Patent Application
20050221285
The invention relates to anti-viral agents which may be effective for treating, for example, respiratory infections by Respiratory Syncytial Virus (RSV). More particularly, the invention relates to the generation of a three-dimensional model of the RSV-F protein and the use of that model to identify, screen, and/or develop anti-viral agents.
RSV is the predominant cause of acute lower respiratory tract infection in children, with symptoms ranging from severe pneumonia and bronchiolitis to much milder infections similar to the common cold (reviewed in Cane, P. A., Rev. Med. Virol., 11:103-116, 2001). While primary infection does generate a long term immune response this is not fully protective and recurrent infection can occur throughout life with annual re-infection rates of 3-6% in adults (Monto, A. S. and Kim, S. K., Am. J. Epidemiol., 94:290-301, 1971). Subsequently the virus is also a major cause of morbidity and mortality in institutionalised elderly and immunocompromised adults.
RSV is a member of the Pneumovirus genus of the family Paramyxoviridae, which includes other human pathogens such as mumps, measles and the parainfluenza viruses. The RSV genome is a single strand of negative-sense RNA encoding ten viral proteins, three of which are present on the surface of the infected cells as well as on the virion itself. These three proteins are the G, F and SH proteins; G and F are responsible for binding of the virus to target cells and fusion of the viral membrane with the target cell membrane, respectively. The F protein is apparently necessary and sufficient for viral infection to occur as mutant RSV lacking G and SH protein are still able to infect cells in vitro, albeit at a reduced level (Techaarpornkul, S. et al., J. Virology, 75:6825-6834, 2001). F is also expressed on the surface of infected cells and syncytia formation is a result of fusion of neighbouring cells mediated by the F protein.
RSV-F is expressed as a single precursor of 574 amino acids with several sites of N-linked glycosylation. This precursor molecule F0 oligomerises in the endoplasmic reticulum and is proteolytically processed at two sites in each monomer, resulting in a trimer of two disulphide-linked fragments: F2 (the smaller N-terminal fragment) and F1. The protein is anchored to the virion membrane through a hydrophobic peptide in the C-terminal region of F1, and is believed to adopt a metastable prefusogenic conformation until triggered in the presence of a target membrane and/or receptor. Triggering the protein is thought to result in a conformational change that exposes a very hydrophobic stretch of residues at the extreme N-terminus of F1, known as the fusion peptide. It is generally believed that the fusion peptide associates with the target cell membrane upon exposure. The present inventors hypothesise that the change from the prefusogenic structure to the final fusogenic conformation of the protein brings the fusion peptide into close proximity to the anchor sequence, promoting the fusion of the virus (and/or infected cell) membrane with the target cell membrane.
The structure of a fragment of the fusion protein of Newcastle Disease Virus (NDV) was recently reported at a resolution of 3.3 Å (Chen, L. et al., M.C. Structure, 9:255-266, 2001). This protein has some similarity in both sequence and mode of action to the RSV fusion protein. Only 355 of the 553 residues in NDV fusion protein were present in the final crystal form (residues 33-105 and 17144), such that important regions of the protein structure were absent. Another structure was subsequently reported for the core fragment of RSV-F itself (Zhao, X. et al., Proc. Nat Acad. Sci. USA, 97:14172-14177, 2000). This structure included a large section of the protein absent from the NDV-F structure. In the examples provided below, the present inventors describe the structure of the majority of the RSV-F protein produced by molecular modelling on the basis of a chimeric model devised by combining these two crystal structures. It has been found that the surface of the protein contains a region (“Site A”) vital for the activity of the protein and into which compounds known to inhibit viral fusion probably bind. The present inventors suggest that the region is important in the sequestration of parts of the RSV-F protein in its prefusogenic form and that it plays a key role in the process of cellular invasion by the virus.
The process of rationalised drug design requires no explanation or teaching for the person skilled in the art, but a brief description is given here of computational design for the lay reader. A person skilled in the art may use one of several methods to screen chemical entities or fragments for their ability to associate with a target molecule. For example, the screening process may begin by visual inspection of the target molecule, or a portion thereof, on a computer screen, generated from a machine-readable storage medium. Selected fragments or chemical entities may then be positioned in a variety of orientations, or docked, within identified or possible binding pockets (ie target sites). Docking may be accomplished using software such as Quanta (Accelrys, Inc., Burlington, Mass., USA. COPYRGT. 1994) and Sybyl, (Tripos Associates, St. Louis, Mo., USA) followed by energy minimization and molecular dynamics with standard molecular mechanics force fields, such as CHARMM (Accelrys, Inc., Burlington, Mass., USA. COPYRGT. 1994) and AMBER (Weiner, S. J. et al., J. Am. Chem. Soc., 106:765-784, 1984; Kollman, P. A., University of California at San Francisco, Calif., USA. COPYRGT. 1994).
Specialised computer programs may also assist in the process of selecting fragments or chemical entities. These include:
Once suitable chemical entities or fragments have been selected, they are 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 RSV-F. This is generally followed by manual model building using software such as Quanta or Sybyl.
Useful programs to aid the person skilled in the art in connecting the individual chemical entities or fragments include:
As is well known to the person skilled in the art, instead of proceeding to build a single compound or complex for the target site in a step-wise fashion, one fragment or chemical entity at a time as described above, inhibitory or other target-binding compounds may be designed as a whole or de novo. Methods for achieving such include:
Other molecular modelling techniques may also be employed. See, eg, Cohen, N. C. et al., J. Med. Chem., 33: 883-894, 1990). See also, Navia, M. A. et al., Current Opinion in Structural Biology, 2:202-210, 1992).
Once a single compound or chemical complex has been designed or selected by the above methods, the efficiency with which that entity may bind to a target site may be tested and optimized by computational evaluation. For example, an effective entity will preferably demonstrate a relatively small difference in energy between its bound and free states (ie a small deformation energy of binding). Thus, the most efficient entities should preferably be designed with a deformation energy of binding of not greater than about 10 kcal/mole, and preferably, not greater than 7 kcal/mole. Further, some entities may interact with the target site in more than one conformation that is similar in overall binding energy. In those cases, the deformation energy of binding is taken to be the difference between the energy of the free entity and the average energy of the conformations observed when the entity binds to the target site.
A compound or chemical complex designed or selected so as to bind to a target site may be further computationally optimized so that in its bound state it would preferably lack repulsive electrostatic interaction with the target enzyme. Such non-complementary (eg electrostatic) interactions include repulsive charge-charge, dipole-dipole and charge-dipole interactions. Specifically, the sum of all electrostatic interactions between the entity or other entity and the target site, when the entity is bound to the target site, preferably make a neutral or favourable contribution to the enthalpy of binding.
Specific computer software is available in the art to evaluate compound deformation energy and electrostatic interaction. Examples of programs designed for such uses include: Gaussian 92, revision C (Frisch, M. J., Gaussian, Inc., Pittsburgh, Pa. USA. COPYRGT. 1992); AMBER, version 4.0 (Kollman, P. A., University of California at San Francisco, Calif., USA. COPYRGT. 1994); QUANTA/CHARMM; and Insight 11/Discover (Accelrys, Inc. Burlington, Mass. USA. COPYRGT. 1994). These programs may be implemented, for instance, using a Silicon Graphics O2 workstation or Intel CPU based Linux duster. Other hardware systems and software packages will be known to the person skilled in the art.
Once a compound or chemical complex has been optimally designed or selected, as described above, modifications may be made to, for example, improve or modify its binding properties. Thus, for a compound, substitutions may be made in some of its atoms or side groups. Generally, initial substitutions of this kind will be conservative, that is the replacement group will have approximately the same size, shape, hydrophobicity and charge as the original group. It should, of course, be understood that components known in the art to alter conformation should be avoided. Such substituted chemical compounds may then be analyzed for efficiency of fit to a specific RSV-F target site by the same computer methods described in detail above.
Another approach is the computational screening of small molecule databases for compounds or chemical complexes that can interact in whole, or in part, to a target site. In this screening, the quality of fit of such entities to the target site may be judged either by shape complementarity or by estimated interaction energy (see for example: Meng, E. C. et al., J. Comp. Chem., 13: 505-524, 1992).
In a first aspect, the present invention provides a method for identifying a candidate anti-viral agent, said method comprising the steps of:
In a second aspect, the present invention provides a method for screening compounds and/or chemical complexes for a candidate anti-viral agent, said method comprising the steps of:
In a third aspect, the present invention provides a method for modifying a candidate anti-viral agent to provide an agent with improved anti-viral activity, said method comprising the steps of:
The methods of the first to third aspects of the invention permit the identification of candidate anti-viral agents and their possible development into anti-viral agents suitable for therapeutic use. Such candidate anti-viral agents and/or developed anti-viral agents preferably show inhibitory activity against at least RSV, but may also show inhibitory activity against other viruses with fusion proteins homologous to RSV-F protein such as all other known members of the Paramyxoviridae family of viruses. These viruses may be pathogenic to species other than humans such as Bovine RSV infecting cattle and NDV infecting poultry. Viral-inhibitory activity shown by candidate anti-viral agents and/or developed anti-viral agents is caused, preferably, by an ability to bind to the said target site and thereby inhibit fusion activity of RSV-F protein to block cellular invasion and/or cell to cell transfer of the virus.
The three-dimensional structure model generated in step (i) of each of the above methods comprises, at least, the three-dimensional structure of a target site to which a candidate anti-viral or developed anti-viral agent may interact (eg bind) to inhibit the activity of RSV-F protein. Preferably, the target site comprises all or part of each of the following amino acids of RSV-F protein; Tyr33, Cys37, Ser38, Ala39, Val40, Ser41, Lys42, Gly43, Leu48, Arg49, Thr50, Lys315, Leu316, His317, Thr318, Ser319, Pro320, Leu321, Cys322, Thr323, Ser330, Asn331, Ile332, Cys333, Leu334, Thr335, Arg336, Thr337, Asp338, Arg339, Phe352, Pro353, Gln354, Ala355, Glu356, Thr357, Cys358, Phe366, Cys367, Asp368, Thr369, Met370, Asn371, Ser372, Leu373, Lys394, Ile395, Met396, Thr397, Ser398, Lys399, Thr400, Asp401, Val402, Ser403, Ser404, Ser405, Val406, Ile407, Thr408, Ser409, Leu410, Gly411, Ala412, Ile4113, Val414, Ser415, Lys419, Lys421 and Asp440 (as derived from the amino acid sequence of F protein from RSV strain A2 provided SEQ ID NO:1), or the corresponding amino acids of F protein from variant RSV strains (eg a known RSV mutant which includes Leu in place of Ser at position 398; Andries, K. et al., AntiviralRes., 50: A29, 2001). The atomic coordinate data for those amino acids within the three-dimensional structure model of RSV-F protein is provided in Table 1 hereinafter. Thus, the three-dimensional structure model generated in the methods of the first to third aspects is preferably generated using at least the atomic coordinate data of Table 1. However, more preferably, the three-dimensional structure model is generated from the atomic coordinate data of Table 2.
The methods of the first to third aspects of the invention are preferably in silico methods.
In the method of the first aspect, the step of identifying a candidate anti-viral agent (ie step (ii)) may be achieved by methods described above for designing and selecting compounds or chemical complexes with three-dimensional structures that fit and interact with a target site.
The method of the first aspect may further comprise a step of assessing the deformation of energy of the candidate anti-viral agent when brought from the free state to the target site-interacting state (eg bound state). Preferably, the deformation of energy is not greater than 10 kcal/mole and, more preferably, not greater than 7 kcal/mole. Additionally or alternatively to the step of assessing the deformation of the candidate anti-viral agent, the method of the first aspect may comprise a step assessing the enthalpy of the interaction (eg binding) of the candidate anti-viral agent with the target site. Preferably, the candidate anti-viral agent shall make a neutral or favourable contribution to the enthalpy of the interaction.
In the method of the second aspect, the step of screening compounds and/or chemical complexes to identify any compound(s) or chemical complex(es) with a three-dimensional structure enabling interaction with the target site (ie step (ii)) may be achieved by methods described above. The screened compounds and/or chemical complexes may belong to a library or database of suitable compounds and/or chemical complexes (eg ACD-SC (Available Chemicals Directory Screening Compounds), MDL Inc., San Leandro, Calif. USA.).
In the method of the third aspect, the step of modifying a candidate anti-viral agent (ie Step (ii)) may be achieved by methods described above such as substituting one or more groups (eg functional groups) on compounds.
In a fourth aspect, the present invention provides a computer for producing a three-dimensional structure model of RSV-F protein or a portion thereof, said structure model comprising the three-dimensional structure of a target site to which an anti-viral agent may interact and thereby inhibit the activity of said RSV-P protein, wherein said computer comprises:
Preferably, the machine-readable data storage medium (eg a magnetic or optical storage medium such as a hard drive, floppy disc or a CD-ROM) comprises the atomic coordinate data of Table 1 or Table 2.
The computer may further comprise:
The atomic coordinate data for the range of chemical components and substituents and the atomic coordinate data for the range of compounds and/or chemical complexes, can be obtained from suitable databases.
In a fifth aspect, the present invention provides a machine-readable data storage medium comprising the atomic coordinate data of the following amino acids of RSV-F protein; Tyr33, Cys37, Ser38, Ala39, Val40, Ser41, Lys42, Gly43, Leu48, Arg49, Thr50, Lys315, Leu316, His317, Thr318, Ser319, Pro320, Leu321, Cys322, Thr323, Ser330, Asn331, Ile332, Cys333, Leu334, Thr335, Arg336, Thr337, Asp338, Arg339, Phe352, Pro353, Gln354, Ala355, Glu356, Thr357, Cys358, Phe366, Cys367, Asp368, Thr369, Met370, Asn371, Ser372, Leu373, Lys394, Ile395, Met396, Thr397, Ser398, Lys399, Thr400, Asp401, Val402, Ser403, Ser404, Ser405, Val406, Ile407, Thr408, Ser409, Leu410, Gly411, Ala412, Ile413, Val414, Ser415, Lys419, Lys421 and Asp440 (as derived from the amino acid sequence shown as SEQ ID NO:1), or the corresponding amino acids of F protein from variant RSV strains.
Preferably, the machine-readable data storage medium comprises the atomic coordinate data of Table 1 or Table 2.
In a sixth aspect, the present invention provides a candidate anti-viral agent identified in accordance with the method of the first or second aspect, or an anti-viral agent produced in accordance with the third aspect.
The agent of the sixth aspect may be used to prepare a medicament for treating a viral infection, especially an infection by RSV or another virus with a fusion protein homologous to RSV-F protein such as all other known members of the Paramyxoviridae family of viruses which may infect species other than humans.
Thus, in a seventh aspect, the present invention provides the use of the agent of the sixth aspect in the preparation of a medicament for use in treating a viral infection.
And, in an eighth aspect, the present invention provides a method of treating a viral infection in a subject, said method comprising administering a medicament comprising an agent of the sixth aspect.
Preferably, the viral infection treated by the medicament of the seventh aspect or the method of the eighth aspect, is an infection by RSV or another virus with a fusion protein homologous to RSV-F protein such as all other known members of the Paramyxoviridae family of viruses. Most preferably, the viral infection is a respiratory infection by RSV.
Studies of the mechanism of infection by RSV both in vitro and in animal models show that the RSV fusion protein (RSV-F) is of critical importance in both initial cellular invasion and the formation of syncytia. In order to develop an understanding of the mode of action of RSV-F at the molecular level, a homology model of the post-fusogenic RSV-F structure has been built, including residues 29 to 98 and 160 to 515 of the intact protein. The atomic coordinate data for this structure is provided in Table 2.
According to the homology model, the structure of the RSV-F trimer can be divided into three regions; an extended stalk (commonly referred to as the “core” region of the protein) consisting of a 6-membered helical coiled-coil, a relatively globular head region, characterised by an unusual central cavity connected to the exterior of the protein by symmetrically located radial channels and a single axial opening and the neck region, an interface between the head and core. While not wishing to be bound by theory, it is hereby proposed that the RSV-F protein sequesters very hydrophobic fusion peptides in the internal cavity of the head domain in the pre-fusogenic form of the protein, and that within the cavity are three, symmetrically positioned grooves that link the top of the central channel with each of the radial channels. It is considered that these grooves are ideal sites of interaction for the three fusion peptides in the pre-fusogenic form and it is shown hereinafter that, surprisingly, they are also the likely primary sites of binding for three known RSV-F fusion inhibitors of quite different molecular structure. The mechanism of action for these inhibitors, which to the best of the present inventors' knowledge has not been previously determined at the molecular level, is attributed to their ability to interfere with the fusion peptide function through binding within the groove. Again, while not wishing to be bound by theory, it is considered likely that the inhibitors either prematurely displace the fusion peptides from sequestration in the head region of the protein prior to attachment of the F protein to a cell, or prevent egress of the fusion peptides once the F protein has bound to a cell. This, at a minimum, reduces the ability of the virus to infect cells or form syncytia and may also result in the aggregation of viral particles due to the premature exposure of extensive hydrophobic regions of the fusion protein.
This information provides a rational basis for the design and development of compounds and/or chemical complexes that strongly interact (eg bind with high affinity) to one or more of the abovementioned grooves (Site As) in RSV-F and thereby disrupt RSV-F function. Accordingly, these compounds are likely to have anti-viral activity and thereby therapeutic value in the treatment or prevention of viral infections such as a respiratory infection by RSV.
Each of the three Site As of RSV-F protein are defined by the following amino acids: Tyr33, Cys37, Ser38, Ala39, Val40, Ser41, Lys42, Gly43, Leu48, Arg49, Thr50, Lys315, Leu316, His317, Thr318, Ser319, Pro320, Leu321, Cys322, Thr323, Ser330, Asn331, Ile332, Cys333, Leu334, Thr335, Arg336, Thr337, Asp338, Arg339, Phe352, Pro353, Gln354, Ala355, Glu356, Thr357, Cys358, Phe366, Cys367, Asp368, Thr369, Met370, Asn371, Ser372, Leu373, Lys394, Ile395, Met396, Thr397, Ser398, Lys399, Thr400, Asp401, Val402, Ser403, Ser404, Ser405, Val406, Ile407, Thr408, Ser409, Leu410, Gly411, Ala412, Ile413, Val414, Ser415, Lys419, Lys421 and Asp440 of the monomers of RSV-F protein. It is envisaged that homologous fusion proteins from other viruses (eg other viruses from the family Paramyxoviridae) may include a site similar to Site A which may also be useful for identifying or screening anti-vital agents. Thus, it is to be understood that the present invention also extends to the generation of three-dimensional structure models of homologous fusion proteins (especially those showing at least 20%, preferably at least 30%, more preferably at least 50%, and most preferably at least 80% (as calculated using the CLUSTAL algorithm; Thompson et al., Nucleic Adds Research, 24:4876-4882,1997), amino acid similarity to the amino acid sequence of RSV-F protein from RSV strain A2, and/or which have an amino acid backbone with a root mean square deviation from the amino acid backbone of RSV-F protein of not greater than 1.5 Å), and the use of such three dimensional structure models to identify and screen for candidate anti-viral agents.
Agents resulting from the use of a method of the first to third aspects of the present invention, may be assayed for anti-viral activity by a number of in vitro assays. For example, the agents can be assayed for their effect on viral activity by culturing virus in the presence of said agents. The efficacy of an agent will be reflected in the reduction of viral infectivity in the presence of increasing levels of said agent. Analysis of in vitro anti-viral activity of the agents can also be carried out using a recombinant RSV-F fusion assay in which suitable mammalian cells (eg 293 cells) are transfected with plasmids or vectors carrying a polynucleotide sequence for expressing RSV-F protein. Expression of the RSV-F protein results in cellular fusion and the formation of syncytia, which effect may be suppressed by an agent which is an inhibitor of RSV-F protein activity. The relative levels of syncytia formation in the presence of increasing levels of an inhibitor provides an assessment of the efficacy of the agent.
Medicaments comprising anti-viral agents may be prepared in accordance with methods and materials (eg excipients and carriers) well known to a person skilled in the art As such, the medicament may be formulated for various routes of administration including oral, nasal, vaginal, rectal, topical, intramuscular and intravenous routes.
Oral formulations may be in the form of tablets, capsules or emulsions, and may include one or more carriers such as lactose, saccharose, starch, calcium stearate and methylcellulose. Alternatively, oral formulations (and nasal formulations) may be in the form of a gas or aerosol for inhalation (particularly suitable for treatment of respiratory infections). Vaginal and rectal formulations may be in the form of pessaries and suppositories respectively. Topical formulations may be in the form of ointments, creams or lotions. Injectable formulations for administration by intramuscular or intravenous routes may include well known carriers and/or excipients.
It is routine for a person skilled in the art to determine an effective dosage of an anti-viral agent and he/she will understand that the amount may vary according to, for example, the type of virus to be treated and the severity of the infection. However, generally a medicament comprising an anti-viral will be administered to a subject to provide a dose of the anti-viral agent in the range of 0.1 to 10 mg/kg/day.
The homology model of the post-fusogenic RSV-F structure described herein may also be used for the rational selection of regions within the RSV-F protein which, when appropriately presented, may be antigenic and capable of eliciting RSV-specific antibodies which may be useful for, for example, RSV diagnostic assays and/or methods of prevention or treatment of RSV infection.
Thus, the present invention also relates to a method of identifying a region within the RSV-F protein comprising an antigenic determinant which may be used to elicit antibodies capable of binding to RSV, said method comprising the steps of:
The three-dimensional structure model used to identify a region within the RSV-F protein comprising an antigenic determinant capable of eliciting antibodies capable of binding to RSV, is preferably generated from the atomic coordinate data of Table 2.
The identification of a surface-located region of the RSV-F protein which may include an antigenic determinant may simply be made by the eye of a person skilled in the art or, in a preferred in silico method, the identification may be made a computer. The surface-located region will typically comprise 5 to 100 amino acids, more preferably 10 to 50 amino acids. The amino acids will preferably be contiguous, however the present invention also contemplates non-linear antigenic determinants comprising non-contiguous amino acids. A “surface-located” region of the RSV-F protein is to be regarded as consisting of any region of the RSV-F protein which is accessible and may be bound to by an anti-RSV-F protein.
The antibodies elicited by the identified region preferably binds to RSV in a specific manner (ie the antibodies show negligible, if any, binding to other agents). The elicited antibodies may be RSV-neutralising antibodies (ie antibodies able to inhibit the infectivity of RSV).
An identified surface-located region of the RSV-F protein may be tested for its ability to elicit antibodies capable of binding to RSV by any of the standard methods known to a person skilled in the art For example, the surface-located region may be isolated or synthesised, conjugated to a suitable carrier protein (eg bovine serum albumin), and used to raise monoclonal antibodies by routine methods. The elicited monoclonal antibodies may then be tested using, for example, an appropriate ELISA assay to determine whether the monoclonal antibodies bind to RSV.
The capacity of raised monoclonal antibodies to neutralise RSV may be assessed by conducting standard virus neutralisation tests in vivo using susceptible animals, or otherwise, in tissue cultures.
Thus, the method of identifying a region within the RSV-F protein comprising an antigenic determinant which may be used to elicit antibodies capable of binding to RSV, may additionally comprise the steps of:
The present invention further relates to RSV specific antibodies or fragments thereof (eg Fab, Fc or scFv fragments), which bind to a region of the RSV-F protein comprising a novel antigenic determinant identified by the abovementioned method. Where such antibodies or fragments thereof are capable of neutralising RSV, the antibodies or fragments thereof may be used as an anti-viral agent, especially against an infection by RSV or another virus with a fusion protein homologous to RSV-F protein such as all other known members of the Paramyxoviridae family of viruses. Therefore, the present invention also relates to a medicament comprising such antibodies and its use in a method of treating or preventing a viral infection in a subject. Antibodies in accordance with the present invention may also be used in methods to detect RSV or another virus with a fusion protein homologous to RSV-F protein (eg all other known members of the Paramyxoviridae family of viruses), such as diagnostic assays for subjects suspected to be suffering from RSV infection.
Moreover, the present invention relates to the use of novel antigenic determinants identified from the three-dimensional structure model of RSV-F protein, as vaccine agents (optionally in combination with a suitable adjuvant or carrier) which may be administered with a suitable adjuvant or carrier to a subject to treat or prevent a viral infection. Alternatively, suitable recombinant viral vectors may be engineered, by routine methods, to express an antigenic determinant identified from the three-dimensional structure model of RSV-F protein. Such viral vectors may be administered with a suitable adjuvant or carrier to a subject to treat or prevent a viral infection. Suitable viral vectors are well known to a person skilled in the art, and include vaccinia and adenovirus vectors.
Throughout this specification the word “comprise”, or variations such as “comprises” or “comprising”, will be understood to imply the inclusion of a stated element, integer or step, or group of elements, integers or steps, but not the exclusion of any other element, integer or step, or group of elements, integers or steps.
Any discussion of documents, acts, materials, devices, articles or the like which has been included in the present specification is solely for the purpose of providing a context for the present invention. It is not to be taken as an admission that any or all of these matters form part of the prior art base or were common general knowledge in the field relevant to the present invention as it existed in Australia or elsewhere before the priority date of each claim of this application.
The present invention is further described below with reference to the accompanying figures and following, non-limiting examples.
Homology Modelling of RSV-F
As a preliminary step in modelling the RSV-F protein, a chimeric construct of the NDV-F structure and the RSV-F core structure was produced. This chimera was built using SwissPDB Viewer (SPDBV) (Guex, N. and Peitsch, M. C., Electrophoresis, 18:2714-2723, 1997) by superimposing the coordinates of G184-Q202 of chain A from the RSV-F core onto G171-Q189 of chain A in the NDV-F structure. These residues were selected after conducting alignment of the amino acid sequence of the two proteins from which it was determined that they were structurally equivalent regions (
The chimera was then used as the basis for building a model of RSV-F. The primary alignment used for model building was created using Clustal (Thompson et al., Nucleic Adds Research, 24: 4876-4882, 1997) and was adhered to except for one loop region in F2 (
After insertion of the required loops, the separate RSV-F1 and RSV-F2 structures were combined and the complete trimeric RSV-F structure formed by superimposing the initial dimer onto the coordinates of the other dimers in the chimera. Minimisation of the complete RSV-F model was then carried out with Discover in the InsightII package (Accelrys) under the cvff forcefield to remove undesirable interactions that had been generated by the modelling process. Firstly, the residues surrounding each inserted loop were minimised with the rest of the structure held fixed to correct any minor mismatches at insertion points. Then the backbone atoms of the whole protein were held fixed and the sidechain atoms minimised using Steepest Descent minimisation (see the InsightII Manual distributed with the software and references therein, all of which are to be regarded as incorporated herein by reference) until the maximum gradient value dropped to below 10. The backbone constraints were then removed and the whole structure allowed to minimise for a further 100 steps of Steepest Descent minimisation.
Ligand Docking
Consideration of the intricate surface and three-fold symmetry of the RSV-F structure suggested that there would be two classes of potential target sites for small molecules: 1) sites where ligand binding can occur simultaneously in three non-overlapping symmetrical places on the protein; 2) sites at points where symmetrically equivalent interactions overlap, meaning each trio of sites could only be occupied by a single ligand at any time, but in one of three potential orientations. This complication, and the very extensive surface area of the protein (more than 40,000 Å2), indicated that locating the appropriate binding site for RSV-F inhibitors would not be trivial and would require care to ensure that the site is defined as objectively as possible. Subsequently, a two-stage docking analysis, described in detail below, was used with three different RSV-F inhibitors to try and identify sites of action for different classes of RSV-F ligands.
Docking analysis of RSV-F was carried out using the program AutoDock 3.0.3. The algorithm employed by AutoDock is described in detail in the AutoDock User Guide distributed with the software and references therein (Morris, G. M. et al., J. Comp. Chem., 19: 1639-1662, 1998), all of which are to be regarded as incorporated herein by reference. In brief, the region of space around the protein target selected for analysis is divided into a grid of points spaced a user-definable distance apart. At each of these points, the potential energy of every type of atom in the ligand being docked is calculated and stored in a potential map. During the actual docking process, the ligand is inserted into the assigned region of space and translated, rotated and allowed to alter user-selected torsion angles and the energy of the ligand rapidly calculated by reference to the stored maps of potentials, rather than direct calculation during the docking itself. After a number of steps of optimisation determined by the user, the best conformation is saved and the process repeated. After the desired number of conformations are generated they are grouped into “clusters” of similar structures, defined by a user selected RMS (Root Mean Square deviation in atomic position) tolerance.
For docking with RSV-F, two different levels of docking analysis were used. A relatively coarse docking analysis over the entire protein was carried out first to aid in identifying potential binding sites on the protein without the intention of defining details of the interaction at such sites. This analysis, was partly automated by a simple set of in-house scripts that take, as input, a coordinate file for a protein in PDB (protein data bank) format then define a set of overlapping cubic regions that completely encompass the protein surface. These regions are then used for separate docking analyses using AutoDock. Each region was defined as a cube of 121×121×121 grid points separated by a grid step of 0.5 Å, making a cube of 60.5 Å each side. The cubes overlapped by 10 Å, to ensure adequate coverage of every point on the protein surface. Typically, a total of 16 cubes was needed for complete coverage of the RSV-F surface. A total of 50 docked conformations was calculated for each cube, with the conformations clustered using an RMS tolerance of 1.8 Å. Thus for each ligand used for full-protein docking runs, 800 docked conformations were calculated.
The top 1% of docked conformations (as ranked by the docking algorithm's calculated energy for the best structure in each cluster) were subjected to detailed visual inspection. Of these eight conformations, 4 were selected as candidate physiological binding sites on the basis of the interactions observed and the position of the site on the protein surface. These candidate sites were then used as the centre of detailed docking analyses using a higher-resolution grid (61×61×61 points, 0.375 Å grid step size for a grid 22.9 Å each side, 100 structures calculated and clustered with a tolerance of 1.2 Å). Visual analysis of the best structures in each site was then used to discriminate the most probable site of physiological ligand binding.
Three published RSV-F inhibitors with quite different molecular structures were submitted to the docking regime described above. The first is a compound in development by the company Viropharma, described in International patent specification No. WO 99/38508 (
Binding Site Characterisation
The conformations determined by AutoDock do not take flexibility of the protein into account. Subsequently, to optimise the accuracy of the docked structures, the ligand and all residues within 15 Å of the ligand centroid were subject to energy minimisation using Discover (Accelrys, USA). Details of the contacts made between the ligand and protein were measured using a simple in-house script that measures the distance between every atom in the ligand and the atoms in the neighbouring residues of the protein then reports those atoms separated by less than a user-defined cutoff value. This data was then interpreted to derive two pieces of information; 1) delineation of the residues that made up the binding site for the particular docked conformation of the ligand; 2) a list of potential sites for mutagenesis studies useful for the selection of the actual physiological binding site.
Viral Assays: Assay for Compounds Active Against RSV
Graded concentrations of test compounds were prepared by serial dilution in 96 well tissue culture plates containing 100 μL per well of assay media (Eagles Minimal Essential Medium (MEM) supplemented with 2% foetal bovine serum and Glutamine alanyl-L-glutamine). Human laryngeal carcinoma (HEp2) cells were trypsinised and diluted in assay media to a concentration of 1.5×105 viable cells/mL according to standard methods. 1.5×104 of the diluted cells were added to several individual wells of each 96 well plate which were designated uninfected control wells. The remaining cell suspension was exposed to infectious RSV by addition of sufficient virus to the suspension to yield a concentration of 0.4 plaque forming units per cell. The cell-virus suspension was mixed well by inversion and 100 μL immediately dispensed into the remaining wells of the 96 well plates. Assay plates were incubated in a 37° C. incubator for 5 days. Viral replication was quantified by determination of cell viability substantially as described by Watanabe and colleagues (Watanabe, W. et al., J. Virol. Methods, 48: 257-265, 1994). Briefly, 100 μL of 3 mg/mL MTT (3(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) was added to each well and plates incubated at 37° C. in a 5% CO2 incubator for 2 hours. Wells were aspirated to dryness and the crystalline product of MIT metabolism dissolved by addition of 200 μL of 100% isopropanol. The extent of metabolism was measured via determination of absorbance using a computer-controlled microplate reader at 540 nm with a reference wavelength of 690 nm. The metabolism of MTT is proportional to the number of viable HEp2 cells (Watanabe et al., supra). The 50% cytotoxic concentration (CC50) and the 50% effective concentration (EC50) were determined using computer curve fit software (eg XLFit® (ID Business Solutions, Guildford, UK) and Ascent® (MTX Lab Systems, VA, USA)).
Viral Assays: Selection of Drug Resistant Virus and Cross-Resistance Testing
RSV was grown in the presence of graded concentrations of COMPOUND 1 or COMPOUND 2 diluted in assay medium. Putative resistant virus (evident by the presence of viral cytopathology in these cultures) was recovered, RNA isolated and the F gene sequence determined via RT-PCR according to standard methods. Changes in nucleotides resulting in amino acid sequence changes were noted and the respective amino acids were mapped on to the known F protein sequence. In addition, resistant viruses were grown, titred, F gene re-sequenced and assayed for drug resistance using the Assay described above.
Syncytia Formation Assays
293 cells (Graham, F. L. et al., J. Gen. Virol. 36: 59-72, 1977) were transfected with plasmid pCICO F.FL.opt which contains a gene encoding RSV-F. Cells transfected with this plasmid were incubated at 37° C. in a 5% CO2 incubator and observed 24-48 hours post transfection. The F-transfected cells are identical in appearance to RSV-F infected cells, being characterised by the presence of many large syncytia and dying cells due to the fusion of cellular membranes by the expressed F protein. Control cells were confluent at the same time point. Fusion inhibitors were assayed by transfecting the plasmid into 293 cells as above in the presence of varying amounts of inhibitors. Syncytia were scored by microscopic inspection and cell viability determined using MTS (3(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, Promega) via standard methods. The reduction in scorable syncytia formation and cell viability was used as a direct measurement of the efficacy of the inhibitor used. Site specific mutants of RSV-F were produced (see below) and assayed in an identical manner.
Mutagenesis
Mutagenesis was used to introduce specific amino acid changes in recombinant RSV-F using the pCICO F.FL.opt plasmid carrying the optimised wild-type RSV-F gene and the synthetic oligonucleotide PCR method. Where necessary, particularly with mutant constructs that showed poor expression in the syncytia formation assay, the entire modified RSV-F gene was sequenced to ensure that only the desired mutations had been introduced into the sequence.
Results
Molecular Modelling
The final modelled structure of RSV-F residues 29 to 98 and 160 to 515 is shown in
Docking of RSV-F Inhibitors
The docking results for the three different inhibitors are summarised diagrammatically in
While the docking analysis described above was carried out in a unbiased manner, there are some published data concerning the residues that may be involved in interaction with two of the known RSV-F inhibitors. Viral escape mutants resistant to inactivation by COMPOUND 2 were reported by Andries and colleagues (Andries, K. et al., supra) and have been isolated by us using COMPOUND 1 and COMPOUND 2 in separate experiments. Andries reported that two residues in RSV-F were mutated in COMPOUND 2-resistant virus: S398 (to Leucine) and D486 (to Asparagine). In COMPOUND 1-resistant viral cultures F488 to Valine or Isoleucine were commonly isolated.
The results of the docking studies were correlated with this known resistance mutant data and it was with some surprise that all three compounds were found to dock best in a site immediately adjacent to the position of one of these escape-mutant residues, S398.
The best docked conformations for each of the three ligands are shown in
This groove was assigned the label Site A, such that there are three Site As in each RSV-F molecule. Visual inspection of Site A shows that it consists of a hydrophobic pocket and groove, with flanking polar residues including a series of potential salt-bridge pairs (
Confirmation of Site A
Having identified a single site that appears to be responsible for binding three very different types of fusion inhibitor, two separate approaches were employed to try and confirm the relevance of the site and thus the accuracy of the study as a whole. Firstly, a series of specific mutations were inserted into recombinant RSV-F and their effect on fusion activity and susceptibility to fusion inhibitors examined in a syncytia formation assay. The second approach was to screen RSV cultures for resistance to COMPOUND 1 and COMPOUND 2 to try and identify new escape mutants which could then be correlated with the proposed site.
A total of 14 mutations were suggested on the basis of the contacts observed in the docking experiments (summarised in Table 3). Nine of these suggested mutants were in Site A, the other five were at other points in the protein to act as site-definition controls. Some of the mutations were minor changes, such as serine to threonine, specifically intended to cause slight alterations in susceptibility of RSV-F to inhibitors without disrupting function. More drastic mutations, such as serine to leucine, were also proposed as changes guaranteed to have a substantial effect on inhibition, though probably also on activity. Of the 14 mutants, 8 were inactive in the assay apparently due to misfolding and degradation of the protein before it reaches the cell surface. Two of the active mutant proteins showed clear resistance to inhibitors. These were S319L (a Site A mutant) and T326L. The S319L mutant is fully active in the syncytia formation assay and almost completely resistant to inhibition by either COMPOUND 1 or COMPOUND 2, showing fusion activity at concentrations more than two orders of magnitude more than needed to cause complete inhibition in wild-type RSV-F. The resistance seen in T326L is weaker than that for S319L (about 10 fold only) and restricted to COMPOUND 1. T326L, however, is not part of the defined Site A, being more on the outside face of the head region. It is, however, in close proximity to some Site A residues (the closest, Ile395 which interacts only with COMPOUND 1, being just 3.8 Å away) and a substantial change in local conformation at T326 could be expected to influence these residues.
Isolation of escape mutants through viral assay has also produced supporting evidence for the relevance of Site A. Three mutants were isolated by culturing virus in the presence of COMPOUND 1. These are F488V, T400A and T337A, the latter two of which are in RSV-F Site A. The F488V mutant is a variation on the F488L/I mutants that have been identified elsewhere. The other two mutations are at the top of Site A, flanking the position occupied by the three ligands in their best docking sites. A double mutant D489Y, F488I has also been identified. This F488 mutation has been seen previously but it is the first identification of a mutant at D489. Interestingly, there is a known D486 resistance mutant for COMPOUND 2, suggesting that the aspartic add residues flanking F488 may have a role in stabilisation of the pre-fusogenic form of RSV-F where it is hypothesised that F488 interacts with Site A. A possible role for these aspartates is interaction with some of the positively charged residues that flank the Site A groove.
Two new COMPOUND 2 escape mutants have also been identified. One contains the same D489Y mutation mentioned above for COMPOUND 1 and another new mutation, L1381. Unfortunately L138 is outside the RSV-F model in F2, so it is difficult to comment on its potential role in RSV-F activity. The other COMPOUND 2 escape mutant is a triple mutant, containing the L138I substitution, H516N (which has been shown to have no effect on activity or resistance when recreated alone by mutagenesis of F and expression via the fusion assay—data not shown) and, most intriguingly, S319F. Thus, a major substitution at S319 correlates with COMPOUND 2 resistance in a viral background. The correlation between the escape mutations identified for COMPOUND 2 and COMPOUND 1 and cross-resistance assays provide strong support for their acting at the same site and in a very similar manner. Some of the resistance mutant isolation and cross-resistance data is summarised in Table 4.
These results provide dear evidence of cross-resistance to COMPOUNDS 1 and 2 via fusion peptide and stalk mutations. This strongly suggests an identical site and mechanism of action for both compounds.
The escape mutant data, therefore, supports the importance of Site A in the function of RSV-F and the action of RSV-F inhibitors. Identified mutants are located either within Site A itself, or in a duster well removed from the Site and associated with the coiled coil. This second region appears to be conserved amongst widely divergent viral fusion-protein sequences. Its precise role in inhibitor action is unclear; while several escape mutations have been identified in this region, they are always restricted to the external HR-B helix of the coiled-coil. No mutations have been observed that alter the pocket within the internal HR-A helices where the HR-B residues found to mutate in escape mutants bind. Although not wishing to be bound by theory, it is hypothesised that mutations in this site are actually altering the inhibitor susceptibility of the RSV-F protein in the pre-fusogenic structure, rather than disrupting coiled-coil formation as is the currently accepted mode of action. According to the hypothesis, the HR-B residues shown to give rise to escape mutations must interact with parts of Site A in the pre-fusogenic structure. This is a simpler explanation of the widely separated sites of inhibitor binding identified by escape mutations than the only other possibility; that there are two separate sites (Site A and this region in the coiled-coil), both of which must be bound by inhibitor to cause inhibition.
Characterisation of Site A
The region of the RSV-F structure identified by this docking procedure is highlighted in
aSite-specific mutation made in RSV-F sequence.
bPosition of the mutation either in Site A itself or elsewhere within the RSV-F structure.
cActivity of the recombinant RSV-F mutant in the standard syncytia formation assay detailed herein.
dResistance or sensitivity of recombinant RSV-F protein carrying the specific mutation in the standard syncytia formation assay.
N/A: not active.
NT: not tested.
*_(1-D-ribofuranosyl-1H-1,2,4-triazole-3-carboxamide)
The following provides test examples of the application of the present invention.
Computational Prediction of the Binding Orientation of Fusion Inhibitors
Three known fusion inhibitors that act against RSV-F, ie COMPOUNDS 1-3 (
The algorithm employed by AutoDock is described in detail in the AutoDock User Guide distributed with the software and references therein (Morris, G. M. et al., supra), all of which are to be regarded as incorporated herein by reference. In brief, the region of space around the protein target selected for analysis is divided into a grid of points spaced a user-definable distance apart. At each of these points, the potential energy of every type of atom in the ligand being docked is calculated and stored in a potential map. During the actual docking process, the ligand is inserted into the assigned region of space and translated, rotated and allowed to alter user-selected torsion angles and the energy of the ligand rapidly calculated by reference to the stored maps of potentials, rather than direct calculation during the docking itself. After a number of steps of optimisation determined by the user, the best conformation are saved and the process repeated. After the desired number of conformations are generated they are grouped into “clusters” of similar structures, defined by a user selected RMS (Root Mean Square deviation in atomic position) tolerance.
Each of the three inhibitors was manually constructed in the Builder module of InsightII (Accelrys, Inc., Burlington, Mass., USA). After geometry optimisation within InsightII the ligands were saved as Accelrys CAR format files. These were converted to PDB format using the AutoDock utility program CARTOPDBQ. Rotatable bonds in the molecule were defined using the AutoDock utility program AUTOTORS to generate an AutoDock ligand input file given the extension PDBQT, containing atom charge information and definitions of rotatable bonds as well as the position and type of each atom. The minimised RSV-F model in Accelrys CAR format was converted to an AutoDock PDBQ file using the CARTOPDBQ utility. AutoDock solvation parameters were then added to the PDBQ format file using the AutoDock utility program ADDSOL, creating an AutoDock macromolecule input file given the extension PDBQS, containing atom charge and simple solvation model information as well as the position and type of each atom. Parameter files for generating the appropriate atom type grids for each ligand were then created using the AutoDock MKGPF3 program. This takes a ligand in PDBQT format and a macromolecule in PDBQS format and generates a grid parameter file (GPF) containing the required information to define the various energy grids needed to match the ligand with the target macromolecule. The GPF also contains positional information, defining the extent (in number of grid points), size (in angstrom per grid point) and precise position (as a set of XYZ coordinates in the same reference frame as the target macromolecule) of the chosen docking site. As the three ligands being analysed all contain the same set of atom types (C, N, O, and H) only a single GPF was required. The target site in RSV-F was selected as the center point in Site A, determined by averaging the coordinates of residues contributing to the site. A grid size of 0.375 Å, with 60 grid points in each dimension was used to fully define the grid position. The grid files were then generated using the AutoDock utility program AUTOGRID3, which takes as input a GPF file and the associated macromolecular PDBQS file and produces a set of grid maps (MAP files) defining the interaction energy at every point within the grid, one for each atom sort and one for electrostatic interactions.
A docking parameter file (DPF) was created for each ligand using the program MKDPF3. This takes a ligand in PDBQT format and a macromolecule in PDBQS format and generates a DPF which contains the required information for carrying out the docking analysis, including the names of various input and output files and the desired details of the docking process. The docking process was set to use the Genetic Algorithm option, with a total of 100 docked structures determined. Docking was carried out using the AUTODOCK3 program, which takes the DPF file and the associated PDBQT, PDBQS and MAP files and carries out the actual docking process, generating a docking log (DLG file) containing complete details of the docking process, the energies and final coordinates of the docked ligand structures.
The 100 docked structures for each ligand were clustered by AutoDock, as part of the docking process, into groups of structures with an RMSD less than a user-defined tolerance value. The clusters were then ranked on the energy of the best dock within each duster.
Structure-Activity Relationship (SAR) Analysis by Characterization of Drug Binding Orientation
Detailed analyses of the contacts arising between RSV-F inhibitors and the protein can be determined by careful examination of the docked conformation of the inhibitor to the model. Correlation of the contacts between the protein atoms and regions of the drug with traditional SAR arising from activity in an in vitro assay can be made, allowing rationalisation of the observed SAR properties of the compounds and prediction of other SAR information. For this procedure, a model of the ligand—protein complex must be prepared in some manner such as that described above in Test Example 1. Interactions between the protein and ligand can then be measured in a variety of ways including, but not limited to, the following.
Firstly, the distances between each atom of the ligand and those atoms of the protein within a distance corresponding to effective physical contact, for example 3 Å, can be measured with a simple computer program written to calculate the distances between all atoms in two bound molecules and produce a list of adjacent atoms as output Secondly, the nature of the interactions between the ligand and protein in the complex can be defined by a graphical representation such as that generated by the program LIGPLOT (Wallace, A C et al., Protein Eng. 2:127-34, 1995). Thirdly, the interactions between the molecules can be examined visually using a computer with appropriate visual display hardware, preferably including the ability to display three-dimensional images using an appropriate stereo-viewing adaptor, and molecular graphics software such as Sybyl (Tripos Associates, St. Louis, Mo., USA), InsightII (Accelrys, Inc., Burlington, Mass., USA), VMD (Humphrey, W et al., J. Molec. Graphics. 14:33-38,1996), Vida (OpenEye, Santa Fe, N. Mex., USA) or the like which can calculate and display important contacts such as hydrogen bonds between the protein and ligand.
On the basis of one or more of these methods for analysing contacts between the ligand and protein details of the types of interactions that stabilise the complex, for example hydrogen bonds, favourable ionic interactions and the burial of exposed hydrophobic surfaces, or that destabilise the complex, for example repulsive ionic interactions, steric hindrance and restriction of mobile groups can be determined. Modifications of the chemistry of the ligand can then be proposed to increase the stabilising interactions and/or decrease the unfavourable interactions. These modified molecules can also be subjected to the same process of docking and examination independently, to ascertain whether the proposed modifications are likely to have the desired effect or instead cause the ligand to bind to the protein in a different, and potentially undesirable, manner.
De novo Design of Novel Fusion Inhibitors by Property-Based Analysis of Drug Binding Sites
A structure for the site of action of known inhibitors of RSV-F activity allows the de novo design of novel compounds with complementary properties to the site. This allows new fusion inhibitors to be devised without recourse to extensive in vitro screening of libraries of compounds. This process can be carried out with, for example, the program LUDI (Accelrys, Inc., Burlington, Mass., USA) or LeapFrog (Tripos Associates, St. Louis, Mo., USA) or other similar programs written for the purpose of de novo drug design.
LUDI is described in detail in the original paper (Bohm, H.-J. supra) and the supplied manual (Accelrys), all of which is to be regarded as incorporated herein by reference. In brief, LUDI functions by fitting small molecular fragments from an in silico library to optimal positions within a defined site on the surface of a target protein. These separate fragments are then joined by appropriate chemical linkers, again selected from an in silico library of potential linking chemistries. Alternately, new fragments can be added to existing fragments in such a way as to allow the two fragments to fuse into a single larger entity without the need for an additional linker.
Site A of RSV-F can be used as a target for LUDI to design novel fusion inhibitors. For instance, a LUDI run can be initiated by selecting a point within the site of interest, for example the gamma-2 carbon of residue Valine 406 in the RSV-F model, as the centre point for a sphere of user-defied radius within which LUDI identifies appropriate positions for fragments from the library. A series of parameters can then be set by the user including the scoring function used to rank the fragment positions, the number of bonds allowed to rotate during conformational searches of flexible ligands, the minimum and maximum values for various interactions such as hydrogen bonds and contact surface areas and the like. A set of initial fragment hits can then be generated, ranked by the scoring function selected by the user. After this, the structure of the protein containing each of these initial hits can be used as the basis for additional rounds of LUDI screening, building a larger ligand at each step. When a molecule of appropriate size for a lead candidate has been achieved (eg 500 to 1000 Daltons), the molecule can be saved and provided to appropriately skilled chemists as a synthetic target.
Screening in Silico Compound Libraries for Novel Fusion Inhibitors
Selection of compounds from an in silico library with complementarity to a site on a target protein can be carried out using a variety of software such as Cerius2 (available from Accelrys, Inc., Burlington, Mass., USA), FlexX (available from Tripos Associates, St. Louis, Mo., USA), FRED (available from OpenEye, Santa Fe, N. Mex., USA), or other similar programs written for the purpose of in silico screening. Libraries of small molecules can be obtained from a number of small molecule supply companies such as Maybridge (Tintagel, Cornwall, UK), Chembridge (San Diego, Calif., USA), InterBioscreen (Chernogolovka, Russia) or SPECS (Rijswijk, The Netherlands). These libraries provide a source of potential ligands which can be screened against the target site.
Site A of RSV-F can be employed for the purposes of in silico screening to identify potential inhibitors of RSV-F activity. For example, in Cerius2, the screening process involves creating a series of “queries” in a sphere centred on a position selected by the user. These queries are generated using the Query Builder subprogram which examines the volume of space contained within the sphere to locate and cluster particular properties, such as hydrogen bond donors and acceptors and lipophilic areas. A set of volume exclusions are also prepared to prevent selected ligands overlapping with atoms from the protein. The property clusters produced by Query Builder are used for screening the library, such that compounds selected from the library must satisfy a user-defined number of the interactions apparent in the property clusters. The compounds selected in this manner can then be scored using a number of scoring functions to rank their fit to the protein site. Compounds selected in this way can then be obtained from their respective suppliers and tested for anti-fusion activity in in vitro assays of RSV-F activity.
Identification of antigenic regions likely to give rise to neutralizing antibodies Another use for the RSV-F structure model of the present invention is the rational selection of regions within the RSV-F protein that are likely to give rise to RSV-binding and/or RSV-neutralising antibodies if employed as antigens. Visual inspection of the structure model allows regions of protein that are both exposed and associated with the putative mechanism of action of the protein to be identified. Peptides corresponding to these regions of the protein can then be synthesised and employed as antigenic agents for the generation of monoclonal antibodies by routine methods. Elicited monoclonal antibodies may be tested using, for example, an appropriate ELISA assay to determine whether the monoclonal antibodies bind to RSV. Further, the capacity of elicited monoclonal antibodies to neutralise RSV may be assessed by conducting standard virus neutralisation tests in vivo using susceptible animals, or otherwise, in tissue cultures.
Antigenic sites on the RSV-F protein have been found to duster in immunogenic hotspots (Collins, P. L. et al., p. 1313-135I In B. N. Fields, D. M. Knipe, P. M. Howely (ed.) Fields Virology, 3rd Edition, Lippincott-Raven, Phil., 1996; Lopez, J. A. et al., J. Gen. Virol. 77:649-660, 1996), and the positions of some of these sites investigated by electron microscopy of RSV-F-antibody complexes (Calder, L. J. et al, Virology 271:122-131, 2000). Several of these were mapped on the RSV-F structure model (
A discrepancy is apparent between the RSV-F structure model and the predicted location of another immunogenic region comprising residues 205 to 237 (
It will be appreciated by persons skilled in the art that numerous variations and/or modifications may be made to the invention as shown in the specific embodiments without departing from the spirit or scope of the invention as broadly described. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive.
| Number | Date | Country | Kind |
|---|---|---|---|
| PR 8784 | Nov 2001 | AU | national |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/AU02/01522 | 11/8/2002 | WO | 4/9/2004 |
