Patenting proteins: an Australian perspective

Biological molecules, including polypeptides and, in some cases, nucleic acids, can be patented under Australian law.  The level of disclosure that is required to support and enable a claim to a polypeptide will depend on several factors, including:

 

i.    the identity of the person skilled in the art;

ii.   the information disclosed in the specification;

iii.  the function of the polypeptide; and

iv.  the specific sequence being claimed.

 

This article briefly describes how each of these factors contributes to the determination of a claim’s scope under Australian practice.

 

The person skilled in the art

In considering whether a claim is sufficiently enabled, the Australian Patent Office will ask:

i.   whether it is plausible that the invention could be worked across the full scope of the claim; and

ii.  whether the invention can be performed across the full scope of the claim without undue burden.

The answer to both questions will depend on the identity of the hypothetical person skilled in the art.  The skilled person may carry out ordinary methods of trial and error and may use their common general knowledge to supplement the information contained in the patent specification.[1]  Where the subject matter of the invention is inherently complex – protein engineering may be one example – a substantial amount of work may be required, and indeed expected, for a person skilled in the art to perform the claimed invention.  When seeking to claim a large number of polypeptides, a patent applicant will therefore be well served if they can frame the invention as one which lies within a complex technical field, where the “ordinary” skilled person is highly qualified and capable of complex experimental techniques.

The skilled person played a prominent role in Evolva SA [2017] APO 57 (“Evolva”), a case which concerned the use of uridine-5′-diphospho-dependent glycosyltransferases (UGTs) to glycosylate mogrol.  The specification described only five UGTs, but the claims encompassed the use of polypeptides having at least 90% sequence identity to any one of them.  The examiner rejected the claims, alleging that an undue burden would be placed on the skilled person who would have to produce every polypeptide encompassed by the claims and then perform an assay on each to identify those which exhibited the desired activity.  In arriving at the decision, the examiner characterised the skilled person as being “a single synthetic chemist, perhaps a PhD candidate”.

The matter was heard before a Delegate of the Commissioner of Patents who ultimately allowed the application including claims that defined polypeptides having at least 90% sequence identity to any one of the five UGTs described in the specification.  Crucial to the Delegate’s decision was the characterisation of the person skilled in the art.  Contrary to the examiner’s characterisation, the Delegate considered that a team, comprising biotechnologists and synthetic chemists, would be the “person” to whom the specification is addressed.[2]  Needless to say, the team of people described by the Delegate was considered to have a much wider breadth of skills and knowledge at their disposal when it comes to assessing the plausibility of, and to performing, the invention as claimed.

 The information disclosed in the specification

Australian law does not prescribe a level of experimental data that must be disclosed by a patent specification to enable a polypeptide claim of a certain scope.  Of course, an abundance of experimental data will, in many cases, help to enable a claim that encompasses a broad range of polypeptides, particularly when it comes to demonstrating the plausibility of the claimed invention.  However, plausibility can also be demonstrated by a priori reasoning.[3]  In the absence of extensive experimental data, a broadly drafted claim may be enabled where the specification discloses a technical rationale that makes the claimed invention plausible to a person skilled in the art and provides guidance for the skilled person to perform the invention across the full scope of the claim.

To take a simple example, a claim which defines an antibody solely by reference to its target-binding specificity will typically be enabled provided the target is new and the specification sufficiently describes the epitope and has shown the antibodies can be raised against it.  In such cases, the specification need only disclose one such antibody since raising antibodies against the epitope is considered to be a principle that can be generally applied to produce antibodies across the entire scope of the claim.[4]  In each case, the question to ask is whether, based on the disclosure of the specification, including any a priori reasoning provided therein, a person skilled in the art would find the claimed invention plausible and could work the invention across the full scope of the claim.

The function of the polypeptide

A novel polypeptide may be defined in a claim either structurally, such as by its amino acid sequence, or functionally, for example, by its target-binding specificity or its enzymatic activity.  But somewhat paradoxically, a narrower claim which includes both structural and functional limitations can, in some cases, be found insufficiently enabled.

Take, for instance, our antibody example above.  While an antibody defined solely by its target-binding specificity may be fully enabled, further defining the antibody by reference to one or more of its complementarity determining regions (CDRs) will often render the claim insufficiently enabled.  The Patent Office takes the position that the disclosure of the epitope is not a principle of general application that can be applied to generate the antibodies in the claimed subset, to the exclusion of other antibodies not encompassed by the claim.[5]  In performing the claimed invention, a person skilled in the art would have to raise antibodies against the epitope and then perform an additional sequencing step to identify those that fall within the claimed subset.  This, in the Patent Office’s view, would place an undue burden on the person skilled in the art.

There is no general rule, however, prohibiting the inclusion of functional and structural limitations in a claim.  In Evolva, the claims in question defined a method of producing a mogroside compound using a UGT having at least 90% sequence identity to any one of five UGTs described in the specification.  Given what was known about the structure and function of UGTs, including the information disclosed by the specification, the Delegate found that the skilled person would consider it plausible that functional variants to a level of identity of at least 90% could be identified and would be useful in the claimed methods.[6]  On the issue of “undue burden”, the Delegate observed that the specification provided sufficient information about the manner in which UGT variants may be generated and tested.  Although such tests may involve time-consuming experimentation, they did not, it was found, present difficulties that would require the skilled person to undertake any prolonged research or experimentation that would be considered an undue burden in the relevant field of technology.[7]

The decision in Evolva contrasts with the more recent decision in Gary B Cox v MacroGenetics, Inc. [2019] APO 13, where the Patent Office rejected claims directed to deimmunized therapeutic proteins having an extended serum half-life. The therapeutic proteins comprised an albumin-binding domain (ABD) that could bind to serum albumin and thereby extend the half-life of the entire protein.

Claim 1 of the application defined a polypeptide comprising a variant ABD, the variant ABD having an amino acid sequence that differs from wild-type ABD by comprising one of two combinations of amino acid substitutions.  Importantly, the claim also included two functional limitations, namely that the ABD is deimmunized and that it is albumin-binding.

The use of the inclusive term “comprising” in claim 1 allowed for the variant ABD to have any number of mutations, provided it included those specifically recited in the claim. While acknowledging that the level of skill possessed by the relevant person skilled in the art would be high, it was not apparent to the Delegate that the skilled person would appreciate what mutations could be combined with those specified in the claim whilst achieving the functional limitations of the claims.[8]  To the contrary, evidence suggested that combinations of mutations would have an unpredictable effect on albumin-binding and could even introduce immunogenic regions.[9]  The Delegate found that the work required to identify variants having the desired properties, particularly the ability to bind albumin, while systematic and iterative, is unpredictable, with limited guidance provided by the specification.  The effort required to perform the invention across the full scope of the claims was described as being “in the nature of a research project” and therefore constituted an undue burden.[10]

The specific sequence being claimed

Certain polypeptides, and indeed certain regions within a polypeptide, may be more likely than others to tolerate amino acid variation without compromising the functionality of the polypeptide.  Large structural proteins, for instance, may tolerate a greater level of variation compared to smaller antigen-binding proteins or enzymes.

When it comes to antibodies, for example, one might expect that an amino acid change within a CDR would be more likely to affect binding specificity compared with an amino acid change within a framework region. (Of course, amino acid changes within the framework region can still impair the folding within, or the pairing between, the variable domains.)  Given the significant role played by the CDRs in dictating binding specificity, the Patent Office may allow a greater degree of sequence variability within the variable domain than it would within the CDRs themselves.  For claims which define an antibody solely by its heavy and light chain variable regions, the Patent Office has allowed 90% sequence identity.[11]

In some cases, mutations within a polypeptide may have unpredictable effects on the polypeptide’s function.  In Gary B Cox v MacroGenetics, Inc. [2019] APO 13, evidence was adduced showing that mutations would have an unpredictable effect on certain properties of the claimed polypeptide; properties that were included in the claims as functional limitations.[12]  Given the unpredictability in the art, and the lack of guidance provided by the specification, the Delegate found that the work required to perform the invention across the full scope of the claims would constitute an undue burden, and so the claims were found to be insufficiently enabled.[13]

Conclusions

When patenting polypeptides, there is no general rule that dictates the level of sequence variation that may be encompassed by any particular claim.  Certain proteins may be more resilient than others to amino acid changes, or the changes may have predictable consequences on the protein’s function, in which case, a broad range of protein variants may be validly claimed.  A large number of variants may also be claimed in circumstances where the person skilled in the relevant art is highly qualified and capable of complex experimental techniques that could be employed to perform the invention across the full scope of the claims.  Examples in the specification which demonstrate the functionality of protein variants can also help, but so too can a technical rationale which makes the claimed invention plausible and workable across the full scope of the claims.


[1] Eli Lilly & Co v. Human Genome Sciences, Inc [2008] RPC 29, [239]. Evolva SA [2017] APO 57, [24].

[2] Evolva, [54].

[3] Warner-Lambert v. Generics (t/a Mylan) & Actavis [2018] UKSC 56.

[4] The Australian Patent Office Manual of Practice and Procedure, s 2.11A Annex A.

[5] The Australian Patent Office Manual of Practice and Procedure, s 2.11A Annex A.

[6] Evolva, [62].

[7] Evolva, [68].

[8] Gary B Cox v MacroGenetics, Inc. [2019] APO 13, [95].

[9] Gary B Cox v MacroGenetics, Inc. [2019] APO 13, [97]-[99].

[10] Gary B Cox v MacroGenetics, Inc. [2019] APO 13, [107].

[11] See, for example, Australian Patent No. 2015242981.

[12] Gary B Cox v MacroGenetics, Inc. [2019] APO 13, [97]-[99].

[13] Gary B Cox v MacroGenetics, Inc. [2019] APO 13, [107].

笔者: