Are the Broad Institute’s CRISPR patents too broad?

The ability of the CRISPR/Cas9 system to make precisely targeted mutations in genes, including human genes, has taken the biotech industry by storm. It has been described by some as being the most important development in biology since the discovery of the structure of DNA by Watson and Crick in 1953.

But whereas Watson and Crick famously celebrated their discovery in the traditional way in The Eagle pub in Cambridge, England, today’s inventors reach first for the telephone to call their patent attorneys.

So it was that Feng Zhang, a professor at the Broad Institute, and his group filed a number of US provisional applications to the CRISPR/Cas9 technology in 2012–2013 after realising the value of that technology for genome editing.

Subsequently, in May 2014, a European patent application (EP 2771468 A)was filed; a number of European patent divisional applications based on that first application have also been filed and four patents have now been granted by the European Patent Office (EPO) (see Table 1).

Table 1: Broad Institute’s granted CRISPR patents

The language used in the main claims of all of these patent applications is very similar. One example of this—claim 1 from EP 2771468 B1—is given in Table 2. In Table 2, the individual features have been separated; each will be discussed in turn below.

Table 2: Claim 1 of EP 2771468 B1

Feature A

The first feature of claim 1 is one of the most obscure: “A non-naturally occurring or engineered composition ...”. These terms are defined in the description as indicating “the involvement of the hand of man”. The terms are also said to mean “that the nucleic acid molecule or the polypeptide is at least substantially free from at least one other component with which they are naturally associated in nature and as found in nature.”

“The broader the claims are, the easier they are to attack. For example, broad claims might inadvertently encompass accidental prior art or they might be open to enablement objections.”

However, the use of these terms in claim 1 places an extraordinary burden on the reader to check whether or not a particular composition exists anywhere in nature or has been “engineered” in some way (compared to what?). While such language is not uncommon in US patents, it is rarely allowed in European patents.

Feature B

The expansion of the CRISPR acronym in feature B seems to have been added over the original claims in order to ensure that there is no confusion over the meaning of term CRISPR. While this creates some rather tortuous wording, the aim appears to be reasonable given the fact that the technology is still relatively new. But strangely, the terms “tracr mate” and “tracrRNA” in features D and E are left entirely undefined. If the basic term CRISPR needed clarifying, then surely the terms “tracr mate” and “tracrRNA” should also have been clarified?

Stripping out the brackets and repetitions from feature B, it can be seen that claim 1 relates essentially to a composition comprising a RNA polynucleotide. The reference to “Cas” appears only to help to define the term “system”; a Cas protein does not have to be present in the composition.

This RNA polynucleotide has three components: features C, D and E.

Feature C

The first component is a “guide sequence” which must be “capable of hybridising to a target sequence in eukaryotic cell”. The sequence of this component is left entirely undefined; the type of eukaryotic cell is undefined (must the reader test their RNA sequence against the entire database of eukaryotic sequences?); and the stringency of hybridisation is undefined.

Ordinarily, EPO examiners require all references to “hybridisation” to be clarified by reference to some defined stringency conditions. Why not here? Under appropriate low stringency conditions, essentially any piece of RNA can hybridise to any other. Therefore term “guide sequence” covers essentially any RNA sequence—but it must be 10 to 30 nucleotides in length.

Features D and E

Component (b) relates to “a tracr mate sequence”; component (c) relates to a tracrRNA sequence. These are defined in the description as follows:

“In general, a tracr mate sequence includes any sequence that has sufficient complementarity with a tracr sequence to promote one or more of: (1) excision of a guide sequence flanked by tracr mate sequences in a cell containing the corresponding tracr sequence; and (2) formation of a CRISPR complex at a target sequence, wherein the CRISPR complex comprises the tracr mate sequence hybridised to the tracr sequence.”

The above—rather convoluted—paragraph seems to say little more than the tracr mate sequence must be capable of hybridising to the tracrRNA sequence; no details are given as to how the tracr mate sequence achieves (on its own? without Cas9?) the excision of a guide sequence. The hybridisation of tracr mate and tracrRNA is possibly also required in feature G, but it gives the reader no further clues as to the sequences of the tracr mate or the tracrRNA.

Feature F

This requires the guide RNA, tracr mate and tracrRNA sequences to be “arranged in a 5’-to-3’ orientation”. But it does not require them to be near one another. This feature possibly covers embodiments wherein 10,000 nucleotides of extraneous RNA is inserted between each of the guide RNA, the tracr mate and tracrRNA sequences.

Feature G

This feature is also curious: “wherein when transcribed, the tracr mate sequence hybridises to the tracrRNA sequence”. It must be remembered that claim 1 is directed to a product—not a method—and hence this feature is arguably out of place in a product claim. Is this feature implying that the tracr mate sequence must be transcribed from a DNA template? What if the RNA is made synthetically? If it is not transcribed, does the tracr mate not need to be capable of hybridising to the tracrRNA sequence?

Feature H

The “guide sequence directs sequence-specific binding of a CRISPR complex to the target sequence”. This is the first reference in the claim to a CRISPR complex. The final part of claim 1 defines this CRISPR complex as comprising a Type II Cas9 protein complexed with the guide sequence (which itself is hybridised to the target sequence) and the tracr mate sequence (which is hybridised to the tracrRNA). It is assumed that the phrase “the guide sequence directs sequence-specific binding of a CRISPR complex to the target sequence” should be interpreted as “the guide sequence is capable of directing, under appropriate conditions, sequence-specific binding of a CRISPR complex to the target sequence”, because the claim relates to a product (ie, an RNA sequence) and not to a method (the drafter appears to have lost sight of this latter point).

Remembering that the claim is for an RNA product, the last two “wherein” clauses in claim 1 must be read as functional limitations on the claimed RNA. That is, the claimed RNA must be one which is capable of directing a Type II Cas9 protein which is complexed with the RNA to a target in a eukaryotic cell.

Hence the claim places the entire onus on the reader to determine whether or not any specific length of RNA is capable of interacting with any Type II Cas9 protein in such a way that the RNA is capable of directing the Cas9 protein to a target sequence. No clues whatsoever are given in the claim regarding the specific sequence of any of the RNA components.

So it seems that the Broad Institute has obtained the grant of a claim of extraordinary scope: a first piece of undefined RNA which is attached to two other pieces of undefined RNA which might or might not hybridise to each other, wherein the three pieces of RNA are capable of directing a Type II Cas9 protein to an undefined target sequence in an undefined eukaryotic cell.

At first sight, congratulations to the Broad Institute might seem in order—after all, broad claims are what every applicant seeks. But the broader the claims are, the easier they are to attack. For example, broad claims might inadvertently encompass accidental prior art or they might be open to enablement objections.

In this case, lack of clarity appears to be a key issue. Although this is not one of the grounds for EPO oppositions, lack of clarity can be smuggled into opposition proceedings under the guise of lack of enablement or obviousness.

All four of the Broad Institute’s granted European patents are currently in the nine-month post-grant opposition period and given the media attention that these patents have attracted, it seems highly likely that one or more of them will be opposed. The opposition period on the first of the granted European patents will expire on November 11, 2015, so we should not have to wait too long to find out.

Philip Webber is a partner at  Dehns Patent & Trademark Attorneys and a member of the UK Chartered Institute of Patent Attorneys’ Life Sciences Committee. He can be contacted at: pwebber@dehns.com