Dr. Chris Mason Responds to Blog Posts on Genome Medicine Article -- Updated

[Ed. Patent Docs recently reported on an article in the current issue of Genome Medicine which contends that due to the non-specificity of sequence uniqueness across the genome and the broad scope of claims to nucleotide sequences, the Supreme Court and Congress should limit the patenting of existing nucleotide sequences (see "Genome Medicine Article Calls for Limits on Patenting of Existing Nucleotide Sequences" and "Revisiting Genome Medicine Article on 'Pervasive Sequence Patents' That 'Cover the Entire Human Genome'".  The article was authored by Jeffrey Rosenfeld of the University of Medicine & Dentistry of New Jersey and Christopher Mason of Cornell University.  Dr. Mason (below) has provided Patent Docs with a response to our posts -- as well as a response to an article posted by Prof. Chris Holman on Holman's Biotech IP Blog (see "A Critique of a Recent Article Which Found That Sequence Patents Cover the Entire Human Genome").  To promote further discussion of the issue of "gene" patenting, we are posting Dr. Mason's response today.  As always, we invite Patent Docs readers to join in the discussion by submitting comments to any of the above posts.  Our reply to Dr. Mason's post will follow tomorrow.]

Several recent posts and commentaries have suggested that the analysis from our recent paper, "Pervasive Sequence Patents Cover the Entire Human Genome," was mistaken or did not embody an accurate characterization of the claims' construction.  I would like to comment on several aspects of the posts from both Holman's Biotech IP Blog and the Patent Docs blog, but foremost I would like to thank both sites for their dialogue and debate on this important issue.

Part One: The Holman Critique:  "A Critique of a Recent Article Which Found That Sequence Patents Cover the Entire Human Genome"

1)  In his post, Prof. Holman states "the authors seem to assume that every patent with a claim mentioning a gene sequence also claims every 15mer present in the sequence, i.e., every contiguous 15 nucleotide sequence appearing in the gene."

We do not.  We performed two distinct analyses that were described in the main text.  First, we examined the uniqueness of 15mers in general, which was shown to be exceedingly non-unique genome-wide (no gene is unique at the 15mer level).  Secondly, we used patents that claimed 15mer sequences in their construction, and we indicated the matches we could find given their sequence composition.  From these, we found many exact matches, ranging from 4% (for BRCA1) to potentially as high 91%.

2)  Prof. Holman continues, "In my experience, claims of this type are extremely rare.  I looked at [a] hundred patents identified as gene patents in the Jensen Murray study and found that most only claim the full-length gene sequence.  . . .  I looked through hundreds of gene patents trying to find another 15mer claim analogous to those in the Myriad patents and could not find one.  The patent claims at issue in the Myriad case will be expiring within the next few years.  . . .  I doubt that this sort of broad 15mer claim has been issued by the patent office in recent years, or if it has it seems to be extremely rare."

After searching for a short time on Google Patents, I was able to find two potential examples of recently published patent applications that directly claim a large subset of genes.

US 20130030040 A1

Methods and compositions for increasing sialic acid production and treating sialic related disease conditions

14. An isolated nucleic acid molecule the sequence set forth in SEQ ID NO:1 or SEQ ID NO:2.

"In some embodiments, the therapeutic product is a polynucleotide, while in other embodiments, the therapeutic product is a polypeptide.  In some embodiments, the polynucleotide is a DNA molecule, which can comprise the full-length coding region for a protein, the coding region for a domain of a protein, or a coding region for a protein fragment, which is shorter than a recognized and identified domain of a protein.  Thus, the polynucleotides disclosed herein can range from oligomers of at least 15 base pairs in length to DNA molecule comprising the full-length coding region for a protein."

This patent application refers to the gene "GNE," which allows it to also claim 15mers.  Given the thousands of nucleotides in this patent application, it turns out that this patent application can cover 1,306 other genes, using the same 15-mer matching algorithm.

US 20130041209 A1

Methods and compositions for improved fertilization and embryonic survival

"An isolated nucleic acid molecule comprising a nucleotide corresponding to a nucleotide at a first polymorphic position selected from the group consisting of positions 85146, 85161, 85216, 85292, and 85300 of the nucleic acid sequence shown in FIG. 1 (SEQ ID NO: 1), and at least 10 contiguous nucleotides of SEQ ID NO: 1 adjacent to the first polymorphic position, wherein position 85146 is guanine, position 85161 is guanine, position 85216 is adenosine, position 85292 is cytosine, or position 85300 is guanine; or an isolated nucleic acid molecule comprising a nucleotide corresponding to a nucleotide at a second polymorphic position selected from the group consisting of positions 35728, 36016, and 38867 of the nucleic acid sequence shown in FIG. 2 (SEQ ID NO: 2), and at least 10 contiguous nucleotides of SEQ ID NO: 2 adjacent to the second polymorphic position, wherein position 35728 is guanine, position 36016 is guanine, or position 38867 is guanine.

2. A nucleic acid molecule according to claim 1, which comprises at least 10, at least 11, at least 12, at least 13, at least 14, or at least 15 contiguous bases of SEQ ID NO: 1 adjacent to the first polymorphic position, or of SEQ ID NO: 2 adjacent to second polymorphic position.

7. A nucleic acid molecule according to claim 1, wherein the first or second polymorphic site is within 4 nucleotides of the center of the nucleic acid molecule.

8. A nucleic acid molecule according to claim 7, wherein the first or second polymorphic site is at the center of the nucleic acid molecule."

Notably, these are highly polymorphic sites in the human genome, so if you look at any piece of DNA anywhere near this gene, you will easily match the specified variants from this gene, and then be entitled to any genes that are on the same fragment.  From simple pipetting, this can be anywhere from 10,000-100,000 bases.

3)  Prof. Holman then says, "DNA does not contain 2'-deoxy-2-fluoro pyrimidine nucleotides and 2'-deoxy purine nucleotides, these are synthetic analogues to the nucleotides that appear in DNA."

This is completely false.  "2'-deoxy purine nucleotides" are simply the "A" and "G" nucleotides in normal DNA, and they indeed appear in normal DNA.  Modifications of these bases are allowed in this patent for "chemical variations" of RNA, even though (in this case) they just turn RNA bases into DNA bases.

4)  Prof. Holman continues, "the authors reported that US7468248 contains 'explicit claims for 15mers that matched 84% of human genes.'  In fact, the '248 patent has only two independent claims, both of them method claims."

We agree that these are method claims, but their specifications allow for a broad interpretation.  In particular, from the patent:

"In one embodiment, the present invention provides an isolated polynucleotide that includes at least 20 contiguous nucleotides of any one of SEQ ID NOS:24493 to 64886, a polynucleotide at least 90% identical to the 20 contiguous nucleotide fragment, or a complement thereof, wherein the isolated polynucleotide includes a nucleotide occurrence of a single nucleotide polymorphism (SNP) associated with a trait, wherein the SNP corresponds to position 300 of SEQ ID NOS:19473 to 21982."

5)  Prof. Holman then says, "the publication of the article highlights the limitations of peer review (assuming Genome Medicine engages in peer review)."

The publication of contentious research or discordant viewpoints does not demonstrate a failure of the entire peer-review system.  Rather, discussion and debate are key drivers of scientific progress through these peer-reviewed publications.  Five scientific reviewers and two patent attorneys reviewed and approved our article before its publication, and the site of the Journal clearly indicates that it conducts a thorough peer review.  We also note that the journal is open-access, allowing for easier dissemination of data and results, and that the Journal hosts a large set of leading researchers on its Editorial Board.

Part Two: The Patent Docs blog post:  "Revisiting Genome Medicine Article on 'Pervasive Sequence Patents' That 'Cover the Entire Human Genome'"

1)  Patent Docs states that "the lone independent claim of the '422 patent is directed to a "chemically modified" double-stranded nucleic acid molecule.  It is therefore difficult to see how the claimed sequences of the '422 patent could "match[] 91.5% of human genes."

It is notable that one of the claimed chemical modifications allowed from their claims is "2′-deoxyribonucleotides," which is the same thing as dexoyribonucleotides, or DNA ("Ribonucleotides versus Deoxyribonucleotides").

Also, here is a listing of six other potential embodiments (of many) from this patent that could easily cover unmodified bases or fragments thereof.  The claims must be interpreted in plain language first, of course, but also in light of their specifications of the patent, which include:

1.1)  In one embodiment, the invention features one or more chemically modified siNA constructs having specificity for HIF1 expressing nucleic acid molecules, such as RNA encoding a HIF1 protein.  In one embodiment, the invention features a RNA based siNA molecule (e.g., an siNA comprising 2′-OH nucleotides) having specificity for HIF1 expressing nucleic acid molecules that includes one or more chemical modifications described herein.

1.2)  Non-limiting examples of such chemical modifications include without limitation phosphorothioate internucleotide linkages, 2′-deoxyribonucleotides, 2′-O-methyl ribonucleotides, 2′-deoxy-2′-fluoro ribonucleotides, "universal base" nucleotides, "acyclic" nucleotides, 5-C-methyl nucleotides, and terminal glyceryl and/or inverted deoxy abasic residue incorporation.

1.3)  In another embodiment, the invention features a double-stranded short interfering nucleic acid (siNA) molecule that down-regulates expression of a HIF1 gene comprising an antisense region, wherein the antisense region comprises a nucleotide sequence that is complementary to a nucleotide sequence of the HIF1 gene or a portion thereof, and a sense region, wherein the sense region comprises a nucleotide sequence substantially similar to the nucleotide sequence of the HIF1 gene or a portion thereof.  In one embodiment, the antisense region and the sense region independently comprise about 15 to about 30 (e.g. about 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, or 30) nucleotides, wherein the antisense region comprises about 15 to about 30 (e.g. about 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, or 30) nucleotides that are complementary to nucleotides of the sense region.

1.4)  In one embodiment, an siNA molecule of the invention comprises no ribonucleotides.  In another embodiment, an siNA molecule of the invention comprises ribonucleotides.

1.5)  In one embodiment, the invention features a chemically synthesized double-stranded RNA molecule that directs cleavage of a HIF1 RNA via RNA interference, wherein each strand of said RNA molecule is about 15 to about 30 nucleotides in length; one strand of the RNA molecule comprises nucleotide sequence having sufficient complementarity to the HIF1 RNA for the RNA molecule to direct cleavage of the HIF1 RNA via RNA interference; and wherein at least one strand of the RNA molecule optionally comprises one or more chemically modified nucleotides described herein, such as without limitation deoxynucleotides, 2′-O-methyl nucleotides, 2′-deoxy-2′-fluoro nucleotides, 2′-O-methoxyethyl nucleotides etc.

1.6)  In any of the above-described embodiments of a double-stranded short interfering nucleic acid (siNA) molecule that inhibits expression of a HIF1 gene, wherein a majority of the pyrimidine nucleotides present in the double-stranded siNA molecule comprises a sugar modification, each of the two strands of the siNA molecule can comprise about 15 to about 30 or more (e.g., about 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, or 30 or more) nucleotides.

2)  Even if you disagree with our analysis of the '422 patent, you can look at patent 8,273,866, which is in the same family as 7,795,422.

Notably, Claim 1 says:

"1.  A short interfering RNA (siRNA) molecule having a sense strand and an antisense strand that mediates RNA interference, wherein:
    (a) each strand is between 18 and 24 nucleotides in length;
    (b) the sense strand comprises 10 or more 2′-deoxy, 2′-O-methyl, 2′-deoxy-2′-fluoro, or universal base modified nucleotides, and a terminal cap molecule at the 3′-end, the 5′-end, or both 3′ and 5′-ends of the sense strand;
    (c) the antisense strand comprises 10 or more 2′-deoxy, 2′-O-methyl, 2′-deoxy-2′-fluoro, or universal base modified nucleotides; and
    (d) 10 or more pyrimidine nucleotides of the sense and antisense strand are 2′-deoxy, 2′-O-methyl or 2′-deoxy-2′-fluoro nucleotides."

The patent also specifies considerable flexibility for the terminal cap molecule, stating "[t]he cap moiety can be an inverted deoxy abasic moiety, an inverted deoxy thymidine moiety, or a thymidine moiety."

Again, given that "2'deoxy nucleotides" and "a thymidine moiety" are simply normal DNA bases, and that any sequence that is 18-24 bases can be constructed from these claims, this may mean that every single 18-24mer is potentially claimed by this patent.

3)  Patent Docs then says, "while the '248 patent recites oligonucleotides -- which the specification states "[i]n certain aspects" can be "at least 15 nucleotides in length" -- it is difficult to see how the authors can conclude that the '248 patent has "explicit claims for 15mers that matched 84% of human genes."

See above for our further analysis on this point.  Our main point is that 15mers from these patents match both human and bovine genomes, and that claim construction becomes non-specific (even across species) at these low k-mer sizes.

4)  "In paragraph 10 of his Declaration, Dr. Mason surprisingly states that "Claim #I and #2 of '282 are so broad that they can include up to 100% of the genes in the human genome."

I made this point for two reasons:

4.1)  The 55% homology is a very low threshold, and it allows matches to many other genes, because "homology" and "identity" are not the same thing in biology; homology allows for far more flexibility in the mismatches.

4.2)  The Myriad patents claim any "isolated DNA having at least 15 nucleotides of the DNA of claim 2," which was my focus.  Notably, this is not the same as claiming 15 contiguous nucleotides.  Most other gene patents I have found will describe contiguous nucleotides, which obviously limits their scope.  But, if we allow any 15 nucleotides, in any order, it is easy to match these sequences to the entire genome and indeed every gene.

5)  "As with the '422 and '248 patents, Dr. Mason appears to be having some difficulty ascertaining the subject matter that is actually encompassed by claims 1 and 2 of the '282 patent."

I would agree, insofar as I am worried about the likely overly broad scope of these claims.  Since I risk liability whenever I perform genetic testing on my own DNA, or the DNA from any of my patients, I welcome the Supreme Court or the legal scholars to clarify the issue.  In the absence of a statutory research exemption for infringement liability or some other guarantee, I am restricted from researching thousands of genes for many years to come.

But, if any lawyer is confident enough about the irrelevance of these patents that he or she would be willing to state, in writing, the willingness to defend me in Court and pay all legal fees or damages if I get sued, then I will happily join you and start working again to develop new tests, tools, and algorithms to ameliorate and eliminate human diseases.

Thank you,

Christopher Mason, Ph.D.

[Ed. At Dr. Mason's request, his response has been revised to indicate that US 20130030040 A1 and US 20130041209 A1 are published patent applications rather than patents as Dr. Mason stated in the original version of his response.]

Comments

[How to put this?]

I'm not trying to be rude, but could somebody please explain to Dr. Mason the difference between a patent application publication and an issued patent? And also how claim construction works? It's unfortunate that authors don't try harder to understand their subject matter before they publish.

[justcurious]

Kevin and Don,

I am waiting to see what you say about Dr. Mason's response. You always have great opinions.

[Stop it man, you're killin' me!]

"[Ed. At Dr. Mason's request, his response has been revised to indicate that US 20130030040 A1 and US 20130041209 A1 are published patent applications rather than patents as Dr. Mason stated in the original version of his response."

LOL. I only saw this post after the revision, so thanks for including that editorial note, which pretty much says it all.

If the claims of these two patent APPLICATIONS make it out of the USPTO, then Dr. Mason can cite to them in support of his contentions. Until then...

[Moocow]

This is a awkward. While much could be said, just a few very limited observations. For example, the claim of the '040 publication: "14. An isolated nucleic acid molecule the sequence set forth in SEQ ID NO:1 or SEQ ID NO:2." is defective in that it doesn't specify whether the isolated DNA comprises, consists of, or consists essentially of the specified SEQ ID. Not the best example, but nonetheless, it's clear that SEQ IDs 1 and 2 are 5.2kb and 6.1kb DNAs,respectively. There is no claim to anything less than a 5,266 bp polynucleotide here, and 15-mers are just not encompassed by this claim.
It doesn't matter if the specification also "discloses" shorter fragments, because "disclosing" something and "claiming" it are very different concepts in patent law, which should not be confused. It is not sound to make broad pronouncements about our impending loss of genomic liberty based on "patents that claimed 15-mers in their construction." A patent that explicitly claims an isolated 5,000 bp DNA doesn't "constructively claim" a 15-mer.
One thing is clear, though. Dr. Mason appears to understand such claims in ways that don't match the conventions used by patent examiners, judges, and patent practitioners. Accordingly, he gives them an almost infinitely broader meaning than a patent examiner (or judge) would. It's true that we're trying to communicate across a broad divide here, but nothing I've seen allows the kinds of conclusions Rosenfeld and Mason are drawing in their paper.

[xtian]

@how to put this? Bump.

For more information on Dr. Mason, See here.
http://physiology.med.cornell.edu/faculty/research_summary.php?id=mason

Does anyone know where he went to law school or what his reg no. is? I couldn't find it on his bio.

Also, I am not a searching wiz, but could someone identify any patent or application that he is listed as an inventor?

[Paul San Quentin]

I can appreciate his effort to engage in a conversation here, but it is just troubling that an amount of his argument regarding "what is claimed" is actually in no way found in any of the issued claims. Regarding any inappropriately broad claims to isolated DNAs in pending patent applications, the prior art will do a fine job of dispensing with them.

And the siRNA claims of the '442 patent - even if you ignore the functional language requiring that the molecule "mediates RNA interference" (which you can't) - the claims are directed to chemically modified RNA, which can in some instances be DNA, but that provides a mixed RNA/DNA molecule. And each strand is 18-24 NTs in length. I'm quite happy and content that no aspect of this claim encompasses any my genes.

At least it'll all be over next week, and sometime in the summer we'll be able to direct our snide comments to the SCt opinion holding that isolated DNA is not patent eligible because, "it's old and natural". I need a drink.

[Karen Canady]

Um, I don't get it. If the claims can really be construed as broadly as Dr. Mason claims (which strikes me as an absurdly twisted interpretation of the claim language), then they claims aren't valid. End of story. Even if the claims he is worried about as encompassing thousands of molecules are actually novel and nonobvious over the relevant prior art, they could be shot down as "overly broad" under section 112.

[Skeptical]

It's missteps like this that only reinforce the notion that those in the sciences without an understanding of law should not try to make a legal position, even one that concerns science.

[Greg Zornetzer]

Let's turn the 15-mer argument on its head, using the BRCA1 Myriad patent as an example.
As has been suggested by Dr Mason, let's assume that there are 15-mer sequences in the BRCA1 coding sequence that are also present in other genes. Let's further assume that the sequencing of some of these other genes is prior art to the BRCA1 discovery and the '282 patent. Could Myriad enforce in court '282 claims 5 or 6 against a 15-mer that could have been derived from previously known sequences? I would think not, since it's well within the standard practice of molecular biology to pull shorter sequences out of a gene sequence.
Thoughts?

Truth be told, I think that '282 claim 16 is much more broad and dangerous. It claims to cover any PCR primer pair that amplifies all or a portion of the BRCA1 gene. It could be construed as covering almost any sequence on that chromosome, including sequences not disclosed in the patent filing. I could imagine a court invalidating this claim if it were challenged.

Fair warning - I'm a scientist with a small amount of understanding of the law. Criticism is welcome.