S Basics for biosystems of the Cell PATENTING OF PROTEIN STRUCTURES AND PROTEOMICS INVENTIONS IN THE EUROPEAN PATENT OFFICE

Transcription

1 S Basics for biosystems of the Cell PATENTING OF PROTEIN STRUCTURES AND PROTEOMICS INVENTIONS IN THE EUROPEAN PATENT OFFICE Riku Rinta-Jouppi, 44448J Written Course Work Presentation given on 1 st December 2004

2 PATENTING OF PROTEIN STRUCTURES AND PROTEOMICS INVENTIONS 1.1 Tertiary and quarternary structure of proteins 1.2 Patentability of protein structures 1.3 Conclusions 1.4 References 1.1 TERTIARY AND QUARTERNARY STRUCTURE OF PROTEINS Protein function is highly dependent on its tertiary ie three-dimensional structure when folded. The prediction of the tertiary structure of a folded protein from the amino acid sequence remains the major unsolved problem in structural molecular biology. Certain threading methods have been used that assign a known fold to a given sequence by threading the sequence through all known folds but this is a very time consuming exercise even when the most modern computer technology. However, significant progress has been made in solving the inverse problem of predicting which amino acid sequences can have a given three-dimensional fold. i The quarternary structure describes how individual polypeptide sequences attach together to form larger functional complexes. A simple example might be the haemoglobin protein which consists of two alpha-type polypeptide sequence sub-units and two beta-type polypeptide sequence sub-units. Proteins can also form major complexes as structural components in cells. However, the protein complexes at the centre of most intense scientific research effort in the recent years have been functional transmembrane protein complexes which execute some of the most important higher functions in cells. One such complex is the synaptic NMDA-receptor complex consisting of 83 proteins and which modulates the long term potentication (LTP) of neurons which forms the biochemical basis of learning and memory.

3 1.2 PATENTABILITY OF PROTEIN STRUCTURES Because of the unpredictability of the tertiary structure, the patenting of the threedimensional (3-D) structures of proteins and protein complexes cannot be considered a simple extension of the patenting of proteins on the basis of their amino acid sequences and it raises some additional patentability issues. Firstly, at the point the 3-D structure is determined the protein is often already well known and possibly even patented which presents a problem in relation to the novelty of the invention. Secondly, the 3-D structure presented in a computer readable format is simply a presentation of information and unpatentable as such under Article 52(2)(d) European Patent Convention (EPC). Thirdly, the industrial applicability of the 3-D structures is in many cases only speculative, as this area, which is the subject of intense scientific research effort, has not yet produced many commercially feasible applications. This complicates matters in respect of the invention meeting the industrial applicability requirement. In addition to meeting these basic requirements of novelty, inventive step and industrial applicability and providing enabling disclosure of the invention, some matters relating to the technical character and support for the invention ought to be considered. The basis for the requirements is Article 84 1, which has been already discussed. That requirement is reflected, with an added emphasis on the use of technical terminology, in the European Patent Convention Rules (EPC Rules) in Rule 29(1): The claims shall define the matter for which protection is sought in terms of the technical features of the invention. There must also be sufficient support for the claims in the description. The Guidelines of the European Patent Office ii stress that: Support must, however, be of a technical character; vague statements or assertions having no technical content provide no basis [to regard the claims as supported by the definition]. There is no European Patent Office (EPO) case law in relation to the patentability of threedimensional protein structures as such. The patentability of protein structures has only officially been considered in a Trilateral Project WM4: Comparative studies in new technologies (biotechnology, business methods, etc.) Report on comparative study on protein 3-dimensional (3-D) structure related claims dated 4-8 November 2002 (the 1 Article 84 EPC: The claims shall define the matter for which protection is sought. They shall be clear and concise and be supported by the description.

4 Trilateral Report). Eight, carefully constructed, fictional cases relating to the patentability of protein structures were presented for discussion with the representatives of the EPO, the United States Patent and Trademark Office and the Japan Patent Office. Each of the cases represents a particular type of invention addressing different sets of patentability issues that I will comment on here. The first case is about 3-D structural data as such in the form of a computer model. According to the EPO: a computer model is not considered to be a patentable invention; it merely presents the atomic coordinates of a single protein molecule in space as such, without any direct technical character (it is non-technical by not solving a technical [emphasis in the original] problem, and it does not have a technical effect in itself. iii The requirement of a solution to a technical problem is useful in excluding from patentability inventions that are solutions to scientific rather than technical problems. The technology to experimentally determine the three dimensional structure has existed for some time and there are special data banks for the structural information. 2 However, there are still only a few technical problems that can be solved with the use of these structures without further research. The second case deals with a computer-readable storage medium encoded with structural data of a protein. The EPO stated that: such a storage medium is not considered to be a patentable invention; it merely specifies the atomic coordinates of a single molecule in space, without any direct technical character; the data merely encode cognitive content in a standard manner. Thus, the subject-matter referring to the presentation of information is not directed to patent eligible subject matter according to Article 52(2) of the EPC. iv In a similar manner a in the first case the lack of a technical character and effect seems to exclude this invention from patentability. However, the EPO also states in relation to the second case that: It is noted that in a particular circumstance a Technical Board of Appeal has decided (T1173/97, OJ 1999, 609) that the information encoded on a computer-readable storage medium can be patentable, ie can be regarded as an invention which is susceptible of industrial application and which is new and which involves an inventive step. The data encoded on the medium 2 See RCSB Protein Data Bank, a worldwide repository for the processing and distribution of 3-D biological macromolecular structure data which currently deposits 29,733 structures.

5 concerned a computer program considered in this special case to have a technical character by having a further technical effect v The further technical effect is defined in the case in the following way: the program brings about, or is capable of bringing about a technical effect which goes beyond the normal physical interactions between the program (software) and the computer (hardware) on which it is run. vi Thus, for example a computer program which when run establishes a graphical representation of complex three-dimensional structures which can be worked on on the screen with the use of a computer mouse might be a patentable invention by virtue of solving a technical problem if it meets the other patentability requirements. In the third case a claim is made for a protein defined by its tertiary structure, ie structural coordinates, and the biological function of lowering blood pressure when administered. The EPO considered that: the claimed subject-matter may satisfy the requirements of industrial applicability and clarity, enablement, support It is assumed that the description also specifies the purification of the protein necessary to perform NMR [Nuclear Magnetic Resonance imaging] at 0,2nm resolution.the prior art teaches a protein from the same source organism having the same specific function and approximately the same molecular weight. It would be the starting position of the EPO to raise an objection for lack of novelty over the cited prior art document. If the applicant provides evidence for the novelty over the prior art protein, novelty and inventive step can be accepted (it is noted that it is presently considered that the structural data at high resolution fully define the protein, including the deducible primary sequence). vii The stand taken by the EPO here in raising an objection for lack of novelty is not surprising. Yet, it has the effect of limiting the scope of patentable protein structure quite drastically if the amino acid sequence predicted on the basis of bioinformatics data is taken to make the protein known in the state of the art. The human genome has now been sequenced to the extent that covers 99 per cent of the gene containing parts at a 99,999 per cent accuracy locating 22,287 genes. viii This means that there is also a predicted protein amino acid sequence for each of the genes discovered, produced on the basis of the gene sequence information which available through various genome browsers such as ENSEMBL provided by the European Bioinformatics Institute. ix Thus, at least the predicted amino acid sequences of the great majority of human proteins already form a part of the state of the art. As it has been indicated in the EPO case law discussed in Chapter 4 a

6 protein can be considered to have been unambiguously identified even prior to the experimental determination of its amino acid sequence if its other relevant characteristics are known. The publication of the predicted amino acid sequence of a protein may not in itself destroy novelty but supposedly only a few additional characteristics such as the atomic weight and the isoelectric point of the protein would unambiguously identify the protein. The strict interpretation of the novelty requirement here would effectively limit the patentability of the three-dimensional structures of isolated and purified human proteins to second generation ie modified proteins and protein complexes. In contrast to the third case, the EPO took a positive stand in relation to the patentability of the crystalline form of a known protein in the fourth case. The EPO found that: The claimed subject-matter may satisfy the requirements of clarity, enablement, and support (the description gives experimental data how to make the crystals and explains that the protein, when active, lowers blood pressure). The prior art did not disclose crystals of protein P and also did not teach how to obtain crystals of protein P. Although protein P in crystalline form is said to be inactive (but by which assay is it possible to determine that the crystallised protein is inactive? Such an assay appears only to be possible for a protein in solution) it may be advantageous to produce the protein in eg a stable form (and high purity) and also to use the crystals for determination of the 3-D structure, the atomic coordinates useful in in silico screening methods and rational drug design. x Thus, the EPO adopts a positive approach to patenting the crystalline form of a known protein even if the amino acid sequence is known. The adopted position is in line with the statements of Dr Siobhan Yeats, the Director of Biotechnology Patent Examination and Opposition in the EPO, who has been reported to have indicated in previous presentations (prior to the publication of the Trilateral Report) that the crystalization of a known protein can be regarded novel solely due to the fact that the 3-dimensional structure represented by the crystal is novel over the art. xi This is in effect is comparable to ice, when found for the first time, being regarded novel over water despite of the chemical identity as the threedimensional structure of ice is different from that of water. However, the patent would need to be formulated in such a way as to exclude the previously known (or native) form of the compound. The next case, case five, is drafted to address the patentability to the three-dimensional structure of protein binding pockets and domains, ie specified sub-sections, of a known

7 and also an unknown protein. The EPO considers, in line with the stand in relation to case three that if the protein is known that is prejudicial also to the novelty of the binding pocket forming a part of the known protein. On the other hand, if the protein is new, there is no objection in relation to the novelty of a sub-section of it. Case six and seven are related as they present in silico screening methods for identifying compounds that bind to a specific, known protein. The EPO states in relation to case six which is presented without working examples that: An in silico screening method is considered to be a patentable invention under Article 52(3) and 52(3) of the EPC; it refers to a method having a link to a technical contribution by the use of technical data. With respect to the sufficiency of disclosure it is noted that, although the applicant may file as support further technical information with examples providing evidence for the correctness of the prediction of the position of the pocket, the method as claimed (in the absence of eg further information on the candidate compounds) is not fully enabled in the absence of any working examples. xii However, case seven provides such working examples and discloses the previously unknown three-dimensional structure of the protein. Here the EPO considers that: [the screening method cannot be] regarded as a presentation of information or as a pure mathematical method, excluded by Article 52(2)(d) or (a) of the EPC, respectively, but to the use of the structural data. Thus the subject-matter is directed to patent eligible subject-matter The prior art did not disclose or suggest the 3-D coordinates of protein P. The claimed method applying the use of the coordinates is therefore considered to be new, non-obvious and industrial applicable. xiii It is expected that in silico screening methods will be a major area of growth and development in the near future. With ever increasing scientific understanding of the molecular basis of disease, the pharmaceutical industry is able to start to shift resources from the laborious and time consuming in vitro testing in laboratory to in silico screening methods utilising the latest computational technology in the name of rational drug development. This is expected to significantly shorten the time needed to bring a new small pharmaceutical molecule or a biopharmaceutical to the market.

8 Finally, case eight deals with pharmacophores which provide the three-dimensional map of the biological properties of reactive atoms or domains within a protein. 3 The EPO takes a stand in line with the first case that the pharmacophore is simply a presentation of information as: it merely presents the spatial arrangement of some further characterised atoms within a molecule xiv In addition it is considered in the last case whether a new protein can be defined by means of a pharmacophore and claimed as an independent invention. The case description describes that a novel ligand was designed based on the pharmacophore, and shows experimental results that the ligand binds to the protein with relatively high affinity. The EPO sees that: the claimed subject matter concerns a compound and therefore represents patent eligible subject matter. However, it would appear that a lack of sufficiency of disclosure should be raised (article 83 EPC). The application must contain sufficient information to enable the person skilled in the art, using his common general knowledge, to perform the invention over the whole area claimed without undue burden and without needing inventive skill (EPO Guidelines C-II, 4.9). The application is not sufficiently enabled, taking into account the scope of the claim embracing an excessively large number of compounds and the fact that the description refers to one novel ligand designed on the basis of the pharmacophore. A lack of support should also be raised according to Article 84 EPC in the light of the EPO Guidelines C-iii, 6.3: In order to comply with the requirements of Article 84, there must be sufficient support of technical character in the description that allows the applicant to extend the particular teaching of the description to the whole field claimed. In the case of the pharmacophore it is unclear how for the scope extends: the features of the compound concerning the spatial arrangement of the atoms do not support sufficiently the characterisation of the compound (a meaningful comparison with the prior art cannot be made (Guidelines C-III, 4.7a). It is furthermore noted that the natural agonist of protein P will fulfil the structural requirements of the pharmacophore, and therefore would be prejudicial to novelty (and in fact should result also in an objection for lack of unity of invention). xv 3 The Trilateral Report defines pharmacophores in the following way on page 13 of the report: A pharmacophore is a description of a generalized concept of molecular feature in terms of information on spatial arrangement of chemical elements (eg hydrophobic groups, charged/ionizable groups, hydrogen bond donors/acceptors, and substructures) that are considered to be responsible for a desired biological activity.

9 Similarly to in silico screening methods pharmacophores are used in rational drug development. They are very useful in demonstrating the link between structure and function and in the design of suitable ligands for a particular molecule. The pharmacophore style of presenting information is at a higher level of generalisation than the a simple three-dimensional structural map of a molecule. This generalisation of information is in the root of the central patentability problem with pharmacophores, which is that the presented structure does not individualise a particular molecule but in fact applies to a wide group of molecules. Therefore, a pharmacophore might be easier to claim by the use of some process claim formulation.

10 1.3 CONCLUSIONS Some general principles can be drawn from the positions adopted by the EPO in the Trilateral Report in relation to different types of inventions utilising the structural information relating to proteins. First, mere presentations of structural information are not patentable in the absence of a further technical effect. This appears to a matter of simple application of Article 52(2) EPC. The required further technical effect can be achieved by means of a computer program which somehow acts on the structural data, for example, enabling structures to be manipulated, compared with other structures or screened against a great number of ligands to find ligands which bind to the structure with high affinity and specificity. There is little question of the patentability of novel in silico screening methods when working examples are provided and the technical contribution disclosed in the description surpasses the present state of the art. Sophisticated computer models of the interactions of biomolecules in general will be an area of significant growth in the near future because there is great demand for them in the pharmaceutical and biotechnological industry. Secondly, the three-dimensional structure of an isolated and purified protein, which is characterised in the prior art by its primary structure ie amino acid sequence, is not likely to be patentable. However, the crystalization of even a known protein may well be patentable as the EPO appears to take a stand that as the three-dimensional conformation of the crystalized protein is different from the conformation of the protein in the uncrystalized form. Crystalized proteins are useful, because of their high purity, for example in the process of determining the three-dimensional structure of an unknown protein. Thirdly, the presentation of the claims referencing structural data at a level of sufficient specificity is necessary for them to pass the test of enabling disclosure in the EPO. Pharmacophores provide only the bare bones of the molecular structure by displaying only the atoms and domains which are biologically reactive. That generalised structure can cover a vast range of actual molecules and therefore a patent claiming the molecules would be very easily and unwittingly infringed. This is essentially a problem of insufficient support as the pharmacophore does not characterise the claimed compounds at a level of

11 identifying them unambiguously as the surrounding atoms in the molecular structure always have an effect on the biological reactivity of the selected atoms presented. 1.4 REFERENCES i Branden, Tooze (1998), Introduction to Protein Structure, Garland Publishing Inc, p.370 ii Guidelines (2003), Part C p.33 iii Trilateral Web Site (EPO, JPO, USPTO) Trilateral Project WM4. Comparative studies in new technologies (biotechnology, business methods, etc.). Report on comparative study on protein 3-dimensional (3-D) structure related claims. Vienna, Austria. November 4-8, p.34 iv Trilateral Project WM4, p.35 v Trilateral Project WM4, p.35 vi Decision of the Technical Board of Appeal 1 July 1998, T 1173//97, paragraph 13 of the reasons vii Trilateral Project WM4, p.35 viii The Sanger Institute Web Site. Press Releases: 20th October 2004, The finished human genome: the Gold Standard sequence, p.1 ix The Sanger Institute Web Site. Press Releases: 20th October 2004, The finished human genome: the Gold Standard sequence, p.1 x Trilateral Project WM4, p xi Dr. J. Wachenfield (undated), Patenting of Protein Structures, Vossius & Partner, p.9 xii Trilateral Project WM4, p.37 xiii Trilateral Project WM4, p.37 xiv Trilateral Project WM4, p.38 xv Trilateral Project WM4, p.38-39