Patent Application
20090271163
The present invention relates to human Factor VIII, in particular, B-domain deleted human Factor VIII, methods for its crystallization, crystals, 3-dimensional structures, and uses thereof.
Factor VIII is a protein cofactor that, when activated, forms a complex with Factor IXa on membrane surfaces to activate factor X during blood coagulation. This glycoprotein is encoded by a gene of 186 kb that is divided into 26 exons. Hemophilia A is caused by defects in the Factor VIII gene that leads to diminished or absent Factor VIII activity in blood, for example, missense mutations, nonsense mutations, gene deletions of varying size, inversions and splice junction mutations. The major treatment of the bleeding disorder associated with hemophilia involves the infusion of Factor VIII into the circulation of patients with hemophilia and the correction of hemostasis.
The three dimensional structure of Factor VIII is unknown.
The present invention provides crystals of human Factor VIII, in particular, a B-domain deleted human Factor VIII, and its three-dimensional structure. The analysis of the three dimensional structure provides previously unknown structural information about the human Factor VIII protein which can be used for the design and development of novel, potent and specific therapeutics for the treatment of hemophilia and other thromboembolic disorders.
In one aspect, the present invention provides a composition comprising a human Factor VIII, in a crystalline form. In particular, the human Factor VIII lacks at least a portion of B-domain. In one embodiment, the crystalline human Factor VIII has a space group P41212. In another embodiment, the crystalline human Factor VIII has unit cell dimensions a=b=134.11 Å, c=349.76 Å. In yet another embodiment, the crystalline human Factor VIII diffracts X-rays for a determination of structural coordinates to a resolution of about 4.0 Angstroms or below (e.g., about 3.8 Å or below, about 3.6 Å or below, about 3.4 Å or below, about 3.2 Å or below, about 3.0 Å or below, about 2.8 Å or below, about 2.5 Å or below, about 2.4 Å or below, about 2.3 Å or below, about 2.2 Å or below, about 2.1 Å or below, about 2.0 Å or below, about 1.9 Å or below, about 1.8 Å or below, about 1.7 Å or below, about 1.6 Å or below, about 1.5 Å or below, about 1.4 Å or below). In one particular example, the crystalline human Factor VIII diffracts X-rays for a determination of structural coordinates to a resolution at about 3.98 Angstroms.
In one embodiment, the human Factor VIII suitable for the present invention includes an amino acid sequence at least 70%, 75%, 80%, 85%, 90%, 95%, 98%, or 99% identical to SEQ ID NO:1. In a preferred embodiment, the human Factor VIII suitable for the present invention includes the amino acid sequence of SEQ ID NO:1. In another embodiment, the human Factor VIII suitable for the present invention includes an amino acid sequence at least 70%, 75%, 80%, 85%, 90%, 95%, 98%, or 99% identical to SEQ ID NO:2. In a preferred embodiment, the human Factor VIII suitable for the present invention includes the amino acid sequence of SEQ ID NO:2. More preferably, the human Factor VIII suitable for the present invention includes amino acid sequences of SEQ ID NO:1 and SEQ ID NO:2.
In another aspect, the present invention provides a computer-readable medium containing computer-readable data defined by structural coordinates of atoms of human Factor VIII according to Table 2, +/−a root mean square deviation for alpha carbon atoms of less than 2 Angstroms, or selected coordinates thereof. As used herein, a root mean square deviation for alpha carbon atoms of less than 2 Angstroms includes a root mean square deviation for alpha carbon atoms at about 2 Å or less, about 1.8 Å or less, about 1.5 Å or less, about 1.2 Å or less, about 1.0 Å or less, or about 0.5 Å or less.
In some embodiments, the computer-readable data is defined by structural coordinates of atoms of human Factor VIII according to Table 2, +/−a root mean square deviation for alpha carbon atoms of less than 2 Angstroms, or at least 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% of the coordinates thereof.
In some embodiments, the computer-readable data is defined by structural coordinates of one or more atoms selected from the group consisting of atoms of Lys 107, Glu 110, Asp 116, Glu 122, Asp 126, Asp 125, His 267, Cys 310, His 315, His 1954, Cys 2000, and His 2005. In other embodiments, the computer-readable data is defined by structural coordinates of one or more atoms selected from the group consisting of atoms of residues 558-565, 707-712 and 1811-1819 according to Table 2. In further embodiments, the computer-readable data are defined by three-dimensional structural coordinates of one or more Factor VIII domains selected from the group consisting of domain A1, A2, A3, C1 and C2.
In yet another aspect, the present invention provides a method for constructing a three-dimensional structural representation of a Factor VIIIa-Factor IXa complex, including a complex containing at least a region of Factor VIIIa and at least a region of Factor IXa. The method includes the steps of: (a) providing a three-dimensional structural representation of human Factor VIII, or a region thereof; (b) providing a three-dimensional structural representation of Factor IXa, or a region thereof, and (c) fitting the three-dimensional structural representation from step (a) to the three-dimensional structural representation from step (b).
In some embodiments, the three-dimensional structural representation of step (a) is defined by structural coordinates of atoms of human Factor VIII according to Table 2, +/−a root mean square deviation for alpha carbon atoms of less than 2 Angstroms, or selected coordinates thereof.
In some embodiments, the three-dimensional structural representation of step (a) is defined at least by structural coordinates of one or more atoms selected from the group consisting of atoms of residues 558-565, 707-712 and 1811-1819 according to Table 2.
In other embodiments, step (c) of the method of this aspect of the invention is based on one or more constraints selected from the group consisting of: (1) residues 558-565 of Factor VIIIa interact with residues 330-339 of Factor IXa; (2) residues 707-712 of Factor VIIIa interact with residues 301-303 of Factor IXa; (3) residues 1811-1819 of Factor VIIIa interact with the light chain of Factor IXa; (4) Phe 25 in the Gla domain of Factor IXa is juxtaposed with the light chain of Factor VIIIa; and (5) the Gla domain of Factor IXa is situated within a phospholipid membrane.
The present invention also provides computer-readable media containing computer-readable data defined by a three-dimensional structural representation of a complex comprising at least a region of Factor VIIIa and a region of Factor IXa constructed according to the methods as described in various embodiments above.
In still another aspect, the present invention provides a method for modifying human Factor VIIIa to alter its interaction with Factor IXa. The method includes the steps of: (a) providing a structural representation of human Factor VIIIa, or a region thereof; (b) fitting the structural representation of step (a) to a three-dimensional structural representation of Factor IXa, or a region thereof; and (d) computationally modifying the structural representation of step (a) to increase or decrease its interaction with the three-dimensional structural representation of Factor IXa, or a region thereof.
In some embodiments, the structural representation of step (a) is defined by structural coordinates of atoms of human Factor VIII according to Table 2, +/−a root mean square deviation for alpha carbon atoms of less than 2 Angstroms, or selected coordinates thereof.
In other embodiments, the structural representation of step (a) is defined at least by structural coordinates of one or more atoms selected from the group consisting of atoms of residues 558-565, 707-712 and 1811-1819 according to Table 2.
In some embodiments, the three-dimensional structural representation of Factor IXa, or a region thereof, is defined at least by the Gla domain of Factor IXa. In other embodiments, the three-dimensional structural representation of Factor IXa, or a region thereof, is defined at least by the light chain of Factor IXa.
In some embodiments, the structural representation of step (a) is modified to increase its interaction with the three-dimensional structural representation of Factor IXa, or a region thereof.
The present invention also provides modified human Factor VIII with increased interaction with Factor IXa according to the methods described in various embodiments above.
In one aspect, the present invention provides a method for evaluating the activity of a modified human Factor VIII. The method includes the steps of: (a) providing a structural representation of human Factor VIII, or a region thereof; (b) computationally modifying the structural representation of step (a) to introduce one or more amino acid modifications; and (c) evaluating the activity of the modified human Factor VIII based on the modified structural representation from step (b).
In some embodiments, the structural representation of step (a) is defined by coordinates of atoms of human Factor VIII according to Table 2, +/−a root mean square deviation for alpha carbon atoms of less than 2 Angstroms, or selected coordinates thereof.
In one embodiment, the evaluation of the activity of the modified human Factor VIII is based on an evaluation of its interaction with Factor IXa. In another embodiment, the evaluation of the activity of the modified human Factor VIII is based on an evaluation of its interaction with a phospholipid membrane. In yet another embodiment, the evaluation of the activity of the modified human Factor VIII is based on an evaluation of its interaction with von Willebrand Factor (vWF).
In some embodiments, the method of this aspect of the invention further includes a step of identifying a modified Factor VIII with an improved property. In one particular embodiment, the modified Factor VIII has increased plasma half-life.
In another aspect, the present invention provides a method of predicting a three dimensional structure of a human Factor VIII homologue or analogue of unknown structure. The method includes the steps of: (a) aligning an amino acid sequence of a target human Factor VIII homolog or analog of unknown structure with the amino acid sequence of human Factor VIII defined by coordinates according to Table 2, +/−a root mean square deviation for alpha carbon atoms of less than 2 Angstroms, or selected coordinates thereof, to match one or more homologous regions; (b) modeling the structure of the matched one or more homologous regions of the target human Factor VIII homolog or analog of unknown structure on the corresponding regions of the human Factor VIII as defined by coordinates according to Table 2, +/−a root mean square deviation for alpha carbon atoms of less than 2 Angstroms, or selected coordinates thereof; and (c) determining a structural conformation for said target human Factor VIII homolog or analog of unknown structure which substantially preserves the structure of said matched one or more homologous regions. In one embodiment, the method of this aspect of the invention further includes a step of evaluating the activity of the target human Factor VIII homolog or analog of unknown structure based on the structural conformation determined at step (c).
In still another aspect, the present invention provides a method for designing a mimetic compound of human Factor VIII. The method includes the steps of: (a) providing a selected human Factor VIII structure that is associated with a biological activity of human Factor VIII; (b) superimposing a three-dimensional structure of a compound on the selected human Factor VIII structure; and (c) modifying the three-dimensional structure of the compound such that the modified three-dimensional structure comprises a structural confirmation substantially mimicking the selected human Factor VIII structure. In one embodiment, the mimetic compound includes an antibody structure. In another embodiment, the method further includes the steps of: (d) synthesizing the modified compound from step (c); and (e) evaluating the activity of the modified compound.
The present invention also provides mimetic antibodies of human Factor VIII designed by the methods of this aspect of the invention.
In yet another aspect, the present invention provides a method for rational drug design. The method includes the steps of: (a) providing selected coordinates of a human Factor VIII structure; (b) providing a plurality of moieties; (c) fitting the structure of each of the plurality of moieties to the selected coordinates; (d) selecting one or more moieties that fit into the selected coordinates; and (e) assembling the one or more moieties selected from step (d) into a single molecule to form a candidate modulator molecule. In some embodiments, the selected coordinates of a human Factor VIII structure comprises one or more coordinates as defined in Table 2, +/−a root mean square deviation for alpha carbon atoms of less than 2 Angstroms.
In some embodiments, wherein the moieties suitable for the method of this aspect of the invention are selected from the group consisting of molecular fragments, small molecules, ligands designed de novo, and compounds known to bind Factor VIII or modified compounds thereof.
In other embodiments, the method of this aspect of the invention further includes the steps of: (f) obtaining or synthesizing the candidate modulator molecule; and (g) contacting the candidate modulator molecule with human Factor VIII to determine the ability of the candidate modulator molecule to interact with human Factor VIII.
In a further aspect, the present invention provides a method for producing a computer readable database including a structural representation of at least one compound capable of binding human Factor VIII. The method includes the steps of: (a) introducing into a computer program selected coordinates of a human Factor VIII structure; (b) fitting a three-dimensional model of at least one binding test compound on the selected coordinates; (d) assessing whether said test compound model fits spatially into the selected coordinates; and (e) storing a structural representation of a compound that fits into the selected coordinates.
The present invention also provides a computer readable database produced by the method of this aspect of the invention.
Other features, objects, and advantages of the present invention are apparent in the detailed description, drawings and claims that follow. It should be understood, however, that the detailed description, the drawings, and the claims, while indicating embodiments of the present invention, are given by way of illustration only, not limitation. Various changes and modifications within the scope of the invention will become apparent to those skilled in the art.
The drawings are for illustration purposes only, not for limitation.
The present invention provides crystals of human Factor VIII, in particular, a B-domain deleted human Factor VIII, and its three-dimensional structure. The present invention also provides the structural information of Factor VIII, and methods for identifying compounds that modulate Factor VIII activity, for determining structures of Factor VIII homologs or analogs, and for designing drug candidates for the treatment of hemophilia and other thromboembolic disorders based on the structural information.
Various aspects of the invention are described in detail in the following sections. The use of sections is not meant to limit the invention. Each section can apply to any aspect of the invention. In this application, the use of “or” means “and/or” unless stated otherwise.
Factor VIII is a protein cofactor that, when activated, forms a complex with Factor IXa on membrane surfaces to activate factor X during blood coagulation (Furie, B. & Furie, B. C. (1988) Cell 53, 505-518). This glycoprotein is encoded by a gene of 186 kb that is divided into 26 exons (Gitschier, J., Wood, W. I., Goralka, T. M., Wion, K. L., Chen, E. Y., Eaton, D. H., Vehar, G. A., Capon, D. J. & Lawn, R. M. (1984) Nature 312, 326-30; Toole, J. J., Knopf, J. L., Wozney, J. M., Sultzman, L. A., Buecker, J. L., Pittman, D. D., Kaufman, R. J., Brown, E., Shoemaker, C., Orr, E. C. & et al. (1984) Nature 312, 342-7). Factor VIII is synthesized as a single polypeptide chain, including a 19-residue signal peptide. The mature Factor VIII contains 2,332 amino acid residues arranged within five domains organized as A1-A2-B-A3-C1-C2 (Gitschier, J., Wood, W. I., Goralka, T. M., Wion, K. L., Chen, E. Y., Eaton, D. H., Vehar, G. A., Capon, D. J. & Lawn, R. M. (1984) Nature 312, 326-30; Toole, J. J., Knopf, J. L., Wozney, J. M., Sultzman, L. A., Buecker, J. L., Pittman, D. D., Kaufman, R. J., Brown, E., Shoemaker, C., Orr, E. C. & et al. (1984) Nature 312, 342-7) (
Factor VIII circulates in the blood as a heterodimer composed of two polypeptide chains, a light chain with a molecular weight of about 80,000 and a heterogeneous heavy chain with a molecular weight varying between about 90,000 and 200,000, both derived from the single peptide chain. A region of the C2 domain defines the membrane-binding properties of Factor VIII and the site of interaction with von Willebrand factor (Pratt, K. P., Shen, B. W., Takeshima, K., Davie, E. W., Fujikawa, K. & Stoddard, B. L. (1999) Nature 402, 439-442). Factor VIII is inactive or minimally active as a cofactor in blood coagulation, but is converted into its active cofactor form by proteolytic cleavage. Although active Factor VIII can be formed from cleavage at Arg 372 and Arg 1689, it is generally appreciated that activated Factor VIII is generated from 3 cleavage events (Arg 372, Arg 1689, and Arg 740).
Activated Factor VIII (Factor VIIIa) acts as a cofactor for activated Factor IX (Factor IXa) to accelerate the conversion of factor X to activated factor X (factor Xa). Factor Xa converts prothrombin into thrombin. Thrombin then converts fibrinogen into fibrin and a clot is formed.
Classic hemophilia (also known as hemophilia A) is caused by a defect in the Factor VIII gene that leads to diminished or absent Factor VIII activity in blood. A heterogeneous genetic disease, hemophilia A has been associated with missense mutations, nonsense mutations, gene deletions of varying size, inversions and splice junction mutations (Furie, B. & Furie, B. C. (1990) Semin Hematol 27, 270-85; Graw, J., Brackmann, H. H., Oldenburg, J., Schneppenheim, R., Spannagl, M. & Schwaab, R. (2005) Nat Rev Genet. 6, 488-501). The major treatment of the bleeding disorder associated with hemophilia involves the infusion of Factor VIII into the circulation of patients with hemophilia and the correction of hemostasis (Mannucci, P. M. & Tuddenham, E. G. (2001) N Engl J Med 344, 1773-9; Key, N. S. & Negrier, C. (2007) Lancet 370, 439-48).
Based upon the observation that the B-domain of porcine Factor VIII shows no sequence homology to the B-domain of human Factor VIII yet porcine and human Factor VIII have similar specific coagulant activities when evaluated in human plasma, engineered B-domain deleted human Factor VIII was shown to have full biological activity and could be expressed in heterologous cells with improved expression efficiency relative to that of the full-length molecule (Toole, J. J., Pittman, D. D., Orr, E. C., Murtha, P., Wasley, L. C. & Kaufman, R. J. (1986) Proc Natl Acad Sci USA 83, 5939-42). The structural heterogeneity of B-domain deleted Factor VIII is significantly less than that for full length Factor VIII, with a heavy chain of 90,000 and a light chain of 80,000 molecular weight (
As used herein, a “B-domain deleted human Factor VIII” includes a Factor VIII, or a structural or functional variant, that lacks at least a portion or the entirety of the B-domain. As used herein, the B-domain of Factor VIII corresponds to amino residues 741 thru and including 1648 of human Factor VIII.
Thus, a B-domain deleted human Factor VIII suitable for the invention includes a heavy chain and a light chain. The heavy chain of human Factor VIII includes amino acid residues 1-740 (A1-a1-A2-a2) and the light chain of human Factor VIII includes amino acid residues 1649-2332 (a3-A3-C1-C2). This sequence of the heavy chain of human Factor VIII is shown below as SEQ ID NO:1 and the sequence of the light chain of human Factor VIII is shown below as SEQ ID NO:2.
The heavy chain of human Factor VIII (including residues 1-740) (SEQ ID NO:1)
The light chain of human Factor VIII (including residues 1649-2332) (SEQ ID NO:2):
A B-domain deleted Factor VIII suitable for the invention may also contain modified or mutated heavy and/or light chains. Modified or mutant heavy or light chains may be generated by replacing at least one amino acid residue in a native polypeptide with a different amino acid residue, or by adding or deleting amino acid residues within the native polypeptide or at the N- or C-terminus of the native polypeptide. Preferably, the B-domain deleted Factor VIII including modified heavy and/or light chains has substantially the same three-dimensional structure. By having substantially the same three-dimensional structure is meant having a set of atomic structure coordinates that have a root-mean-square deviation of less than or equal to about 2 Å when superimposed with the atomic structure coordinates of the native protein, or a region thereof, from which the mutant is derived when at least about 50% to 100% of the alpha carbon atoms of the corresponding native protein, or a region thereof, are included in the superposition.
Amino acid substitutions, deletions and additions which do not significantly interfere with the three-dimensional structure of Factor VIII will depend, in part, on the region of Factor VIII where the substitution, addition or deletion occurs. In highly variable regions of the molecule, non-conservative substitutions as well as conservative substitutions may be tolerated without significantly disrupting the three-dimensional, structure of the molecule. In highly conserved regions, or regions containing significant secondary structure, conservative amino acid substitutions are preferred.
Conservative amino acid substitutions are well known in the art, and include substitutions made on the basis of similarity in polarity, charge, solubility, hydrophobicity, hydrophilicity and/or the amphipathic nature of the amino acid residues involved. For example, negatively charged amino acids include aspartic acid and glutamic acid; positively charged amino acids include lysine and arginine; amino acids with uncharged polar head groups having similar hydrophilicity values include the following: leucine, isoleucine, valine; glycine, alanine; asparagine, glutamine; serine, threonine; phenylalanine, tyrosine. Other conservative amino acid substitutions are well known in the art.
For proteins obtained in whole or in part by chemical synthesis, the selection of amino acids available for substitution or addition is not limited to the genetically encoded amino acids. Indeed, the mutants described herein may contain non-genetically encoded amino acids. Conservative amino acid substitutions for many of the commonly known non-genetically encoded amino acids are well known in the art. Conservative substitutions for other amino acids can be determined based on their physical properties as compared to the properties of the genetically encoded amino acids.
In some instances, it may be particularly advantageous or convenient to substitute, delete and/or add amino acid residues to a native protein in order to provide convenient cloning sites in cDNA encoding the polypeptide, to aid in purification of the polypeptide, and for crystallization of the polypeptide. Such substitutions, deletions and/or additions which do not substantially alter the three dimensional structure of Factor VIII will be apparent to those of ordinary skill in the art.
Thus, a human Factor VIII suitable for the present invention includes an amino acid sequence at least 70%, 75%, 80%, 85%, 90%, 95%, 98%, or 99% identical to SEQ ID NO:1. In one particular embodiment, the human Factor VIII suitable for the present invention includes the amino acid sequence of SEQ ID NO:1. In another embodiment, the human Factor VIII suitable for the present invention includes an amino acid sequence at least 70%, 75%, 80%, 85%, 90%, 95%, 98%, or 99% identical to SEQ ID NO:2. In one particular embodiment, the human Factor VIII suitable for the present invention includes the amino acid sequence of SEQ ID NO:2. More particularly, the human Factor VIII suitable for the present invention includes amino acid sequences of SEQ ID NO:1 and SEQ ID NO:2.
In addition, the human Factor VIII suitable for the invention may also contain a linker, for example, at the C-terminus of the heavy chain, that connects the heavy chain and the light chain. A suitable linker may be derived from the amino acid sequence of the B-domain. For example, a linker may contain amino acids derived from the N-terminal region and/or the C-terminal region of the B-domain. One such exemplary linker sequence is as follows: SFSQNPPVLKRHQR (SEQ ID NO:3). Alternatively, a linker may incorporate artificial amino acid sequences other than any naturally-occurring sequences and is generally designed to be flexible or to interpose a structure, such as an alpha-helix, between the two protein moieties.
X-ray crystallography is a method of solving the three dimensional structures of molecules. The structure of a molecule is calculated from X-ray diffraction patterns using a crystal as a diffraction grating. Three dimensional structures of protein molecules arise from crystals grown from a concentrated aqueous solution of that protein. The process of X-ray crystallography can include the following steps: (a) synthesizing and isolating (or otherwise obtaining) a protein; (b) growing a crystal from an aqueous solution comprising the protein with or without a modulator; and (c) collecting X-ray diffraction patterns from the crystals, determining unit cell dimensions and symmetry, determining electron density, fitting the amino acid sequence of the protein to the electron density, and refining the structure.
Preparation of B-Domain Deleted Human Factor VIII
A B-domain deleted human Factor VIII described herein may be chemically synthesized in whole or part using techniques that are well-known in the art (see, e.g., Creighton (1983) Biopolymers 22(1):49-58). Alternatively, methods which are well known to those skilled in the art can be used to construct expression vectors containing the Factor VIII coding sequence and appropriate transcriptional/translational control signals. These methods include in vitro recombinant DNA techniques, synthetic techniques and in vivo recombination/genetic recombination. See, for example, the techniques described in Maniatis, T (1989). Molecular cloning: A laboratory Manual. Cold Spring Harbor Laboratory, New York. Cold Spring Harbor Laboratory Press; and Ausubel, F. M. et al. (1994) Current Protocols in Molecular Biology. John Wiley & Sons, Secaucus, N.J.
B-domain deleted human Factor VIII can be expressed in a variety of mammalian cell lines including, but not limited to, human embryonic kidney (HEK) 293, Chinese hamster ovary (CHO), monkey kidney (COS), HTH080, C10, HeLa, baby hamster kidney (BHK), 3T3, C127, CV-1, HaK, NS/O, and L-929 cells. B-domain deleted human Factor VIII can also be expressed in a variety of non-mammalian host cells such as, for example, insect (e.g., Sf-9, Sf-21, Hi5), plant (e.g., Leguminosa, cereal, or tobacco), yeast (e.g., S. cerivisae, P. pastoris), prokaryote (e.g., E. Coli, B. subtilis and other Bacillus spp., Pseudomonas spp., Streptomyces spp), or fungus. Detailed preparation of B-domain deleted human Factor VIII was described in Sandberg et al. (2001) Seminars in Hematology, Vol. 38, No. 2, Suppl. 4: pp 4-12; and Eriksson et al. (2001) Semin. Hematol., 38:24-31, the teachings of both of which are hereby incorporated by reference.
Crystal Growth
Typically, crystals are grown from an aqueous solution containing the purified and concentrated protein by a variety of techniques. Suitable techniques include batch, liquid, bridge, dialysis, vapor diffusion, and hanging drop methods. McPherson (1982) John Wiley, New York; McPherson (1990) Eur. J. Biochem. 189:1-23; Webber (1991) Adv. Protein Chem. 41:1-36, incorporated by reference herein in their entireties, including all figures, tables, and drawings.
The crystals of the invention are, in general, grown by adding precipitants to the concentrated solution of the polypeptide. The precipitants are added at a concentration just below that necessary to precipitate the protein. Water is removed by controlled evaporation to produce precipitating conditions, which are maintained until crystal growth ceases. For crystals of the invention, exemplary crystallization conditions are described in the Examples. Those of ordinary skill in the art will recognize that the exemplary crystallization conditions can be varied. Such variations may be used alone or in combination. In addition, other crystallizations may be found, e.g., by using crystallization screening plates to identify such other conditions.
Derivative crystals of the invention can be obtained by soaking original crystals in mother liquor containing salts of heavy metal atoms. Heavy metal atoms useful for providing derivative crystals include, by way of example and not limitation, gold, mercury, selenium, etc. It has been found that soaking an original crystal in a solution containing about 0.1 mM to about 5 mM thimerosal, 4-chloromeruribenzoic acid or KAu(CN)2 for about 2 hr to about 72 hr provides derivative crystals suitable for use as isomorphous replacements in determining the X-ray crystal structure of Factor VIII.
Co-crystals of the invention can be obtained by soaking a crystal of Factor VIII in mother liquor containing compound that binds Factor VIII, or can be obtained by co-crystallizing the Factor VIII protein in the presence of a binding compounds. Thus, the co-crystals generally comprise a crystalline Factor VIII, or a region thereof, in association with one or more compounds. The association may be covalent or non-covalent. Such compounds include, but are not limited to, cofactors, substrates, substrate analogues, inhibitors, allosteric effectors, etc.
Generally, co-crystallization of Factor VIII and binding compound can be accomplished using conditions identified for crystallizing the corresponding Factor VIII without binding compound. It is advantageous if a plurality of different crystallization conditions have been identified for Factor VIII, and these can be tested to determine which condition gives the best co-crystals. It may also be beneficial to optimize the conditions for co-crystallization.
Determining Unit Cell Dimensions and the Three Dimensional Structure
Once the crystal is grown, it can be placed in a glass capillary tube or other mounting device and mounted onto a holding device connected to an X-ray generator and an X-ray detection device. Collection of X-ray diffraction patterns are well documented by those in the art. See, e.g., Ducruix and Geige, (1992), IRL Press, Oxford, England, and references cited therein. A beam of X-rays enters the crystal and then diffracts from the crystal. An X-ray detection device can be utilized to record the diffraction patterns emanating from the crystal. Although the X-ray detection device on older models of these instruments is a piece of film, modern instruments digitally record X-ray diffraction scattering. X-ray sources can be of various types, but advantageously, a high intensity source is used, e.g., a synchrotron beam source.
Methods for obtaining the three dimensional structure of the crystalline form of a peptide molecule or molecule complex are well known in the art. See, e.g., Ducruix and Geige, (1992), IRL Press, Oxford, England, and references cited therein. The following are steps in the process of determining the three dimensional structure of a molecule or complex from X-ray diffraction data.
After the X-ray diffraction patterns are collected from the crystal, the unit cell dimensions and orientation in the crystal can be determined. They can be determined from the spacing between the diffraction emissions as well as the patterns made from these emissions. The unit cell dimensions are characterized in three dimensions in units of Angstroms (one Å=10−10 meters) and by angles at each vertices. The symmetry of the unit cell in the crystals is also characterized at this stage. The symmetry of the unit cell in the crystal simplifies the complexity of the collected data by identifying repeating patterns. Application of the symmetry and dimensions of the unit cell is described below.
Each diffraction pattern emission is characterized as a vector and the data collected at this stage of the method determine the amplitude of each vector. The phases of the vectors can be determined using multiple techniques. In one method, heavy atoms can be soaked into a crystal, a method called isomorphous replacement, and the phases of the vectors can be determined by using these heavy atoms as reference points in the X-ray analysis. (Otwinowski, (1991), Daresbury, United Kingdom, 80-86). The isomorphous replacement method usually utilizes more than one heavy atom derivative. In another method, the amplitudes and phases of vectors from a crystalline polypeptide with an already determined structure can be applied to the amplitudes of the vectors from a crystalline polypeptide of unknown structure and consequently determine the phases of these vectors. This second method is known as molecular replacement and the protein structure which is used as a reference must have a closely related structure to the protein of interest. (Naraza (1994) Proteins 11:281-296). For example, the structural information from any isolated Factor VIII domain, or related proteins such as, Factor Va or ceroluplasmin, can be used as references for the molecular replacement analysis.
Once the phases of the vectors describing the unit cell of a crystal are determined, the vector amplitudes and phases, unit cell dimensions, and unit cell symmetry can be used as terms in a Fourier transform function. The Fourier transform function calculates the electron density in the unit cell from these measurements. The electron density that describes one of the molecules or one of the molecule complexes in the unit cell can be referred to as an electron density map. The amino acid structures of the sequence or the molecular structures of compounds complexed with the crystalline polypeptide may then be fitted to the electron density using a variety of computer programs. This step of the process is sometimes referred to as model building and can be accomplished by using computer programs such as Turbo/FRODO or “O” (Jones (1985) Methods in Enzymology 115:157-171), or AMoRe (Navaza, (1994) Acta Cryst. A50:157-163)
A theoretical electron density map can then be calculated from the amino acid structures fit to the experimentally determined electron density. The theoretical and experimental electron density maps can be compared to one another and the agreement between these two maps can be described by a parameter called an R-factor. A low value for an R-factor describes a high degree of overlapping electron density between a theoretical and experimental electron density map.
The R-factor is then minimized by using computer programs that refine the theoretical electron density map. A computer program such as X-PLOR can be used for model refinement by those skilled in the art. Briinger (1992) Nature 355:472-475. Other suitable computer programs such as or REFMAC are well known in the art. Refinement may be achieved in an iterative process. A first step can entail altering the conformation of atoms defined in an electron density map. The conformations of the atoms can be altered by simulating a rise in temperature, which will increase the vibrational frequency of the bonds and modify positions of atoms in the structure. At a particular point in the atomic perturbation process, a force field, which typically defines interactions between atoms in terms of allowed bond angles and bond lengths, Van der Waals interactions, hydrogen bonds, ionic interactions, and hydrophobic interactions, can be applied to the system of atoms. Favorable interactions may be described in terms of free energy and the atoms can be moved over many iterations until a free energy minimum is achieved. The refinement process can be iterated until the R-factor reaches a minimum value.
The three dimensional structure of the molecule or molecule complex is described by atoms that fit the theoretical electron density characterized by a minimum R-value. A file can then be created for the three dimensional structure that defines each atom by coordinates in three dimensions. An example of such a structural coordinate file is shown in Table 2.
The present invention provides three-dimensional structures and atomic structure coordinates of B-domain deleted human Factor VIII as determined by X-ray crystallography. The specific methods used to obtain the structure coordinates are provided in the examples. Exemplary atomic structure coordinates of B-domain deleted human Factor VIII are listed in Table 2.
Those having skill in the art will recognize that atomic structure coordinates as determined by X-ray crystallography are not without error. Thus, it is to be understood that any set of structure coordinates obtained for crystals of B-domain deleted human Factor VIII, or a region thereof, that have a root mean square deviation (“rmsd”) of less than or equal to about 2 Å when superimposed, using backbone atoms (e.g. N, Ca, C or O), on the corresponding structure coordinates listed in Table 2 are considered to be identical with the structure coordinates listed in the Table 2 when at least about 50% to 100% of the corresponding backbone atoms are included in the superposition. As used herein, a root mean square deviation for alpha carbon atoms of less than 2 Angstroms includes a root mean square deviation for alpha carbon atoms at about 2 Å or less, at about 1.8 Å or less, at about 1.5 Å or less, at about 1.2 Å or less, at about 1.0 Å or less, or at about 0.5 Å or less.
The overall three dimensional structure of B-domain deleted Factor VIII is illustrated in
The A1, A2 and A3 domains each consists of two β-barrel structures that resemble the fold of a typical cupredoxin-type domain. All three A domains share high structural homology with each other and the A domains of ceruloplasmin. The A domains form a triangular heterotrimer where A1 and A3 domains serve as the base and interact with the C2 and C1 domains respectively. The C1 and C2 domains are defined by a distorted β-barrel and are structurally homologous with each other. At the base of the Factor VIII structure, both C domains reveal membrane binding features.
The domain organization in Factor VIII is homologous to the activated protein C-inactivated Factor Va (Factor Vai) structure (Adams et al., Proc. Natl. Acad. Sci. USA, 101:8918, 2004) except that it also includes the A2 domain.
The absence of B-domain and the disorder of the loops containing the cleavage sites necessary for Factor VIII activation suggest that the our current structural model resembles the covalent structure of Factor VIIIa. Thus, the structural model of the present invention can be used to construct a three-dimensional structural model of a Factor VIIIa-Factor IXa complex, including a complex containing at least a region of Factor VIIIa and at least a region of Factor IXa. In general, the method includes the steps of: (a) providing a three-dimensional structural representation of human Factor VIII, or a region thereof; (b) providing a three-dimensional structural representation of Factor IXa, or a region thereof; and (c) fitting the three-dimensional structural representation from step (a) to the three-dimensional structural representation from step (b).
The assembly of the Factor IXa-Factor VIIIa complex normally involves the binding of Factor VIIIa and Factor IXa on phospholipid membrane surfaces in the presence of calcium ions. Based upon homology modeling (Autin, L., Miteva, M. A., Lee, W. H., Mertens, K., Radtke, K. P. & Villoutreix, B. O. (2005) J Thromb Haemost 3, 2044-56), the analysis of naturally occurring hemophilia A and B mutations or mutations introduced by site-specific mutagenesis (Mannucci, P. M. & Tuddenham, E. G. (2001) N Engl J Med 344, 1773-9; Jenkins, P. V., Dill, J. L., Zhou, Q. & Fay, P. J. (2004) Biochemistry 43, 5094-101; Nishimura, H., Takeya, H., Miyata, T., Suchiro, K., Okamura, T., Niho, Y. & Iwanaga, S. (1993) J Biol Chem 268, 24041-6; Hughes, P. E., Morgan, G., Rooney, E. K., Brownlee, G. G. & Handford, P. (1993) J Biol Chem 268, 17727-33), cross-linking studies (Blostein, M. D., Furie, B. C., Rajotte, I. & Furie, B. (2003) J Biol Chem 278, 31297-302) and inhibition with synthetic peptides (Lenting, P. J., van de Loo, J. W., Donath, M. J., van Mourik, J. A. & Mertens, K. (1996) J Biol Chem 271, 1935-40), the binding surface on Factor VIIIa is thought to involve the A2 and A3 domains interacting with multiple domains on Factor IXa. Thus, a model of the Factor IXa-Factor VIIIa complex using our Factor VIII structure and the x-ray crystal structure of porcine Factor IXa backbone may be constructed using one or more of the following constraints: (1) residues 558-565 of Factor VIII interact with the 330-339 helix of Factor IXa; (2) 707-712 of Factor VIIIa binds to Factor IXa residues 301-303; (3) residues 1811-1819 of Factor VIII interact with the light chain of Factor IXa; (4) Phe 25 in the Gla domain of Factor IX is juxtaposed with the light chain of Factor VIII; and (5) the Gla domain of Factor IXa is situated within the phospholipid membrane, forming non-covalent interactions between the phosphoserine head group and fatty acid chains of the phospholipid bilayer and the hydrophobic patch and the Gla residues within the Gla domain of Factor IXa.
Structural information of B-domain deleted Factor VIII, or regions thereof (e.g., domain A1, A2, A3, C1, C2), and structural models of the present invention (including various structural models designed by computer-based methods described below) can be represented in many different ways. Particularly useful are electronic representations, as such representations allow rapid and convenient data manipulations and structural modifications. Electronic representations can be embedded in many different storage or memory media, frequently computer readable media. Examples include without limitations, computer random access memory (RAM), floppy disk, magnetic hard drive, magnetic tape (analog or digital), compact disk (CD), optical disk, CD-ROM, memory card, digital video disk (DVD), and others. The storage medium can be separate or part of a computer system. Such a computer system may be a dedicated, special purpose, or embedded system, such as a computer system that forms part of an X-ray crystallography system, or may be a general purpose computer (which may have data connection with other equipment such as a sensor device in an X-ray crystallographic system. In many cases, the information provided by such electronic representations can also be represented physically or visually in two or three dimensions, e.g., on paper, as a visual display (e.g., on a computer monitor as a two dimensional or pseudo-three dimensional image) or as a three dimensional physical model. Such physical representations can also be used, alone or in connection with electronic representations. Exemplary useful representations include, but are not limited to, the following:
Atomic Coordinate Representation
One type of representation is a list or table of atomic coordinates representing positions of particular atoms in a molecular structure, portions of a structure, or complex (e.g., a co-crystal). Such a representation may also include additional information, for example, information about occupancy of particular coordinates.
Energy Surface or Surface of Interaction Representation
Another representation is an energy surface representation, e.g., of an active site or other binding site, representing an energy surface for electronic and steric interactions. Such a representation may also include other features. An example is the inclusion of representation of a particular amino acid residue(s) or group(s) on a particular amino acid residue(s), e.g., a residue or group that can participate in H-bonding or ionic interaction.
Structural Representation
Still another representation is a structural representation, i.e., a physical representation or an electronic representation of such a physical representation. Such a structural representation includes representations of relative positions of particular features of a molecule or complex, often with linkage between structural features. For example, a structure can be represented in which all atoms are linked; atoms other than hydrogen are linked; backbone atoms, with or without representation of sidechain atoms that could participate in significant electronic interaction, are linked; among others. However, not all features need to be linked. For example, for structural representations of portions of a molecule or complex, structural features significant for that feature may be represented (e.g., atoms of amino acid residues that can have significant binding interation with a ligand at a binding site. Those amino acid residues may not be linked with each other.
A structural representation can also be a schematic representation. For example, a schematic representation can represent secondary and/or tertiary structure in a schematic manner. Within such a schematic representation of a polypeptide, a particular amino acid residue(s) or group(s) on a residue(s) can be included, e.g., conserved residues in a binding site, and/or residue(s) or group(s) that may interact with binding compounds.
The x-ray crystallographic structure of a biologically active human Factor VIII provides important structural information of Factor VIII, such as, for example, the domain organization, the metal binding sites and the surface features in this protein. In particular, the invention provides structural coordinates of atoms corresponding to various binding regions of Factor VIII, such as, for example, the low-density lipoprotein receptor-related protein (LRP) binding site (e.g., amino acids 484-509), heparin sulfate proteoglycans (HSPGs) binding site (e.g., amino acids 558-565), Factor IXa binding regions (e.g., amino acids 558-565, 707-712 or 1811-1819), and LRP/vWF/phospholipids (PL) binding region (e.g., amino acids 2303-2332).
Given the critical importance of Factor VIII to normal hemostasis in the blood coagulation cascade, the crystals of the invention, and particularly the atomic structure coordinates obtained therefrom, have a wide variety of uses including the design of improved therapies for hemophilia A. For example, the crystals described herein provides a useful tool for exploring the rich database of missense mutations that characterize many forms of hemophilia A and link them to functional abnormalities in vitro and in vivo. For example, the present invention allows detailed analysis of structure-function role of specific amino acids in Factor VIIIa in binding to the enzyme Factor IXa, activation of Factor VIII to Factor VIIIa by thrombin or factor Xa, binding of the substrate, factor X, and interaction of the complex with membrane surfaces. Understanding the structure-function relationship of the Factor IXa-Factor VIIIa complex and its interaction with membrane surfaces is critically important to detailed understanding of normal hemostasis within the context of the blood coagulation cascade.
The structures described herein can be used as a starting point in methods for modifying Factor VIII to improve its interaction with, for example, Factor IXa, vWF, or phospholipids membranes, resulting in modified Factor VIII with improved plasma half-life, improved functional activity (e.g., increased activation, or resistance to inactivation), reduced antigenicity or immunogenicity.
The structure coordinates described herein can also be used as phasing models for determining the crystal structures of Factor VIII homologs or analogs of unknown structure, as well as the structures of co-crystals of Factor VIII with ligands such as inhibitors, agonists, antagonists, and other molecules.
The structure described in herein can be used to design compounds that mimic Factor VIII cofactor activity. For example, mimetic antibodies can be designed based on the structure of Factor VIII described herein to substitute for Factor VIII cofactor activity. Such antibodies can be used as long-acting therapeutics for hemophilia.
In addition, the crystals and structure coordinates provided by the present invention are particularly useful for identifying compounds or molecules that modulate Factor VIII activity as an approach towards developing new therapeutic agents. In particular, the crystals and structural information are particularly useful in methods based on rational drug design.
A human Factor VIII structure suitable for various methods of these aspects of the invention includes structures defined by structural coordinates of atoms of human Factor VIII according to Table 2, +/−a root mean square deviation for alpha carbon atoms of less than 2 Angstroms, or selected coordinates thereof, as well as the models of Factor VIII homologs or analogs obtained by the methods of the invention.
In preferred embodiments, the methods of these aspects of the invention described above are computer-based methods. Various exemplary computational techniques suitable for the methods of these aspects of the invention are described below.
Homology Modeling
Homology modeling is a method of applying structural coordinates of a polypeptide of known structure to the amino acid sequence of a polypeptide of unknown structure. This method is accomplished using a computer representation of the three dimensional structure of a polypeptide or polypeptide complex, the computer representation of amino acid sequences of the polypeptides with known and unknown structures, and standard computer representations of the structures of amino acids. Homology modeling generally involves (a) aligning the amino acid sequences of the polypeptides with and without known structure to match one or more homologous regions or amino acids; (b) modeling the structure of the matched one or more homologous regions or amino acids of the polypeptide of unknown structure on the corresponding regions of the known structure; and (d) determining a structural confirmation for the polypeptide of unknown structure which substantially preserves the structure of the matched homologous regions. Methods for matching homologous regions or amino acids are well known in the art. For example, the programs BLAST, gapped BLAST, BLASTN, PSI-BLAST and BLAST2 (provided by the National Center for Biotechnology Information) are widely used in the art for this purpose, and can align homologous regions of two amino acid sequences. These may be used with default parameters.
The above method is well known to those skilled in the art. (Greer (1985) Science 228:1055; Blundell et al. A(1988) Eur. J. Biochem., 172:513. An exemplary computer program that can be utilized for homology modeling by those skilled in the art is the Homology module in the Insight II modeling package distributed by Accelerys Inc.
Alignment of the amino acid sequence may be accomplished by first placing the computer representation of the amino acid sequence of a polypeptide with known structure above the amino acid sequence of the polypeptide of unknown structure. Amino acids in the sequences are then compared and groups of amino acids that are homologous (e.g., amino acid side chains that are similar in chemical nature—aliphatic, aromatic, polar, or charged) are grouped together. This method will detect conserved regions of the polypeptides and account for amino acid insertions or deletions.
Once the amino acid sequences of the polypeptides with known and unknown structures are aligned, the structures of the conserved amino acids in the computer representation of the polypeptide with known structure are transferred to the corresponding amino acids of the polypeptide whose structure is unknown. For example, a tyrosine in the amino acid sequence of known structure may be replaced by a phenylalanine, the corresponding homologous amino acid in the amino acid sequence of unknown structure.
The structures of amino acids located in non-conserved regions are to be assigned manually by either using standard peptide geometries or molecular simulation techniques, such as molecular dynamics. The final step in the process is accomplished by refining the entire structure using molecular dynamics and/or energy minimization. The homology modeling method is well known to those skilled in the art and has been practiced using different protein molecules. For example, the three dimensional structure of the polypeptide corresponding to the catalytic domain of a serine/threonine protein kinase, myosin light chain protein kinase, was homology modeled from the cAMP-dependent protein kinase catalytic subunit. (Knighton et al. (1992) Science 258:130-135.)
In addition, specific computer software is available in the art to evaluate compound deformation energy and electrostatic interactions. Examples of programs designed for such uses include: Gaussian 94, revision C (M. J. Frisch, Gaussian, Inc., Pittsburgh, Pa. C) 1995); AMBER, version 4.1 (P. A. Kollman, University of California at San Francisco, C) 1995) QUANTA/CHARMM (Molecular Simulations, Inc., San Diego, Calif. © 1995); Insight II/Discover (Molecular Simulations, Inc., San Diego, Calif. © 1995); DelPhi (Molecular Simulations, Inc., San Diego, Calif. © 1995); and AMSOL (Qunatum Chemistry Program Exchange, Indiana University). These programs may be implemented, for instance, using a Silicon Graphics workstation such as an Indigo2 with “IMPACT” graphics. Other modem hardware systems and software packages will be known to those skilled in the art.
Molecular Replacement
Molecular replacement is a method of applying the X-ray diffraction data of a polypeptide of known structure to the X-ray diffraction data of a polypeptide of unknown sequence. This method can be utilized to define the phases describing the X-ray diffraction data of a polypeptide of unknown structure when only the amplitudes are known. X-PLOR is a commonly utilized computer software package used for molecular replacement. Brunger (1992) Nature 355:472-475. AMORE is another program used for molecular replacement. Navaza (1994) Acta Crystallogr. A50:157-163. Preferably, the resulting structure does not exhibit a root-mean-square deviation of more than 3 Å. Specific steps of molecular replacement are described below.
A goal of molecular replacement is to position the atomic coordinates of a structure model into the unit cell of Factor VIII crystal. A program such as X-PLOR can involve four steps. A first step can be to determine the number of molecules in the unit cell. A second step can involve rotating the structure model to define the orientation of the molecules in the unit cell. A third step can be to translate the structure model in three dimensions to correctly position the molecules in the unit cell. Once the amplitudes and phases of the X-ray diffraction data is determined, an R-factor can be calculated by comparing X-ray diffraction data calculated experimentally from the reference data set and calculated from the new data set. An R-factor between 30-50% indicates that the orientations of the atoms in the unit cell are reasonably determined by this method. A fourth step in the process can be to decrease the R-factor to roughly 25% or lower by refining the positioned structure model using iterative refinement techniques described herein and known to those or ordinary skill in the art.
Designing Mimetic Compounds of Factor VIII
In general, a mimetic compound of human Factor VIII may be designed by: (a) providing a selected human Factor VIII structure that is associated with a biological activity of human Factor VIII; (b) superimposing a three-dimensional structure of a compound on the selected human Factor VIII structure; and (c) modifying the three-dimensional structure of the compound such that the modified three-dimensional structure comprises a structural confirmation substantially mimicking the selected human Factor VIII structure.
A less biased approach involves computer algorithms for searching databases of three dimensional structures suitable compounds. By this method, one can generate compounds for which the bioactive conformation is heavily populated, i.e., compounds which are based on particularly biologically relevant conformations of the target protein. Algorithms for this purpose are implemented in programs such as Cast-3D (Chemical Abstracts Service), 3DB Unity (Tripos, Inc.), Quest-3D (Cambridge Crystallographic Data Center), and MACCS/ISIS-3D (Molecular Design Limited). These geometric searches can be augmented by steric searching, in which the size and shape requirements of the binding site are used to weed out hits that have prohibitive dimensions. Programs that may be used to synchronize the geometric and steric requirements in a search applied to the FRB of FRAP include CAVEAT (P. Bartlett, University of California, Berkeley), HOOK (MSI), ALADDIN (Daylight Software) and DOCK (I. D. Kuntz, University of California, San Francisco; see e.g. http://www.cmpharm.ucsf.edu/kuntz-/kuntz.html and references cited therein). All of these searching protocols may be used in conjunction with existing corporate databases, the Cambridge Structural Database, or available chemical databases from chemical suppliers.
In addition, mimetic compounds of Factor VIII may be developed from the bound conformation of Factor VIII by design, by searching databases for replacements of one or more structural segments of Factor VIII, or by enhancement of existing ligand-protein interactions (i.e., by replacing a component moiety of a ligand with a substitute moiety capable of greater interaction with the target protein, whether through accessible protein contact points or by extrusion of otherwise sequestered waters). Knowledge of the bound conformation of a ligand can suggest avenues for conformational restriction and replacement of atoms and/or bonds of Factor VIII.
Computer programs such as those described in the homology modeling section above can be used to superimpose a three-dimensional structure of a compound on the selected human Factor VIII structure; and to modify the three-dimensional structure of the compound such that the modified three-dimensional structure includes a structural confirmation substantially mimicking the selected human Factor VIII structure.
Modified compounds may be synthesized and the Factor VIII mimetic activity is tested by in vitro or in vivo methods known in the art.
In particular, the present invention contemplates mimetic antibodies of human Factor VIII designed by the methods described herein. For example, mimetic antibodies can be designed to mimic a binding activity of Factor VIII to Factor IXa and/or Factor X. Such mimetic antibodies may substitute for Factor VIII cofactor activity and could potentially be used as therapeutic agents for hemophilia.
Structure-Based Rational Drug Design
A particular aspect of the invention relates to computer-based rational drug design methods to identify candidate modulators of Factor VIII function that interact with human Factor VIII structures of the present invention.
Determination of the three-dimensional structure of B-domain deleted human Factor VIII provides important information about the binding sites of Factor VIII, particularly when comparisons are made with similar proteins, such as Factor Va, ceroluplasmin. This information may then be used for rational design, e.g., by computational techniques which identify possible binding ligands for the binding sites, by enabling linked-fragment approaches to drug design, and by enabling the identification and location of bound ligands using X-ray crystallographic analysis. The suitable techniques are discussed in detail, for example, by Walters et al. Drug Discovery Today, Vol. 3, No. 4, (1998), 160-178; and Abagyan, R.; Totrov, M. Curr. Opin. Chem. Biol. 2001, 5, 375-382. In particular, automated ligand-receptor docking programs is discussed, for example, by Jones et al. in Current Opinion in Biotechnology, Vol. 6, (1995), 652-656 and Halperin et al. Proteins 2002, 47, 409-443).
The aspects of the invention described herein which utilize the human Factor VIII structure in silico may be equally applied to both the human Factor VIII structure defined by coordinates of Table 2, or selected coordinates thereof, and the models of Factor VIII homologs or analogs obtained by other aspects of the invention. For rational drug design, the coordinates of atoms corresponding to one or more binding regions of Factor VIII are particularly useful. Such binding regions of Factor VIII include, but are not limited to, the LRP and Factor IXa binding region (e.g., amino acids 484-509), the HSPGs and Factor IXa binding region (e.g., amino acids 558-565), the Factor IXa binding regions (e.g., amino acids 707-712, and amino acids 1811-1819), and the LRP, vWF, PL binding region (e.g., amino acids 2303-2332).
Accordingly, the invention provides a computer-based method for the analysis of the interaction of a molecular structure with a human Factor VIII structure of the invention, which generally includes the steps of: providing selected coordinates of a human Factor VIII structure; providing a plurality of moieties to be fitted to said human Factor VIII structure; fitting the structure of each of the plurality of moieties to the human Factor VIII structure; selecting one or more moieties that fit into the selected structure; and, optionally, assembling the fitted one or more moieties into a single molecule to form a candidate modulator molecule.
The moieties suitable for the method of this aspect of the invention can be selected from the group consisting of molecular fragments, small molecules, ligands designed de novo, and compounds known to bind Factor VIII or modified compounds thereof.
In some embodiments, such moieties may be selected from publicly available databases include, for example: a) ACD from Molecular Designs Limited; b) NCI from National Cancer Institute; c) CCDC from Cambridge Crystallographic Data Center; d) CAST from Chemical Abstract Service; e) Derwent from Derwent Information Limited; f) Maybridge from Maybridge Chemical Company LTD; g) Aldrich from Aldrich Chemical Company; h) Directory of Natural Products from Chapman & Hall.
The availability of the structure of B-domain deleted human Factor VIII will allow the generation of highly predictive pharmacophore models for virtual library screening or compound design. The modeling software can be used to determine Factor VIII binding surfaces and to reveal features such as van der Waals contacts, electrostatic interactions, and/or hydrogen bonding opportunities. These binding surfaces can be used to model docking of ligands with Factor VIII, to arrive at pharmacophore hypotheses, and to design possible therapeutic compounds de novo. The term “pharmacophore” refers to a collection of chemical features and three-dimensional constraints that represent specific characteristics responsible for a ligand's activity. The pharmacophore includes surface-accessible features, hydrogen bond donors and acceptors, charged/ionizable groups, and/or hydrophobic patches, among other features.
A pharmacophore can be defined for the Factor VIII ternary complex that includes surface-accessible features, hydrogen bond donors and acceptors, charged/ionizable groups, and/or hydrophobic patches, among other features. These features can be weighted depending on their relative importance in conferring activity (see, e.g., Computer-Assisted Lead Finding and Optimization, Testra & Folkers, 1997).
Pharmacophores can be determined using software such as CATALYST (including HypoGen or HipHop, available from Molecular Stimulations Inc.), CERIUS2, or constructed by hand from a known conformation of a lead compound. The pharmacophore can be used to screen structural libraries, using a program such as CATALYST. The CLIX program (Davic & Lawrence, Proteins 12:31-41, 1992) can also be used, which searches for orientations of candidate molecules in structural databases that yield maximum spatial coincidence with chemical groups which interact with the receptor. The DISCO program (available from Tripos) uses a method of clique detection to identify common pharmacophoric features in each structure, produce optimally aligned structures, and extract the key features of the pharmacophore. The GASP program (available from Tripos) uses a genetic algorithm to automatically find pharmacophores with conformational flexibility.
The binding surface or pharmacophore of the Factor VIII ternary complex can be used to map favorable interaction positions for functional groups (e.g., protons, hydroxyl groups, amine groups, acidic groups, hydrophobic groups and/or divalent cations) or small molecule fragments. Compounds can then be designed de novo in which the relevant functional groups are located in the correct spatial relationship to interact with Factor VIII.
There are many de novo ligand design methods including:
1. LUDI (H.-J. Bohm, “The Computer Program LUDI: A New Method for the De Novo Design of Enzyme Inhibitors,” J. Comp. Aid. Molec. Design, 6, pp. 61-78 (1992)). LUDI is available from Molecular Simulations Incorporated, San Diego, Calif.
2. LEGEND (Y. Nishibata et al., Tetrahedron, 47, p. 8985 (1991)). LEGEND is available from Molecular Simulations Incorporated, San Diego, Calif.
3. LeapFrog (available from Tripos Associates, St. Louis, Mo.).
4. SPROUT (V. Gillet et al., “SPROUT: A Program for Structure Generation,” J. Comput. Aided Mol. Design, 7, pp. 127-153 (1993)). SPROUT is available from the University of Leeds, UK.
In order to provide a three-dimensional structure of moieties to be fitted to a human Factor VIII structure of the invention, the moiety structure may be modeled in three dimensions using commercially available software for this purpose or, if its crystal structure is available, the coordinates of the structure may be used to provide a representation of the compound for fitting to a human Factor VIII structure of the invention.
By “fitting”, it is meant determining by automatic, or semi-automatic means, interactions between at least one atom of a molecular structure and at least one atom of a human Factor VIII structure of the invention, and calculating the extent to which such an interaction is stable. Interactions include attraction and repulsion, brought about by charge, steric considerations and the like. Various computer-based methods for fitting are available in the art, for example, docking program such as GOLD (Jones et al., J. Mol. Biol., 245, 43-53 (1995), Jones et al., J. Mol. Biol., 267, 727-748 (1997)), GRAMM (Vakser, I. A., Proteins, Suppl., 1:226-230 (1997)), DOCK (Kuntz et al., J. Mol. Biol. 1982, 161, 269-288, Makino et al., J. Comput. Chem. 1997, 18, 1812-1825), AUTODOCK (Goodsell et al., Proteins 1990, 8, 195-202, Morris et al., J. Comput. Chem. 1998, 19, 1639-1662.), FlexX, (Rarey et al., J. Mol. Biol. 1996, 261, 470-489) or ICM (Abagyan et al., J. Comput. Chem. 1994, 15, 488-506). This procedure can include computer fitting of a moiety to a human Factor VIII Structure to ascertain how well the shape and the chemical structure of the moiety will bind to human Factor VIII.
Also computer-assisted, manual examination of the structure of human Factor VIIIa may be performed. The use of programs such as GRID (Goodford, J. Med. Chem., 28, (1985), 849-857)—a program that determines probable interaction sites between molecules with various functional groups and an enzyme surface—may also be used to analyse the active site to predict, for example, the types of modifications which will alter the rate of catabolism of a substrate.
Computer programs can be employed to estimate the attraction, repulsion, and steric hindrance of the two binding partners.
Following the fitting of the molecular structures, a person of skill in the art may seek to use molecular modeling to determine to what extent the structures interact with each other (e.g., by hydrogen bonding, other non-covalent interactions, or by reaction to provide a covalent bond between parts of the structures) or the interaction of one structure with human Factor VIII is altered by the presence of another structure.
Once suitable moieties (such as, for example, chemical entities or fragments) have been selected, they can be designed or assembled into a single compound or complex. Assembly may be preceded by visual inspection of the relationship of the fragments to each other on the three-dimensional image displayed on a computer screen in relation to the structure coordinates of human serum albumin. This would be followed by manual model building using software such as Quanta or Sybyl (Tripos Associates, St. Louis, Mo.).
If more than one human Factor VIII region is characterized and a plurality of respective smaller moieties are designed or selected, a candidate modulator may be formed by linking the respective small moieties into a larger molecule, which maintains the relative positions and orientations of the respective smaller moieties at the respective binding regions. The candidate modulator may be formed as a real molecule or by computer modeling.
Useful programs to aid one of skill in the art in connecting the individual chemical entities or fragments include:
1. CAVEAT (P. A. Bartlett et al., “CAVEAT: A Program to Facilitate the Structure-Derived Design of Biologically Active Molecules,” in Molecular Recognition in Chemical and Biological Problems, Special Pub., Royal Chem. Soc., 78, pp. 182-196 (1989); G. Lauri and P. A. Bartlett, “CAVEAT: a Program to Facilitate the Design of Organic Molecules,” J. Comput. Aided Mol. Des., 8, pp. 51-66 (1994)). CAVEAT is available from the University of California, Berkeley, Calif.
2. 3D Database systems such as ISIS (MDL Information Systems, San Leandro, Calif.). This area is reviewed in Y. C. Martin. “3D Database Searching in Drug Design,” J. Med. Chem., 35, pp. 2145-2154 (1992).
3. HOOK (M. B. Eisen et al., “HOOK: A Program for Finding Novel Molecular Architectures that Satisfy the Chemical and Steric Requirements of a Macromolecule Binding Site,” Proteins: Struct., Funct. Genet., 19, pp. 199-221 (1994). HOOK is available from Molecular Simulations, San Diego, Calif.
Detailed structural information can then be obtained about the binding of the candidate modulator to human Factor VIII, and in the light of this information adjustments can be made to the structure or functionality of the candidate modulator, e.g., to alter its interaction with human Factor VIII. The above steps may be repeated and re-repeated as necessary.
A newly designed candidate modulator molecule may be obtained or synthesized and its interaction with human Factor VIII may be determined by binding assays. Various binding assays are well known in the art.
Such modulators may affect the interactions between Factor VIII and its binding partners, such as, LRP, Factor IXa, HSPGs, vWF, and PL, resulting in changed pharmacokinetics and functional activity for Factor VIII. Modulators of human Factor VIII identified by rational drug design may be developed as potential therapeutic agents for hemophilia.
Various molecular analysis and rational drug design techniques are further disclosed in, for example, U.S. Pat. Nos. 5,834,228, 5,939,528 and 5,856,116, as well as in PCT Application No. PCT/US98/16879, published as WO 99/09148, the teachings of all of which are hereby incorporated by reference.
The invention is illustrated by the following non-limiting examples.
The mature B-domain deleted human Factor VIII used for crystallization contains residues 1-740 (SEQ ID NO:1) that comprise the heavy chain (A1 and A2 domains), a short peptide linker (residues 741-754) (SEQ ID NO:3) and residues 1649-2332 (SEQ ID NO:2) that comprise the light chain (A3, C1 and C2 domains) (
Crystals were obtained by hanging drop vapor diffusion at 25° C. using the Hampton Screen (Hampton Research). The drop contained 1 μl of Factor VIII at 10 mg/ml mixed with 1 μl of reservoir solution. The optimal condition for crystallization was found to be 100 mM Tris-HCl (pH 8.5), 10% ethanol, and 7% PEG 3350 in the reservoir. All crystals were cryo-protected by sequential addition of 10%, 15% and finally 20% ethylene glycol (v/v) in the presence of the reservoir solution and flash frozen in liquid nitrogen prior to data collection. Factor VIII crystallized in a P41212 space group (a=b=134.11 Å, c=349.760 Å and α=β=γ=90°) with one molecule per asymmetric unit (Table 1). The crystal contains an unusually high solvent content of 75%.
X-ray diffraction data of Factor VIII crystals were collected at the NE-CAT synchrotron beamline ID-24 of the Advanced Photon Source (APS) at Argonne National Laboratory. All xray data were processed using program HKL2000 (Otwinowski, Z. & Minor, W. (1997) in Methods Enzymol (Academic Press, New York), Vol. 276, pp. 307-326) (Table 1).
aRsym = Σ|I − <I>|/ΣI
bRcrys = Σ∥Fobs| − |Fcalc∥/Σ|Fobs|, where Fobs and Fcalc are the observe and calculated structure factors, respectively.
cRfree was calculated with 5% of the data excluded from the refinement calculation.
For structural determination using the molecular replacement method, a homology structure model of Factor VIII was constructed from the known primary sequence of Factor VIII (McCoy, A. J. (2007) Acta Crystallogr D Biol Crystallogr 63, 32-41). Using the template structures of Factor Vai (PDBID: 15DD) (Sandberg, H., Almstedt, A., Brandt, J., Castro, V. M., Gray, E., Holmquist, L., Lewin, M., Oswaldsson, U., Mikaelsson, M., Jankowski, M. A., Bond, M. & Scoble, H. A. (2001) Semin Hematol 38, 4-12) and ceruloplasmin (PDBID: 1KCW) (Adams, T. E., Hockin, M. F., Mann, K. G. & Everse, S. J. (2004) Proc Natl Acad Sci U SA 101, 8918-23), the structure of the A1 and A2 domains of Factor VIII were initially determined with AMoRe (Murshudov, G. N., Vagin, A. A. & Dodson, E. J. (1997) Acta Crystallogr D Biol Crystallogr 53, 240-55) using the structure of the A1 and A2 domains of the homology model. This yielded clear rotation function and translation function solutions. The A3 domain was then solved by AMoRe after fixing the solution of the A1 and A2 domains and using the A3 domain of ceroluplasmin as a search model (PDBID: 1KCW) (Adams, T. E., Hockin, M. F., Mann, K. G. & Everse, S. J. (2004) Proc Natl Acad Sci USA 101, 8918-23). After fixing all three A domains together, the positions of the C1 and C2 domains were determined with the program PHASER using a polyalanine model built from the high resolution structure of the C2 domain of Factor VIII (PDBID: ID7P) (Pratt, K. P., Shen, B. W., Takeshima, K., Davie, E. W., Fujikawa, K. & Stoddard, B. L. (1999) Nature 402, 439-442; Potterton, E., Briggs, P., Turkenburg, M. & Dodson, E. (2003) Acta Crystallogr D Biol Crystallogr 59, 1131-7). The structure was refined with several cycles of manual refitting and refinements using REFMAC of the CCP41 suite (Bihoreau, N., Pin, S., de Kersabiec, A. M., Vidot, F. & Fontaine-Aupart, M. P. (1994) Eur J Biochem 222, 41-8; Messerschmidt, A. & Huber, R. (1990) Eur J Biochem 187, 341-52). The Rcryst and Rfree for the Factor VIII model were 25.70% and 33.06% respectively for data from 50-3.98 Å (Table 1).
Structural analysis was conducted by methods described above and those known in the art. Our model indicates that Factor VIII is a heterodimer consisting of the heavy chain (A1-A2 domains) and light chain (A3-C1-C2 domains). Several regions within the structure are poorly ordered and were not modeled, including residues 17-43, 334-376, and 714-754 within the heavy chain, and residues 1649-1690 and 1714-1724 of the light chain.
The overall structure of Factor VIII can be described as a triangular heterotrimer of the A domains stacked on two smaller globular C domains (
Factor VIII is a copper binding protein (Bihoreau, N., Pin, S., de Kersabiec, A. M., Vidot, F. & Fontaine-Aupart, M. P. (1994) Eur J Biochem 222, 41-8), and we identified two copper ions and their binding sites internally within the A1 domain and the A3 domain. These are prototypic copper binding sites, with nitrogen and sulhydryl ligands (Messerschmidt, A. & Huber, R. (1990) Eur J Biochem 187, 341-52). In the A3 domain, the copper ion is liganded by His 1954, Cys 2000, His 2005 whereas the copper binding site in the A1 domain is defined by His 267, His 315 and Cys 310 (
A single calcium binding site was located in the A1 domain. This site is defined by carboxyl groups of Glu 110, Asp 116, Asp 126, Asp 125 and the carbonyl 0 of Lys 107 and Glu 122 (
Oligosaccharides was observed linked to Asn 239 in the A1 domain, Asn 1810 in the A3 domain and Asn 2118 in the C1 domain.
The three A domains form a triangular heterotrimer around a pseudo-three-fold symmetry where A1 and A3 domains serve as the base and interact with the C2 and C1 domains respectively (
Although the C2 domain is connected to the C1 domain and located adjacent to the A1 domain, there are few direct contacts of the C2 with either domain (
Factor VIII and Factor V are procofactors that show approximately 40% sequence similarity and a parallel domain arrangement in their primary structure (Kane, W. H. & Davie, E. W. (1986) Proc Natl Acad Sci USA 83, 6800-4). Both cofactors are activated to their active cofactor forms, Factor VIIIa and Factor Va, by thrombin-mediated limited proteolysis (Pittman, D. D. & Kaufman, R. J. (1988) Proc Natl Acad Sci USA 85, 2429-33; Nesheim, M. E. & Mann, K. G. (1979) J Biol Chem 254, 1326-34). Both cofactors are inactivated by activated protein C-mediated limited proteolysis to yield Factor VIIIai and Factor Vai. The three dimensional structure of Factor V has not been solved, thus precluding direct comparison with our current structure of human B-domain deleted Factor VIII. However, a high resolution x-ray crystal structure of activated protein C-inactivated bovine Factor Va, Factor Vai, allows partial comparison since activation and subsequent inactivation of Factor V is associated with the removal of the B-domain and the A2 domain (Adams, T. E., Hockin, M. F., Mann, K. G. & Everse, S. J. (2004) Proc Natl Acad Sci USA 101, 8918-23). The spatial arrangement of the A1 and A3 domains in B-domain deleted Factor VIII are nearly identical to that of Factor Vai, while the C1 and C2 domains show slightly different conformations when comparing our Factor VIII structure and that of Factor Vai. Despite the differences in the C1 and C2 domains, structures of Factor VIII and Factor Vai can be superimposed with a rmsd of 1.54 Å for the carbons of the 524 residues that span the A1, A2, C1 and C2 domains (
The activation of Factor VIII by thrombin requires cleavage of a peptide bond in the A2 domain, after Arg 372, and the removal of the B-domain linked to the 41 residue N-terminal region of the A3 domain, typically referred to as the a3 acidic region comprising residues 1649-1689, by thrombin cleavage after Arg 1689. This exposes the Factor VIII surfaces that are important for Factor IXa binding. Activation is also associated with but does not require cleavage after arginine 740 (Pittman, D. D. & Kaufman, R. J. (1988) Proc Natl Acad Sci USA 85, 2429-33). In B-domain deleted Factor VIII, conversion of Factor VIII to Factor VIIIa requires cleavage at arginine 1689 to remove the acidic a3 region adjacent to the A3 domain as well as cleavage after Arg 372. These regions are located on the front surface of Factor VIII (
The assembly of the Factor IXa-Factor VIIIa complex involves the binding of Factor VIIIa and Factor IXa on phospholipid membrane surfaces in the presence of calcium ions. Based upon homology modeling (Autin, L., Miteva, M. A., Lee, W. H., Mertens, K., Radtke, K. P. & Villoutreix, B. O. (2005) J Thromb Haemost 3, 2044-56), the analysis of naturally occurring hemophilia A and B mutations or mutations introduced by site-specific mutagenesis (Mannucci, P. M. & Tuddenham, E. G. (2001) N Engl J Med 344, 1773-9; Jenkins, P. V., Dill, J. L., Zhou, Q. & Fay, P. J. (2004) Biochemistry 43, 5094-101; Nishimura, H., Takeya, H., Miyata, T., Suchiro, K., Okamura, T., Niho, Y. & Iwanaga, S. (1993) J Biol Chem 268, 24041-6; Hughes, P. E., Morgan, G., Rooney, E. K., Brownlee, G. G. & Handford, P. (1993) J Biol Chem 268, 17727-33), cross-linking studies (Blostein, M. D., Furie, B. C., Rajotte, I. & Furie, B. (2003) J Biol Chem 278, 31297-302) and inhibition with synthetic peptides (Lenting, P. J., van de Loo, J. W., Donath, M. J., van Mourik, J. A. & Mertens, K. (1996) J Biol Chem 271, 1935-40), the Factor VIIIa binding surface for Factor IXa is thought to involve the A2 and A3 domains that interact with multiple domains on Factor IXa. The A3 domain of the light chain contains a high affinity binding site (Kd˜2-15 nM) for Factor IXa (Lenting, P. J., Donath, M. J., van Mourik, J. A. & Mertens, K. (1994) J Biol Chem 269, 7150-5). Inhibition studies with synthetic peptides directed against the A3 domain have located this site to residues 1811-1818 (Lenting, P. J., van de Loo, J. W., Donath, M. J., van Mourik, J. A. & Mertens, K. (1996) J Biol Chem 271, 1935-40). Potential Factor IXa binding sites on the A2 domain include residues 558-565 (Fay, P. J., Beattie, T., Huggins, C. F. & Regan, L. M. (1994) J Biol Chem 269, 20522-7) and the region around Asp 712 (Jenkins, P. V., Dill, J. L., Zhou, Q. & Fay, P. J. (2004) Biochemistry 43, 5094-101). We have mapped these putative binding regions onto our structure of Factor VIII (
The absence of B-domain and the disorder of the loops containing the cleavage sites necessary for Factor VIII activation suggest that the our current structural model resembles the covalent structure of Factor VIIIa. Human Factor IXa was constructed by homology modeling using the x-ray structure of porcine Factor IXa (PDB ID: 1PFX.pdb) and the program SWISS-MODEL (McCoy, A. J. (2007) Acta Crystallogr D Biol Crystallogr 63, 32-41; Wakabayashi, H., Koszelak, M. E., Mastri, M. & Fay, P. J. (2001) Biochemistry 40, 10293-300). The structure of Factor VIIIa was docked with Factor IXa using the program HADDOCK (Saenko, E. L., Scandella, D., Yakhyaev, A. V. & Greco, N. J. (1998) J Biol Chem 273, 27918-26). In this approach, residues previously reported to be important for Factor VIIIa-Factor IXa interactions were defined as the ambiguous interaction constraints. Initially, 600 structures for the complex were generated by docking Factor VIII and Factor IXa as rigid bodies during iterations of energy minimization and 150 of those structures were refined and analyzed by using the default settings.
In particular, we have constructed a model of the Factor IXa-Factor VIIIa complex using our Factor VIII structure and the x-ray crystal structure of porcine Factor IXa backbone using as constraints: (1) residues 558-565 of Factor VIII interact with the 330-339 helix of Factor IXa; (2) 707-712 of Factor VIIIa binds to Factor IXa residues 301-303; (3) residues 1811-1819 of Factor VIII interact with the light chain of Factor IXa (Lenting, P. J., van de Loo, J. W., Donath, M. J., van Mourik, J. A. & Mertens, K. (1996) J Biol Chem 271, 1935-40); (4) Phe 25 in the Gla domain of Factor IX is juxtaposed with the light chain of Factor VIII (Blostein, M. D., Furie, B. C., Rajotte, I. & Furie, B. (2003) J Biol Chem 278, 31297-302); and (5) the Gla domain of Factor IXa is situated within the phospholipid membrane, forming non-covalent interactions between the phosphoserine head group and fatty acid chains of the phospholipid bilayer and the hydrophobic patch and the Gla residues within the Gla domain of Factor LXa (Freedman, S. J., Blostein, M. D., Baleja, J. D., Jacobs, M., Furie, B. C. & Furie, B. (1996) Journal of Biological Chemistry 271, 16227-16236; Huang, M., Rigby, A. C., Morelli, X., Grant, M. A., Huang, G., Furie, B., Seaton, B. & Furie, B. C. (2003) Nature Structural Biology 10, 751-756) (
The model of the Factor IXa-Factor VIIIa complex illustrates that the light chain of Factor IXa, which includes the phospholipid-binding Gla domain, is wrapped across the side of the A3 domain and oriented almost perpendicularly to the Factor VIII molecule, and is distally located from the membrane-binding interface of the C2 domain, which is important for Factor VIIIa interaction with the membrane (
We have identified the four putative phospholipid binding sites in the Factor IXa-Factor VIIIa complex. Within the context of the domain organization of C2 in the Factor IXa-Factor VIII complex, Arg 2215 and Lys 2249 play a special role in phospholipids binding. The hairpin loops in the Factor VIII C1 domain, including Lys 2092, Arg 2090, Arg 2169, likely play a similar role. In addition, a well-defined loop extends downward from the backside of the Factor VIII A3 domain. This loop, which includes a β-turn, is held together by Cys 1899 and Cys 1903, and thrusts Arg 1900 into a favorable position for electrostatical interaction with acidic phospholipid headgroups. The polypeptide backbone of this loop is structurally identical to and co-planar with the omega loop of the Gla domain of Factor IXa. Using the prothrombin Gla domain as a prototype, we have previously established that the phosphoserine head group in lysophosphatidylserine interacts with conserved residues within the Gla domain of vitamin K-dependent proteins, including Factor IX (Huang, M., Rigby, A. C., Morelli, X., Grant, M. A., Huang, G., Furie, B., Seaton, B. & Furie, B. C. (2003) Nature Structural Biology 10, 751-756). Trp 4 within the omega loop of the Gla domain of prothrombin is located 5 to 7 Å below the membrane surface in the interfacial membrane region (Falls, L. A., Furie, B. C., Jacobs, M., Furie, B. & Rigby, A. C. (2001) J Biol Chem 276, 23895-902). Based on this observation, the Factor IXa in this model was similarly positioned.
After the priority date of the present application, a crystal structure of a recombinant form of Factor VIII which consists of a heterodimer of peptides, respectively containing the A1-A2 and A3-C1-C2 domains, was disclosed on Apr. 15, 2008 (PDB ID: 2R7E). See Shen B. W. et al. “The tertiary structure and domain organization of coagulation Factor VIII,” Blood, 2008 Feb. 1; 111(3):1240-7, Epub 2007 Oct. 26. Overlaid of our B domain-deleted FVIII crystal structure (PDB ID: 3CDZ) and 2R7E shows that both structures are maintained in similar conformations with a few exceptions (
Firstly, the loop formed by residues 558 to 565 in 3CDZ is solvent-exposed and appears to be accessible for the interaction with FIXa. In contrast, the model of 2R7E contains extra residues between amino acids 360-376 and 715-725, which form two large extensions and bury the 558-565 loop. Importantly, the extension formed by residues 360 to 376 contains the activating cleavage site (Arg 372). Cleavage at this position during FVIII activation may help to expose the 558-565 loop for FIXa binding. The difference between the two structures also suggests that this loop is flexible and may adopt different conformations upon binding with FIXa. Furthermore, the two extensions in 2R7E appear to be more ordered than those in 3CDZ because of the presence of B-domain regions in 2R7E. Therefore, the burial of the 558-568 loop may be a feature of the full length FVIII before activation and removal of the B-domain.
Secondly, residues 1712 to 1725 are disordered in 3CDZ but they form an extended loop in 2R7E. This loop is comprised mostly of basic and hydrophobic residues and is oriented in a similar manner as the other A3 domain loop (amino acids 1895-1907). Based on the proposed orientation of our FVIIIa:FIXa complex model, this region of the A3 domain of FVIII is also likely to interact with the acidic phospholipids membrane.
Exemplary electrostatic surface potential of Factor VIII is illustrated in
As also shown in
All publications and patent documents cited in this application are incorporated by reference in their entirety to the same extent as if the contents of each individual publication or patent document were incorporated herein.
This application claims priority to U.S. Provisional Patent Application Ser. No. 61/005,887, filed on Dec. 6, 2007; U.S. Provisional Patent Application Ser. No. 61/009,061, filed on Dec. 20, 2007; and U.S. Provisional Patent Application Ser. No. 61/072,777, filed on Mar. 31, 2008, the entirety of each of which is hereby incorporated herein by reference.
| Number | Date | Country | |
|---|---|---|---|
| 61005887 | Dec 2007 | US | |
| 61009061 | Dec 2007 | US | |
| 61072777 | Mar 2008 | US |
