The Blood Coagulation Factors and Inhibitors: Their Primary Structure, Complementary DNAs, Genes, and Expression

Dominic W. Chung

Wenfeng Xu

Earl W. Davie

A BRIEF INTRODUCTION TO THE BLOOD COAGULATION CASCADE

When tissue damage occurs, platelets become activated, aggregate, and adhere to the site of vascular injury where they form a thrombus. During this process, the blood coagulation cascade is initiated when the tissue factor (TF) comes in contact with factor VII (or factor VIIa) in the flowing blood and forms a potent enzyme complex at the site of injury (FIGURE 9.1). This complex initiates a series of reactions involving coagulation proteins circulating in blood in a precursor or inactive form.¹ The complex of factor VIIa and TF catalyzes the conversion of factor X to factor Xa. Factor Xa, in turn, converts prothrombin to thrombin in the presence of factor Va, phospholipid, and calcium. Thrombin then converts fibrinogen to fibrin, leading to an insoluble clot of activated platelets and fibrin polymers. Thrombin also activates the intrinsic pathway of blood coagulation by converting factor XI to factor XIa by a mechanism that is poorly understood. Factor XIa, in turn, converts factor IX to factor IXa in a reaction requiring calcium. Factor IXa then converts factor X to factor Xa in the presence of factor VIIIa, phospholipid, and calcium. This reaction provides a second mechanism for the activation of factor X in blood. Thrombin plays a critical role in blood coagulation in that it converts fibrinogen to fibrin, factor V and factor VIII to active cofactors, factor XI and protein C to serine proteases, factor XIII to a transglutaminase, and finally, it activates platelets (FIGURE 9.1). All these activation reactions involve minor proteolysis at arginine peptide bonds. The phospholipid participating in several of these steps is provided by the activated platelets and cell membranes. Divalent calcium is also required in several steps of the coagulation cascade, and the normal level ranges from about 3.5 to 5 mM in blood.

FACTOR VII

Factor VII is highly homologous in its amino acid sequence and gene organization with the other vitamin K-dependent proteins, including factor IX, factor X, and protein C (Table 9.1). In the presence of TF, factor VII initiates the extrinsic pathway. Factor VII is a single-chain glycoprotein (Mr 50,000) that is synthesized in the liver and secreted into the blood as a zymogen composed of 406 amino acids (FIGURE 9.2).²^,³^,⁴^,⁵ It contains 10 γ-carboxyglutamic acid residues that are localized in the Gla domain of the protein (amino acids 1 to 35). The Gla region is followed by two epidermal growth factor (EGF)-like domains and a serine protease domain (FIGURE 9.2). The γ- carboxyglutamic residues require vitamin K for their biosynthesis.⁶ Factor VII also contains a residue of β-hydroxyaspartic acid (amino acid 63), located in the first of the two EGF-like domains.⁷ The first EGF-like domain stabilizes the factor VIIa-TF complex and participates in calcium binding.⁸^,⁹^,¹⁰ The β-hydroxylation of the Asn63 is not essential for the coagulant activity of factor VIIa since recombinant factor VII is not β-hydroxylated, and functions as well as the plasma-derived counterpart.¹¹ The first EGF-like domain contains novel O-linked carbohydrates at residues Ser52 and Ser60.¹²^,¹³^,¹⁴ The Ser52 is O-glycosylated, being linked to either a disaccharide (Xyl-Glc) or a trisaccharide (Xyl2-Glc) in approximately equal amounts. The Ser60 contains one residue that is conjugated to fucose by O-glycosylation. Recombinant factor VII/VIIa missing either one or both O-linked Ser residues exhibits decreased TF binding but has similar calcium binding relative to the wildtype.⁹ Factor VIIa has another potential carbohydrate attachment site located at Asn322. Factors VII and VIIa contain two high-affinity calcium-binding sites located in the protease domain and the first EGF1 domain as well as six to seven lowaffinity calcium-binding sites in the Gla domain.¹⁵

Factor VII is converted to a serine protease (factor VIIa) by minor proteolysis. Factor VIIa, however, has little if any physiologic activity until it combines with TF. The factor VIIa-TF complex then converts factor X to factor Xa in the presence of phospholipids and calcium ions.¹ Factor VIIa also converts factor IX to factor IXa in the presence of TF and calcium ions.⁹ The physiological importance of the latter pathway is unclear. In vitro experiments evaluating the initiation phase of blood coagulation with ultrasensitive fluorescent markers indicate that factor Xa is generated almost exclusively by the factor VIIa-TF complex during the initiation phase of coagulation.¹⁶ The conversion of human factor VII to factor VIIa is catalyzed by thrombin and factor Xa as well as factor IXa, and factor XIIa.³^,⁵^,¹⁷^,¹⁸^,¹⁹^,²⁰

The activation of factor VII is due to the cleavage of a single peptide bond (Arg152-Ile) (FIGURE 9.2). This leads to the formation of a salt bridge between the α-ammonium group of Ile153 and the carboxylate group of Asp343 and a reregistration of a β-strand in the protein,²¹ resulting in a conformational change and formation of an active charge-relay system. Factor VIIa may also be formed by an autocatalytic mechanism,²²^,²³ but this pathway is very slow and may have
little physiological importance.²⁴ Factor VIIa is a serine protease composed of a light chain (152 amino acids) and a heavy chain (254 amino acids) held together by a single disulfide bond between Cys135 and Cys262. The light chain contains the Gla domain followed by the two EGF domains, whereas the heavy chain contains the catalytic domain with the active site residues of His193, Asp242, and Ser344. The active site Ser344 is located in the same sequence of Gly-Asp-Ser-Gly-Gly-Pro that is present in all the other serine protease clotting factors, including thrombin, factor IX, factor X, factor XI, and activated protein C (APC).

FIGURE 9.1 An abbreviated blood coagulation cascade. (Modified from Davie EW, Fujikawa K, Kisiel W. The coagulation cascade: initiation, maintenance, and regulation. Biochemistry 1991;30:10363.)

The crystal structure of human factor VIIa lacking the N-terminal γ-carboxyglutamic acid domain is shown in FIGURE 9.3.²⁵ Factor VIIa has an elongated shape similar to that of a tulip, in which the Gla and EGF-like domains form the stem while the catalytic domain forms the blossom.

The cDNA and gene for factor VII have been isolated, and their sequence determined.²⁶^,²⁷ The sequence of the mRNA coding for human factor VII contains approximately 2,450 nucleotides (nts) that code for a prepro leader sequence of 38 amino acids and 406 amino acids present in the mature protein circulating in blood (FIGURE 9.2). A noncoding region of 1,026 nts plus a poly(A) tail follows the stop codon of TAG. The noncoding region also contains the polyadenylation recognition sequence of AATAAA, which are located 40 nts upstream from the poly(A) tail.²⁸

A second clone has been identified for human factor VII, containing a prepro leader sequence of 60 amino acids.²⁶ This leader sequence contains an additional 22 amino acids inserted between Val at -17 and Ala at -18 in the 38-amino acid prepro leader sequence encoded by an additional exon in intron A, suggesting that the two mRNA species result from alternative splicing.²⁷ The removal of the prepro leader sequence requires signal peptidase in addition to a second processing protease that cleaves the peptide bond following the arginine at position -1.²⁹

The gene for factor VII (F7) spans approximately 12.8 kb and consists of eight exons interrupted by seven introns. The positions of the introns with respect to the amino acid sequence and the types of intron-exon boundaries are almost exactly the same as the genes coding for the other vitamin K-dependent proteins (FIGURE 9.1 and Table 9.1). The gene coding for factor VII also contains five regions of tandem repetitive sequences, and more than a quarter of the intron sequences consist of minisatellite DNA sequences, which vary in the number of copies among individuals.³⁰

The absence of a CCAAT box in the factor VII promoter contrasts with the promoter structures of factor IX, which has a functional CCAAT box, and factor X, which contains a putative CCAAT box that binds the ubiquitous transcription factor NF-Y.³¹^,³²^,³³ The major transcription start site for factor VII is located approximately 50-bp upstream from the first initiation Met and is close to the binding sites for a Sp1-like transcription factor and HNF-4.³⁴^,³⁵ The G-C rich Sp1-like site is located at -100 to -94, whereas the HNF-4 site is located at -63 to -58. The factor VII HNF-4 recognition sequence, ACTTTG, is also present in the promoters of factor X and factor IX. A naturally occurring mutation in the factor IX HNF-4 site causes the hemophilia B Leiden phenotype, whereas a similar mutation in the factor VII promoter causes lifelong bleeding and virtually no detectable factor VII.³³^,³⁶

Genetic defects leading to an abnormal factor VII include a replacement of Arg304 or Arg353 with Gln, resulting in reduced clotting activity,³⁷^,³⁸^,³⁹ presumably by a conformational change near the catalytic site. A substitution of Phe38 to Ser (factor VII central)⁴⁰ and factor VII variant Gln100 to Arg⁴¹ result in reduced TF binding and impaired activation of factors IX and X.⁴⁰ A polymorphism originating from a decanucleotide (CCTATATCCT) inserted at position -323 relative to the first Met-initiating codon⁴² and a substitution of G for T at nt -81 disrupts an HNF-4-binding site in the factor VII promoter,³⁶ resulting in a modest reduction in factor VII levels.⁴²

Pharmaceutical preparations of recombinant factor VIIa have been very helpful in the treatment of patients with factor VIII deficiency and inhibitors as well as in patients with other coagulopathies.⁴³^,⁴⁴

Table 9.1 Comparison of intron location and splice junction type and size for human factors VII, IX, X, protein C, and prothrombin

Intron	Protein	Location (Amino Acid)	Splice Type	Size (bp)
A	Prothrombin	-17	1	386
	Factor VII	-17	1	2,574
	Factor IX	-17	1	6,206
	Factor X	-17	1	6,542
	Protein C	-17	1	1,263
B	Prothrombin	37/38	0	659
	Factor VII	38/39	0	1,919
	Factor IX	38/39	0	188
	Factor X	37/38	0	8,836
	Protein C	37/38	0	1,462
C	Prothrombin	46	1	242
	Factor VII	46	1	68
	Factor IX	47	1	3,689
	Factor X	46	1	874
	Protein C	46	1	92
D	Factor VII	84	1	1,908
	Factor IX	85	1	7,163
	Factor X	84	1	1,447
	Protein C	92	1	102
E	Factor VII	131	1	971
	Factor IX	128	1	2,565
	Factor X	128	1	2,798
	Protein C	137	1	2,668
F	Factor VII	15/16^a	0	595
	Factor IX	15/16	0	9,473
	Factor X	15/16	0	3,224
	Protein C	15/16	0	873
G	Factor VII	57	1	816
	Factor IX	54	1	668
	Factor X	55	1	1,418
	Protein C	55	1	1,129
bp, base pair
^aNumbering of amino acids prior to intron F begins with residue 1 of the heavy chain of the active molecules.
Data from O’Hara PJ, Grant FJ, Haldeman BA, et al. Nucleotide sequence of the gene coding for human factor VII, a vitamin K-dependent protein participating in blood coagulation. Proc Natl Acad Sci U S A 1987;84:5158; Leytus SP, Foster DC, Kurachi K, et al. Gene for human factor X: a blood coagulation factor whose gene organization is essentially identical with that of factor IX and protein C. Biochemistry 1986;25:5098; Degen SJ, Davie EW. Nucleotide sequence of the gene for human prothrombin. Biochemistry 1987;26:6165; Foster DC, Yoshitake S, Davie EW. The nucleotide sequence of the gene for human protein C. Proc Natl Acad Sci U S A 1985;82:4673; Yoshitake S, Schach BG, Foster DC, et al. Nucleotide sequence of the gene for human factor IX (antihemophilic factor B). Biochemistry 1985;24:3736.

TISSUE FACTOR

Human TF (Mr 44,000) is a transmembrane glycoprotein synthesized in adventitial fibroblasts. When blood comes into contact with the subendothelium following vascular injury,⁴⁵^,⁴⁶^,⁴⁷^,⁴⁸ factor VII binds to TF to form a bimolecular complex in the presence of calcium ions,¹^,⁴⁹^,⁵⁰^,⁵¹^,⁵²^,⁵³^,⁵⁴ initiating coagulation. TF is a single-chain protein containing 263 amino acids and is synthesized with a signal peptide of 32 amino acids (FIGURE 9.4).⁵⁵^,⁵⁶^,⁵⁷ The extracellular or cellular surface domain of TF is 219 residues and contains three repeating sequences of Trp-Lys-Ser. It also contains two disulfide bonds linking Cys49 with Cys57, and Cys186 with Cys209.⁵⁸ The membrane-spanning region of TF is 23 amino acids (residues 220 to 242), whereas the cytoplasmic portion of the protein at the carboxyl end of the molecule is 21 residues in length. The cytoplasmic portion also contains a half-Cys residue (Cys245) that is acylated by palmitic or stearic acid.⁵⁸ Three potential glycosylation sites with a sequence of Asn-X-Thr/Ser (Asn11, Asn124 and Asn137) are also present in the molecule.⁵⁵^,⁵⁶^,⁵⁷

The crystal structure of TF,⁵⁹ and factor VIIa-TF complex has been determined in several different laboratories.⁸^,⁶⁰ The extracellular region of TF (TF-219) consists of two immunoglobulin-like domains that are formed by two antiparallel β-sheets (FIGURE 9.5). The crystal structure of factor VIIa in a complex with a mutant of bovine pancreatic trypsin inhibitor (BPTI) lacks about two-thirds of the amino-terminal Gla domain of factor VIIa (FIGURE 9.6, left). The complex of TF (TF-219) and the factor VIIa-BPTI mutant (FIGURE 9.6, right),⁶⁰ enhanced by calcium about 150,000 fold,¹⁵ is >100 Å long and 40 to 60 Å wide. In the absence of the inhibitor, the catalytic domain of factor VIIa, bound to the BPTI mutant at the top of the complex, is readily available to bind factor X and convert it to factor Xa.

The mRNA for human TF is 2.3 kb,⁵⁵^,⁵⁶^,⁵⁷ including a 5′ noncoding region of 75 bp, 885 bp of coding sequence, a stop codon, and a 3′ noncoding region of 1,141 bp, followed by a poly(A) tail. The gene for human TF (F3) spans 12.4 kb on chromosome 1,⁵⁶^,⁶¹ contains six exons separated by five introns (A to E, Fig. 9.4), three full Alu sequences, one partial Alu sequence, and a typical TATA promoter element present 26-bp upstream from the cap site. TF in vascular endothelium and monocytes can be induced with several agents, such as phytohemagglutinin and endotoxin, interleukin 1, TNF, phorbol esters, and thrombin.⁶²^,⁶³^,⁶⁴^,⁶⁵^,⁶⁶^,⁶⁷^,⁶⁸ Functional studies indicate that basal TF transcription is controlled by the transcription factor Sp1. A distal enhancer (-227 to -172 bp) containing two AP1 sites and an NFkB site mediates the induction of the TF transcription in monocytic and endothelial cells.⁶⁹

FACTOR X

Factor X (Mr 58,800) is a vitamin K-dependent glycoprotein (15% carbohydrate) that is synthesized in the liver and secreted into the plasma. Human factor X is composed of a light chain of 191 amino acids (Mr 16,200) and a heavy chain of 254 amino acids (Mr 42,000) held together by a single disulfide bond (FIGURE 9.7).⁷⁰^,⁷¹^,⁷²^,⁷³ The light chain of human factor X contains 11 residues of γ-carboxyglutamic acid, one residue of β-hydroxyaspartic acid (residue 63),⁷¹ and two EGF-like domains,⁷³^,⁷⁴^,⁷⁵ the first of which contains one high-affinity calcium-binding site.⁷⁶
The heavy chain of factor X contains the activation peptide and the catalytic domain.⁷⁷ The activation peptide also contains two potential N-linked carbohydrate-binding sites, including an Asn39-Gln-Thr sequence and an Asn49-Leu-Ser sequence. In bovine factor X, there is an N-linked (Asn36) as well as an O-linked (Thr300) carbohydrate chain.⁷⁷

FIGURE 9.2 Amino acid sequence for human prepro factor VII. The prepro leader sequence (-38 to -1) is removed during biosynthesis by signal peptidase and a processing protease that hydrolyze the Arg(-1)-Ala bond. The single cleavage site by factor Xa is shown by a solid arrow. The Gla domain and potential growth factor domains are located within residues 1 to 152, which constitute the light chain of factor VIIa. The heavy chain or catalytic domain contains 254 residues, including the three principal amino acids participating in catalysis. The amino acids (His193, Asp242, and Ser344) are circled. Two potential carbohydrate attachment sites are shown by solid diamonds. The proposed disulfide bonds have been placed by analogy to those established in bovine prothrombin and epidermal growth factor. The single-letter code for amino acids is as follows: A, Ala; R, Arg; N, Asn; D, Asp; C, Cys; Q, Gln; E, Glu; G, Gly; H, His; I, Ile; L, Leu; K, Lys; M, Met; F, Phe; P, Pro; S, Ser; T, Thr; W, Trp; Y, Tyr; V, Val; γ, γ-carboxyglutamic acid; β, β-hydroxyaspartic acid.

Factor X is synthesized with a prepro leader sequence that requires two processing steps for its removal (FIGURE 9.8). These reactions are catalyzed by a signal peptidase as well as by a furin-like enzyme that cleaves the Arg residue on the carboxyl end of the propiece. Additional processing in the single-chain precursor occurs between Arg139 and Ser143, which is the N-terminal residue in the activation peptide. This protease activity results in the removal of a basic tripeptide of Arg-Lys-Arg and the formation of the two-chain circulating molecule that is held together by a single disulfide bond (FIGURE 9.7).

The crystal structure of factor Xa lacking the Gla domain is very similar to that of factor VIIa and factor IXa (FIGURE 9.9), an elongated molecule of 100 Å in length.⁷⁸ In the intact molecule,
the Gla region of factor X is partially buried in the phospholipid surface, which is about 61 to 69 Å from the active site.⁷⁹

FIGURE 9.3 Ribbon structure of Gla domainless factor VIIa starting with Gln 49. The protein takes the shape of a tulip with the EGF1 and EGF2 domains (colored cyan) forming the stem, while the catalytic domain (colored orange) is the flower. The structure on the right shows the location of seven known mutations (black dots) in the human protein. (Modified from Pike AC, Brzozowski AM, Roberts SM, et al. Structure of human factor VIIa and its implications for the triggering of blood coagulation. Proc Natl Acad Sci U S A 1999;96:8925.)

Activation of factor X involves the cleavage of an Arg-Ile1 peptide bond in the amino-terminal end of the heavy chain, liberating a small activation peptide of 52 amino acids (FIGURE 9.7).⁷⁰ During the activation reaction, the new amino-terminal Ile1 residue turns and flips into the interior of the catalytic domain. The new α-ammonium group of Ile1 then forms a salt bridge with the carboxylate group of Asp184 (circled in green, FIGURE 9.9),⁸⁰ generating a serine protease with an active catalytic site involving His42, Asp88, and Ser185 (circled in blue, FIGURE 9.9).

The gene for human factor X (F10) contains eight exons and seven introns on chromosome 13 at q32-qter in approximately 27 kb of DNA,⁷²^,⁸¹ only 2.8 kb downstream from the gene coding for factor VII.³² The introns are located between amino acid residues -17 (intron A), 37, and 38 (B), at residue 46 (C), and at residue 84 (D) between the two potential growth factor domains (FIGURE 9.7 and Table 9.1). Intron E is located at residue 128 following the second growth factor domain and just before the disulfide bond connecting the light and heavy chains. The last two introns (F and G) are located in the heavy chain or catalytic domain of the molecule and are present between residues 15 and 16 and at residue 55. The rest of the catalytic chain is free of introns.

The first patient identified with factor X deficiency⁸²^,⁸³^,⁸⁴ had a single-nucleotide change of G to A (GTG to ATG) resulting in a Val148Met change in the catalytic domain of the protein (red arrow, FIGURE 9.8) (Asakai, Roberts, Davie, unpublished data). Other molecular aberrations include partial deletions of exons VII and VIII, and amino acid substitutions of Arg for Gly(-20), Lys for Gla14, Ser for Pro343, and Cys for Arg366.⁸⁵^,⁸⁶^,⁸⁷^,⁸⁸^,⁸⁹^,⁹⁰ Amino acid changes that result in a loss of biological activity have been found throughout the catalytic domain as well as the EGF domains (FIGURE 9.10, right panel).

The mRNA for human factor X includes 1,475 nts that code for a prepro leader sequence of 40 amino acids, a light chain of 139 amino acids, a connecting tripeptide, and 303 amino acids that constitute the heavy chain.⁷³^,⁷⁵ The processing or polyadenylation sequence of ATTAAA is unusual in that it is located in the coding sequence and precedes the stop codon by one nucleotide.²⁸

Two transcription start sites, a major one at -16-bp and a minor one at -10-bp upstream from the initiation Met codon, are present in the human factor X 5′-flanking region.⁹¹ Full liver-specific promoter activity is contained within a 457-bp region upstream from the translation start site.³² An apparent CCAAT sequence, present at -120 to -116 bp, binds the ubiquitous transcription factor NF-Y.³⁴ An HNF-4 functional element has been localized between -63 and -42 bp, and HNF-4 binding is required for liver-specific expression. Two additional positive regulatory regions have been identified at -215 to -149 bp and at -457 to -351 bp.³²

FACTOR V

Factor V (Mr 330,000) is a glycoprotein that is synthesized as a single-chain molecule of 2,196 amino acids in liver and megakaryocytes and circulates in blood as an inactive cofactor
at 7 µg/mL.⁹² Approximately 20% of the factor V in blood is present in the α-granules of platelets.⁹³ Factor V is comprised of six domains (A1-A2-B-A3-C1-C2) that are highly homologous to those in factor VIII and ceruloplasmin (FIGURE 9.11).⁹⁴^,⁹⁵^,⁹⁶^,⁹⁷ The second and third A domains of factor V and factor VIII are separated by a large connecting B domain. In factor V, this domain is 836 amino acids in length, is located between amino acids 710 and 1,545, and has two tandem repeats of 17 amino acids and 31 tandem repeats of nine amino acids with a consensus sequence of [TNP]LSPDLSQT (FIGURE 9.12). The B domain in factor V shows no similarity in amino acid sequence to that present in factor VIII. Interestingly, chimeric cDNA expression constructs in which the connecting regions of factor V and factor VIII were exchanged show that sequences within the factor V connecting region increase the expression of the factor VIII chimera and its corresponding mRNA by twofold in COS1 cells.⁹⁸ In another
study, the connecting region or B domain of factor V did not require the chaperone proteins calnexin or calreticulin for efficient secretion, whereas the connecting region or B domain of factor VIII was required for both chaperone protein binding and proper secretion of factor VIII in a Chinese hamster ovary (CHO) cell line.⁹⁹ The third A domain in factor V, as well as factor VIII, is followed by two C domains. Each of the C domains is approximately 150 amino acids in length, with sequence identity of 35% to 50%.

FIGURE 9.4 Amino acid sequence of human TF. The solid arrows indicate the position of the five introns (A to E). The three potential carbohydrate-binding sites are indicated by open diamonds. Cys 245 is acylated by palmitic or stearic acid.

FIGURE 9.5 Ribbon structure of the extracellular domain (219 N-terminal residues) of TF. The two immunoglobulin-like domains TF-1 and TF-2 (magenta) are formed by two antiparallel β-sheets. The N-terminal Ser is located in TF-1. (Modified from Harlos K, Martin DM, O’Brien DP, et al. Crystal structure of the extracellular region of human tissue factor. Nature 1994;370:662.)

The membrane-binding C2 domain of factor Va contains a β-barrel motif composed of eight antiparallel strands (FIGURE 9.13),¹⁰⁰ which form two tightly packed β-sheets of five and three strands. The bottom portion of the β-barrel is rich in basic amino acids and consists of three β-hairpin loops that form the calcium-independent membrane- binding site of factor Va.

Before participation in the coagulation cascade, the single-chain factor V undergoes minor proteolysis⁹⁵ by thrombin at Arg709, Arg1018, and Arg1545. Factor V can be activated by factor Xa, but the physiological importance of this activation pathway has not been established.¹⁰¹

Factor Va is composed of a heavy chain (A1-A2 domains, Mr 110,000) derived from the amino-terminus of the protein and a light chain (A3-C1-C2 domains, Mr 78,000) derived from the carboxyl-terminus of the molecule (FIGURE 9.11). The fragments corresponding to the central connecting B region are released from factor V during its conversion to factor Va by thrombin. The newly generated heavy and light chains associate and are held together by calcium that is bound to a high-affinity site
formed when the two chains combine. During the activation of prothrombin by factor Xa, the C1 and C2 domains of the light chain of factor Va are bound to phosphatidylserine in the membranes.¹⁰² Important amino acid residues in the C2 domain of factor Va involved in the phospholipid binding include Trp2063 and Trp2064.¹⁰³ Modeling experiments and peptide inhibition studies suggest that the factor Xa binding to factor Va occurs via the A2 domain.

FIGURE 9.6 Ribbon structure of factor VIIa-inhibitor and factor VIIa-inhibitor-TF complex (Zhang E, St Charles R, Tulinsky A. Structure of extracellular tissue factor complexed with factor VIIa inhibited with a BPTI mutant. J Mol Biol 1999;285:2089). Factor VIIa (cyan and orange) complexed with pancreatic trypsin inhibitor (green) is shown in the left panel. The factor VIIa-inhibitor complex bound to TF (magenta) is shown in the panel on the right.

FIGURE 9.7 Amino acid sequence of human prepro factor X and the location of the seven introns (A to G). The tripeptide of Arg-Lys-Arg that connects the light chain to the heavy chain is not shown. The prepro leader sequence (-40 to -1) is removed during biosynthesis by signal peptidase and a processing protease that cleaves the Arg(-1)-Ala bond. The Gla domain and potential growth factor domains are located in the light chain within residues 1 to 139. The activation peptide of 52 amino acids is released from factor X during its conversion to factor Xa. The serine protease or catalytic domain of factor X contains 254 residues, including the catalytic triad of His42, Asp88, and Ser185, which are circled. Two N-linked carbohydrate attachment sites in the activation peptide are shown by solid diamonds. The proposed disulfide bonds in factor IX have been placed by analogy to those in bovine prothrombin and epidermal growth factor. The single-letter code for amino acids is given in FIGURE 9.2.

The cDNA for factor V is approximately 7 kb in size and codes for a leader peptide of 28 amino acids and a mature protein of 2,196 amino acids. The 3′ noncoding sequence also contains the typical sequence of AATAAA that functions as a polyadenylation signal.²⁸

The gene coding for human factor V (F5) is located on chromosome 1q21-35 within 300 kb of the genes for the selectin family of leukocyte adhesive molecules.¹⁰⁴^,¹⁰⁵ The factor V gene spans more than 80 kb of DNA and consists of 25 exons¹⁰⁶ (72 to 2,820 bp) and 24 introns (400 bp to >11 kb) of DNA.

The organization of the gene for human factor V shows remarkable similarity to that of human factor VIII. The factor VIII gene, however, contains one additional exon (i.e., 26 rather
than 25). The gene for factor VIII is also much larger than the gene for factor V, being approximately 180 kb in size. A comparison of the genomic DNA sequences for factor V and factor VIII indicated that 21 of the intron-exon boundaries occur at exactly the same location in the amino acid sequences coded by the two genes.¹⁰⁶ Of particular interest, the connecting B region factors V and VIII are both coded by a single very large exon. In the factor V gene, this exon is located between the 12th and 13th introns, whereas in factor VIII, it is between the 13th and 14th introns.

FIGURE 9.8 Prepro leader sequence of the vitamin K-dependent human plasma proteins. The putative hydrophobic core of each signal sequence is shaded. Identical and conserved amino acid residues within the propeptide region are also shaded. Numbering is relative to the mature amino-termini of the proteins.

Recently, a homozygous factor V Leiden mutation (A for G at nt 1,691) has been shown to cause APC resistance and is the major known cause of hereditary thrombophilia.¹⁰⁷^,¹⁰⁸ This mutation results in the replacement of Gln for Arg at ammo acid residue 506, and this change disrupts the cleavage site for the inactivation of factor Va by APC,¹² and confers a lifelong risk of thrombosis.¹⁰⁹

FIGURE 9.9 Ribbon structure of Gla domainless factor Xa starting with Asp 46. The catalytic triad of His 42, Asp88, and Ser185 is circled in blue, while the salt bridge formed between the carboxylate group of Asp184 and the α-ammonium group of Ile1 is circled in green. The EGF1 and EGF2 domains are colored cyan, while the catalytic domain is yellow. (Modified from Padmanabhan K, Padmanabhan KP, Tulinsky A, et al. Structure of human des(1-45) factor Xa at 2.2 A resolution. J Mol Biol 1993;232:947.)

PROTHROMBIN

Prothrombin (Mr 71,600) is a glycoprotein containing 8.2% carbohydrate that is synthesized in the liver and secreted into the blood, where it circulates as a precursor to a serine protease at a plasma concentration of 100 µg/mL.¹ The amino acid sequence of human prothrombin is 579 amino acid residues (FIGURE 9.14).¹¹⁰^,¹¹¹^,¹¹²^,¹¹³^,¹¹⁴^,¹¹⁵^,¹¹⁶ The sequence of human prothrombin¹¹⁷^,¹¹⁸ is identical with about 46% within the heavy chain and has a similarity of about 75% with that of other vertebrates. Prothrombin contains an amino-terminal Gla domain of about 40 amino acids followed by two kringle domains, each containing approximately 80 amino acids and a carboxyl-terminal region with a typical serine protease domain homologous to pancreatic trypsin (FIGURE 9.14).¹¹⁴^,¹¹⁹^,¹²⁰

During biosynthesis in the rough endoplasmic reticulum, prothrombin undergoes removal of a signal peptide sequence by signal peptidase, carboxylation of amino-terminal glutamic acid residues, cleavage of the propiece by a furin-like proprotein convertase, and the addition of three carbohydrate chains (FIGURE 9.14). The carboxylation of 10 glutamic acid residues located within the first 40 amino-terminal residues in prothrombin is catalyzed by γ-glutamyl carboxylase, which recognizes the propeptide sequence and is responsible for the vitamin K-dependent conversion of the amino-terminal glutamic acid residues to γ-carboxyglutamate. The γ-carboxyglutamyl residues coordinate Ca²⁺ ions, leading to the binding of the Gla region to an anionic phospholipid surface. The addition of the three N-linked carbohydrate chains occurs at Asn78 and Asn100 in the first kringle domain while the third carbohydrate chain is added to the serine protease domain at Asn53 in the catalytic chain.¹¹¹

During the final stages of blood coagulation, prothrombin is converted to thrombin. Thrombin is a serine protease composed of a light chain of 49 amino acids (Thr272 to Arg320) and a catalytic heavy chain of 259 residues (Ile1 to Glu259)
(FIGURE 9.14), held together by a single disulfide bond. Human thrombin undergoes some additional autolysis at the Arg13Thr bond in the light chain, resulting in the removal of a 13-amino acid fragment, reducing the light chain of α-thrombin to 36 residues.

FIGURE 9.10 Ribbon structure of Gla domainless factor Xa starting with Asp46 and the Gla domainless factor Xa showing the location of missense mutations in 34 patients with factor X deficiency. The protein has a shape of a tulip with the EGF1 and EGF2 domains (colored cyan) forming the stem while the catalytic domain (colored yellow) contributes the flower. (Modified from Padmanabhan K, Padmanabhan KP, Tulinsky A, et al. Structure of human des(1-45) factor Xa at 2.2 A resolution. J Mol Biol 1993;232:947; Mutations are mainly from Peyvandi F, Menegatti M, Santagostino E, et al. Gene mutations and three-dimensional structural analysis in 13 families with severe factor deficiency. Br J Haematol 2002;117:685.)

Thrombin generation is due to the cleavage of two internal peptide bonds in prothrombin catalyzed by factor Xa present in the prothrombinase complex.¹²¹ The first cleavage at the Arg320-Ile bond generates a protease called meizothrombin (FIGURE 9.15), followed by cleavage of the Arg 271-Thr bond forming thrombin and fragment 1.2, a polypeptide containing the Gla domain and two tandem kringle domains. The replacement of the Arg271 by Cys results in a defective molecule (prothrombin Barcelona), demonstrating the importance of this cleavage site.¹²²

In the absence of factor Va, the first cleavage of prothrombin by factor Xa occurs at the Arg271-Thr peptide bond generating fragment 1.2 and prethrombin-2. In a second step, prethrombin-2 is cleaved at Arg320. The catalytic efficiency of the prothrombin-2 pathway, however, is slow and insufficient for physiological clot formation.¹²¹

FIGURE 9.11 Structural domains in factor V, factor VIII, and ceruloplasmin. Thrombin cleavage sites are indicated by solid arrows. The identity of the domains is indicated by letters A, B, and C inside boxes. The 31 tandem repeats in the connecting region of factor V are indicated by vertical bars. The A domains in factor V correspond approximately to amino acids 1 through 331, 337 through 711, and 1,649 through 2,196. The C domains in factor VIII correspond approximately to amino acids 2,020 through 2,172 and 2,173 through 2,332. The A domains in ceruloplasmin correspond approximately to amino acids 1 through 338, 348 through 699, and 711 through 1,047.

Thus far, there is no crystal structure available for prothrombin. In 1989, Bode et al.¹²³^,¹²⁴ published the first detailed crystallographic structures of thrombin and compared its structure with pancreatic trypsin, with which it shares considerable structural similarity. The amino acid sequence of thrombin including the catalytic residues of His43, Asp99, and Ser205 is shown in FIGURE 9.16. These residues correspond to His57, Asp102, and Ser195 as described in the charge-relay system present in chymotrypsinogen.¹²⁵ These three important amino acids (His43, Asp99, and Ser205) are enclosed by a red-dashed oval in thrombin (FIGURE 9.16A). The His, Asp, and Ser in the catalytic site of thrombin are also located in the middle of an equatorial cleft that separates the adjacent upper and lower β-barrels or hemispheres of a roughly spherical thrombin molecule (FIGURE 9.16, A,B).¹²³ The upper lip of the substrate-binding cleft contains the 60 loop
(magenta shading, FIGURE 9.16), which is primarily hydrophobic and makes contact with hydrophobic residues present amino-terminal to the scissile bond of the substrate. The lower lip or γ-loop in the substrate-binding cleft (green shading, FIGURE 9.16) is more hydrophilic than the 60 loop and plays a role in determining the substrate specificity at the carboxyl side of the scissile bond in the substrate.

FIGURE 9.12 Thirty-one tandem repeats in the connecting region of factor V. The consensus sequence for the nine amino acid repeats is shown at the bottom. Residues identical to the consensus sequence are shaded.

Thrombin cleaves peptide bonds immediately following Arg, a specificity that is due to a conserved Asp201 present six amino acids prior to Ser205 and located in the bottom of the S1 primary specificity pocket. This specificity of thrombin is particularly evident for macromolecular substrates and inhibitors and is largely due to two surface regions that are distinct from the catalytic triad of Ser, His, and Asp.¹²⁶ These two surface regions are located on opposite ends of the substrate-binding groove in thrombin and have been called the anion-binding exosite I and exosite II (FIGURE 9.16C-E). Exosite I contains mainly basic amino acids including Lys21, 106, and 107 and Arg62, 68, 70, and 73 (FIGURE 9.16C-E).

FIGURE 9.13 Model of human factor Va obtained by molecular dynamics simulations (Macedo-Ribeiro S, Bode W, Huber R, et al. Crystal structures of the membrane-binding C2 domain of human coagulation factor V. Nature 1999;402:434). The A1 domain is represented in green, A2 in cyan, A3 in magenta, C1 in yellow, C2 in pink. Ca²⁺ and Cu²⁺ are in orange and blue, respectively.

Exosite II (heparin-binding site) consists mainly of the basic amino acids Arg89, 98, and 245 and Lys248 and 252. The sulfated glycosoaminoglycans present in heparin bind to thrombin by an electrostatic interaction and accelerates the rate of thrombin inhibition by antithrombin III.

The cleavage of the Arg320-Ile peptide bond in prothrombin catalyzed by factor Xa results in a new free amino-terminal Ile1 in the catalytic chain of the activated enzyme. The newly generated Ile1 turns and flips into the activation pocket of thrombin where its α-ammonium group forms a salt bridge with the carboxylate group of Asp204 (dashed blue oval, FIGURE 9.16A). This Asp is located just prior to the catalytic Ser205. The salt bridge induces a conformational rearrangement in the protein, leading to an active charge-relay system in the serine protease (dashed red oval, FIGURE 9.16A), essentially identical to that occurring in
the activation of pancreatic trypsin and all the other coagulation factors that form serine proteases.

FIGURE 9.14 Amino acid sequence for human prepro prothrombin and the location of the 14 introns (A to M) (Degen SJ, Davie EW. Nucleotide sequence of the gene for human prothrombin. Biochemistry 1987;26:6165). The prepro leader sequence (-43 to -1) is removed during biosynthesis by signal peptidase and a processing protease that hydrolyzes the Arg(-1)-Ala bond. The Gla domain and the kringle domains are located within residues 1 through 271, which constitute fragment 1. This fragment is released from prothrombin during its conversion to thrombin by factor Xa. The light chain in thrombin is generated by the cleavage of the Arg319-Ile bond by factor Xa, and this chain is attached to the catalytic domain by a single disulfide bond. The serine protease or catalytic domain of thrombin contains 259 residues, including the three principal amino acids participating in catalysis. These three amino acids (His363, Asp419, and Ser525) are circled. Three potential carbohydrate attachment sites are shown by solid diamonds. The proposed disulfide bonds in human prothrombin have been placed by analogy to those in the bovine molecule.

The gene for human prothrombin (F2) is present in approximately 21 kb of DNA,¹¹¹^,¹¹²^,¹²⁷^,¹²⁸ and it contains 13 introns (A through M) (FIGURE 9.14) (84 to 9,447 bp) and 14 exons (25 to 315 bp) located in the coding and 3′ noncoding portions of the gene (Table 9.2). The first intron (A) is in the prepro leader sequence at residue -17 while the second intron (B) follows the Gla domain between residues 37 and 38 in the mature protein. The third intron (C) is nine residues later at residue 46. These three introns are located in positions analogous to the first three introns in the coding regions of the genes for factor VII, factor IX, factor X, and protein C (Table 9.1).²⁷^,⁷²^,¹²⁹^,¹³⁰^,¹³¹ The fourth intron (D) is just before the first kringle, whereas the next intron (E) is present within the first kringle. The fifth intron (F) is located immediately following kringle 1 (residue 144) and the sixth (G) immediately after kringle 2 (residue 249). The seventh intron (H) occurs in the region of prothrombin that becomes the light chain and the remaining five introns (I through M) are located in the catalytic domain at positions 14, 70, 128,
189, and between residues 212 and 213. The sequences at the splice junctions agree with the GT-AG rule of Breathnach and Chambon¹³²^,¹³³ and the consensus sequence of Mount,¹³³ except for one splice site at the 5′ end of intervening sequence L.¹¹²

FIGURE 9.15 The processing and biosynthesis of prothrombin and the generation of thrombin. (Reprinted from Davie EW, Kulman JD. An overview of the structure and function of thrombin. Semin Thromb hemost 2006;32(Suppl 1):3, with permission.)

The gene for prothrombin also contains 30 copies of Alu repetitive sequences, which make up 39% of the gene.¹¹² This family of DNA sequences is composed of approximately 300 nts, and the human haploid genome contains roughly 350,000 copies of Alu repetitive sequences.¹³⁴ In the prothrombin gene, many Alu sequences are tightly clustered and include five sets of tandem repeats. Intervening sequence L (9.5 kb) contains 20 Alu repeats, five of which occur in head-to-tail orientation with no additional DNA between them. The prothrombin gene also contains two copies of partial KpnI repeats (170 bp and 326 bp) located in intervening sequence L.

The mRNA for human prothrombin includes 1,866 nts that code for a prepro leader sequence of 43 amino acids and a mature polypeptide chain of 579 amino acids (FIGURE 9.14).¹¹¹^,¹¹² The prepro leader sequence includes a hydrophobic stretch of amino acids (residues -37 to -26) and ends with an arginine residue just before the amino-terminal alanine that is present in the circulating protein. The Arg-Ala peptide bond, however, is not cleaved by signal peptidase.¹³⁵ This enzyme cleaves the prepro polypeptide chain near the middle of the prepro leader sequence at one of the small amino acid residues such as alanine cysteine, or serine, leaving a propiece (18 to 24 amino acids) still attached to the polypeptide; the propiece is then cleaved by a second processing protease with a substrate preference for a basic residue at -4, -2, and -1.²⁹

A prepro leader sequence, typical of the vitamin K- dependent coagulation factors (FIGURE 9.8), plays a role in the carboxylation reaction that occurs on the lumen side of the rough endoplasmic reticulum.¹³⁶^,¹³⁷ The conserved Phe and Ala residues at positions -16 and -10 appear to play an important role in the recognition sequence for this carboxylase.¹³⁸^,¹³⁹

FIGURE 9.16 The catalytic and anion-binding exosites I and II in thrombin. The ribbon structure of thrombin (A) shows the catalytic triad of His43, Asp99, and Ser205 (enclosed in a red-dashed circle) and the salt bridge formed between Ile1 and Asp204 (enclosed in a blue dashed circle). The surface representations of thrombin are shown in the left (C), standard (B and D), and right (E) orientations. The substrate-binding cleft in thrombin runs from left to right (B,D) and is bordered by an upper lip or the 60 loop (magenta shading) and a lower lip or γ-loop (green shading). Residues constituting exosite I (dark blue) and exosite II (light blue) are enclosed by dashed ovals in (C), (D), and (E), while the Na-binding loop is shown in orange. (From Davie EW, Kulman JD. An overview of the structure and function of thrombin. Semin Thromb Hemost 2006;32(Suppl 1):3.)

FIGURE 9.16 (Continued)

Like the other vitamin K-dependent coagulation factors, the liver-specific expression of prothrombin is transcriptionally regulated. The immediate 5′-flanking sequence does not contain TATA or CCAAT boxes, and the two major transcription start sites are located at -36- and -23-bp upstream of the initiator codon.¹⁴⁰ Full tissue-specific promoter activity is located within 1,000 bp of the transcription start sites.¹⁴⁰^,¹⁴¹^,¹⁴² A weak positive element lies in the region 400-bp upstream of the mRNA coding sequence, accounting for approximately 5% of the total promoter activity in HepG2 cells. A liver-specific enhancer element is located in the region between -940 and -860 bp. DNA protein-binding studies and functional reporter gene analysis with mutant promoter constructs have demonstrated that this 80-bp enhancer contains an HNF-1-binding site flanked by a G-C-rich motif that binds a ubiquitous Sp1-like transacting factor.¹⁴⁰^,¹⁴¹^,¹⁴³ All together, six different transcription factors bind to the prothrombin enhancer region and at least three (HNF-4-alpha, HNF-3-beta, and Sp1/Sp3) are important in the regulation of prothrombin expression.¹⁴³

A common G to A substitution at nucleotide position 20,210 in the 3′-untranslated end of the prothrombin gene results in an elevated level of circulating prothrombin and a 2.8-fo1d increased risk for venous thrombosis.¹⁴⁴^,¹⁴⁵

Table 9.2 Location and size of exons and introns in the human prothrombin gene (F2)

Exon	Nucleotide Positions	Length (bp)	Amino Acids	Intron	Type^a	Nucleotide Positions	Length (bp)	Number of Alu Repeats
I	+1-79	79+^b	-43 to-17	A	1	80-465	386	—
II	466-626	161	-17 to 37	B	0	627-1,285	659	—
III	1,286-1,310	25	38-46	C	1	1,311-1,552	242	—
IV	1,553-1,603	51	46-63	D	1	1,604-3,929	2,326 4
V	3,930-4,035	106	63-98	E	2	4,036-4,131	96	—
VI	4,132-4,268	137	98-144	F	1	4,269-6,606	2,338	3
VII	6,607-6,921	315	144-249	G	1	6,922-7,245	324	—
VIII	7,246-7,374	129	249-292	H	1	7,375-7,458	84	—
IX	7,459-7,585	127	292-334	1	2	7,586-8,742	1,157	2
X	8,743-8,910	168	334-390	J	2	8,911-9,407	497	—
XI	9,408-9,581	174	390-448	K	2	9,582-10,123	542	1
XII	10,124-10,305	182	448-509	L	1	10,306-19,752	9,447	20^c
XIII	19,753-19,823	71	509-532	M	0	19,824-19,969	146	—
XIV	19,970-20,210	241	533—poly(A)	—	—	—	—	—
bp, base pair.
^aIntron placement as discussed by Mount SM. A catalogue of splice junction sequences. Nucleic Acids Res 1982;10:459, in which a type 0 indicates placement between two codons, a type 1 interrupts a codon between the first and second bases, and a type 2 occurs between the second and third bases of the codon.
^bThe length of the 5′ noncoding region of the messenger RNA for human prothrombin is unknown; therefore, the length of exon I is measured from the initiator methionine.
^cThis intron also has two copies of partial Kpn repeats.
Reprinted from Degen SJ, Davie EW. Nucleotide sequence of the gene for human prothrombin. Biochemistry 1987;26:6165, with permission.

FIGURE 9.17 Amino acid sequence of human prepro factor XI. The signal sequence (-18 to -1) is removed during biosynthesis by signal peptidase by cleavage of the Gly(-1)-Glu bond. Factor XI circulates in plasma as a homodimer connected by a single disulfide bond linking Cys321 in the fourth apple domains of both subunits. The location of the 14 introns (A to N) is shown by solid arrows. The four apple domains (of 90 or 91 amino acids) are labeled A1, A2, A3, and A4. The site of cleavage (Arg369-Ile) catalyzed by thrombin during the conversion of factor XI to factor XIa is shown with a small arrow. The three members of the catalytic triad (His413, Asp462, and Ser557) are circled. The four N-linked carbohydrate attachment sites (Asn72, 108, 432, and 473) are shown by solid diamonds.

FACTOR XI

Factor XI (Mr 143,000) is a plasma glycoprotein (5% carbohydrate) composed of 1,214 amino acids synthesized in the liver and secreted into the plasma as a zymogen that circulates as a complex with high molecular weight kininogen (HMWK).¹⁴⁶ Factor XI is an unusual zymogen to a serine protease in that it contains two identical polypeptide chains, each with a catalytic site,¹⁴⁷ linked by a single disulfide bond in the fourth apple domain (Cys321) (FIGURE 9.17).¹⁴⁸^,¹⁴⁹ Each of these chains contains four tandem repeats of 90 (or 91) amino acids (“apple” domains) that range in identity from 23% to 34% (FIGURE 9.17)¹⁵⁰ and are linked by disulfide bonds between the first and sixth, second and fifth, and third and fourth half-Cys.¹⁴⁸ An extra half-Cys (Cys11) present in apple 1 forms a disulfide bond with another half-Cys residue, whereas Cys321 in each fourth apple domain links the two identical polypeptide chains of the protein together by a disulfide bond. The four apple domains in factor XI are also highly homologous to the four tandem repeats present in plasma prekallikrein but have not been identified in any other protein.¹⁵¹^,¹⁵² Present evidence indicates that factor XI is bound to HMWK through apple 1, whereas apple 2 is involved in the interaction of factor XI with factor IX.¹⁵³^,¹⁵⁴^,¹⁵⁵^,¹⁵⁶ The two catalytic domains in factor XI contain amino acid sequences typical of the pancreatic trypsin family of serine proteases and are located in the carboxyl end of the protein (FIGURE 9.17). The four apple domains and the catalytic domains in the dimer of factor XI are readily observed in the crystal structure of the protein (FIGURE 9.18A,B).¹⁵⁷ Each apple domain (60 × 60 × 20 Å) consists of a single α-helix surrounded by seven antiparallel β-strands. The four tandem apple domains form a shape that resembles a flat saucer. In the dimer, the two saucers are located between the two catalytic domains where they form an inverted V shape.

FIGURE 9.18 Ribbon structures of the factor XI dimer (A) and the four apple domains of factor IX (B). The two catalytic sites are on the outside of the catalytic domains (colored magenta), while the two sets of four apple domains are in the middle where they are linked by a single disulfide bond between Cys321 from each of the apple four domains (A). Apple 1 is colored gray, apple 2 is blue, apple 3 is orange, and apple 4 is yellow. (Reprinted from Papagrigoriou E, McEwan PA, Walsh PN, et al. Crystal structure of the factor XI zymogen reveals a pathway for transactivation. Nat Struct Mol Biol 2006;13:557, with permission.)

Factor XIa is a glycoprotein with five potential N-glycosylation sites in each chain. There are four carbohydrate chains, including two on the heavy chain (Asn72 and Asn108) and two on the catalytic or light chain (Asn432 and Asn473), but none on Asn335 of the heavy chain.

The conversion of factor XI to factor XIa catalyzed by thrombin readily occurs in the presence of a negatively charged material such as dextran sulfate, sulfatide, or heparin.¹⁵⁸^,¹⁵⁹ Factor XI can also be activated in vitro by factor XIIa in the presence of HMWK and a negatively charged surface.¹⁶⁰ Both of these activation reactions are due to the cleavage of an internal Arg369-Ile peptide bond in each of the two polypeptide chains,¹⁵⁰ resulting in the formation of factor XIa, a serine protease composed of two heavy chains (each 369 amino acids) and two light chains (each 238 amino acids) held together by three disulfide bonds. Each of the two light chains contains a serine protease domain that starts with Ile370. During the activation of factor XI, the newly generated Ile370 turns and flips into the activation pocket where its α-ammonium group forms a salt bridge with the carboxylate group of Asp556. This generates an active serine protease that includes the catalytic triad of His413, Asp462, and Ser557 in the light chain of each factor XIa subunit.

The cDNA for human factor XI has been isolated from a γgt11 expression library prepared from human liver. This mRNA was approximately 2,100 nts long.¹⁵⁰ These data also indicated that factor XI is synthesized as a single polypeptide chain with a typical leader sequence of 18 amino acids (FIGURE 9.17). Each of the two chains present in the mature molecule contains 607 amino acids. Also, the cDNA for factor XI contains a potential polyadenylation or processing sequence of AACAAA rather than the typical AATAAA.²⁸ This sequence is located 21 nts upstream from the poly(A) tail and is present in the 166 nts that constitute the 3′ noncoding sequence of the mRNA.

The gene for human factor XI (F11) is approximately 23 kb, located on the distal end of the long arm of chromosome 4 (4q35)¹⁶¹^,¹⁶² and contains 15 exons interrupted by 14 introns (FIGURE 9.17). The first exon codes for the 5′-untranslated region, whereas exon II codes for the signal peptide. The four apple domains are coded by the next eight exons. Each apple domain is coded by two exons interrupted by a single intron, and these introns are located in essentially the same position within each of the four apple domains. The carboxyl-terminal region of factor XIa containing the catalytic chain is coded by five exons, four of which are located in the same positions as those in the genes for human tissue plasminogen activator and human urokinase.

Factor XI deficiency is an unusually mild bleeding tendency that occurs in either sex.¹⁶³ Factor XI deficiency is found primarily in the Ashkenazi Jewish population,¹⁶⁴ and mutations occurring almost entirely in two regions of the gene coding for apple 2 and apple 4. The first mutation (type II) results in a stop codon at residue 117 in apple 2, where GAA coding for Gln117 is replaced by a stop codon of TAA (FIGURE 9.18B), leading to the synthesis of a truncated polypeptide and loss of biological activity. The second principal mutation (type III) results in an amino acid substitution at Phe283 in the fourth apple domain by Leu,¹⁶⁵ resulting in reduced dimerization of the molecule and a lowered secretion.¹⁶⁶ The third, far less common mutation (type I), disrupts normal mRNA splicing and changes a nucleotide sequence of GTAAC to ATAAC at the last intron-exon boundary.¹⁶⁵

In a study of 43 patients with severe factor XI deficiency, 49% were due to the type II mutation and 47% to type III.¹⁶⁷^,¹⁶⁸ More than 152 different mutations have been published that occur in all four apple domains as well as in the catalytic domain.¹⁶⁹

FACTOR IX

The gene for factor IX (F9), like factor VIII, is located in the distal region of the long arm of the X chromosome. No link, however, exists between the genes for these two proteins.¹⁷⁰^,¹⁷¹^,¹⁷² The factor IX gene is in region Xq27 and is closely linked to the fragile X site.¹⁷³

Factor IX (Mr 57,000) is a single-chain glycoprotein (17% car bohydrate) composed of 415 amino acids (FIGURE 9.19).¹⁷⁰^,¹⁷¹^,¹⁷² Like prothrombin and factor VII, it is a vitamin K-dependent protein and contains 12 residues of γ-carboxyglutamic acid
(Gla domain) located in the amino-terminal region of the protein. The Gla domain is followed by two EGF-like domains, which show considerable homology with the corresponding regions of factor VII, factor X, protein C, and protein S.⁷³^,⁷⁴^,⁷⁵^,¹⁷⁴ The two EGF domains in factor IX are followed by an activation glycopeptide and a catalytic domain of 235 amino acids (FIGURE 9.19). The individual domains are readily identified in tulip-shaped crystal structure (FIGURE 9.20),¹⁷⁵ which has a stem composed of the amino-terminal Gla domain and two EGF domains attached to the catalytic domain that forms the blossom (FIGURE 9.20).

Only gold members can continue reading. Log In or Register to continue