the volume of the sample. The flask was immersed in a constant temperature bath at 56°C and agitated constantly during the heating procedure. The heat denaturation procedure was timed from the time that the sample reached 56°C. After heating, the sample was rapidly cooled to 0 to +4°C. The soluble material was decanted from the precipitated fibrinogen into 50 m1 conical centrifuge tubes and clarified by centrifugation for 10 minutes at 1100 x G to remove insoluble materials that were carried over during decantation. Cohn's Method 6 29 Plasma 8.0% Ethanol pH 7.2 1 0.1 0 to -4°C Stir 30 minutes Supernatant I (discard) —4—J.—————_—__—-——_— Blomback Procedure Fraction I __________ J_____-_ Extract with k Plasma volume of Glycine Citrate Ethanol Buffer pH 6.8 Stir 30 minutes r Extract 1 (E-l) (discard) I11 Fraction I-o J Extract with %_Plasma volume of Glycine Citrate Ethanol Buffer pH 6.8 Stir 30 minutes 1 Extract 2 (E-2) (discard) 7]] Fraction I-O J Suspend in 1/8 Plasma volume of 0.055M Trisodium Citrate pH 6.8. Stir 30 minutes at 30°C. Dilute to 0.75% protein with 0.055M Trisodium Citrate pH 6.8. Add 2 volumes 0.45M Glycine? 001 to 0°C. Add 10%(v/V)Ethanol to final concentration of 0.5%. Stir 15 minutes. I Supernatant I-1A (discard) Fig. 2. The scheme for the Blomback procedure ll Fraction I-lA Dissolve in 1/8 Plasma volume of 0.055M Trisodium Citrate pH 6.8 u = 0.33. or CCD Buffer pH 6.8. preparation of Factor VIII by the (Blomback, 1958). 30 3. Assay of Factor VIII Activity Factor VIII activity was assayed by the method of Hardisty and MacPherson (1962). This is a modification of the one stage partial thromboplastin assay of Langdell e£_§l.(l952). The assay consists of a system utilizing hemophilic plasma as the clotting substrate which should supply all of the necessary coagulation factors except Factor VIII which is supplied in variable amounts by the test material. To insure complete surface activation, an optimal concentration of kaolin (5 mg/ml) is supplied to the reaction mixture, along with a partial thromboplastin soya bean phospholipid (Inosithin 0.2 mg/ml). The procedure consisted of preincubation of a mixture of equal volumes (0.1 m1) of hemophilic plasma, kaolin, inosithin, and a dilution of the test material (Factor VIII source) for 10 minutes at 37°C. This incubation period was required to insure complete activation of the coagulation components. The incubated mixture was recalcified by the addition of 0.1 m1 of .06M calcium chloride and the coagulation time following recalcification was recorded in seconds. Throughout the course of this study, all clotting assays were performed on a Mechrolab model 202clot timer. The procedure was repeated for four dilutions of the test material ranging from 1:10 to 1:100. A standard curve was prepared, using a standard Factor VIII preparation with a potency of 1.52 units/mg protein. The clotting times obtained were plotted on a log scale setting the clotting time in seconds on the ordinate and dilution on the abscissa. The 31 concentration of Factor VIII in the test sample was determined according to the following calculation by selecting a clottihg time common to the standard and all unknown samples. units/m1 of Std = units/m1 of Unknown dilution of Std dilution of Unknown The activity of the reference standard was based on Kjeldahl protein determination on each standard preparation (Ma et_gl., 1942). The activity of the individual standard was determined as activity units/m1 by multiplying the protein (mg/ml) by specific activity (units/mg protein). 4. Preparation of Antisera Against Human Plasma Bovine Plasma and Bovine Factor VIII , a. ‘Aptisera Against Soluble Antigens Antisera were prepared in rabbits against 1.0% (w/v) bovine plasma and 1.0% (w/v) human plasma and 0.1% (w/v) bovine Factor VIII preparations by repeated intravenous injection of 1 m1 of aqueous antigen per injection for a total of nine injections over an 18 day period. The immunized rabbits were held for one week before testing. b. lAntisera Against Bovine Factor VIII - Freunds Complete Adjuvant Emulsions i. Preparation of Antigen Emulsion--Fraction I-lAA 56°C was concentrated about eight fold by dialysis against a concen- trated solution of Carbowax. This was accomplished by placing mois- tened ground Carbowax in a visking dialysis tubing and immersing the tubing into the solution to be concentrated. The solution was held at 0 to +4°C during the concentration procedure. The concentrated 32 fraction was emulsified with an equal volume of Freund's complete adjuvant. Care was taken to emulsify small quantities at a time with the total desired volume of adjuvant until the total antigen volume was used. This initial emulsification was accomplished using a syringe with a 15 gauge needle. .After the initial emulsification step was completed, the crude emulsion was passed several times through a small bore (20 - 24 gauge) needle to produce a stable emulsion suitable for injection. Elimination of the initial step made it impossible to produce emulsions of suitable consistency for injection. ii. Immunization--Norma1 rabbits were injected by various routes (subcutaneous, deep intramuscular, foot pads) with 1.0 ml of emulsified antigen at five injection sites (0.2 ml/site). The immunized animals were held for 40 days without further injection before testing. iii. Test B1eeding--A11 rabbits were bled by heart puncture, removing between 5 and 10 ml of blood. The blood was allowed to clot at room temperature and incubated for one hour at 37°C to assure complete clot retraction. The clots were removed and the serum clarified by centrifugation at 1100 x G for 10 minutes to remove red cells. The clarified serum was carefully aspirated from the cell pack and transferred into clean vials. The clarified sera were heat treated for one hour at 56°C in a constant temperature bath to denature nonspecific reactants. 33 iv. Titration of Antisera-~The sera were titered using the.interfacia1 ring precipitation test described~by Campbell _et_aL (1963). This test was carried out on a micro scale using 4 mm 0D x 45 mm precipitin tubes and capillary pipettes drawn from 8 mm OD glass tubing. The test antigens were diluted (based on protein) over a range of 1:1000 to 1:128,000 by serial two fold dilutions of the soluble injection antigen in borate-saline buffer pH 8.5. Sera were used full strength or at a 1:4 dilution in borate buffer depending on the expected titer. For each titration, eight 4 x 45 mm precipitation tubes were set up in a labeled test tube rack that indicated the dilution range. Small tubes made individual labeling difficult. About 0.025 ml of the serum to be tested was pipetted into each of the tubes carefully to avoid inclu- sion of air bubbles at the meniscus. Starting with the highest antigen dilution, using a micro pipette, the tube of antiserum corresponding to the dilution was carefully overlaid with an equal volume of the antigen dilution. Extreme care was taken to avoid mixing at the interface of the antiserum and antigen dilution. Using the same pipette, rinsed well with the next lower dilution, the process was repeated, overlaying each tube of serum with its corres- ponding antigen dilution until all dilutions were used. The tubes were allowed to stand for 20 minutes and observed under indirect light against a black background. A posi- tive test was observed as a sharp zone of precipitate formed at the 34 interface of the serum and antigen. Variations in the thickness and sharpness of the precipitation ring were caused by relative mixing at the interface and titer of serum. Each antiserum tested was tested against its homologous antigen and a heterologous antigen, i.e., antibovine plasma ‘was tested against bovine plasma and human plasma. The results of the titration were reported as the highest dilution of antigen that gave a positive reaction with the antiserum. In cases where the anti— serum had been diluted before testing, the final titer of the antiserum was obtained by multiplying the apparent titer (highest positive antigen dilution) by the dilution factor of the antiserum. 5. Immunoelectrophoretic Analysis Immunoelectrophoresis was performed by a modification of the procedure of Grabar and Williams (1953) in 1% (w/v) buffered agar on 25 x 75 mm microscope slides in the Spinco Durrum cell. Molten 1% buffered agar (Oxoid Ionagar) was layered evenly on pre-coated microscope slides at 2 ml agar per slide. The agar was allowed to cool for five minutes and transferred to a humid chamber. Using one of two templates, (two spot - one slit, or one spot - two slit) sample wells and antiserum slits were cut in the agar. The agar was removed from the sample wells by gentle suction with a Pasteur pipette. The slides were placed in the Spinco cell, the buffer (Barbital pH 8.2 ionic strength 0.5) was poured into the chambers and the filter paper wicks placed in position. Contact between the 35 slides and wicks was made by pipetting on molten buffered agar at the juncture between the wicks and slides. The sample wells of the slides were filled with the desired antigens to be tested using melting point capillaries opened at both ends fitted with a small rubber bulb. The cover was placed on the cell and a current of 40 ma. was applied across the cell for a period of 1.5 hours. At the completion of the electrophoresis the current was turned off and the power source disconnected from the cell, the cover was removed and the contact between slides and wicks was broken. The first slide was removed and the cover replaced on the cell. The agar in the precut antiserum slit(s) was removed by gentle suction with a disposable Pasteur pipette, and the slide returned to the Spinco cell. Each slide was thus treated individ- ually, keeping the remaining slides in the humid atmosphere of the cell until all slits had been prepared. The slits were then filled rapidly with the desired antisera, with the aid of a Pasteur pipette drawn to a fine tip. Double diffusion of the antigen and antiserum was allowed to develop for 24 hours at()t014%3in the closed Spinco cell. At the termination of the diffusion period, the precipitin lines were readily visible. The developed slides were dialyzed against several changes of 0.9% saline over a 24-hour period, followed by several changes of distilled water for a further period of 24 hours. The agar layer was allowed to air dry. This was expedited by flooding the slide 36 with distilled water and overlaying the agar with a 25 x 75 mm strip of Whatman #1 filter paper. Then the slides were stained for two minutes in triple stain and differentiated with several changes of 2% (v/v) acetic acid, the final acid bath containing 2% (v/v) glycerol in addition to the acid. The slides were air dried in a dust-free chamber. A staining procedure helped visualize and differ- entiate minor lines not readily seen in the unstained preparation. The stained precipitin lines were readily observed under low power magnification and permanent photographic records could be made using the slides as a negative in a photographic enlarger. 6. Immunodiffusion Techniques During the course of this study, two different micro- immunodiffusion techniques varying only in the support medium were used. The first procedure was a modification of the procedure of Beale and Mason (1962) using agar as the support medium. The second was according to the procedure of Johnson et_al. (1964) using cellu- lose acetate as the support medium. The buffer system used in both procedures was based on the work of Thorne and Belton (1957) and was a phosphate-saline buffer made as two solutions (Solutions A and B). a. ‘Agar Gel Procedures 0.5 gm of Oxoid Ionagar was dissolved in 25 ml of Solution A, and to the dissolved agar solution was added 25 ml of Solution B pre- warmed to 56°C. The agar was mixed by swirling and kept molten in a hot water bath c.a. 60°C. On a clean glass microscope slide, 0.7 ml of molten buffered agar was distributed evenly over an area 2.5 x 4.5 cm 37 and allowed to cool for two minutes. A perspex micro immunodiffusion template was gently placed on the surface of the agar, avoiding air bubbles between the template and agar. In any case where agar was seen rising in the wells of the template, the template was removed and the slide discarded. The slide with template in place was immed- iately placed in a humid chamber until all necessary slides were prepared. The desired antisera were placed in the center well of each template and the peripheral wells filled with the antigens to be tested. The slides were incubated for 40-48 hours in a sealed humid chamber. After incubation, the templates were removed. At this point, the precipitin lines were viewed under low magnification. The slides were made permanent by dialysis, drying and staining as described for immunoelectrophoresis. b. Cellulose Acetate Procedure The method for immunodiffusion on cellulose acetate strips was performed as follows: Strips of cellulose acetate 22 x 45 mm (Oxoid or Millipore cellotate) were soaked in phosphate-saline buffer according to manufacturer's suggested procedure and laid centrally on the surface of a clean glass slide, one strip per slide. A clean perspex microdiffusion template was placed on the surface of the cellulose acetate membrane. By placing the template in the buffer layer at one end of the membrane and sliding it to the center of the membrane, air bubbles were avoided at the interface of the membrane and template. The protruding ends of the membrane were blotted with filter paper to remove excess buffer until the membrane at the bottom 38 of the wells changed from a shiny to matte appearance. The slide was immediately placed in a humid chamber and the sample and anti- sera wells were charged without further delay. Leakage, if any, between the adjacent wells was detected immediately by changes in the appearance of the membrane in the unfilled wells. If leaks were detected, the preparation was discarded. When all slides were thus prepared and filled, the humid chamber was sealed and left undisturbed for a period of 40-48 hours. At the end of the incubation period, the slides were removed from the chamber and the templates washed from the slides in flowing tap water. The membranes were removed and washed by immersion in several changes of 0.9% saline over a period of 4-8 hours followed by a two-hour wash in several changes of distilled water. The washed membranes were stained by immersion in 0.1% (w/v) Thiazine Red-R in 1% (v/v) acetic acid, for 2-10 minutes. The stained mem- branes were blotted to remove excess dye and differentiated in several changes of 4% (v/v) acetic acid. The stained strips were dried by pressing between several layers of paper towels. Clearing of the stained preparations was effected by soaking in a commercial preparation of cottonseed oil or suitable organic solvent clearing agent. The cleared strips were blotted to remove excess oil and placed on clean microscope slides. The reaction area was covered by a clean overslip with gentle pressure to express entrapped air bubbles. The exposed ends of the membrane were trimmed flush with the edges of the coverslip. The oil was wiped off the slide and the coverslip sealed with a suitable sealant (Canada balsam 39 or permount). This gave a permanent preparation of the immunodiffu- sion pattern that could be viewed with low power magnification or photographed. Membranes cleared with organic solvents were dried on the slide and covered with a coverslip sealed to the membrane with permount. 7. Starch Gel Electrophoresis Starch gel electrophoresis was carried out by the method of Smithies (1959). Starch was prepared in 0.026M pH 8.6 borate buffer and molded in a vertical gel mold. Electrophoresis was carried on for 16 hours at room temperature at 16 milliamperes. The gel was sliced and stained in Amido black 10B 0.1% (w/v) in 10% (v/v) acetic acid 50% (v/v) methanol and 40% (v/v) water. The stained gel was differentiated in 50% (v/v) methanol, 40% (v/v) water, 10% (v/v) acetic acid and sealed in Saran Wrap. 8. Urea Gel Electrophoresis Urea gel electrophoresis was carried out by a modification of the procedure of Wake and Baldwin (1961), using an EDTA-Tris- Borate buffer system introduced by Boyer e£_§1. (1963). The modifi- cations were necessary to effect the conversion of the urea gel from a horizontal to a vertical electrophoresis system. The discontinu- ous buffer system described for horizontal urea gel electrophoresis was found unsuitable in that during electrophoresis, there was local- ized shrinkage of the gel along the front of buffer migration, resulting in decreased mechanical strength of the gel, causing fracture of the gel at the point of sample insertion. The change 40 to the mixed buffer described by Boyer (1963), which was more defin— itive in alleviation of the shrinkage problem, eliminated the problem of gel fracture. The urea starch gel prepared in a 1:20 dilution of the stock mixed buffer, contained 60 gm of starch and 147 gm of urea in 360 m1 of buffer. The molten gel was degassed under vacuum, and poured into the gel mold. The gel was allowed to harden overnight at 0 to +4°C. Electrophoresis was carried out for 16—24 hours at 16 m amps. The gel was sliced and stained in Amido black 10B for 15 minutes and destained electrolytically in 7% acetic acid overnight (Ferris e£_§1. 1963). The stained gel was wrapped and sealed in Saran Wrap and photographed. 9. Tyrosine-Tryptophane Ratios Tyrosine-tryptophane ratios were determined on bovine Factor VIII preparations at various stages of purification according to the methods of Beavan and Holiday (1952). Samples were prepared by dilu- tion in 1.0M NaOH to give a final concentration of 0.1M NaOH. Absorbancy of the samples was determined in quartz cells in the Beck- man DU at 280 and 294.4 mu against a blank containing the original solution buffer made 0.1M in NaOH. Ratios were calculated from the data using the following formula: M tyrosine _ 0'592 D294.4 ' 0'263 D280 M tryptophane 0.263 D280 - 0.170 D294.4 41 where: M tyrosine = moles of tyrosine M tryptophane = moles of tryptophane D294.4 = absorbancy of sample at 294.4 mu D280 = absorbancy of sample at 280 mu Calculations of molar absorptivity "a” were made according to the formula: I = lo-abc where: I =‘Transmitted light Io IO = Incident light = 100 a = absorptivity b = optical path = 1 cm c = concentration of protein As rephrased: log I = _ac b = 1 I0 or log I I0 = -a c Molar absorption constant K was derived from ”a” as 2.303a = K. The K's for each sample were determined at both wave lengths and used to calculate molar concentration of trysine and tryptophane/mg protein as follows: M tyrosine = (0.592 K294.4 - 0.263 K280) x 10'3 M tryptophane = (0.263 K280 - 0.170 K294,4) x 10-3 42 10. Amino Acid Analysis a. Preparation of Samples“ Due to the concentration of free glycine carried over during the purification procedure, methods for removal of free glycine were developed. The first method used was a TCA precipitation and washing method that turned out to be laborious and ineffective resulting in mechanical loss of Factor VIII protein, and incomplete removal of glycine. Glycine was readily removed by gel filtration of Factor VIII preparations through a 625 Sephadex column equilibrated with 0.033M NaCl. The gel filtration procedure accomplished a two-fold purpose of removing free glycine and reducing the salt concentration 10 fold, the salt concentration of the original samples being 0.33M NaCl. The protein peak in the column eluates was detected by absorbancy measurements on selected eluate fractions at 280 mu. Glycine was determined by a spot test method on filter paper using 0.1% Ninhydrin in pH 5 citrate buffer. The protein containing glycine-free fractions were pooled. The protein was concentrated by precipitation in 5% (w/v) final concentrations of trichloroacetic acid. The precipitation step was carried out on three aliquots of equal volume. Two were used for protein determination, and the third submitted for amino acid analysis. b. Hydrolysis The samples submitted for analysis were hydrolyzed with 6M HCl, dried and redissolved in buffer for analysis in the Beckman 1203 amino acid analyzer. Amino acid identification was based on 43 the ”standard peak position” of the individual amino acid established for the particular conditions by Dr. B. H. Olson (Michigan Department of Public Health). Individual amino acid content was determined from the chroma- tograms using the half height method described in Instruction Manual AIM 2 for the Beckman 120B amino acid analyzer. The CHW constants used were determined by Dr. Olson for this particular instrument. Tryptophane was estimated by dividing the tyrosine content derived from the chromatographic analysis by the tyrosine-tryptophane ratios determined for the particular sample. IV. EXPERIMENTAL Introduction: The most direct method for the separation of the closely associated proteins fibrinogen (Factor I) and antihemophilic globulin (Factor VIII), is the denaturation (precipitation) of fibrinogen from a mixed solution of the two proteins by heating at 56°C for a period of three to five minutes. This is a simple, rapid method, but it has the disadvantage of variable losses of Factor VIII activity ranging from 50% - 90% incurred during the heating process. Simonetti_e£_al. (1961) observed that in the absence of citrate, these heating losses are minimized. The first experiment conducted in this work involved a comparative study of this citrate effect. A. Modification of Blomback Procedure 1. First Modification, A,Comparative Study During the fractionation and partial purification procedures described by Blomback, citrate containing buffers are used, and the final product, Fraction I-lA is redissolved in a citrate buffer. The first modification of the Blomback procedure involved the 10- fold reduction of citrate in the buffers used, and replacement of citrate with saline of equivalent ionic strength. At the Fraction I-O stage of the modification, citrate was completely replaced by saline of equivalent ionic strength. 44 45 Bovine blood was collected, using a 3.8% solution of trisodium citrate as anticoagulant, 9 volumes of blood, and 1 volume of 3.8% trisodium citrate. The resulting plasma obtained was frac- tionated to Fraction I by Cohn's method 6. (Cohn e£_§l., 1946). The Fraction I paste obtained was divided into two equal portions. One-half was carried through the standard Blomback procedure to Fraction I-lA, and the other half carried through the modified pro- cedure (IO-fold reduced citrate) to Fraction I-O', and further to Fraction_I-1A' in the absence of citrate. To avoid confusion, the materials obtained from the modified procedure were designated by the use of the prime designation (i.e., Fraction I-0', Fraction I- lA', etc.) All fractions obtained from this experiment were assayed in the Blomback assay system which consisted of a recalcified clotting time of Factor VIII deficient plasma using the test material as the source of Factor VIII. The test material was diluted 1:10, 1:20, 1:50, and 1:100. Equal volumes of diluted test material and Factor VIII deficient plasma were mixed, recalcified and the clotting time determined. The unknown samples were quantitated from a standard curve prepared in the same way with a reference standard Factor VIII preparation of known potency. This assay system was replaced with the Hardisty assay system as described earlier for all subsequent experiments. Figure 3 gives a comparison of the two procedures on the basis of Factor VIII activity (units/ml). The greatest variations appeared in Fraction I-0 and Fraction I-O'. These data confirm the work of 46 Blomback [:1 400‘_’ Modified HS 300 .1. . 200 _, H E \ 0‘) .. l . "-4 S % 100.m U -.-4 > "-1 4..) 8 H 50 _. r: a > m m u a. o m U U m b“ 20 1, 10 ' I l§ I-O I-lA I-lA A56° A 56° Fraction Fig. 3. Comparison of Blomback procedure and modified procedure (Mercer) onathe basis of Factor VIII activity units/ml. (First comparative study). 47 Simonetti et a1. (1961) on heat stability of Fraction I-0 in the absence of citrate, as seen by the greater than two-fold difference in activity of the heat-treated materials. Minimal differences were observed during subsequent purification to Fraction I-lA. Examination of the data comparing the specific activity (Factor VIII units/mg protein) in Figure 4 showed the true signifi- cance of the modified procedure. The differences are due mainly to the variations in protein content of the fractions, and not in the activity on a unit/m1 basis as seen from Figure 3. Thus, in the actual purification, the cardinal feature on the modification was based on altered protein content of the purified fractions Fraction I-lA and Fraction I-lA' before and after heat treatment. There was a 4.7-fold difference after heat denaturation. Figure 5 summarizes the data of protein mg/ml (Figure 5A), activity units/ml (Figure 5B), and activity units/mg (Figure 5C) of the fractions showing the interrelationship of these three variables. Starch gel electrophoretic analysis of the fractions before heating showed that two components, one major assumed to be fibrinogen, and one minor component assumed to be the Factor VIII containing compon- ent, were visualized. After heat treatment at 56°C, the major component of the Fraction I-0 56°C was diminished in size. In the Fraction I-A 56°C, the major component was absent, leaving essen- tially one component. 48 3:332:22 g 222:1 A .C 200 -w §§§ f? 100 .“ Egg Egg 5° -> \ \ \ y a y ( L" 10 _. g .. ‘7 \ \ 5 If: F§ . a \\ . Fig. 4. Comparison of Blomback procedure and modified procedure (Mercer) on the basis of Factor VIII activity units/mg protein. (First comparative study). Fig. 5-A 60 Blomback 1:] Modified 11 40 0.3 a ' r: 10" 5- 0.1 a . 0.05-1 2H 1 , *l'_l 0.02 & E-l I-lA 2/7 PCl Protein mg/ml Plasma Super I E-2 I-o I-o I-lA I-lA 456° 056° Fraction 400- 400- ’_ Fig. 5-3 Fig. s-c t: "-1 (D 4.) 8 s c» . 1': on I 2100- £100- 1 ."3 H -.-a t S 50‘ g > .‘ $t :- .. . E 50 8 :> U H :33 s U .. 8 E 0‘) a! '5‘. lO-L \ . f 10.1 I-0 I-0 I-lA I-lA I-O I-O I-lA I-lA 456° 456° A56° 656° Fraction Fraction Fig. 5. Comparison of the Blomback and the modified procedure (Mercer) on the basis of protein mg/ml (Fig. 5-A), Factor VIII activity units/m1 (Fig. S-B), and Factor VIII activity units/mg protein (Fig. 5-C) 50 2. Complete Modification Based on these findings, the Blomback procedure was further modified by complete elimination of citrate from the extraction buffers in the stage leading to Fraction I-O', resulting in a modi- fied Blomback procedure devoid of citrate, save that carried over from the anticoagulant of the starting plasma. Figure 6 gives a schematic presentation of the modified procedure. Four fractionation runs were carried out by this modified procedure to the Fraction I-lA’ stage. The fractions obtained were lyophilized and stored at ~20°C pending further studies. Table II summarizes the data obtained from these four runs, giving the average and range of the results. 3. Second Comparative Study of Mercer Modification A second run was performed comparing the standard Blomback procedure and the modified procedure (Mercer). The Fraction I obtained from 3.9 liters of bovine plasma was divided into two equal portions. One was carried through the standard Blomback pro- cedure and the other through the modified procedure. All fractions obtained were sampled for potency and protein determinations, and the remainder lyophilized and stored at -20°C pending further studies. Potency determinations were made using the Hardisty assay system as previously described. The results of this and the preceding study are comparable with some minor variations. Figure 7 gives comparison of activities on a unit/ml basis. Figure 8 compares specific activities, and Cohn's Method 6 51 Plasma 8.0% Ethanol pH 7.2 f 0.1 0 to —4°C Stir 30 minutes Supernatant I (discard) -——_—h-—_’—————. Modified Procedure (Mercer) ll Fraction I Extract with % Plasma volume of Glycine Saline Ethanol Buffer pH 6.8 Stir 30 minutes 1 Extract 1' (E-l) (discard) Fraction I-o' 1 Extract with % Plasma volume Glycine Saline Ethanol Buffer pH 6.8 Stir 30 minutes I Extract 2' (E-2) (discard) [1 Fraction I-O' Suspend in 1/8 Plasma volume of 0.33M Sodium Chloride pH 6.8 Stir 30 minutes at 30°C. Dilute to 0.75% protein with 0.33M Sodium Chloride pH 6.8. Add 2 volumes of 0.45M Glycine. Cool to 0°C. Add 10% (v/v) Ethanol to final concentration of 0.5%. Stir 15 minutes. I Supernatant I-lA' H Fraction I-lA' Dissolve in 1/8 Plasma volume of 0.33M Sodium Chloride pH 6.8 u = 0.33. Fig. 6. The scheme for the preparation of Factor VIII by the modified procedure. 52 H~.Oe-ae.OO HH.mH-m.NO HH.NO-O.ONO HOOH-H.meO HON.O-ee.eO AOO.O-eeN.OO HHO.N-em.HO HN O.O O.em e.~O O.m emm.O HH.H ooom Eoum N huo>ooom HHH>HHd< HHH> nodded - - dedoem Aumopozv ousmmuoum ooflmwmoz use LHHB mesa coHumcofluumum Hoom Eouw wocwmuno sumo mo xumEEsm .HH manme 53 Blomback Cl Modified \\\\' 200‘L r-| E \ U) U .8 ' :1 100* >. U "-1 > w-l U 3 SOL“ H #4 H >. H 0 U 0 _ m hi ZO-y m E U) m H n. 10 , I-0 I-0 I-lA I-lA A56° A56° Fig. 7. Comparison of Blomback procedure and modified procedure (Mercer) on the basis of Factor VIII activity units/m1. (Second comparative study). 54 200 ‘L BlombaCk D 5 Modified m, o 4.) o '5. 100 a. 00 E \ (I) L) -.-4 a s H 50 -. H H > H o 4..) 0 ti“. 20 q: lO-‘r I-o' I-0 .I.-.1A I-lA A 56° A 56° Fraction. Fig. 8. ComparisoneqfiBlomhack procedure and.modified procedure (Mercer) on the basis of Factor VIII activity units/mg protein. (Second comparative study). 55 Figure 9a, b, and c summarize the three variables, protein content, activity, and Specific activity of the various fractions. With this, as in the first run, the results showed that the differences in activity are significant, but not as striking, as in the first com- parative study. Table III compares the data of the two runs according to variations between individual comparable fractions from each run. There does not seem to be any set pattern to the variations between the individual materials. It must be remembered that strict com- parisons of the two experiments were not possible due to the different assay systems used. The Blomback assay system was affected more by factors other than Factor VIII. Table IV compares the two procedures from each study. Such comparison is not affected by variation in assay systems. The general trend is the same in both studies, showing that the material obtained by the modified procedure was superior to that obtained by the standard Blomback procedure, both on the basis of heat stability and specific activity of the partially purified Factor VIII preparation. B. Antisera 1. Preparation of Antigens Bovine and human plasma were diluted to a concentration of 1% protein with sterile 0.9% saline, and used for intravenous injec- tion at this concentration. Fig. Factor VIII units/m1 56 Fig. 9-A 6O 50 Q Blomback». D ‘ Modified \ 10,- N W . . 5— H 3 E \ on E e #4 OJ L) o H a. 1- H m e H U) (D c“ s 0.5- E m E-l E-2 I-o 'I-o I-lA I-lA 456° A56° Fraction Fig. 9-B .E Fig. 9-C 100“l mIOO- .LJ 0 H CL 50: 532°50- \ (I) U 'E' a H #1 H >. m u E o a - 3 E a “N ’ I-0 I-0 I-0 I-0 I-lA I-lA A56° A56° 456° . -Fraction Fraction 9. Comparison of the Blomback and the modified.procedure (Mercer) on the basis.of protein mg/ml (Fig. 9eA), Factor VIII activity units/m1 (Fig. 9-B), and Factor VIII activity units/mg protein (Fig. 9-C). 57 Table III. Comparison Between Two Comparative Studies (Runs 1 and 6) J Blomback Procedure _A Activity Spec. Activity Protein Material Run u/ml % diff. yu/mg % diff. _mg/ml % diff. Plasma 1 358 94.5 5.9 94.5 60.6 0.5 6 19.6 0.326 60.3 Fraction I-O 1 145 11 9.24 10 15.7 20 6 129 10.3 12.55 Fraction I-0 1 31.9 13 11.55 78 2.76 75 A 56°C 6 36.7 52.2 0.69 Fraction I-lA 1 41.5 49 35.8 21 1.15 48.5 6 81.5 45.5 2.23 Fraction I-lA. 1 14.6 2 62.4 53 0.234 54 A 56°C 6 14.9 29.1 0.512 Modified Procedure (Mercer) Plasma 1 358 94.5 5.9 94.5 60.6 0.5 6 19.6 0.326 60.31 Fraction 1-0’ 1 171 24.5 11.25 5 15.2 20.5 6 129 10.63 12.1 Fraction I-o’ 1 79 0.5 28.8 76 2.74 76 A 56°C 6 78.6 120 0.655 Fraction I-Inf 1 42.8 21 173 76 0.247 80 6 54 43.2 1.245 Fraction I-lAf 1 12.28 59.5 361 79.5 0.033 92 A 56°C 6 30.3 74 0.409 58 mood 3K m.om mmo.o Sm wN.NH z Doomfi om Nam.o Ho H.mN Hm m.¢H ow omm.o mm «.mo 0H o.qa m wuo< .ooam >ua>wuo< awououm. >ua>fluo< .ooam huw>fluo< 11 o cam H cam Auoouozv opsooooum wowwflwoz ocu vcm ousuoooum xomnEon ocu mo acmwummEoo .>H oHan S9 Bovine Factor VIII was prepared by methods described, and consisted of heat-treated Fraction I-lA'. Due to the low protein concentration of the Factor VIII preparation, it was concentrated two fold by dialysis against Carbowax, giving a final protein concen- tration of .091. 2. Immunization and Testing Twenty-one normal rabbits (7 rabbits per antigen) were each injected with 2 ml of soluble antigen in the marginal ear vein every 48 hours for a total of nine injections. One week after the last injection, the rabbits were test bled by heart puncture, removing S to 10 ml of blood. The sera obtained were clarified by centrifugation and heat treated for one hour at 56°C. The heat- treated sera were titered using the interfacial precipitin test, using the corresponding antigen and the heterologous antigen to check for cross reactions. All rabbits showing the highest anti- body titers were bled by heart puncture, removing 40 to 50 ml of blood. The sera obtained were processed and stored frozen at -40°C in small aliquots for testing in immunoelectrophoretic procedures. At the end of the seventh week, the high titer rabbits were again test bled by heart puncture, and the sera obtained processed and titered. Due to the significant drop in antibody titer, a series of five booster injections (2 ml of antigen per injection) was given over a lO-day period. 60 One week after the last injection, the rabbits were test bled and the resulting sera processed and titered. There was a significant rise in antibody titer after the booster injections. At the end of the thirteenth week, a booster injection series was initiated (2 ml of antigen per injection). Four injections were given over a period of eight days. One week after the last injection, the rabbits were test bled and titered. Rabbits having the highest titers were bled on the same day by heart puncture, removing 40 to 50 ml of blood. The sera obtained were clarified, heat treated at 56°C for one hour, and stored frozen at -40°C. Figure 10 summarizes the variations in the highest antibody titers obtained over the lS-week period. 3. Preparation of Anti-Bovine Factor VIII by Repository Immunization The relatively low anti-Factor VIII titers obtained from the injection of soluble antigens were probably due to the low circu- lating level of antigen, and required the use of more effective methods. a. Preparation of Antigens Two-hundred fifty ml of bovine Factor VIII (Fraction I-lA £3 56°C) were concentrated 8.4 fold by dialysis against Carbowax. The concentrated Factor VIII was emulsified with an equal volume of Freund's complete adjuvant and injected immediately after emulsifi- cation. 61 ‘ \VNNNQ. \\\\\\\\\\\\\\\\\\ mmN\WNH\C\\f\.\\\HL\.\\\Wr\\\.\fl\.\WL\.\flMlMV.\p\_\.\\\fib\fl\‘\\ V‘VWU\.\r\.\. ma mop cu Hm >mo noumoon uonm q N\\\K\\\V\\\\\W\\\\\ KV\VNK\\\\N\XI\k\\ “DZ/A em camp on wq chop nonwoon uocmm a m m s T. F S 3T. WQ\\\\\N\\\\\\ a .1 I :\\\s 1 PV P , e r . n n O _ a .1 t m : _ 323223on 3 am. 1 E; .1 .1.1 t t t - n n n , r A. 3% ._ . —\\\\KK\\\E\\\\\\\\\\\\\\\ 1 \.\‘\‘\\.HL\ \ \‘\\LN.\ \ \‘\\\\ m a» a _q q d _._ r. . . 6 5 A. O 0 O 1 1 1 Hooflo aoocfloca 34 48 61 105 Days after initial immunization 25 y titers over a fifteen- Variations in.i§ vivo antibod week period. 10. Fig. 62 b. Immunization and Testing Five normal rabbits were each injected once with 1 ml of antigen-adjuvant emulsion by various routes. These routes were subcutaneous neck and ears, subcutaneous foot pads and ears, and deep intramuscular. The injections were made at five sites, with 0.2 m1 of emulsion per injection site. Six weeks after injection, the rabbits were bled by heart puncture, removing 5 to 10 ml of blood. The sera obtained were clarified and heat treated at 56°C for one hour. Antibody titers were determined against bovine Factor VIII, bovine plasma and human plasma, using the interfacial precipitin test. The titers obtained were somewhat better than those obtained by immunizations with sol- uble antigen preparations. There was no observable cross immuniza- tion to human material detected. Table V summarizes the titers obtained using bovine Factor VIII and bovine plasma as the test antigens. _Table V. Antibody Titers from Injection of Adjuvant Enriched Bovine Factor VIII Six Weeks After Injection Titers Injection Route vs. Bovine vs. Bovine Plasma Factor VIII Subcutaneous 1:3000 l:20,000 Foot Pads l:64,000 l:80,000 Deep Intramuscular l:l6,000 1220,000 l:32,000 l:80,000 63 C. Immunoelectrophoretic Analysis of Bovine Factor VIII Immunoelectrophoresis in agar was performed on several of the partially purified fractions of bovine Factor VIII. Due to the low concentration of the original material and the dilution suffered during electrophoresis, the precipitin lines formed were quite faint, but still visible. The best patterns obtained were those developed with antibovine plasma antisera. Studies were made on fractions at various stages of purification to demonstrate the effect of the purification steps on the immuno;, logically identifiable components present at the various stages of the purification procedures. Table VI summarizes the results of the immunoelectrophoretic analysis on the various fractions tested. Table VI. Results of Immunoelectrophoretic Analysis of Plasma and Factor VIII Preparations at Various Stages of Purification A A Antisera Patterns Antigen vs. Anti-bovine vs. Anti-bovine Plasma Factor VIII Plasma Plasma Pattern 2 lines Fraction I-O' 3 lines N,fi,X 2 lines 3 absent Fraction I-lA' 2 lines O(,,6 2 lines 0(“6 Fracgion I-lA 2 lines 0(,,6 2 lines d,fl 5 o Supernatant I-lA 1 line 1 line 64 D. Gel Electrophoretic Studies Starch gel electrophoresis was performed on Fraction I-0' and Fraction I-lA' before and after heat denaturation of fibrinogen. The stained gel showed the presence of two components in the mater- ials before heat denaturation. The major component (probably fibrinogen) was present in both fractions before heat denaturation. After denaturation, the major component was diminished, but not absent from Fraction I-O', but was completely removed from Fraction I-lAL Figure 11 is a graphic representation of starch gel electro- phoresis patterns. The inclusion of urea in starch gels used for electrophoresis was reported to improve the resolution of protein components during electrophoresis (Wake and Baldwin, 1961). Starch gels were prepared to contain a final concentration of 7M urea, and used to determine the number of protein components present in purified bovine Factor VIII preparations in various stages of purification. There were many mechanical difficulties encountered in the urea gel electro- phoresis procedure, some caused by improper design of the gel mold. The major problem involved the cross contamination of samples in adjacent sample wells. These difficulties rendered the results of some analyses useless. The important finding from these determinations was that Fraction I-lA'A56°C was resolved into three electrophoretic com- ponents of approximately equal concentrations. Figure 12 is a graphic representation of urea starch gel electrophoresis patterns. 65 Anode O O a so ~o \0 Ln Ln LO Q q d - - ‘<: ‘<: ‘4: ‘4 O O H .—4 .—4 .—4 I l I l I I H H +4 H P4 H i: C: c: c: C.‘ c: o o I o o o o ...4 ".4 .,.q ".4 .H -H U u 4.1 U 4.1 U U U U U U U m m m m m m u u u u u p In in in In La In [ -_-- L--- ._-- Origin Cathode Fig. 11. Starch gel electrophoresis patterns of bovine Factor VIII at various stages of purification. 66 Anode 0 O o \D \D so in m Ln 4 <1 <1 ~ 4 ”a: ‘4: ‘4: O O H r—I H I I I I I H H H H H c: c: r: c: c: O O O O 0 0H OH .H OH 0H u u u u u U U u o 0 CU CU CU CU CU H H H H $4 In Lu In In In —-—~ —--- Origin .Cathode Fig. 12. Urea starch gel electrophoresis patterns of bovine Factor VIII at various stages of purification. 67 No attempt was.made to determine at this point which component or components contained Factor VIII activity. We must tentatively conclude from these results that further purification beyond the Fraction I-lA'l5 56°C stage is needed to reduce, if possible, bovine Factor VIII to a single electrophoretic component. E. Gel Filtration of Bovine Factor VIII Urea gel electrophoretic and immunoelectrophoretic studies showed that highly purified bovine Factor VIII preparations (Fraction I-lA'A56°C) consisted of at least three components. No attempts have been made to show at this point in which of these components the Factor VIII activity is concentrated. For this pur- pose, the components must be separated in sufficient quantity to study this aspect. Separation was attempted by gel filtration studies using G-200 Sephadex in the hope that the molecular size difference of the components would be sufficient to permit separa- tion in this resin. Three gm of dry G-200 Sephadex was allowed to hydrate for four days in 150 ml of pH 6.8 Imidazole-saline buffer at room temperature. A column was established and equilibrated in a 1 meter by 1.2 cm glass tube to a total bed height of 0.71 meters or a total bed volume of 80.5 cc. The void volume of the column was estimated at 24 cc based on the manufacturer's data for this gel. A 10 ml sample of bovine Factor VIII (Fraction I-lA‘A 56°C for 5 minutes) was applied to the column, and eluted with the 68 equilibration buffer (Imidazole-NaCl). Fractions were collected in the amount of 1 ml per fraction at an average flow rate of 0.1 ml per minute. The protein peaks were located by determining the absorbancy of selected. fractions at 280 mu. The protein peaks were visualized by plotting absorbancy (ordinate) vs fraction (abscissa). Due to the optical interference of the Imidazole ring structure at 280 mu, subsequent gel filtration studies were performed using the column equilibrated against(lOSSM citrate buffer at pH 6.8. All fractions containing protein were assayed for Factor VIII activity, and selected fractions analyzed by immunodiffusion and urea gel electrophoresis. In the initial gel filtration studies with G-200 Sephadex (in Imidazole buffer), there was a considerable loss of protein. This loss seemed to be due to a fraction insoluble in the cold, precipi- tating and accumulating at the top surface of the resin bed. The nature of this material was studied to see what effect its removal would have on the activity of the purified preparation. This was accomplished in the following experiment. Fraction I-lA A 56°C was cooled to the range of 0-4°C over a period of 30 minutes to precipitate the cryoglobulins, which were then removed by centrifugation at 20,000 x G for 30 minutes at a temperature of 0-2°C. The supernatant material was decanted and the remaining precipitate redissolved to one-fourth the original sample volume with saline. The starting material, the supernatant fraction, and the redissolved cryoglobulin fraction were assayed for 69 Factor VIII activity, and analyzed by immunodiffusion techniques. The supernatant material was used as the starting material in gel filtration studies. Assays performed on the materials obtained from the cryoglobulin removal studies are summarized in Table VII. The data showed that there was an average loss of 17% of the Factor VIII activity in the cryoglobulin removal step that was not recovered in the redissolved precipitate. This is possibly explained by the poor resolution of the precipitate, or by irrever- sible denaturation. The interesting aspect of these data was the fact that there was no detectable loss of protein with the removal of the cryo- globulin. This cold insoluble material had the characteristics of a cryoglobulin, but there was insufficient material available for analysis and identification as a cryoglobulin. Figure 13 shows a typical elution pattern of protein from the G-200 Sephadex column. Superimposed on this is the pattern of assayable Factor VIII activity in the eluate fractions, the dotted portions were plotted to visualize the two closely associated peaks appearing in the eluate fractions. It can be readily seen from examination of Figure 13 that the starting material was partially separated into at least two compon- ents. The slight shouldering tendency of the trailing edge of the major peak might indicate a possible third component. Second, the Factor VIII activity peak is coincident with the major protein peak. 70 oEDHo> HmcHwHuo mHH.o n mm.o . H¢.o I mmm.o ou wouoouuoo wo>Hommeom ouMuHmHooum - - ww.o o - - a-.H o oaaacoflwoauo ucmumauomsm :HHsnono%uo m.Hw mmm mmH me. m.¢w mHN 0.00 mum. .4H1H aowuomum omom ooH owe HmH me. ooH oom m.Hm mum. «AIH cowuomum N cwmuoum HE\muHc: HE\wE N cHououm HE\muH:: HE\wE Hmwuoumz zuo>ooom wE\muHc: :Hmuoum muo>ooom wE\muH:3 cHououm mu3>3uo< HHH> scuomm c uoH 4 zo3>3oo< HHH> soooom N oog Hm>oEom cHHsnonozuu umum¢ paw unawom uH>Huo< HHH> HOuomm Ho comfludeou .HH> oHan 71 1.2 ‘ 1.0 ‘ 0.8 ‘- :3 E O m N .LJ CU f; 0.6 «J P 0.3 :1 CU 43 H 3 E U} U) .0 u 4 a :1 0.4 - i— 0‘2 >‘ U -:--I > w-I .LJ 0 CU H H 002 ‘ P 0.1; H 0 4..) 0 CU Flt () a 0 Eluate Fraction Fig. 13. Elution patterns of protein and Factor VIII activity form gel filtration of Factor VIII (Fraction I-lA 56°) on G-ZOO Sephadex. 72 Immunodiffusion studies were performed on the fractions from the cryoglobulin study, and on selected eluate fractions from the gel filtration studies. The support medium for immunodiffusion was cellulose acetate OHillipore Cellotate), and the precipitin lines were developed with antibovine Factor VIII serum. Table VIII summar- izes the results of the immunodiffusion studies. Table VIII. Results of Immunodiffusion Studies on Factor VIII Preparations After Heat Treatment, Cryoglobulin Removal, and Gel Filtration Number of Material Precipitin Lines Fraction I-lA' 3 Fraction I-lA' A 56°C 3 Fraction I-lA'A 56°C 3 Cryoglobulin Supernatant Cryoglobulin Fraction 1 Eluate Fractions: Fraction 38 fast component Fraction 41 mixed peaks Fraction 42 slow component peak Fraction 43 slow component HHNH These data confirm that there are two distinct components present in the eluate fractions, and that their separation, however incomplete, has been affected. Complete separation might be possible by applying a smaller sample volume to the column. Thus, assuming the gel filtration studies are valid, bovine Factor VIII has been isolated in small quantities as a single immunologic entity. More 73 extensive studies of this aspect of the purification of bovine Factor VIII preparation are being planned, using refinements and extension of these techniques. F. Tyrosine-Tryptophane Ratios Tyrosine-tryptophane ratios were determined on Fraction I-0' and Fraction I-lA' from all but the first lot by the method of Beavan and Holiday (1952). Table IX gives the values of the calcu- lated constants, molar ratios and molar concentrations of tyrosine and tryptOphane for all determinations. The data show there was greater variation in molar concentra- tions of tyrosine and tryptophane than in the molar ratios. These variations would indicate that the individual lots are signifi- cantly different. C. Amino Acid Analysis l. Glycine Removal Four lots were filtered through G-25 Sephadex to separate the protein and contaminant glycine carried over during the frac- tionation procedure. This step was necessary to insure that glycine content, determined by amino acid analysis, was not in error due to the presence of free glycine. Figure 14 shows a typical elution pattern obtained from gel filtration of Fraction I-lA'A 56°C through G-25 Sephadex. The eluates representing the protein peak (Fractions 65-110) were pooled, and the protein concentrated by precipitation with 5% (v/v) final concentration of trichloroacetic acid, in three 74 00.0 00.0 0H0.0 0H0.0 000.0 005.0 500.H 000.H o0muo>< oo00 00.0 0.0 H00.0 H0.H 0.0 0.0 050.H 000.H 00.0 < 00.H H00.H 000.0 .005.0 050.0 00.0 000.H 000.H 000.0 01H .uh 0 00.H 000.H 000.0 005.0 505.0 50.0 50H.H 000.H 000.0 : 0 00.H 5H0.H 500.0 005.0 000.0 50.0 H00.H H00.H 500.0 : 0 00.H 0H0.H H50.0 005.0 000.0 00.0 000.H 000.H 000.0 : 0 500.H 050.H 000.0 H05.0 000.0 00.0 5H0.H 0H0.H 000.0 .0:H .um 0 no Eoum M Eoum .coocoo .Coocoo :E :E :5 DE HE\0E Hmwumumz .oz . . . umHoz umHoz 0.000 000 0.000 000 :Hmuou uo mxue Z nuke z I 1.- m H 41.09 .H z IluNlllou .H E: mgafiofig oFfiwobflH M m unnumcoo :OHu >u0>0umuom0d nanom0< umHo: umHoz . oosooz kooflaom ooo cm>mom osu >0 cofiumcHEuoqu ucoucoo 0cm 000mm ocmzaanzuHumcHwouxH mo muHSmmm .xH mHan Absorbancy at 280 mu 75 1 0.3'- 0.2“' 0.1“ -Strong . PMOdu -Weak .04 I ~ — ///‘/// I ' Neg..i I l 60 80 100 120‘ 180 E Eluate Fraction No. Fig. 14. _A.typical elution pattern obtained from the separation of Factor VIII and free glycine by gel filtration through G-25 Sephadex. 76 aliquots. Two aliquots were used for protein determinations by micro-Kjeldahl, and the other was submitted for amino acid analysis. 2. Amino.AQid Analysis Three of the four samples submitted for amino acid analysis were analyzed for amino acid content. The other sample was ruined during drying of the hydrolyzed sample. The resulting chromatograms of the three lots were analyzed by the half height method to deter- mine the content of the individual amino acids. The amino acid contents (mg amino acid/gram of protein) were derived from the pro- tein content of the original sample before hydrolysis. Corrections for buffer dilution of the hydrolyzed samples were made to determine protein content of the samples applied to the resin columns of the amino acid analyzer. In the case of one sample (Lot 2), the total weight of recovered residue was 30% lower than expected recovery based on protein content of the original sample. It was thus assumed that the original protein value was in error, and the calculation of amino acid content (mg/gram protein) was based on actual recovery. Table X summarizes the data of amino acid analysis. Nitrogen content of the three lots was calculated from the amino acid analysis data. The calculations were made after making the necessary corrections for the nitrogen content of lysine, histi- dine, arginine and tryptophane, and were based on the total recovered amino acids. The results of these calculations are summarized in Table XI. 77 Table X. Results of Amino Acid Analysis of Three Lots of Bovine Factor VIII Lot 2 Lot 6' Lot 6 mg/gm* m-mole mg/gm m-mole mg/gm m-mole Amino ACid Prot. per gm Prot. per gm Prot. ,per gm Lysine 40.79 .279 41.1 .281 46.6 .319 Histidine 20.17 .13 20.35 .131 19.5 .126 Ammonia 16.14 .948 14.9 .875 18.3 .08 Arginine 60.62 .348 62.6 .359 64.1 .378 ASpartic Acid 88.91 .668 94.5 .711 97.4 .73 Threonine 83.67 .702 85.7 .72 89.1 .747 Serine 66.84 .636 63.25 .602 69.6 .663 Glutamic Acid 125.8 .885 125.3 .852 131.1 .89 Proline 74.03 .643 63.2 .55 67.1 .583 Glycine 44.87 .597 47.6 .635 48.8 .651 Alanine 31.36 .352 30.75 .345 31.4 .353 Half Cystine 32.56 .271 21.45 .179 24.1 .201 Valine 72.51 .691 66.9 .562 71.6 .61 Methionine 11.79 .079 13.5 .091 11.95 .08 Isoleucine 38.83 .295 38.8 .296 40.5 .309 Leucine 58.24 .444 51.3 .10 59.2 .453 Tyrosine 57.07 .315 54.75 .302 57.1 .316 Phenylalanine 33.0 .201 31.9 .195 31.4 .191 Tryptophane** 31.65 .155 26.6 .13 27.5 .135 Totals 988.8 954.45 1006.35 * Lot 2 data based on calculated recovery. **Ca1culated from molar ratio data. 78 Table XI. Nitrogen Content of Three Lots of Factor VIII Calculated from Amino Acid Analysis Data Total Amino Acids Total Nitrogen Lot No. Micrograms Micrograms Z Nitrogen 2' 98808 13935 1411 6' 1,222,4 174.5 14.25 6 1,762.3 251.3 14.25 Comparison of the data of amino acid analysis (Table X) with the results of tyrosine-tryptophane ratio and content obtained by the method of Beavan and Holiday (1952), in Table IX, show signi- ficant differences in regard to tyrosine and tryptophane content. Table XII summarizes these variations in results obtained by the two methods. Table XII. Comparison of Tyrosine and Tryptophane Content from Amino Acid Analysis and the Beavan and Holiday Technique Amino Acid Anal. Beavan and Holiday Difference 1 2 3 4 Lot Tyro. Tryp. Tyro. Tryp. Tyro. Tryp. No m-moles m-moles m-moles m-moles Q, .3 ' per gm per gm per gm per gm 1 2 2 .315 .155 .529 .255 1.67 1.64 6' .302 .13 .945 .418 3.13 3.21 6 .316 .135 1.01 .421 3.2 3.12 79 Because of these differences, another spectrophotometric method for the determination of tyrosine and tryptophane molar ratios and content was investigated. In the past, this procedure had shown good correlation with tyrosine content determined on the amino acid analyzer. The method was first described by Bencze and Schmid (1957) as an improvement on the earlier methods of Holiday (1936 and 1938), and Goodwin and Morton (1946), by eliminating errors caused by the bathochromic shift of absorption spectra of tyrosine and tryptophane. The method consists of determining the absorption spectrum between 270 mu and 350 mu. The slope of a line tangent to the two characteristic maxima of the absorption curve was calculated as follows: Slope (S) = (A A/A mu) x 103 where: A = absorbance A max A max = maximum absorbance mu = wave length Figure 15 is a typical example of the spectrum and tangent line used in this method. The extinction coefficient (Elzcm) and molar ratio of tyrosine-tryptophane (R) were derived from a table of values pre- pared from determinations on known mixtures of the two amino acids. Where necessary, the values were obtained by interpolation for values of S not in the table. The total concentration of tyrosine and tryptophane was cal- culated as follows: A.max 1% 1 cm C tyrosine + tryptophane% = 80 1.0 - 0.9-— Amax s =LAA/ Amu) x 103 Amax 0.8'- 3 >3 U 5?} g 0.7-J O U) .0 4: 0.64— 0.5“ 0.4“ ‘ I fl I 260 280 300 320 Wave Length mu Fig. 15. ”A portion of a typical absorption spectrum as used to calculate tyrosine-tryptophane ratios by the method of Bencze and Schmid (1957). 81 The molar concentrations of tyrosine and tryptophane were calculated using molar ratios and total concentration data. Table XIII summar- izes the calculation of tyrosine-tryptophane ratios and content by the above method. Inspection of the data shows even greater disagree- ment than that obtained by the method of Beavan and Holiday (1952). The lack of agreement of the three methods indicates that one of two situations exists: 1) one method is correct and the other two are in error, or 2) all three methods are in error. It must be pointed out, however, that when the original calculations of amino acid analysis data were made, the molar ratios determined by the Beavan and Holiday (1952) technique were used to determine trypto- phane content. Re-examination of these data shows that the values obtained for tryptophane, assuming the protein determinations were accurate, gave good correlation with the total protein recovered. For example, in the case of Lot 6, 1.75 mg of protein was placed on the column and 1.762 mg was recovered. According to the calcu- lations of the chromatograms, this total includes the calculated value of tryptophane based on the molar ratio. error of 0.6%. For Lot 6', the recovery error 1.1%. Thus, it seems that the ratios obtained Holiday (1952) technique are reasonable, while tions calculated are not in agreement with the data. Due to the limited amounts of material, This represents an was 4.5%, and Lot 2, by the Beavan and the molar concentra- amino acid analysis this disagreement could not be resolved by determination of tyrosine content by other available techniques (i.e., the Folin-Ciocalteau phenol method). 82 0H 0H 00.0 050.0 000.H 5.00H 05.0 1 050.0 0 05.0 5.0 00.0 000.0 000.0 0H.00H 50.0 I 000.0 .0 0.0H 00.0 05.0 005.0 H00.0 0.HOH 50.0 . 0H0.0 .0 00.0 00.0 HH.0 000.0 05.0 H00 00.0 1 000.0 .0 0.0H 00.0 00.0 0H0.0 0H0.0 0.000 00.0H- 000.0 .0 00H x 00H x 80\0E me< umHoz Eu H 0 HE\0E .oz uOH 80 you 60 you .m0ua+.ouxe oGMLQOumzuH IIqu omon chuoum moHoEIE moHoEuE o ocww0p0H o mcmanHSHHV ochouhH 0 Eu H NHm a 0% z me< . maon xm Noam: cu neumchOH H x E u 24 0 m2 3 \4 V mcowumuucmocou 0cm mowumm pmHoz oGMLQOumzuHumchouxH mo coHumHnono .HHHx mHan V. DISCUSSION A. Modification of the Blomback Procedure The purification of Factor VIII from human or bovine plasma requires that it initially be isolated in crude form. The isola- tion of the crude factor is probably the simplest of all the steps required in the purification of this protein. The initial prepara- tion is contaminated with most of the other plasma proteins either as trace components occluded in the precipitate, (i.e., albumins;(, fl and U globulins) or as coprecipitates in large quantities, (i.e., fibrinogen). Blomback (1958) devised methods for removal of the trace contaminants by extraction of the crude precipitates with glycine-containing buffers, and developed methods for the partial separation of Factor VIII and fibrinogen (Factor I) which is truly the major component of the crude preparation. The product (Fraction I-lA) resulting from this partial separa- tion step is still grossly contaminated with fibrinogen, containing about 15% of the fibrinogen initially precipitated in the crude fraction (Blomback, 1958). The removal of fibrinogen and the other contaminants from Fraction I-lA while maintaining Factor VIII activity has been the prime goal of the present research. The removal of fibrinogen from protein solutions is readily accomplished by heat treatment at 56°C for a period of 2 to 5 minutes, depending on the fibrinogen content. 83 84 When this procedure was used on fractions prepared according to the Blomback procedure, significant loss of Factor VIII potency was encountered. At the Fraction I-O stage, the loss of activity due to heating varied from 72 to 78%, and resulted in a 1.25 to 5 fold increase in specific activity. At the Fraction I-lA stage, there was loss of activity due to heating, varying from 65 to 82%, and an 0.8 to 1.7 fold increase in specific activity. The observations of Simonetti g£_a1. (1961) that citrate ions have an adverse effect on the heat stability of Factor VIII preparations, seemed to be a logical starting point for modification of the Blomback procedure. The first comparative study of the Blomback procedure and the initial modification, was made to test the validity of Simonetti's observations. The results of this study showed that there was a reduction in activity loss due to heat treatment at 56°C for two minutes. This aSpect of the modification was overshadowed by an unex- pected, yet more significant, gain in apparent purification of the heat-treated material as seen in the increased specific activity. Table XIV summarizes the differences of the Blomback and initial modified procedure in regard to the effect of heat treatment. Table XIV. Comparison of the Effects of Heat Treatment for Two Minutes at 56°C of Fraction I-0 and Fraction I-lA, Prepared by the Blomback Procedure and the Initial Modified Procedure (Mercer) Procedure Blomback Init. Mod. (Mercer) Fraction Factor VIII Specific Factor VIII Specific Heat Treated Activity Activity Activity Activity at 56°C Loss % Increase Loss % Increase Fraction I-0 78 1.25 54 2.6 Fraction I-1A 65 1.7 71 4.8 85 Thus, it is apparent that although the data on Fraction I-O substantiates the observation of Simonetti, the more important aspect is the actual increased purification obtained by reduced citrate levels in the initial modified procedure. The complete modification, which entailed the elimination of citrate from the purification procedure, resulted in even smaller heating losses of Factor VIII activity. Table XV summarizes the differences in the Blomback procedure and the modified procedure (second comparative study) in regard to the effects of heat treatment of fractions at 56°C for five minutes. Table XV. Comparison of the Effects of Heat Treatment for Five Minutes at 56°C of Fraction I-0 and Fraction I-1A Prepared by the Blomback Procedure and the Modified Procedure (Mercer) Procedure Blomback Modified (Mercer) Fraction Heat Specific Specific Treated 56 C Activity Activity Activity Activity 5 Minutes Loss % Increase Loss % Increase Fraction I-O 72 5 29 11.25 Fraction I—lA 82 0.8 33 1.7 As in the first study, the data showed that the significant advantages of citrate reduction or elimination were decreased heating losses and increased purification of heat-treated materials. The significance of the modification is further substantiated by examination of the data in Tables III and IV which compare the two studies. Both studies show 86 the same trends of increased activity and purification of Factor VIII by the modified procedure. The variations between results of the two experiments appear to be, in part, a function of the varia- tion in assay systems. Eflnnowara (1966) studied thermal denaturation of Factor VIII activity in crude (Fraction I) and highly purified Factor VIII preparations. His experiments were carried out in the presence of citrate-phOSphate buffer at pH 7.3, ionic strength .092. Although the citrate concentration used by Shinowara was lower than that used in the Blomback procedure, the results of thermal denaturation com- pared with those obtained with the Blomback Factor VIII preparations (See Tables XIV and XV) are strikingly similar. Shinowara reported that after five minutes at 56°C, only 15% of the original Factor VIII activity remained. It would be interesting to know what effect the absence of citrate would have had on his findings. From his studies, he has postulated the presence of two forms of Factor VIII having different rates of thermal denaturation; the less labile D form had a thermal half life of six minutes and the labile M form a thermal half life of 0.6 minutes. The data obtained in this present study on bovine Factor VIII shows that thermal stability of Factor VIII is greatly dependent upon the ionic species present in the solution. Although no attempts were made to determine the "thermal" half life of Factor VIII, it has been established that the presence or absence of citrate certainly influences thermal stability of Factor VIII, and that in the absence of citrate, the half life is certainly longer than five minutes. 87 The results of the four fractionation runs with the modified procedure (See Table II) show some variations in protein content of the various fractions. These variations are based on variation in the starting plasma (individual variations of animals). The small variations of protein distribution in the fractions, as indicated best by protein recovery values, show that there was only minor vari- ability between runs. The greatest variation seemed to be found in the fluctuation of Factor VIII activity from run to run, which again can be partly attributed to individual variations among animals. This is indicated by variations in the Factor VIII activity of the starting plasmas. There is, however, one disturbing fact shown by the data. That is the poor recovery of activity obtained in the final product. These great losses in activity during the purification procedure can be attributed to any number of causes, i.e., mechanical loss in pre- cipitation of the crude fraction in that 9% is lost in Super I and only 45% is recovered at the Fraction I-O stage. The goal of this present work was to obtain significant purification regardless of such losses. B. Preparation of Antisera Antisera prepared against bovine plasma and human plasma were of good quality and titer. These antisera proved valuable in purity determinations of Factor VIII preparations. The antiserum against bovine plasma was the most valuable in determining the number of com- ponents present in purified Factor VIII preparations. In many cases, 88 these same determinations were not possible with the specific antisera prepared against Factor VIII. The preparation of antisera against bovine Factor VIII was ham- pered by the low protein content of the injected antigen, resulting in low circulating levels of antigen. Distribution studies were not made on the injected antigen, but had such studies been made, the results would probably have shown that the disappearance rate of antigen was such that inadequate levels of antigen were maintained for full antibody response. The antisera obtained following the injection of adjuvant- Factor VIII emulsions were considerably better. These were valuable for two reasons. Although as prepared they were not specific for a single antigen, the titer was sufficient to develop precipitin patterns against antigens of low concentration. The second advan- tage of preparing antisera against partially purified proteins was that the number of components present in the original antigen could be qualitatively determined indirectly by reacting the antiserum against bovine plasma by immunoelectrophoretic or immunodiffusion techniques. C. Immunoelectrophoretic Analysis The initial immunoelectrophoretic studies performed were made using commercially prepared antisera. These antisera were unsuitable due nathe high degree of cross reaction with heterologous proteins. Immunoelectrophoresis performed with antisera prepared in our labora- tory was helpful in determining the degree of purification obtained 89 at various stages of the procedure. There was onelimitation to its use that was overcome in immunodiffusion studies, that is, the protein dilution suffered during electrophoresis prevented the development of strong precipitin patterns. The results of several studies, (summarized in Table VI), show that there were still two components present at the Fraction I-lA' .AS6°C stage of purification. These two components were present in patterns developed either against bovine plasma antiserum or Factor VIII antiserum. An interesting aspect of these data was the differ- ence in the patterns obtained against Fraction I-O with the two antisera. Three components were present when the pattern was devel- oped against antibovine plasma. The pattern developed with Fraction I-O against Factor VIII antiserum showed only two components. From this it was concluded that the ‘6 precipitin line represented fibrin- ogen, since this antigen was present in plasma but not in the Factor VIII antigen used in the initial preparation of the antiserum. D. Gel Electrophoresis Starch gel electrophoresis of plasma and serum gives much greater resolution than that obtained with the more conventional methods (i.e., moving boundary electrophoresis and electrophoresis on paper strips). Serum is resolved into seven components by moving boundary electrophoresis, whereas in starch gel, at least twice this number can be visualized. 90 Bovine Factor VIII preparations before fibrinogen denaturation were resolved into two components in starch gel. The major com- ponent was assumed to be fibrinogen, and the minor component assumed to be Factor VIII. With the removal of fibrinogen by heat denatura- tion, the major component was absent on subsequent electrophoretic analysis. Thus, it was assumed that Factor VIII was resolved to one component. The inclusion of urea in gel electrophoresis was reported to increase resolution. The suggested mechanisms for this increased resolution are partial denaturation with the subsequent splitting of the subunits. The presenaaof urea further prevents aggregation of the denatured protein subunits GWake and Baldwin, 1961). Bovine Factor VIII preparations found to contain one component by conventional starch gel electrophoresis were further resolved into three components of approximately equal concentrations when 7M urea was included in the gel. It was thus concluded from these results that further purification was required to obtain Factor VIII as a single component. E. Gel Filtration Studies The results obtained from immunoelectrophoresis and urea-starch gel electrophoresis gave evidence that further purification of Factor VIII was required. Earlier observations on attempted small scale purification of human Factor VIII on G-100 Sephadex showed that the exclusion limit (MW=100,000) of this gel was too low to effect detectable separation of Factor VIII and its contaminants 91 (Mercer §£_§1., 1963). The estimated molecular weight of human Factor VIII has been reported as 196,000 (Shulman §£_él°’ 1960)£nd later by Aronson g£_§l. (1962) as 180,000. If these estimates are good, then gel filtration through a resin with an exclusion limit of 200,000 M.W. should prove fruitful for separation provided that there are signifi- cant differences in the molecular weights of Factor VIII and its contaminants. Such was the rationale for selecting G-200 Sephadex. The results of the gel filtration studies show that separation of the components of Factor VIII was accomplished. The separation was only partial in that there was mixing of the peaks (Figure 13). Immunodiffusion studies on various fractions representing the peaks confirm the presence of single immunologic components repre- sented by the peaks and the presence of two components in the mixed- peaks fractions. Theaapreliminary data lend great promise for future research. It was concluded that, based on these data, the two components separated have molecular weights under 200,000 with the major peak (Factor VIII) having the lower molecular weight of the two components. The data, in part, agree with earlier estimates of Factor VIII molecular weight (Shulman gt_al., 1960; Aronson g£_§1., 1962). F. Cryoglobulin Studies The removal of cryoglobulins (cold insoluble proteins) from purified Factor VIII preparations was done for two reasons. First, it was apparent from earlier observation with gel filtration that material was precipitating at the surface of the resin bed. This 92 insoluble material at the surface was undesirable in that it could, if allowed to accumulate in the column, tend to slow and eventually stop the gel filtration procedure. Second, it was essential to know what this protein loss represented in terms of Factor VIII activity. Was it contributing to significant loss of Factor VIII in the eluates? The results showed that cryoglobulin removal represented a loss in Factor VIII activity of 15.5% to 18.5%, with no detectable loss of protein in either case. The total material lost by visual observa- tion of the precipitate was slight, and it is thus possible that it was below the sensitivity of the protein determination. The loss of activity could have been due to the loss of a small quantity of high potency Factor VIII in the precipitate, that was not detected due to the poor resolution of the precipitate. A purely speculative, but not too probable answer to this problem is found in some work on the subfractionation of Fraction I-lA reported by Blomback et a1. (1961). They reported that when Fraction I-lA was adsorbed on tricalcium citrate, and the bulk of the protein eluted with 0.1M EDTA, there was a phospholipid remaining in the tricalcium citrate precipitate. The precipitate, when extracted with organic solvents, exhibited a low level of Factor VIII activity in the extract. This lipid could not, in itself, account for the total activity loss from the eluate. When the lipid and eluate fractions were recombined, the resulting activity was greater than the sum of the two separate activities. In the present case, this explanation would not be valid in that if the cryoglobulin 93 were purely phospholipid, it would have risen to the surface instead of sedimenting. Secondly, lipids are not readily separated by cooling alone as are certain lipoproteins. G. Amino Acid Analysis Analyses of three lots of Factor VIII (Fraction I-lA A 56°C Sephadex eluate) for amino acid composition and content showed that, with the exception of some amino acids, there was good correlation between the individual lots (Table X). Mihalyi §£_§1. (1964) reported on the amino acid composition of bovine fibrinogen. The preparation analyzed was 94% clottable pro- tein. The differences between their data and the data presented here for bovine Factor VIII are obvious. From their data, tyrosine- tryptophane molar ratios were calculated with the resulting ratio of 1.23. Mihalyi reported the data of Tristram (1949) on human fibrin- ogen. The molar ratio calculated from thesadata was 1.85. These data are in fair agreement with the molar ratios reported here (Table IX) for bovine Fraction I-O which is mainly fibrinogen. The data for total nitrogen content of the individual lots showed that, in this respect, the agreement was even better. If the results for nitrogen content are valid, then it must be pointed out that in the case of Factor VIII at this degree of purification, the factor,6,normally used Unconvert nitrogen to protein from Kjeldahl analysis data is in error. Based on the average nitrogen content of 14.2%, this conversion factor should then be the reciprocal of 14.2, or 7.04. 94 H. Tyrosine-Tryptophane Ratio and Content The determination of tyrosine-tryptophane ratios and the molar concentration of these two amino acids was done by two different methods (Beavan and Holiday, 1952; Bencze and Schmid, 1957). The results of these two methods (Table IX and Table XIII) are in com- plete disagreement with each other, and neither method agrees with the molar concentrations as determined by amino acid analysis (Table X). As mentioned earlier, the tryptophane content for the amino acid analysis data was derived using the molar ratio determined by the Beavan and Holiday procedure. The resulting values gave close corre- lation between the recovery of total amino acids and the total protein analyzed. The Beavan and Holiday values for tyrosine and tryptophane concentrations, although not in agreement with amino acid analysis values, are still much closer than the Bencze-Schmid values. Since the calculation of these values was dependent on determination of protein concentration, any error in this later determination would have a great effect on the calculated concentrations of the two amino acids in question. Since it was not in the scope of this thesis to resolve disagreements of these two methods, the data obtained from the Beavan and Holiday procedure was used because of closer agreement to amino acid analysis. 95 I. Significance The modification of the Blomback procedure described has resulted in great increases in the purification of Factor VIII, by changing protein distribution during the purification procedure and reducing the heat losses of Factor VIII during the heat denatur- ation of fibrinogen. The best purification obtained with this modified procedure was 230 fold on a protein basis compared to plasma. This purification was 2.6 fold greater than that obtained by the Blomback procedure from the same starting material. Gel filtration of partially purified Factor VIII has yielded a fraction containing the Factor VIII activity in a single immuno- logic component. Further confirmation of these results and refinements of the techniques used will permit the production of purified Factor VIII on a larger scale. There are many areas yet to be investigated that will be greatly aided by this present study. With a highly purified Factor VIII preparation, it will be possible to prepare a specific antiserum for Factor VIII. This specific antiserum will be used basically in four ways: 1) to develop immunologic assays of Factor VIII in an attempt to resolve the discrepancies of present in yitgg clotting assays, 2) to study 12.Xl££2 distribution of Factor VIII during the purification procedures in an attempt to improve yield of Factor VIII obtained from these procedures, 3) to utilize the antiserum coupled to cellulose derivatives for purification by immuno- absorbant techniques, 4) to use the specific antiserum to study in vivo 96 distribution and degradation of Factor VIII and determine the sites of Factor VIII synthesis and degradation. The eventual aim is to achieve a method of producing Factor VIII of animal origin that will be suitable for less limited clinical use in man. rmm‘““**nluav VI. SUMMARY The purpose of this study was to isolate highly purified bovine blood clotting Factor VIII (antihemophilic globulin). A. This required modifying presently available purification tech- . W- €911“: my niques to assure separation of Factor VIII from fibrinogen with minimal losses of Factor VIII activity. The Blomback procedure was ,. I‘ modified by replacement of sodium citrate with sodium chloride of equivalent ionic strength. In connection with the modification of the purification procedure, it was found that: 1. The elimination of citrate from the reagents used in the purification procedure developed by Blomback resulted in increased purification of Factor VIII. 2. The observed citrate effect was two fold. First, the reduction or elimination of citrate caused changes in protein distribution in the fractionation and purification procedure, re- sulting in increased purification. Second, the reduction or elimin- ation of citrate reduced the losses of Factor VIII incurred during heat denaturation of contaminant fibrinogen at 56°C. 3. The mechanism of the citrate effect was not elucidated by this study, but the possibility of heavy metal contamination of the citrate would be an unlikely cause due to the reported low con- centrations of heavy metals (0.0001% as Pb) in the citrate used. 97 98 B. Studies on purified Factor VIII obtained from the modified procedure were performed to determine the actual purification ob- tained. From these studies it was found that: 1. Starch gel electrophoretic studies on Factor VIII showed that denaturation of fibrinogen by heating resulted in a preparation purified to a single electrophoretic component. 2. Antisera prepared against the purified Factor VIII prepar- ation showed that the preparation was not homogeneous. a. Immunoelectrophoretic studies showed the presence of two immunologic components in the purified preparations. b. Immunodiffusion studies on cellulose acetate membranes visualized three immunologic components present in the purified preparations. 3. Urea-starch gel electrophoretic analysis of the purified preparations confirmed the presence of three protein components. 4. Amino acid analysis of several preparations showed that the preparations were quite uniform on the basis of amino acid composition. C. Further purification of Factor VIII preparations was accomplished on a small scale by cryoglobulin precipitation followed by gel filtra- tion with the following results: 1. A cryoglobulin precipitate was removed from the fibrinogen- free Factor VIII preparations that resulted in minor losses of Factor VIII activity. 99 2. Gel filtration of the cryoglobulin-free Factor VIII preparation on G-200 Sephadex resulted in a partial separation of two components. 3. Assay of the eluate fractions from gel filtration showed that the peak of Factor VIII activity was coincident with the major protein component. 4. Immunodiffusion studies on the redissolved cryoglobulin precipitate and eluate fractions representing the components from gel filtration showed that each material was a single immunologic component. REFERENCES Addis, T. 1910. The Pathogenesis of Hereditary Hemophilia. J. Path. and Bact., 15, 427. Adelson, E., J. J. Rheingold, 0. Parker, M. Steiner, and h. C. Kirby. 1963. The Survival of Factor VIII (Antihemophilic Globulin) and Factor IX (Plasma Thromboplastin Component) in Normal Humans. J. Clin. Invest., 42, 1040. Aggeler, P. M., S. G. White, M. B. Glendening, E. W. Page, T. B. Leake, and G. Bates. 1952. Plasma Thromboplastin Component (PTC) Deficiency, A New Disease Resembling Hemophilia. Proc. Soc. Exper. Bio. Med., 12, 691. Alexander, B., and G. Landwehr. 1948. Studies of Hemophilia. II. The Assay of Antihemophilic Clot-Promoting Principle in Normal Human Plasma With Some Observations on the Relative Potency of Certain Plasma Fractions. J. Clin. Invest., 27, 98. Anderson, H.D. 1964. .A Program Which Coordinates the Activities of Blood Banks, Maternity Hospitals, Physicians and the Department of Health Laboratories to Provide All Useful Derivatives in Adequate Quantity with a Minimum of Cost to the Recipient. Proc. 9th Congr. int. Soc. Blood Trans., Mexico 1962. p. 664. Anderson, H.D. 1964. The Use of Human Blood Derivatives in a Population Where Such Products are Readily Available in Adequate Quantity Without Cost. Proc. 9th Congr. int. Soc. Blood Trans., Mexico 1962. p. 669. Aronson, D. L., J. W. Priess, and M. W. Mosesson. 1962. Molecular Weights of Factor VIII (AHF) and Factor IX (PTC) by Electron Irradiation. Thromb. Diath. Haemorrhag., 8, 270. .Astrup, T., P. Brakman, P. Ollendorff, and J. Rasmussen. 1960. Haemostasis in Haemophilia in Relation to the Haemostatic Balance in the Normal Organism and the Effect of Pea Nuts. Thromb. Diath. Haemorrhag., 5, 329. Astrup, T., P. Brakman, and K.-E. delin. 1962. Haemophilia and the Protease Inhibitor in Peanuts. Nature, 194, 980. Beale, A.J., and P. J. Mason. 1962. Measurement of D-Antigen in Poliovirus Preparations. J. Hyg.,,§Q, 113. 100 101 Beavan, G. E., and E. R. Holiday. 1952. Ultraviolet Absorption Spectra of Proteins and Amino Acids. In: Advances in Protein Chemistry, 1, 319. Academic Press, New York. Bendien, W. M., and S. van Creveld. 1939. On Some Factors of Blood Coagulation Especially with Regard to the Problem of Hemophilia. Acta. Med. Scand., 22, 12. Bencze, W. L., and K. Schmid. 1957. Determination of Tyrosine and Tryptophane in Proteins. Analyt. Chem., 22, 1193. Bidwell, E. 19553. The Purification of Bovine AHG. Brit. J. Haematol., 2, 35. Bidwell, E. 1955b. The Purification of Antihemophilic Globulin from Animal Plasma. Brit. J. Haematol., 2, 386. Biggs, R., A. S. Douglas, R. G. MacFarlane, J. V. Dacie, W. R. Pitney, A C. Mersky, and J. R. O'Brien. 1952. Christmas Disease; A Condition Previously Mistaken for Hemophilia. Brit. Med. J., 2, 1378. Biggs, R. 1960. The Treatment of Hemophilic Patients With Animal Antihaemophilic Globulin. Acta Haemat., 22, 124. Bizzozero, G. 1882. Uber ein neuen Formbestandtheil des Blutes und desen Rolle bei der Thrombose und der Blutgerinnung. Virchows Arch. f. path. Anat., 22, 261. Blomback, M. 1958. Purification of Antihemophilic Globulin. I. Some Studies on the Stability of Antihemophilic Globulin Activity in Fraction I-0 and a Method for its Separation From Fibrinogen. Arkiv. f. kemi, 22, 387. Blomback, B., M. Blomback, and I Struwe. 1961. Studies on Blood Clotting Factor VIII. Arkiv. f. kemi, 22, 223. Boudreaux, H. D., and V. L. Frampton. 1960. A PeanUt Factor for Haemostasis in Haemophilia. Nature, (Lond.) 185, 469. Boyer, S. H., D. C. Fainer, and M. A. Naughton. 1963. Myoglobin Inherited Structural Variation In Man. Science, 140, 1228. Brakman, P., K.-E. Sjélin, and T. Astrup. 1962. Is There A Hemo- static Effect of Peanuts in Hemophiloid Disorders? Thromb. Diath. Haemorrhag., 2, 442. Breckenridge, R. T., and O. D. Ratnoff. 1962. Studies On The Nature of a Circulating Anticoagulant Directed Against Antihemophilic Factor: With Notes On An Assay of Antihemophilic Factor. Blood, 22, 137. 102 Brinkhous, K. M. 1947. Clotting Defect in Hemophilia: Deficiency in a Plasma Factor Required for Platelet Utilization. Proc. Soc. Exper. Biol. and Med., 22, 117. Brinkhous, K. M. 1964. Hemophilia - Pathophysiologic Studies and the Evolution of Transfusion Therapy. Am. J. Clin. Path., 22, 342. Campbell, D.H., et a1. 1963. Methods in Immunology, W. A. Benjamin Co., New York. Cohn, E. J., L. E. Strong, W. L. Hughes, Jr., D. J. Mulford, J. N. .Ashworth, M. Melin, and H. L. Taylor. 1946. Preparation and Properties of Serum and Plasma Proteins. IV. A System for the Separation Into Fractions of the Protein and Lipoprotein Components of Biological Tissues and Fluids. J. Am. Chem. Soc., 22, 459. Collingwood, B. J., and M. T. MacMahon. 1912. The Anticoagulants in Blood and Serum. J. Physiol., 22, 119. Davie, E. W., and O. D. Ratnoff. 1964. Waterfall Sequence for Intrinsic Blood Clotting. Science, 145, 1310. Douglas, A. S. 1958. Antihemophilic Globulin Assay Following Plasma Infusions in Hemophilia. J. Lab. Clin. Med., 22, 850. Duckert, F.,P. Fluckiger, M. Matter, and F. Koller. 1955. Clotting Factor X Physiologic and Physico-Chemical Properties. Proc. Soc. Exper. Biol. and Med., 22, 17. Ferris, T. G., R. E. Easterling, and R. E. Budd. 1963. Electro- phoresis of Serum Proteins in Acrylamide Gel. I. Removal of Background Stain by Means of Electric Current. Am. J. Clin. Path., 22, 383. Field, R. C., W. F. Enneking, and G. Rothstein. 1963. Elective Surgery in Hemophilia. I. Correction of Talipes Equinus Deformity. J. Amer. Med. Assoc., 185, 628. Frank, E., and E. Hartmann. 1927. Uber das Wesen und die therapeu- tische Korrektur der hamophilen Gerinnungsstorung. Klin. Wchnschr., g, 435. Frankel, G. J., and G. E. Honey. 1955. Gunshot Wounds in a Hemo- philic Patient: Successful Treatment by Animal Antihaemo- philic Globulin and_Surgery-, Lancet, 2, 1117. 103 Fuld, E., and K. Spiro. 1904. Der Einfluss einiger gerinnung- shemmender Agentein auf das Vogelplasma. Beitr. Chem. Physiol. u. Path., 2, 171. Gallick, H. 1965. The Development of an Assay and Method of Puri- fication for Human Blood Clotting Factor XI. Ph.D. Thesis. Michigan State University. Gaston, L. W. 1964. The Blood Clotting Factors (In Two Parts). New Eng. J. Med., 270, 236 and 290. Goodwin, T. W., and R. A. Morton. 1946. The Spectrophotometric f Determination of Tyrosine and Tryptophane In Proteins. * Biochem. J., 39, 628. Grabar, P., and C. A. Williams, Jr. 1953. Méthode Permettant 1'etude conjugee des proprietés electrophoretiques et immunochimiques d'un anelange de proteines. Application au serum sanguin. Biochem. Biophys. Acta., 29, 193. W»: .2 Gugler, E. 1961. Zur therapeutischen Anwendung der Fraktion I nach Cohn. Bibl. haemat., 22, 270. ’ Handley, D. A., N. S. Painter, and M. R. P. Hall. 1961. Surgical Treatment of Duodenal Ulceration in Two Haemophiliacs. Lancet, 2, 482. Hardisty, R. M., and J. C. MacPherson. 1962. A One-Stage Factor VIII (Antihemophilic Globulin) Assay and Its Use on Venous and Capillary Plasma. Thromb. Diath. Haemorrhag., 2, 215. Hayem, G. 1896. Du caillot nairétractile: suppression de la formation du serum sanguin dans quelques états pathologique. Comp. rend. Acad. d. Sc., 123, 894. Holiday, E. R. 1936. Spectrophotometry of Proteins. 1. Absorption Spectra of Tyrosine, Tryptophane and Their Mixtures. II. Estimation of Tyrosine and Tryptophane in Proteins. Bioch. J., 22, 1795. Holiday, E. R., and A. G. Ogston. 1938. The Ultraviolet Absorp- tion and Potentiometric Titration Curves of Serum Proteins and Some Others. Biochem. J., 22, 1166. Hougie, C., E. M. Barrow, and J. B. Graham. 1957. Stuart Clotting Defect. I. Segregation of an Hereditary Hemorrhagic State from the Heterozygous Group Heretofore Called"Stable Factor". (SPCA, Proconvertin, Factor VII). Deficiency. J. Clin. Invest., 22, 485. 104 Ingram, G. I. C. 1962. Major Surgery in Haemophilia Use of Animal Antihaemophilic Factor in Britain. Transfusion, 2, 88. Janiak, A., and J. P. Soulier. 1962. Preparation d'une fraction anti-hemophilique A.a partir de plasma ACD standard. Thromb. Diath. Haemorrhag., 2, 406. Johnson, C.M., J. C. N. Westwood, and M. Beaulieu. 1964. A Contin- uous-flow Micro Double Diffusion technique with Cellulose Acetate. Nature, 204, 1321. Jones, A. F. D. 1805. A treatise on the Process Employed by Nature 1 in Suppressing its Haemorrhage of Divided and Punctured Arteries and the Use of the Ligature. Longman, Hurse, Rees, Orme and Brown, London. 1810. Joules, H., and R. G. MacFarlane. 1938. Pseudo-haemophilia in a a Woman. Lancet, 2, 715. Leitner, A., E. Bidwell, and G. W. R. Dike. 1963. An Antihaemo- philic Globulin (Factor VIII) Inhibitor. Purification Characterization and Reaction Kinetics. Brit. J. Haemat., _9_, 245. Lozner, E. L., L. S. Jolliffe, and F. H. L. Taylor. 1940. Hemorrhagic Diathesis with Prolonged Coagulation Time Associated with Circulatory Anticoagulant. Am. J. M. Sc., .222, 318. Lane, S. 1840. Haemorrhagic Diathesis: Successful Transfusion of Blood. Lancet, 2, 185. Cited by Brinkhous, K. M. Am. J. Clin. Path., 32, 342. 1964. Langdell, R. D., R. H. Wagner, and K. M. Brinkhous. 1953. Effect of Antihemophilic Factor on One Stage Clotting Tests. J. Lab. Clin. Med., 32, 637. Lorand, L., and K. Laki. 1954. A Simple Method for Purifying an Activator of Prothrombin (Antihemophilic Factor?). Biochem. Biophys. Acta, 22, 448. Ma, T. S., and G. Zuazaga. 1942. Micro Kjeldahl Determination of Nitrogen. Ind. Eng. Chem., Anal. Ed., 23, 280. MacFarlane, R. G., R. Biggs, and E. Bidwell. 1954. Bovine Anti- haemophilic Globulin in the Treatment of Hemophilia. Lancet, 2, 1316. 105 MacFarlane, R. G., P. Mallam, L. Witts, E. Bidwell, R. Biggs, G. Frankel, G. Honey, and K. Taylor. 1957. Surgery in Hemophilia: The Use of Animal Antihaemophilic Globulin and Human Plasma in Thirteen Cases. Lancet, 2, 251. MacFarlane, R. G., and R. Biggs. 1959. The Treatment of Haemo- philic Patients with Animal Antihemophilic Globulin. In: Brinkhous, K. M.: Hemophilia and Other Hemorrhagic States. Chapel Hill. New York, Van Rees Press, 1959. p. 19. MacFarlane, R. G. 1964. An Enzyme Cascade in the Blood Clotting Mechanism, and its Function as a Biochemical Amplifier. Nature, 202, 498. Mammen, E. F. 1963. Investigations on the Nature of Hemophilia A. Thromb. Diath. Haemorrhag., 2, 30. Mammen, E. F. 1964. Concentration of Platelet Cofactor I (AHG) from Bovine Serum. Thromb. Diath. Haemorrhag., 22, 127. Maycock, W. D'AJ, S. Evans, L. Vallet, B. Combridge, P. Wolf, B. H. MacGibbon, E. E. French, L. H. Wallet, J. V. Dacie, R. Biggs, D. Handley, and R. G. MacFarlane. 1963. Further Experience With a Concentrate Containing Human Antihemophilic Factor. Brit. J. Haemat., 2, 215. McLester, W., and R. Wagner. 1965. Antibody to Antihemophilic Factor and its Lack of Reaction with Hemophilic Plasma. Amer. J. Physiol., 208, 499. McMillan, C. W., L. K. Diamond, and D. M. Surgenor. 1961. Treatment of Classic Hemophilia: The Use of Fibrinogen Rich in Factor VIII for Hemorrhage and Surgery. New Eng. J. Med., 265, 224. Mercer, J. E., H. Gallick, and L. A. Hyndman. Unpublished observations. Michael, S. W., and G. W. Tunnah. 1963. The Purification of Factor VIII (Antihaemophilic Globulin). Brit. J. Haemat., 2, 236. Mihalyi, E., P. A. Small, Jr., and J. P. Cooke. 1964. The Amino Acid Composition of Bovine Fibrinogen. Arch. Biochem. and Biophys., 106, 229. Morand, S. F. 1736. Mem. d. Acad. Roy de Chirurgie. V Ed., p. 8. Morawitz, P. 1904. Bietrage zur Kenntnis der Blutgerinnung. Beitr. Chem. Physiol. u Path., 2, 131. 106 Munro, F. L., and H. W. Jones. 1943. The Detrimental Effect of Frequent Transfusions in the Treatment of a Patient with Hemophilia. Am. J. M. Sc., 206, 710. Niemetz, J.,C. Weilland, and J. P. Soulier. 1961. Preparation d'une Fraction Plasmatique Humaine Riche en Facteur VIII (Antihemophilique A) et pauvre en Fibrinogene. Nouvelle Rev. Francaise d'Hematologie, 1, 880. Nolf, P. 1938. Coagulation of the Blood. Medicine, 21, 381. Otto, J. C. 1803. An Account of an Hemorrhagic Disposition Existing in Certain Families. Med. Reposit., Q, 1. Owen, C. A., and J. L. Bollman. 1948. Prothrombin Conversion Factors of Dicoumarol Plasma. Proc. Soc. Exper. Biol. and Med., 32, 231. Owren, P.A. 1947. The Coagulation of Blood: Investigation on a New Clotting Factor. Acta. Med. Scand., 194, 223. Patek, A. J., and R. P. Stetson. 1936. Hemophilia I. The Abnormal Coagulation of the Blood and its Relation to the Blood Platelets. J. Clin. Invest., 22, 531. Patek, A. J., and F. H. L. Taylor. 1937. Hemophilia II: Some Properties of a Substance Obtained from Normal Human Plasma Effective in Accelerating the Coagulation of Hemophilic Blood. J. Clin. Invest., 22, 113. Pavlovsky, A., C. Simonetti, G. Casillas, L. J. Bergna, A. Andino, and A. E. Bachman. 1961. The Use of Factor VIII Freed from Fibrinogen in the Treatment of Hemophilic Patients. Brit. J. Haemat., l, 365. Petit, J. E. 1731. Mem. de l'Acad. Roy. des Sci. p. 90. Pool, J. G., E. J. Hershgold, and A. R. Pappenhagen. 1964. High Potency Antihemophilic Factor Concentrate Prepared from a Cryoglobulin Precipitate. Nature, 203, 312. Quick, A. J. 1943. On the Constitution of Prothrombin. Am J. Physiol., 140, 212. Quick, A. J. 1947. Studies on the Enigma of the Hemostatic Dysfunction of Hemophilia. Am. J. Med. Sci., 214, 272. Quick, A. J. 1957. Hemorrhagic Diseases. Lea and Febiger, Phila- delphia. 107 Ratnoff, O. D., and J. E. Colopy. 1955. A Familial Hemorrhagic Trait Associated with a Deficiency of a Clot-Promoting Fraction of Plasma. J. Clin. Invest., 23, 602. Rosenthal, R. L., O. H. Dreskin, and N. Rosenthal. 1953. New Hemophilia-Like Disease Caused by Deficiency of a Third Plasma Thromboplastin Factor. Proc. Exper. Biol and Med., 22, 171. Rosenthal, R. L., L. H. Dreskin, and N. Rosenthal. 1955. Plasma Thromboplastin Antecedent (PTA) Deficiency: Clinical Coagu- lation, Therapeutic Aspects of a New Hemophilia-Like Disease. Blood, 22, 120. Sahli, H. 1905. Uber das Wesen der Hamophilie. Ztschr. klin. Med., ‘3 29,264. 11 Seegers, W. H., R. Landaburu, and R. L. Fenichel. 1957. Isolation of Platelet Cofactor I from Plasma and Some Properties of the Preparation. Am. J. Physiol., 190, 1. Seegers, W. H., E. Mammen, J. M. Lee, R. H. Landaburu, M. H. Cho, W. J. Baker, and R. S. Shepard. Further Purification of Platelet Cofactor I. In: Brinkhous, K. M. Hemophilia and Other Hemorrhagic States, Van Rees Press, 1959. New York. Shinowara, G. Y. 1966. Kinetic Studies on the Thermal Denaturation of Human Antihemophilic Globulin (AHG). Abstracts of 14th Annual Symposium on Blood. Wayne State Univ. School of Med., Detroit, Mich., Jan. 22, 1966. Shulman, S., R. H..Landaburu, and W. H. Seegers. 1960. Biophysical Studies on Platelet Cofactor I Preparations. Thromb. Diath. Haemorrhag., 3, 336. SimonettL C., G. Casillas, and A. Pavlovsky. 1961. Purification du Facteur VIII Antihemophilique (FAH).Hemostase, 2, 57. Smith, P. H. 1965. Management of Splenectomy in Hemophilia: Case of Spontaneous Rupture of Spleen Due to Glandular Fever in a Hemophiliac. Brit. Med. Jour., 2, 296. Smithies, O. 1959. An Improved Procedure for Starch-gel E1ectro~ phoresis: Further Variations in the Serum Proteins of Normal Individuals. Biochem. J., 12, 585. Spaet, T. H., and B. G. Kinsell. 1953. Properties of Bovine Anti- hemophilic Factor. Proc. Soc. Exper.Biol. and Med., 22, 314. 108 Strauss, H. K., S. V. Kerry, L. K. Diamond. 1965. Ineffectiveness of Prophylactic EACA in Severe Hemophilia. New Eng. J. Med., 273(6), 301. Taylor, F. H. L., C. S. Davidson, H. J. Tagnon, M. A. Adamis, A. H. MacDonald, and G. R. Minot. 1945. Studies in Blood Coagu- lation. The Coagulation Properties of Certain Globulin Fractions of Normal Human Plasma in Vitro. J. Clin. Invest., 24, 698. Thorne, C. B., and F. C. Belton. 1957. An Agar Diffusion Method for Titrating B. Anthracis Immunizing Antigen and its Appli- cation to a Study of Antigen Production. J. Gen. Microbiol., 17, 505. Tristram, G. R. 1949. Amino Acid Composition of Purified Proteins. Advan. in Protein Chem.,.é, 83. van Creveld, S., H. A. Veder, C. N. Pascha, and W. F. Kroeze. 1959. The Separation of AHF from Fibrinogen I. Thromb. Diath. Haemorrhag.,‘2, 572. van Creveld, S., H. A. Veder, and C. N. Pascha. 1960. The Separa- tion of AHF from Fibrinogen II. Thromb. Diath. Haemorrhag., 4, 211. van Creveld, S., C. N. Pascha, and H. A. Veder. 1961. The Separation of AHF from Fibrinogen III. Thromb. Diath. Haemorrhag., Q, 282. Verstraete, M. 1957. Excess of Anticoagulant as a Cause of the Clotting Defect in Hemophilia. In” Brinkhous, K. M. Hemophilia and Hemophiloid Diseases. Chapel Hill. Van Rees Press, New York. 1957. Wagner, R-, W. McLester, M. Smith, and K. Brinkhous. 1964. Purification of Antihemophilic Factor (Factor VIII) by Amino Acid Precipitation. Thromb. Diath. Haemorrhag. 22, 64. Wake, R. G., andIbL.Baldwin. 1961. Analysis of Casein Fractions by Zone Electrophoresis in Concentrated Urea. Biochem. et Biophys. Acta., 21, 225. 111' 3129"“1'1'111I11111 111111111