Textbook of Hemophilia

By Christine A. Lee, Erik E. Berntorp and W. Keith Hoots

The Textbook of Hemophilia has become a definitive resource for all those managing hemophilia patients. It covers all the common and rare bleeding disorders, both in terms of clinical management as well as the genetic, laboratory, financial and psychological aspects.

This second edition covers all the latest developments in the field of hemophilia, with new chapters on:

• the genetic and molecular basis of inherited blood disorders
• how to manage adolescent and older patients
• emergency medicine and inherited blood disorders
• national hemophilia databases

Drawing on the vast experience of the authors, the aim of this textbook remains the same - to improve the care of patients suffering from hemophilia. The book is full of detailed guidance and advice on everyday clinical questions making it invaluable to all trainee and practicing hematologists.

• Language: English
• Publisher: Wiley
• Release date: Jul 5, 2011
• ISBN: 9781444347708

Book preview

Part I: Introduction

1

Overview of hemostasis

Kathleen Brummel Ziedins and Kenneth G. Mann

University of Vermont, College of Medicine, VT, USA

Introduction

The maintenance of blood fluidity and protection from blood leakage provide major biophysical challenges for the organism. Nature has evolved a highly complex, integrated, and dynamic system which balances the presentations of procoagulant, anticoagulant, and fibrinolytic systems. These systems function collectively to maintain blood within the vasculature in a fluid state while at the same time providing potent leak attenuating activity which can be elicited upon vascular perforation to provide the rapid assembly of a thrombus principally composed of platelets and fibrin to attenuate extravascular blood loss. The dynamic control of this system is such that the coagulation response is under the synergistic control of a variety of blood and vascular inhibitors, resulting in a process that is regionally restricted to the site of vascular damage and does not propagate throughout the vascular system. The rapid coagulation response is also tightly linked to the vascular repair process during which the thrombus is removed by the fibrinolytic system which also is activated regionally to provide clot removal coincident with vascular repair.

A list of important procoagulant, anticoagulant, and fibrinolytic proteins, inhibitors, and receptors can be seen in Table 1.1.

Importance of complex assembly to coagulation

Laboratory data combined with clinical pathology lead to the conclusion that the physiologically relevant hemostatic mechanism is primarily composed of three procoagulant vitamin K - dependent enzyme complexes (which utilize the proteases factor VIIa, factor IXa, and factor Xa) and one anticoagulant vitamin K - dependent complex (which utilizes the protease thrombin) [1,2] (Figure 1.1). These complexes—extrinsic factor Xase (tissue factor–factor VIIa complex), intrinsic factor Xase (factor VIIIa–factor IXa complex) [3], and the protein Case complex (thrombin–thrombomodulin) [4]—are each composed of a vitamin K - dependent serine protease, a cofactor protein and a phospholipid membrane; the latter provided by an activated or damaged cell. The membrane binding properties of the vitamin K - dependent proteins are a consequence of the post - translational γ- carboxylation of these macromolecules [5]. The cofactor proteins are either membrane binding (factor Va, factor VIIIa), recruited from plasma, or intrinsic membrane proteins (tissue factor, thrombomodulin). Cofactor–protease assembly on membrane surfaces yields enhancements in the rates of substrate processing ranging from 10⁵–10⁹- fold relative to rates observed when the same reactions are limited to solution - phase biomolecular interactions between the individual proteases (factor VIIa, factor IXa, and factor Xa) and their corresponding substrates [6–8] (Figure 1.2a). Membrane binding, intrinsic to complex assembly, also localizes catalysis to the region of vascular damage. Thus, a system selective for regulated, efficient activity presentation provides for a regionally limited, vigorous arrest of hemorrhage.

Additional complexes associated with the intrinsic pathway are involved in the surface contact activation of blood [3]. However, the association of the contact - initiating proteins (factor XII, prekallikrein, high - molecular- weight kininogen) with hemorrhagic disease is uncertain [9].

Of equal importance to the procoagulant processes is regulation of anticoagulation by the stoichiometric and dynamic inhibitory systems. The effectiveness of inhibitory functions are far in excess of the potential procoagulant responses. These inhibitory processes provide activation thresholds, which require presentation of a limiting concentration of tissue factor prior to significant thrombin generation [10]. Antithrombin and tissue factor pathway inhibitor [11] are the primary stoichiometric inhibitors while the thrombin–thrombomodulin–protein C system (protein Case, Figure 1.1) is dynamic in its function.

Extrinsic pathway to blood coagulation

The initiating event in the generation of thrombin involves the binding of membrane - bound tissue factor with plasma factor VIIa [12]. The latter is present in blood at ~0.1 nM [~1–2% of the factor VII concentration (10 nM)] [13]. Plasma factor VIIa does not express proteolytic activity unless it is bound to tissue factor; thus factor VIIa at normal blood level has no significant activity toward either factor IX or factor X prior to its binding to tissue factor. The inefficient active site of factor VIIa permits its escape from inhibition by the anti-thrombin present in blood. Vascular damage [14] or cytokine related presentation of the active tissue factor triggers the process by interaction with activated factor VIIa, which increases the k cat of the enzyme and increases the rate of factor X activation by four orders of magnitude [15]. This increase is the result of the improvement in catalytic efficiency and the membrane binding of factor IX and factor X.

Table 1.1 Procoagulant, anticoagulant, and fibrinolytic proteins, inhibitors, and receptors.

+, presence of phenotype; −, absence of phenotype; ±, some individuals present with the phenotype and others do not; H, hemorrhagic disease/hemophilia; T, thrombotic disease/thrombophilia; VKD; vitamin K -dependent proteins.

aClinical phenotype; the expression of either hemorrhagic or thrombotic phenotype in deficient individuals.

The tissue factor–factor VIIa complex (extrinsic factor Xase) (Figure 1.2) catalyzes the activation of both factor IX and factor X, the latter being the more efficient substrate [16]. Thus, the initial product formed is factor Xa. Feedback cleavage of factor IX by membrane - bound factor Xa enhances the rate of generation of factor IXa in a cooperative process with the tissue factor–factor VII complex [17].

The initially formed, membrane - bound factor Xa activates small amounts of prothrombin to thrombin [18]. This initial prothrombin activation provides the thrombin essential to the acceleration of the hemostatic process by serving as the activator for platelets [19], factor V [20], and factor VIII [21] (Figure 1.1). Once factor VIIIa is formed, the factor IXa generated by tissue factor–factor VIIa combines with factor VIIIa on the activated platelet membrane to form the intrinsic factor Xase (Figure 1.2a), which becomes the major activator of factor X. The factor VIIIa–factor IXa complex is 10⁹- fold more active as a factor X activator and 50 times more efficient than tissue factor–factor VIIa in catalyzing factor X activation [22,23]; thus, the bulk of factor Xa is ultimately produced by the factor VIIIa–factor IXa complex (Figure 1.2).

Figure 1.1 Overview of hemostasis. Coagulation is initiated through two pathways: the primary extrinsic pathway (shown on the right) and the intrinsic pathway (historically called the contact or accessory pathway , shown on the left). The components of these multistep processes are illustrated as follows: enzymes (open circle ), inhibitors (hatched circles ), zymogens (open boxes ), or complexes open ovals ). Fibrin formation is also shown as an oval. The intrinsic pathway has no known bleeding etiology associated with it, thus this path is considered accessory to hemostasis. Upon injury to the vessel wall, tissue factor, the cofactor for the extrinsic factor Xase complex, is exposed to circulating factor VIIa and forms the vitamin K - dependent complex, the extrinsic factor Xase. Factor IX and factor X are converted to their serine proteases factor IXa (FIXa) and factor Xa (FXa), which then form the intrinsic factor Xase and the prothrombinase complexes, respectively. The combined actions of the intrinsic and extrinsic factor Xase and the prothrombinase complexes lead to an explosive burst of the enzyme thrombin (IIa). In addition to its multiple procoagulant roles, thrombin also acts in an anticoagulant capacity when combined with the cofactor thrombomodulin in the protein Case complex. The product of the protein Case reaction, activated protein C (aPC), inactivates the cofactors factors Va and VIIIa. The cleaved species, factors Vai and VIIIai, no longer support the respective procoagulant activities of the prothrombinase and intrinsic Xase complexes. Once thrombin is generated through procoagulant mechanisms, thrombin cleaves fibrinogen, releasing fibrinopeptide A and B (FPA and FPB) and activate factor XIII to form a cross - linked fibrin clot. Thrombin–thrombomodulin also activates thrombin activatable fibrinolysis inhibitor (TAFIa) that slows down fibrin degradation by plasmin. The procoagulant response is downregulated by the stoichiometric inhibitors tissue factor pathway inhibitor (TFPI) and antithrombin (AT). TFPI serves to attenuate the activity of the extrinsic factor Xase trigger of coagulation. Antithrombin directly inhibits thrombin, factor IXa and factor Xa. The intrinsic pathway provides an alternative route for the generation of factor IXa. Thrombin has also been shown to activate factor XI. The fibrin clot is eventually degraded by plasmin - yielding soluble fibrin peptides. Modified from [32].

As the reaction progresses, factor Xa generation by the more active intrinsic factor Xase complex exceeds that of the extrinsic factor Xase complex [24]. In addition, the extrinsic factor Xase complex is subject to inhibition by tissue factor pathway inhibitor (Figure 1.2 b) [25]. As a consequence, most (>90%) of factor Xa is ultimately produced by the factor VIIIa–factor IXa complex in the tissue factor- initiated hemostatic processes. In hemophilia A and hemophilia B the intrinsic factor Xase complex cannot be assembled, and amplification of factor Xa generation does not occur [26]. Factor Xa combines with factor Va on the activated platelet membrane receptors and this factor Va– factor Xa prothrombinase catalyst (Figure 1.2a) converts pro-thrombin to thrombin. Prothrombinase is 300 000 - fold more active than factor Xa alone in catalyzing prothrombin activation [6].

Figure 1.2 Vitamin K - dependent complex assembly. (a) The factor Xa generated by the tissue factor–factor VIIa complex activates a small amount of thrombin which activates factor V and factor VIII leading to the presentation of the intrinsic factor Xase (factor VIIIa–factor IXa) and prothrombinase (factor Va–factor Xa) complexes. At this point in the reaction factor IXa generation is cooperatively catalyzed by membrane - bound factor Xa and by the tissue factor–factor VIIa complex. The thick arrow representing factor Xa generation by the intrinsic factor Xase illustrates the more efficient factor Xa generation by this catalyst. (b) The tissue factor pathway inhibitor (TFPI) interacts with the tissue factor–factor VIIa–factor Xa product complex to block the tissue factor - initiated activation of both factors IX and factor X. Inhibition of the extrinsic factor Xase complex results in the factor VIIIa–factor IXa complex (intrinsic factor Xase), becoming the only viable catalyst for factor X activation. Used with permission from the Dynamics of Hemostasis Haematologic Technologies, K.G. Mann, 2002.

Attenuation of the procoagulant response

The coagulation system is tightly regulated by the inhibition systems. The tissue factor concentration threshold for reaction initiation is steep and the ultimate amount of thrombin produced is largely regulated by the concentrations of plasma procoagulants and the stoichiometric inhibitors and the constituents of the dynamic inhibition processes [24]. Tissue factor pathway inhibitor blocks the tissue factor–factor VIIa–factor Xa product complex, thus effectively neutralizing the extrinsic factor Xase complex (Figure 1.2 b) [27]. However, tissue factor pathway inhibitor is present at low abundance (~2.5 nM) in blood and can only delay the hemostatic reaction [28]. Antithrombin, normally present in plasma at twice the concentration (3.2μM) of any potential coagulation enzyme, neutralizes all the procoagulant serine proteases primarily in the uncomplexed state [11].

The dynamic protein C system is activated by thrombin binding to constitutive vascular thrombomodulin (Protein Case). This complex activates protein C to its activated species activated protein C (Figure 1.1) [4]. Activated protein C competes in binding with factor Xa and factor IXa and cleaves factor Va and factor VIIIa, eliminating their respective complexes [20]. The protein C system, tissue factor pathway inhibitor, and activated protein C cooperate to produce steep tissue factor concentration thresholds, acting like a digital switch, allowing or blocking thrombin formation

[10].

In humans, the zymogen factor XI which is present in plasma and platelets has been variably associated with hemorrhagic pathology [29]. Factor XI is a substrate for thrombin and has been invoked in a revised pathway of coagulation contributing to factor IX activation (Figure 1.1) [30]. The importance of the thrombin activation of factor XI is evident only at low tissue factor concentrations [26].

Factor XII, prekallekrein, and high-molecular-weight kininogen (Figure 1.1) do not appear to be fundamental to the process of hemostasis [31]. The contribution of these contact pathway elements to thrombosis remains an open question and further experimentation is required to resolve this issue [31].

Conclusion

Advances in genetics, protein chemistry, bioinformatics, physical biochemistry, and cell biology provide an array of information with respect to normal and pathologic processes leading to hemorrhagic or thrombotic disease. The challenge for the 21st century will be to merge mechanism - based, quantitative data with epidemiologic studies and subjective clinical experience associated with the tendency to bleed or thrombose and with the therapeutic management of individuals with thrombotic or hemorrhagic disease. In vitro data and clinical experience with individuals with thrombotic and hemorrhagic disease will ultimately provide algorithms which can combine the art of clinical management with the quantitative science available to define the phenotype vis á vis the outcome of a challenge or the efficacy of an intervention [28–34].

Acknowledgment

The authors were supported by HL46703, HL34575, and HL07594 from the National Institutes of Health National Heart, Lung, and Blood Institute.

References

1 Mann KG, Nesheim ME, Church WR, Haley P, Krishnaswamy S. Surface - dependent reactions of the vitamin K - dependent enzyme complexes. Blood 1990; 76: 1–16.

2 Brummel - Ziedins K, Orfeo T, Jenny NS, Everse SJ, Mann KG. Blood coagulation and fibrinolysis. In: Greer JP, Foerster J, Lukens J, Rodgers GM, Paraskevas F, Glader B (eds.) Wintrobe’s Clinical Hematology . Philadelphia: Lippincott Williams & Wilkins, 2003: 677–774.

3 Davie EW, Ratnoff OD. Waterfall sequence for intrinsic blood clotting. Science 1964; 145: 1310–12. 4 Esmon CT. The protein C pathway. Chest 2003; 124: 26S–32S. 5 Stenflo J. Contributions of Gla and EGF - like domains to the function of vitamin K - dependent coagulation factors. Crit Rev Eukaryot Gene Expr 1999; 9: 59–88.

4 Esmon CT. The protein C pathway. Chest 2003 ; 124 : 26S–32S.

5 Stenflo J. Contributions of Gla and EGF–like domains to the function of vitamin K - dependent coagulation factors. Crit Rev Eukaryot Gene Expr 1999 ; 9 : 59–88.

6 Nesheim ME, Taswell JB, Mann KG. The contribution of bovine Factor V and Factor Va to the activity of prothrombinase. J Biol Chem 1979; 254: 10952–62.

7 van Dieijen G, Tans G, Rosing J, Hemker HC. The role of phospholipid and factor VIIIa in the activation of bovine factor X. J Biol Chem 1981; 256: 3433–42.

8 Komiyama Y, Pedersen AH, Kisiel W. Proteolytic activation of human factors IX and X by recombinant human factor VIIa: effects of calcium, phospholipids, and tissue factor. Biochemistry 1990; 29: 9418–25.

9 Davie EW. A brief historical review of the waterfall/cascade of blood coagulation. J Biol Chem 2003; 278: 50819–32.

10 van’t Veer C, Golden NJ, Kalafatis M, Mann KG. Inhibitory mechanism of the protein C pathway on tissue factor- induced thrombin generation. Synergistic effect in combination with tissue factor pathway inhibitor. J Biol Chem 1997; 272: 7983–94.

11 Olson ST, Bjork I, Shore JD. Kinetic characterization of heparin catalyzed and uncatalyzed inhibition of blood coagulation proteinases by antithrombin. Methods Enzymol 1993; 222: 525–59.

12 Nemerson Y. Tissue factor and hemostasis. Blood 1988; 71: 1–8.

13 Morrissey JH, Macik BG, Neuenschwander PF, Comp PC. Quantitation of activated factor VII levels in plasma using a tissue factor mutant selectively deficient in promoting factor VII activation. Blood 1993; 81: 734–44.

14 Osterud B. The role of platelets in decrypting monocyte tissue factor. Semin Hematol 2001; 38: 2–5.

15 Bom VJ, Bertina RM. The contributions of Ca²+, phospholipids and tissue - factor apoprotein to the activation of human blood coagulation factor X by activated factor VII. Biochem J 1990; 265: 327–36.

16 Osterud B, Rapaport SI. Activation of factor IX by the reaction product of tissue factor and factor VII: additional pathway for initiating blood coagulation. Proc Natl Acad Sci USA 1977; 74: 5260–64.

17 Lawson JH, Mann KG. Cooperative activation of human factor IX by the human extrinsic pathway of blood coagulation. J Biol Chem 1991; 266: 11317–27.

18 Butenas S, DiLorenzo ME, Mann KG. Ultrasensitive fluorogenic substrates for serine proteases. Thromb Haemost 1997; 78: 1193–201.

19 Brass LF. Thrombin and platelet activation. Chest 2003; 124: 18S–25S.

20 Mann KG, Kalafatis M. Factor V: a combination of Dr Jekyll and Mr Hyde. Blood 2003; 101: 20–30.

21 Fay PJ. Subunit structure of thrombin - activated human factor VIIIa. Biochim Biophys Acta 1988; 952: 181–90.

22 Mann KG, Krishnaswamy S, Lawson JH. Surface - dependent hemostasis. Semin Hematol 1992; 29: 213–26.

23 Ahmad SS, Rawala - Sheikh R, Walsh PN. Components and assembly of the factor X activating complex. Semin Thromb Hemost 1992; 18: 311–23.

24 Hockin MF, Jones KC, Everse SJ, Mann KG. A model for the stoichiometric regulation of blood coagulation. J Biol Chem 2002; 277: 18322–33.

25 Girard TJ, Warren LA, Novotny WF, Likert KM, Brown SG, Miletich JP, Broze GJ, Jr. Functional significance of the Kunitz type inhibitory domains of lipoprotein - associated coagulation inhibitor. Nature 1989; 338: 518–20.

26 Cawthern KM, van’t Veer C, Lock JB, DiLorenzo ME, Branda RF, Mann KG. Blood coagulation in hemophilia A and hemophilia C. Blood 1998; 91: 4581–92.

27 Baugh RJ, Broze GJ, Jr, Krishnaswamy S. Regulation of extrinsic pathway factor Xa formation by tissue factor pathway inhibitor. J Biol Chem 1998; 273: 4378–4386.

28 Novotny WF, Brown SG, Miletich JP, Rader DJ, Broze GJ, Jr. Plasma antigen levels of the lipoprotein - associated coagulation inhibitor in patient samples. Blood 1991; 78: 387–93.

29 Seligsohn U. Factor XI deficiency. Thromb Haemost 1993; 70: 68–71.

30 Gailani D, Broze GJ, Jr. Factor XI activation in a revised model of blood coagulation. Science 1991; 253: 909–12.

31 Colman RW. Contact activation pathway: Inflammatory, fibrinolytic, anticoagulant, antiadhesive and antiangiogenic activities. In: Colman RW, Hirsh J, Marder VJ, Clowes AW, George JN (eds.) Hemostasis and Thrombosis: Basic Principles & Clinical Practice . Philadelphia: Lippincott Williams & Wilkins, 2001: 103–121.

32 Undas A, Brummel - Ziedins KE, Mann KG. Antithrombotic properties of aspirin and resistance to aspirin: beyond strictly antiplatelet actions. Blood 2007; 109: 2285–92.

33 Brummel - Zeidins K, Vossen CY, Rosendaal FR, Umezaki K, Mann KG. The plasma hemostatic proteome: thrombin generation in healthy individuals. J Thromb Haemost 2005; 3: 1472–81.

34 Mann GK, Brummel - Ziedins K, Undas A, Butenas S. Does the genotype predict the phenotype? Evaluations of the hemostatic proteome. J Thromb Haemost 2004; 2: 1727–34.

2

Cellular processing of factor VIII and factor IX

Michael U. Callaghan¹ and Randal J. Kaufman²

¹Children’s Hospital of Michigan, Wayne State University, Detroit, MI, USA

²Howard Hughes Medical Institute, University of Michigan Medical School, Ann Arbor, MI, USA

Factor VIII

Factor VIII and h emophilia A

Blood coagulation is a finely controlled process which enables plasma and cellular blood components to perform their functions in a fluid phase. However, upon damage to the lining of a blood vessel an insoluble clot must be formed at the site of injury to minimize loss of blood components. This process is initiated by activation of platelets and the formation of a primary platelet plug followed by the coordinated and highly regulated formation of a stable fibrin mesh. Maintenance of hemostasis relies on the regulated interaction of the vitamin K-dependent proteases, protease cofactors, membrane surfaces and receptors, calcium ions, and protease inhibitors. This process is characterized by the rapid and sequential activation of three separate vitamin K-dependent serine pro-teases — factors VII, factor IX, and factor X — and their cof actor complexes — tissue factor, factor VIII, and factor V — that make up the intrinsic, extrinsic, and common coagulation pathways, respectively. These pathways act to rapidly and efficiently cleave the vitamin K-dependent zymogen pro-thrombin to its active serine protease form, thrombin, at the site of injury, which leads to cleavage of soluble fibrinogen to insoluble fibrin and clot formation. Factor VIII acts as an essential cof actor for factor IX in the intrinsic coagulation cascade, amplifying factor IX activity by several orders of magnitude. The physiologic significance of these pathways is evident from genetic deficiencies that result in bleeding disorders. In the absence of factor VIII clot formation is impaired leading to prolonged bleeding. Mutations in F8 , the gene coding for coagulation FVIII leading to deficiency of factor VIII or impaired factor VIII function, result in the clinical disease hemophilia A. Hemophilia A has been recognized for over 2000 years as an X-linked bleeding disorder characterized by spontaneous bleeding into joints and muscles and severe bleeding from traumas. Treatment of hemophilia A has steadily improved since the discovery in the nineteenth century that whole-blood transfusion improved coagulation in patients with hemophilia. In the 1980s the gene for factor VIII was cloned and this discovery led quickly to production of recombinant factor VIII in mammalian cells for replacement therapy in patients. All the proteins involved in the coagulation cascade require post-translational modifications for appropriate secretion, plasma half-life, and function-recombinant DNA technology has provided the ability to produce safe and efficacious preparations of factor VIII replacement therapy. Gene therapy approaches for hemophilia are rapidly approaching, and need to consider, the requirement for proper post-translational modification in protein secretion and function.

Domain structure of factor VIII

Factor VIII and factor V are homologous glycoproteins that serve as cofactors for proteolytic activation of factor X and prothrombin, respectively. These cofactors act to increase the V max of substrate activation by four orders of magnitude. They have a conserved domain organization of A1–A2–B–A3–C1–C2 (Figure 2.1) [1]. The A domains of factors V and VIII are homologous to the A domains of the plasma copper-binding protein ceruloplasmin. Copper has been detected in factor VIII and its presence is associated with functional factor VIII activity [2]. One mole of reduced Cu(I) was detected in recombinant factor VIII and likely resides within a type 1 copper ion-binding site within the A1 domain [3]. The C domains are homologous to phospholipid-binding proteins such as milk-fat globule protein, suggesting a role in phospholipid interaction. While the amino acid sequences in the A and C domains are 40% identical between factors V and VIII, there is only limited homology between the B domains. However, the B domains of both proteins have conserved the addition of a large number of asparagine-linked oligosaccharides as well as a large number of serine/threonine-linked oligosaccharides, suggesting a role of the carbohydrate in cof actor function.

Recently, the crystal structure of a B-domain-less factor VIII was solved revealing a triangular heterotrimer composed of the three A domains with the A1 domain interacting with the C2 domain and the A3 domain interacting with the C2 domain [4,5]. This crystal structure and biochemical studies have yielded an in silico model of the activated factor VIII–activated factor IX complex with factor IXa wrapping across the side of factor VIIIa and forming an extended area of interaction including large portions of both the heavy and light chains of factor VIII [4]. Interestingly, these factor VIII structures contain two Cu²+ ions and one or two Ca²+ ions and three asparagine-linked carbohydrate moieties which are essential to the structure [4,5].

Figure 2.1 Domain structure and processing of factor VIII. The structural domains of factor VIII are depicted: A1 domain (1–336), A2 domain (372–740), B domain (740–1648), A3 domain (1690–2020) and the C domains (2021–2332). Above, the pairing of disulfide bonds is shown. Below are represented the potential N-linked glycosylation sites (vertical bars up). Three regions (stippled areas) rich in acidic amino acid residues and lying between domains A1 and A2, A2 and B, B and A3 contain sites of tyrosine sulfation (s). Intracellularly, factor VIII is cleaved within the B domain after Arg1313 and Arg1648 to generate an approximately 200-kDa peptide and the 80-kDa light chain. The two cleavages required for thrombin activation are indicated (**). The sites for aPC cleavage and inactivation are also shown (*).

Factor VIII contains a 19-amino-acid signal peptide that is removed upon translocation into the endoplasmic reticulum (ER). Factor V is secreted from hepatocytes as a single-chain polypeptide of 330 kDa. Factor VIII is processed within the secretory pathway in the cell to yield a heterodimer primarily composed of a heavy chain extending up to 200 kDa (primarily two species from residues 1 to 1313 or 1648, where residue 1 is the amino-terminal amino acid after signal peptide cleavage) in a metal ion-dependent association with an 80 kDa light chain (residues 1649 to 2332) (Figure 2.2). This association is stabilized by noncovalent interactions between the amino-terminal and carboxy-terminal ends of the factor VIII light chain with the amino-terminus of mature von Willebrand factor (VWF). VWF interaction stabilizes factor VIII upon secretion from the cell, inhibits factor VIII binding to phospholipids, and increases the half-life of factor VIII circulating in plasma [6,7]. The ratio of VWF to factor VIII is maintained at 50:1, where an increase or decrease in the plasma VWF level results in a corresponding change in the level of factor VIII.

Factor V and factor VIII circulate in plasma as inactive precursors that are activated through limited proteolysis by either thrombin or activated factor X (Xa). Thrombin activation of factor VIII results in cleavage initially after Arg740 and subsequently after Arg residues 372 and 1689 [8]. Cleavage at both Arg372 and Arg1689 is required for activation of factor VIII procoagulant activity. The cleavage at Arg1689 releases activated factor VIII from VWF, thereby relieving the inhibitory activity of VWF on factor VIII, permitting the activated form of factor VIII to interact with negatively charged phospholipids. Thrombin-activated factor VIII consists of a heterotrimer of a 50-kDa A1-domain-derived polypeptide, a 43-kDa A2-domain-derived polypeptide, and a 73-kDa derived light-chain fragment [9,10]. Upon thrombin activation, the B domains of both factors V and VIII are released. The amino-terminal sides of the thrombin cleavage sites within factors V and VIII are rich in acidic amino acids and contain the post-translationally modified amino acid, tyrosine sulfate.

Disulfide bond formation

Factor VIII and factor V also have a conserved disulfide bonding pattern in which two disulfide bonds occur within the A1 and A2 domains, whereas only the small disulfide loop is present in their A3 domains. In addition, each C domain in factor V and VIII contains one disulfide bond [11]. There are a number of nondisulfide-bonded cysteine residues within factor VIII; one cysteine residue is not oxidized in each A domain and there are four cysteine residues within the B domain that are also likely not oxidized. Disulfide bond formation occurs in the oxidizing environment of the ER and it is possible that protein chaperones, such as protein disulfide isomerase, are important to ensure proper disulfide bond formation and exchange occurs prior to exit from the ER. Factor VIII contains a total of eight disulfide bonds, seven of which, interestingly, are found in factor V [12]. Replacement of cysteine residues with glycine in any of the seven conserved bonding pairs in factor VIII resulted in impaired secretion while elimination of the disulfide loop pairing residues 1899 to 1903 in factor VIII resulted in improved secretion [12].

Figure 2.2 Synthesis, processing, and secretion of factor VIII in mammalian cells. The factor VIII primary translation product is translocated into the lumen of the endoplasmic reticulum (ER), where N-linked glycosylation occurs. A fraction of factor VIII binds tightly to the protein chaperone BiP and requires adenosine triphosphate (ATP) hydrolysis for release [43]. A portion of factor VIII is retrotranslocated into the cytoplasm and is degraded by the cytosolic 26S proteasome. Another fraction of the molecules interact with the lectins calnexin/calreticulin and then with the protein chaperone complex LMAN1–MCFD2 for transit to the Golgi apparatus. In the Golgi apparatus additional processing occurs that includes complex modification of carbohydrate on N-linked sites, addition of carbohydrate to serine and threonine residues, sulfation of tyrosine residues, and cleavage of the protein to the mature heavy and light chains.

Asparagine -and serine/threonine-linked glycosylation

Addition of N-linked oligosaccharides to many glycoproteins is an obligatory event for the folding and assembly of newly synthesized polypeptides. The presence of oligosaccharides is of ten required for the efficient transport of individual glycoproteins through the secretory pathway [13,14]. In addition, N-linked glycosylation frequently affects the plasma half-life and biologic activity of glycoproteins. The consensus site for N-linked glycosylation is Asn–Xxx–Ser/Thr, where Xxx may be any amino acid except for proline. The utilization of a particular consensus site for N-linked oligosaccharide attachment is determined by the structure of the growing polypeptide. As a consequence, proteins expressed in heterologous cells most frequently exhibit occupancy of N-linked sites very similar to that of the native polypeptide.

After addition of the high-mannose-containing oligosaccharide core structure (composed of glucose3- mannose 9- N-acetylglucosamine2) to consensus asparagine residues, trimming begins with the removal of the three terminal glucose residues that is mediated by the action of glucosidases I and II. Glucosidase I removes the terminal α1–3 glucose and glucosidase II subsequently removes the two α1–2 glucose residues. Glucose trimming is required for binding to the protein chaperones calnexin (CNX) and calreticulin (CRT) within the lumen of the ER. Prolonged association with CNX and/or CRT is observed when proteins are unfolded, misfolded, or unable to oligomerize. CNX and CRT bind most avidly to monoglucosylated forms of the N-linked core structure. Removal of the third glucose from the oligosaccharide core structure correlates with release from CNX and CRT and transport to the Golgi apparatus. The selectivity in binding of unfolded glycoproteins to CRT and CNX is mediated by reglucosylation of the deglucosylated N-linked oligosaccharide. This reglucosylation activity is performed by a UDP glucose:glycoprotein glucosyltransferase (UGT). Only unfolded, mutant, or unassembled proteins are subject to reglucosylation. Reglucosylated proteins rebind CNX and/or CRT and, in this manner, unfolded proteins are retained in the ER through a cycle on CNX–CRT interaction, glucosidase II activity, and UGT activity. Subsequent to glucose trimming in the ER, at least one α1–2- linked mannose is removed by an ER α1–2- mannosidase prior to transport out of the ER. This system acts as a quality control mechanism allowing only properly folded factor V or factor VIII to be translocated out of the ER for secretion while sequestering or sending misfolded proteins to the proteasome through ER associates degradation pathways.

Upon transit through the Golgi apparatus, a series of additional carbohydrate modifications occur that are separated spatially and temporally and involve the removal of mannose residues by Golgi mannosidases I and II and the addition of N-acetylglucosamine, fucose, galactose, and sialic acid residues. These reactions occur by specific glycosyltransferases that modify the high-mannose carbohydrate to complex forms. Also within the Golgi apparatus, O-linked oligosaccharides are attached to the hydroxyl of serine or threonine residues through an O-glycosidic bond to N-acetylgalactosamine. Serine and threonine residues subject to glycosylation are frequently clustered together and contain an increased frequency of proline residues in the region, especially at positions −1 and +3, relative to the glycosylated residue. Galactose, fucose, and sialic acid are frequently attached to the serine/threonine-linked N-acetylgalactosamine. O-glycosylation occurs in the Golgi complex concomitant with complex processing of N-linked oligosaccharides.

Factor V and factor VIII contain a large number of N-linked oligosaccharides. Comparison of the N-linked oligosaccharides present on recombinant factor VIII expressed in mammalian cells to human plasma-derived factor VIII indicated that both proteins display similar occupancy and complexity at the N-linked sites [6]. However, a detailed analysis demonstrated that differences in the microheterogeneity of oligosaccharides present on human plasma-derived factor VIII and recombinant factor VIII produced in baby hamster kidney cells do exist [15]. The light chains of factor VIIIa and factor Va migrate as doublets upon SDS-PAGE because of differences in the complexity of N-linked oligosaccharides present on the light chain [16]. The difference in complexity of the N-linked sugars on the light chain does not affect factor VIII activity. The majority of N-linked oligosaccharides within factor VIII and factor V occur within the B domain. Recent studies indicate that the N-linked oligosaccharides within the factor V and factor VIII B domains may be important to interact with the protein chaperone complex LMAN1–MCFD2 for facilitated transport from the ER to the Golgi compartment [17,18] (Figure 2.2). Mutations in either of the subunits of this heterodimeric complex cause combined deficiency of coagulation factors V and VIII [19,20]. MCFD2 appears to recruit factor V and VIII to the LMAN1 cargo complex and LMAN1 acts to recycle MCFD2 back to the ER. In the absence of LMAN1, MCFD2 is secreted from the cell [21]. In patients with combined factor V and VIII-deficiency plasma this results in lower circulating levels of factors V and VIII for patients with MCFD2 mutations compared with those with LMAN1 mutations [22].

Detailed analysis of recombinant factor VIII demonstrated that 3% of the total sugar chains contain a Galα1–3Gal group on some of the outer chains of the bi, tri, and tetra-antennary complex-type sugar chains that is absent on factor VIII derived from human plasma. This structure was present in Kogenate ® (prepared from baby hamster kidney cells) and not in Recombinate ® [prepared from Chinese hamster ovary (CHO) cells] [15]. The α1–3-galactosyltransferase that produces this structure is expressed in most nonprimate mammalian cells, and primates frequently develop antibodies to this structure. Approximately 1% of immunoglobulin in human plasma is directed toward this moiety, so it is expected that antibodies should be detected. A limited clinical trial did not detect any difference in the efficacy and/or half-life of factor VIII that contains the Galα1–3Gal group. Therefore, there is no evidence of detrimental effects of this structure present on recombinant factor VIII.

Chaperone-assisted factor VIII folding

Factor VIII protein is secreted at markedly lower levels than similar proteins including factor V [23]. Secretion of factors V and VIII in the proper tertiary and quaternary structure requires considerable chaperone assistance in the ER. It has been established that folding of factor VIII is considerably more onerous than even the homologous protein factor V and this accounts for the major difference in secretion [16]. Misfolded, unfolded, or defective proteins are refolded, sequestered in the ER or degraded through ER-associated degradation (ERAD) by retrotranslocation into the cytosol and degradation by the 26S proteasome [24–27]. In addition to the calnexin/calreticulin system that acts in a quality control capacity, the ER has a complex tripartite system to match protein-folding capacity to protein load, termed the unfolded protein response (UPR) [28]. In addition to detection of unfolded proteins in the ER lumen, reactive oxygen species induced by factor VIII expression act a signal to activate the UPR [29]. These reactive oxygen species also impair protein folding and antioxidant treatment improved factor VIII secretion [29]. This pathway acts to decrease general protein translation, increase protein-folding chaperones, and increase degradation of misfolded proteins in order to decrease the protein-folding stress on the ER. The IRE/XBP-1 arm of the UPR is largely responsible for increasing chaperone-assisted protein-folding capacity through spliced XBP-1 mediated transcription of chaperones. Overexpression of spliced XBP-1 with factor VIII resulted in increased BiP/GRP78, an ER lumenal chaperone protein of the heat shock protein 70 (hsp70) family. As factor VIII is translated and translocated into the ER, BiP interacts with factor VIII transiently and retains factor VIII within the ER lumen. Expression of factor VIII induces transcription of the BiP gene [30] and the level of BiP inversely correlates with the factor VIII secretion efficiency. Overexpressed BiP complexes with factor VIII in the ER and results in decreased factor VIII secretion [31]. Site directed mutation of F309S of factor impairs interaction with BiP and improves factor VIII secretion efficiency [32,33]. Dissociation of BiP from wild-type factor VIII requires much more ATP than dissociation from the F309S mutant [32,34].

Tyrosine sulfation

Sulfate addition to tyrosine as an O⁴-sulfate ester is a common post-translational modification of secretory proteins that occurs in the trans-Golgi apparatus and is mediated by tyrosylprotein sulfotransferase that utilizes the activated sulfate donor 3′-phosphoadenosine 5′- phosphosulfate (PAPS). This modification occurs on many secretory proteins including a number of proteins that interact with thrombin, such as hirudin, fibrinogen, heparin cof actor II, α2- antiplasmin, vitronectin, and bovine factor X. In addition, both factor V and factor VIII contain multiple sites of tyrosine sulfation [35–37]. Tyrosine sulfation can modulate the biologic activity, binding affinities, and secretion of specific proteins. For example, tyrosine sulfation at the carboxy-terminus of hirudin increases its binding affinity to the anion-binding exosite of thrombin [38].

Recombinant factor VIII contains six sites of tyrosine sulfation at residues 346, 718, 719, 723, 1664, and 1680 [35]. All sites are sulfated to near completion, so it does not appear that this modification is inefficient in CHO cells. Site-directed mutagenesis was used to change individual or multiple tyrosine residues to the conserved residue phenylalanine in order to identify their role in factor VIII function. Tyrosine sulfation at all six sites was required for full factor VIII activity. In addition, mutagenesis of Tyr1680 to Phe demonstrated that sulfation at that residue was required for high-affinity interaction with VWF [37,39]. In the absence of tyrosine sulfation at 1680 in factor VIII, the affinity for VWF was reduced by fivefold. In contrast, mutation at residue Tyr1664 did not affect VWF interaction. The significance of the Tyr1680 sulfation in vivo is made evident by the presence of a Tyr1680 to Phe mutation that causes a moderate hemophilia A, likely owing to reduced interaction with VWF and decreased plasma half-life [40]. The other sites of tyrosine sulfation within factor VIII affect the rate of cleavage by thrombin at the adjacent thrombin cleavage site. It was suggested that thrombin selectively utilizes the tyrosine sulfate residues adjacent to cleavage sites in factors V and VIII to facilitate interaction and/or cleavage.

Phosphorylation of serine and t hreonine residues

Phosphate has been observed in factor V and factor VIII, although its significance remains unknown. Exposure of factors V and VIII to activated platelets results in phosphorylation of serine residues in factor V and primarily threonine residues in factor VIII [41]. Phosphorylation can occur within the heavy chain of factor Va and both the heavy chain and light chains of factor VIII, possibly within the acid-rich regions. Phosphorylation of factor VIII by casein kinase II is thought to occur within the acidic regions 337 through 372 and 1649 through 1689. Although the kinase responsible for the phosphorylation remains unknown, it may be related to casein kinase II [42]. Partially phosphorylated factor Va was shown to be more sensitive to activated protein C (aPC) inactivation, suggesting that phosphorylation of these cofactors may downregulate their activity.

Proteolytic processing

Factor VIII proteolytic processing within the B domain after arginine residues 1313 and 1648 can saturate the proteolytic machinery of the cell. Both arginine residues at 1313 and 1648 have consensus sites for furin cleavage. In this case, secretion of heavy chains that extend to residue 1648 and secretion of light chains that extend to 1313 can be detected. In addition, some single-chain factor VIII is detected in conditioned medium from transfected mammalian cells and in heparin treated human plasma [6,43]. However, all analyses to date indicate that these partially processed products of factor VIII have identical activity to fully processed factor VIII. For example, double mutation of Arg1313Ile and Arg1648Ile yields a single-chain factor VIII molecule with functional activity similar to wild-type factor VIII [16].

Summary

Eukaryotic cells contain an extensive machinery to modify polypeptides that transit the secretory compartment. In the case of coagulation factors VIII, a large number of post translational modifications occur; many are required for secretion of the polypeptide and others are required for functional activity of the polypeptide. Proper synthesis and secretion of factor VIII requires that the primary translation product is modified by signal peptide cleavage and core high-mannose oligosaccharide addition upon translocation into the lumen of the ER. Within the ER, factor VIII requires trimming of glucose residues on the core N-linked glycols for transport to the Golgi compartment. In the Golgi compartment additional modifications occur, which include: (i) tyrosine sulfation of six residues that are required for efficient activation by thrombin and for high-affinity VWF interaction; (ii) extensive addition of oligosaccharides to many serine/threonine residues within the B domain; (iii) complex modification of N-linked glycols; and (iv) cleavage of single-chain factor VIII to its heavy-and light-chain species. To date, there do not appear to be any specific post-translational modifications that significantly limit secretion and/or functional activity of factor VIII. Further studies are required to elucidate the effect of factor VIII expression in different cell types in order to identify the importance that subtle differences in post-translational modifications may have on their secretion, in vivo half-life, and function. These considerations will be important when considering different cells and tissues as targets for gene therapy.

Factor IX

Factor IX and h emophilia B

Hemophilia B is caused by mutations in the F9 gene leading to deficient or defective factor IX in its role as a serine protease in the intrinsic coagulation cascade. These mutations result in slow clot formation and prolonged bleeding. Hemophilia B was recognized as a clinical entity distinct from hemophilia A when it was noted that mixing plasma from patients with one hemophilia corrected the prolonged clotting times from patients with the other hemophilia [44,45]. The primary structure of human factor IX was determined from affinity-purified factor IX from plasma and the cDNA sequence was cloned using oligonucleotides derived from the bovine factor IX amino acid sequence [46–48]. This work led to the development of recombinant factor IX produced in CHO cells for clinical use in patients with hemophilia B [49,50]. Prior to being secreted from hepatocytes, factor IX undergoes γ-carboxylation, O-and N-linked glycosylation, phosphorylation, sulfation, disulfide bond formation, and β-hydroxylation, as well as cleavage of the signal peptide and propeptide. Gene therapy trials in hemophilia B have shown safety, but, to date, little efficacy [51]. Improved understanding of the mechanisms of factor IX processing and secretion has important implications for both production of recombinant factor IX and future gene therapy trials in hemophilia B.

The domain structures of the vitamin K-dependent coagulation factors factor VII, factor IX, factor X, prothrombin, protein C, and protein S deduced from their cDNA sequences demonstrate that they contain common structural features (Figure 2.3) [52]. All contain a signal peptide that is required for translocation into the lumen of the ER. This is followed by a propeptide that directs vitamin K-dependent γ-carboxylation of the mature polypeptide. Upon transit through the trans-Golgi apparatus, the propeptide is cleaved away. The amino-terminus of the mature protein contains a γ-carboxy glutamic acid-rich region (Gla) that includes a short α- helical stack of aromatic amino acids. Then there are two epidermal growth factor (EGF)-like domains. In factor IX, protein C, and factor X, the amino-terminal EGF domain contains β-hydroxyaspartic acid (Hya) at homologous locations. The next region is the activation peptide (12–52 residues) that is glycosylated on asparagine residues and is released by specific proteolysis accompanying activation. The remainder of the vitamin K-dependent protease comprises the serine protease catalytic triad that is absent in protein S.

Figure 2.3 Domain structure and processing of factor IX. Factor IX is composed of a signal peptide, propeptide, γ-carboxyglutamic acid domain (GLA), epidermal growth factor (EGF)-like domains, activation peptide, and serine protease catalytic domain. Short arrows represent intracellular processing sites that cleave away the signal peptide and the propeptide. The long arrows represent the cleavages required for activation by factor VIIa/tissue factor (VIIa/TF) or factor XIa. The 35 amino acid activation peptide is indicated. γ- represents γ-carboxyglutamic acid and β- represents β-hydroxyaspartic acid. The 330–338 loop that interacts with factor VIII is shown by a dotted line. Also indicated are sites of addition of asparagine linked oligosaccharides (N), serine or threonine-linked oligosaccharides (S and T, respectively), tyrosine sulfation (Y- S), and serine phosphorylation (S-P).

Disulfide bond formation

The vitamin K-dependent coagulation factors, exemplified by factor IX, have conserved disulfide bonds. Generally, three disulfide bonds occur within each EGF domain, and several disulfide bonds occur within the serine protease catalytic domain. In addition, factor IX has a disulfide bond that connects the amino-terminal halfwith the carboxy-terminal half of the protein so that after activation the two portions of the molecule do not dissociate. In factor IX, cysteine residues at 18 and 23 within the Gla domain form a small essential disulfide loop where mutations at either cysteine residue results in severe hemophilia B [53].

Asparagine-and serine/threonine-linked glycosylation

With the development of recombinant factor IX produced in CHO cells for treatment of hemophilia B, a detailed characterization and comparison of the carbohydrate structures was performed between plasma-derived and recombinant-derived factor IX [54]. In both plasma-and recombinant-derived factor IX, Asn157 and Asn167 within the activation peptide are fully occupied with complex-type N-glycans [55]. Recombinant factor IX contains tetra-antennary, tetrasialylated, and core fucosylated glycans at both sites. Plasma derived factor IX contains bi -, tri -, and tetra-antennary, sialylated glycans, with and without fucose. Both molecules have a range of minor structures; however, the glycans present on plasma-derived factor IX are considerably more heterogeneous and diverse. The diversity may be a consequence of the plasma pool.

Both plasma-and recombinant-derived factor IX contain a number of O-linked oligosaccharides. In the first factor IX EGF domain, serine residues 53 and 61 are uniformly O-glycosylated. The EGF1 domain in both recombinant and plasma-derived factor IX contains nonclassical O-linked glycans at Ser53 and Ser61. Ser53 contains Xyl–Xyl–Glc–Ser and Ser61 contains the tetrasaccharide with a terminal sialic acid (NeuAc), NeuAc–Gal–GlcNac–Fuc–Ser61 [56–58]. This indicates that CHO cells (the cells used as a host to produce recombinant factor IX) have the enzymatic machinery to produce the structures present on plasma-derived factor IX that is synthesized in human hepatocytes and that it is not saturated at high expression levels. The carbohydrate structure at Ser61 in factor IX contains fucose-linked tetrasaccharide with a terminal sialic acid. Ser61 within the first EGF domain of factor IX has the consensus sequence (C–X–X–G–G–T/S–C) for fucosyl modification of O-linked sugars and is also found in factor VII, but not in factor X. However, a crystal structure of factor IX demonstrated that both these O-linked modifications reside on the face of the EGF domain that apparently does not interact with other components of the factor Xase complex [59]. In addition to the serine-linked oligosaccharide addition in the first EGF domain, both plasma derived and recombinant-derived factor IX molecules are partially occupied by O-linked glycans at residues Thr159, Thr169, Thr172, and Thr179, as well as as-yet unidentified additional sites [55]. The function of these O-linked glycans remains unknown.

γ-Carboxylation of glutamic a cid residues

The vitamin K-dependent coagulation factors contain the post-translationally modified amino acid γ-carboxy-glutamic acid (Gla). The Gla residues are essential for these proteins to attain a calcium-dependent conformation and for their ability to bind to phospholipid surfaces, an essential interaction for their function. The precursor of the vitamin K-dependent coagulation factors contains a propeptide that directs γ-carboxylation of up to 12 glutamic acid residues at the amino-terminus of the mature protein, all of which is completed prior to translocation out of the ER [60–62]. The propeptides (residues −18 to −1 in factor IX) of these factors share amino acid similarity by conservation of the γ-carboxylase recognition site and the site for cleavage of the propeptide.

The residues that are carboxylated in factor IX are glutamic acid residues 7, 8, 15, 17, 20, 21, 26, 27, 30, 33, 36, and 40. Mutations at residues 6, 7, 17, 21, 27, 30, or 33 result in moderate to severe hemophilia B, indicating their functional importance. High-level expression of the vitamin K-dependent plasma proteins in transfected mammalian cells is limited by the ability of the mammalian host cell to efficiently perform γ-carboxylation of amino-terminal glutamic acid residues and also to efficiently cleave the propeptide [49,63,64]. Analysis of factor IX expressed in CHO cells revealed that the protein had a much lower specific activity compared with the natural human plasma-derived protein. The reduced specific activity was attributed to both the limited ability of CHO cells to cleave the propeptide of factor IX as well to efficiently perform γ-carboxylation [49,64]. Generally, expression of factor IX at levels greater than 1 μg/10⁶ cells/day saturates the activity for most cells studied [63]. Overexpression of the γ-carboxylase did not improve γ-carboxylation of factor IX when coexpressed in transfected mammalian cells [49]. These results suggest that the amount of carboxylase protein is not a limiting factor to direct vitamin K-dependent γ-carboxylation in vivo . Several possibilities exist for the inability of the overexpressed γ-carboxylase to improve factor IX carboxylation in vivo . First, the overexpressed γ-carboxylase may be mislocalized within the secretory pathway. It is possible that another protein, such as a protein chaperone, may be required to utilize a more complex substrate, such as factor IX, as opposed to a small peptide substrate. It is possible that another cof actor, possibly reduced vitamin K, is limiting for factor IX carboxylation in vivo . Further information on the mechanism of γ-carboxylation reaction in vivo is required in order to elucidate the rate-limiting step for γ-carboxylation in vivo .

Recombinant factor IX produced in CHO cells contains 11.8 Gla residues/mole of factor IX, compared with plasma derived factor IX that contains 12 Gla residues/mole. The difference resides in the inefficient carboxylation of residues 36 and 40 within recombinant factor IX [65]. In contrast to the first 10 Gla residues in factor IX, glutamic acid residues 36 and 40 are not conserved in the other vitamin K-dependent coagulation factors. To date, no functional difference is observed between fully carboxylated factor IX and factor IX deficient in Gla at residues 36 and 40.

β--Hydroxylation of a spartic a cid and a sparagine

Blood coagulation factors IX and X, protein C, and protein S contain the modified amino acid erythro-β-hydroxyaspartic acid in the first EGF domain. In addition, one molecule of β-hydroxyasparagine is found in each of the three carboxy terminal EGF domains in protein S. Hydroxylation of both aspartic acid and asparagine is catalyzed by aspartyl β-hydroxylase, requires 2-ketoglutarate and Fe²+, and is inhibited by agents that inhibit 2-ketoglutarate-dependent dioxygenases. This enzyme recognizes a consensus sequence C–X–X–X–X–X–X–X–X–C in the β-sheet and C–X–D/N–X–X–X–X–Y/F–X in the antiparallel β-sheet [66]. β-hydroxylation is unnecessary for high-affinity calcium binding to the first EGF domain [67]. In addition, inhibition of β-hydroxylation of factor IX expressed in mammalian cells did not reduce functional activity in factor IX [68]. It is interesting that only 30% of plasma factor IX is modified by β-hydroxylation at Asp64 and this same amount of β-hydroxylation occurs in recombinant factor IX expressed at high levels in CHO cells [68].

Tyrosine sulfation

Plasma-derived and recombinant-derived factor IX are sulfated on Tyr155. Whereas plasma-derived factor IX is mostly sulfated, recombinant factor IX is approximately 15% sulfated [54,58]. This is one unusual example where a sulfated tyrosine occurs adjacent to an occupied N-linked glycosylation site (at asparagine residue 157). Plasma-derived factor IX and plasma-derived factor IX differ in their in vivo recovery, where the absolute recovery of plasma-derived factor IX is approximately 50% and the recovery of recombinant factor IX is approximately 30%. Studies suggest tyrosine sulfation on factor IX may be responsible for the difference in the recovery of these two sources of factor IX [58]. For example, infusion of recombinant factor IX enriched for full sulfation at Tyr155 demonstrated an equivalent recovery to plasma derived factor IX (approximately 50%). Similarly, removal of the sulfate, as well as phosphate, from plasma-derived factor IX resulted in a molecule having a recovery similar to recombinant factor IX. Finally, administration of recombinant factor IX to hemophilia B dogs and isolation of the circulating factor IX yielded species that were enriched with tyrosine sulfate compared with the starting material. The sum of these observations suggests that tyrosine sulfation at 155 in factor IX can influence in vivo recovery.

Phosphorylation of serine and t hreonine residues

Phosphate has been observed in factor IX, although its significance remains unknown. Plasma-derived factor IX is fully phosphorylated at Ser158 whereas recombinant factor IX contains no phosphate at this position [58,69]. Factor IX produced in myotubes had considerably less phosphorylation at Ser158 compared with plasma-derived factor IX but maintained similar specific activity [70]. The presence or absence of phosphate or sulfate on factor IX has no effect on the in vitro clotting activity.

Proteolytic processing

The requirement for propeptide processing for factor IX function was first made apparent by identification mutations resulting in hemophilia B that prevent processing of the factor IX propeptide. Mutations of the Arg at the P1 or P4 positions inhibit propeptide cleavage and the resultant factor IX is secreted into the plasma, but is nonfunctional because of the presence of the propeptide [71,72]. This mutant is unable to bind phospholipid vesicles and may also display reduced γ-carboxylation of glutamic acid residues [72]. It is likely that the presence of the propeptide yields a molecule that is defective in phospholipid interaction as a result of an inability to undergo a calcium-dependent conformation in the Gla domain.

Characterization of the amino acid requirements around the propeptide cleavage site has identified that both the P1 and P4 arginine are important for efficient processing mediated by furin and PACE4 [73,74]. Overexpression of furin in transfected cells as well as in transgenic animals improves the processing ability to yield fully processed proteins [64]. Recombinant factor IX is produced by coexpression with furin/PACE to ensure complete processing of the propeptide.

Summary

Factor IX undergoes a remarkable number of varied post translational modifications prior to secretion from the hepatocyte. Cotranslational translocation into the lumen of the ER occurs concomitantly with signal peptide cleavage and addition of core high-mannose oligosaccharides to the polypeptide is followed in the ER by glucose trimming of the N-linked oligosaccharide core structures, γ-carboxylation of 12 amino terminal glutamic acid residues, and β-hydroxylation of a portion of molecules on residue Asp64 occurs. Upon transit into the Golgi compartment, additional modifications occur, which include (i) complex modification of N-linked oligosaccharides; (ii) tyrosine sulfation at Tyr155; (iii) serine/threonine glycosylation at residues Ser61 and Ser53, as well as several threonine residues within the activation peptide; and (iv) cleavage of the propeptide. In addition, factor IX isolated from human plasma is phosphorylated at Ser158 within the activation peptide. A majority of the modifications within factor IX occur within the activation peptide and may regulate activation of factor IX. Appropriate γ-carboxylation and propeptide cleavage are essential for functional secretion and activity of secreted factor IX. Both of these activities are easily saturated upon expression of factor IX in heterologous cells. The large number of other modifications likely also affects factor IX activity by mechanisms that are not understood to date. Further studies are required to elucidate the effect of factor IX expression in different cell types in order to identify the importance that subtle differences in post-translational modifications may have on their secretion, in vivo half-life, and function. These considerations will be important when considering different cells and tissues as targets for gene therapy.

References

1 Toole JJ, KnopfJL, Wozney JM, et al. Molecular cloning of a cDNA encoding human antihaemophilic factor. Nature 1984; 312: 342–7.

2 Bihoreau N, Pin S, de Kersabiec AM, Vidot F, Fontaine-Aupart MP. Copper- atom identification in the active and inactive forms of plasma-derived FVIII and recombinant FVIII-delta II. Eur J Biochem 1994; 222: 41–8.

3 Tagliavacca L, Moon N, Dunham WR, Kaufman RJ. Identification and functional requirement of Cu(I) and its ligands within coagulation factor VIII. J Biol Chem 1997; 272: 27428–34.

4 Ngo JC, Huang M, Roth DA, Furie BC, Furie B. Crystal structure of human factor VIII: implications for the formation of the factor IXa-factor VIIIa complex. Structure 2008; 16: 597–606.

5 Shen BW, Spiegel PC, Chang CH, et al. The tertiary structure and domain organization of coagulation factor VIII. Blood 2008; 111: 1240–7.

6 Kaufman RJ, Wasley LC, Dorner AJ. Synthesis, processing, and secretion of recombinant human factor VIII expressed in mammalian cells. J Biol Chem 1988; 263: 6352–62.

7 Weiss HJ, Sussman, II, Hoyer LW. Stabilization of factor VIII in plasma by the von Willebrand factor. Studies on posttransfusion and dissociated factor VIII and in patients with von Willebrand’s disease. J Clin Invest 1977; 60: 390–404.

8 Eaton D, Rodriguez H, Vehar GA. Proteolytic processing of human factor VIII. Correlation of specific cleavages by thrombin, factor Xa, and activated protein C with activation and inactivation of factor VIII coagulant activity. Biochemistry 1986; 25: 505–12.

9 Fay PJ, Haidaris PJ, Smudzin TM. Human factor VIIIa subunit structure. Reconstruction of factor VIIIa from the isolated A1/ A3-C1-C2 dimer and A2 subunit. J Biol Chem 1991; 266: 8957–62.

10 Lollar P, Parker CG, Kajenski PJ, Litwiller RD, Fass DN. Degradation of coagulation proteins by an enzyme from Malayan pit viper (Akistrodon rhodostoma ) venom. Biochemistry 1987; 26: 7627–36.

11 McMullen BA, Fujikawa K, Davie EW, Hedner U, Ezban M. Locations of disulfide bonds and free cysteines in the heavy and light chains of recombinant human factor VIII (antihemophilic factor A). Protein Sci 1995; 4: 740–6.

12 Selvaraj SR, Kaufman RJ, Pipe SW. Incremental improvement in bioengineering of coagulation factor VIII for efficient expression: elimination of a dispensable disulfide loop enhances secretion. Blood (ASH Annual Meeting Abstracts) 2008; 112: 3073.

13 Sousa MC, Ferrero-Garcia MA, Parodi AJ. Recognition of the oligosaccharide and protein moieties of glycoproteins by the UDP Glc:glycoprotein glucosyltransferase. Biochemistry 1992; 31: 97–105.

14 Sousa M, Parodi AJ. The molecular basis for the recognition of misfolded glycoproteins by the UDP-Glc:glycoprotein glucosyltransferase. EMBO J 1995; 14: 4196–203.

15 Hironaka T, Furukawa K, Esmon PC, et al. Comparative study of the sugar chains of factor VIII purified from human plasma and from the culture media of recombinant baby hamster kidney cells. J Biol Chem 1992; 267: 8012–20.

16 Pittman DD, Tomkinson KN, Kaufman RJ. Post-translational requirements for functional factor V and factor VIII secretion in mammalian cells. J Biol Chem 1994; 269: 17329–37.

17 Moussalli M, Pipe SW, Hauri HP, Nichols WC, Ginsburg D, Kaufman RJ. Mannose-dependent endoplasmic reticulum (ER) Golgi intermediate compartment-53-mediated ER to Golgi trafficking of coagulation factors V and VIII. J Biol Chem 1999; 274: 32539–42.

18 Cunningham MA, Pipe SW, Zhang B, Hauri HP, Ginsburg D , Kaufman RJ. LMAN1 is a molecular chaperone for the secretion of coagulation factor VIII. J Thromb Haemost 2003; 1: 2360–7.

19 Nichols WC , Seligsohn U , Zivelin A , et al. Mutations in the ER–Golgi intermediate compartment protein ERGIC-53 cause combined deficiency of coagulation factors V and VIII. Cell 1998; 93: 61–70.

20 Zhang B , Cunningham MA , Nichols WC , et al. Bleeding due to disruption of a cargo-specific ER-to-Golgi transport complex. Nat Genet 2003; 34: 220–5.

21 Nyfeler B , Zhang B , Ginsburg D , Kaufman RJ , Hauri HP. Cargo selectivity of the ERGIC-53/MCFD2 transport receptor complex. Traffic 2006; 7: 1473–81.

22 Zhang B , Spreafico M , Zheng C , et al. Genotype-phenotype correlation in combined deficiency of factor V and factor VIII. Blood 2008; 111: 5592–600.

23 Lynch CM , Israel DI , Kaufman RJ , Miller AD. Sequences in the coding region of clotting factor VIII act as dominant inhibitors of RNA accumulation and protein production. Hum Gene Ther 1993; 4: 259–72.

24 Eriksson KK , Vago R , Calanca V , Galli C , Paganetti P , Molinari M. EDEM contributes to maintenance of protein folding efficiency and secretory capacity. J Biol Chem 2004; 279: 44600–5.

25 Molinari M , Calanca V , Galli C , Lucca P , Paganetti P. Role of EDEM in the release of misfolded glycoproteins from the calnexin cycle. Science 2003; 299: 1397–400.

26 Oda Y , Hosokawa N , Wada I , Nagata K. EDEM as an acceptor of terminally misfolded glycoproteins released from calnexin. Science 2003; 299: 1394–7.

27 Kaufman RJ. Stress signaling from the lumen of the endoplasmic reticulum: coordination of gene transcriptional and translational controls. Genes Dev 1999; 13: 1211–33.

28 Kaufman RJ. Orchestrating the unfolded protein response in health and disease.