E-mail Newsletter

  • Enter your e-mail address below to receive the "Patent Docs" e-mail newsletter.

Enter your email address:

Delivered by FeedBurner

Contact the Docs

Docs on Twitter

About the Authors

  • The Authors and Contributors of "Patent Docs" are patent attorneys and agents, many of whom hold doctorates in a diverse array of disciplines.


  • "Patent Docs" does not contain any legal advice whatsoever. This weblog is for informational purposes only, and its publication does not create an attorney-client relationship. In addition, nothing on "Patent Docs" constitutes a solicitation for business. This weblog is intended primarily for other attorneys. Moreover, "Patent Docs" is the personal weblog of the Authors; it is not edited by the Authors' employers or clients and, as such, no part of this weblog may be so attributed. All posts on "Patent Docs" should be double-checked for their accuracy and current applicability.

  • Law Blogs

Become a Fan

« USPTO News: Pilot PPH Program Between Singapore and U.S. | Main | Amendment to Japanese IP Law Creates New After Final Deadlines for Foreign Applicants »

February 01, 2009


Kevin, you wrote, "Thus, the Office imposes a standard of knowing the function of a protein encoded by a nucleic acid, but considers the mere disclosure of the nucleic acid to anticipate not only the nucleic acid but the protein encoded thereby. Under these circumstances, it is likely that no later-filed claims to proteins encoded by open reading frames identified from the Human Genome Project will be patented, even if the protein has a unique, unappreciated biological activity."

The observation of the first sentence, while often vexing, is unsurprising in view of the CAFC's adoption of the CCPA's reasoning that "enablement" in the 102 context is a lower standard than "enablement" under 112 (see e.g. Rasmusson v SKB, citing In re Hafner).

But the second sentence, and the implication thereof that you spell out, if it comes to pass, makes me shudder, because it leads to the wrong result from a policy perspective by being inconsistent with analogous case law in the small molecule area. With small molecules, it's been the case for years that you can claim in a later application a specific molecule that falls within an earlier-disclosed genus, without worrying about 102, as long as the later-claimed molecule wasn't specifically disclosed in the earlier disclosure. The CAFC recently reiterated this point (and how!) in the Sanofi v Apotex decision. Identifying a protein having a particular biological activity, when the DNA sequence encoding that protein is already known, is no different. Like broad molecular genus claims, there's a huge amount of information available now about the human genome, i.e. the DNA seqeuences, but that's the not the same as having made every protein that could be encoded by those sequences, just as the molecular genus claim doesn't anticipate the later species claim on the specific compound.

Let's hope that the CAFC knocks some sense into the BPAI when this one comes up on appeal.

Dear Dan:

Yes, it is consistent with Hafner, which I recall took me a while to wrap my head around when I was first starting in this business.

I think the distinction (which does not work in our favor) over small molecule IP is that here, disclosure of an open reading frame with disclosure of the predicted amino acid sequence of the encoded protein, is enough. Taken from a PTO perspective, this makes sense, because technically the amino acid sequence IS in the prior art (albeit as merely a recitation of the sequence of an unknown protein). The first leap the BPAI took is in deciding that a claim to an isolated polypeptide did not require the art to disclose an isolated polypeptide, but rather an isolated cDNA, because that constituted an enabling disclosure (1) having the cDNA and 2) being able to use it to express the protein (good luck figuring out that the protein does).

The leap the BPAI did not take, but it is just around the corner, is saying that disclosure of a cDNA, without explicit disclosure of an encoded amino acid sequence, is anticipating. The basis for this would be that the skilled worker could "find" an open reading frame and then express the protein. This is more analogous to your small molecule scenario, and here I think there would be some occasion to push back.

Thanks for the comment.

"The leap the BPAI did not take, but it is just around the corner, is saying that disclosure of a cDNA, without explicit disclosure of an encoded amino acid sequence, is anticipating."


Unless I'm missing something here, disclosure of the cDNA doesn't (or shouldn't) anticipate the isolated polypeptide from the encoded amino acid sequence, unless the polypeptide is actually expressed. But you may be staring at a prima facie case of obviousness instead.

Dear EG:

You would be correct, outside the world of the Patent Office. In Chuang, the Board affirmatively stated, relying on the Donohue case, that the prior art did not have to practice the invention to anticipate, just show that the skilled worker would be enabled for practicing the invention.

So I think the logical extension would be that the existence of a cDNA in a database (it would have to be a cDNA, which would show that something was expressed), with or without annotation of an open reading frame (which might otherwise raise an In re Hall problem) might be enough to anticipate, based on the Chuang reasoning that this would be enough to enable the isolated protein.

I suspect one countervailing argument would be that under those circumstances, the skilled worker would not know how to isolate the protein once she had produced it, but I'm sure there is some answer to that argument that the Board could conjure (this point did not come up, it appears, in Chuang).

Thanks for the comment.

Another fact that the BPAI did not consider is that recombinantly expressed polypeptides invariably have different post-translational modifications, which a simple amino acid sequence cannot dictate and thus anticipate. So, "an isolated polypeptide" requires more than just a recited aa sequence.


As I thought, the BPAI mischaracterized Donohue, or at least cited it out of context. Here's the key quote from Donohue:

"Appellant argues that the Fields affidavit, which states that the authors of Nomura did not make the disclosed dicarboxylic acid TMBP and dimethyl ester TMBP compounds, overcomes the PTO's rejection. It is urged that Donohue I and In re Samour, 571 F.2d 559, 197 USPQ 1 (CCPA 1978), require, inter alia, that a 35 U.S.C. Sec. 102(b) rejection based on a primary reference disclosing a claimed compound in conjunction with one or more references which teach how to make that compound, should be sustained only if the claimed compound was actually made. We disagree."

What the above quote says it that a rejection under 102(b) based on a disclosure of the compound doesn't require that the disclosed compound actually be made (actual reduction to practice) if other references would teach how to make the compound. Conversely, if the compound isn't disclosed by the reference, there is no anticipation, even under Donohue. In fact, the Federal Circuit didn't just rely on 102(b) in Donohue (if I read it correctly), but also 103.

This result does not seem to be a departure from longstanding CCPA and CAFC precedent. Anticipation is not judged against what the reference strictly recites. Rather, we judge anticipation against that which the reference places into the possession of the skilled artisan. After all, it is the skilled biologist who reads the reference, not the skilled janitor.

Dear Evan:

And that's the point. I think there is some logic to the Board's decision in Chuang; I think the extension I propose might be an over-extension of the principle. And I do think we need to consider the policy implications.

Thanks for the comment.

What is the argument that a disclosed amino acid sequence does not anticipate a peptide having that amino acid sequence? Put the other way around, I dont think that anyone would require a patentee to make every species explicitly disclosed in his application in order to claim those individual species. If the sequence is disclosed and the peptide enabled, that is good enough to support a claim, and should be good enough to anticipate.

Dear Grimace:

I think your point explains why Chuang lost. But I think an important issue is that there is a disparity between what is required to enable for patenting purposes (i.e., you need a utility, stemming from the "make and use" language of 112) and the anticipation standard here that the mere recitation of an amino acid sequence in the prior art is enough. Now, in the past the existence of an isolated protein in the art was enough since someone had actually isolated the protein. Here, you have the (anomalous) result that the isolation of a cDNA (a different molecule) anticipates a claim for an isolated protein, which the Board admits was never made in the prior art. The link is the genetic code, wherein knowing the nucleotide sequence of the cDNA lets you "know" the amino acid sequence of the protein, and that's fine as far as it goes. But since there is no "isolated protein" in the art, the purpose of 102, to prevent anyone from making private what is already in the public domain, doesn't seem to be properly implicated here.

I think the result in Chuang is not particularly wrong or significant (it is non-precedential), but it is the top of the slippery slope towards having all the sequences in all the databases in all the world considered as "prior art" no matter how "known" the sequences were. As I said upthread, unrecognized open reading frames will probably have an In re Hall problem, and that may have to be the limits to the application of the Chuang doctrine. But precluding patent protection on isolated proteins based on cDNA sequences in databases may inhibit investment in those proteins as therapeutics, and that could be a problem. Or not - we'll just have to see what happens.

Thanks for the comment.

The comments to this entry are closed.

July 2015

Sun Mon Tue Wed Thu Fri Sat
      1 2 3 4
5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21 22 23 24 25
26 27 28 29 30 31