What is CRISPR?
Clustered regularly interspaced short palindromic repeats (CRISPRs) are a family of DNA sequences found in bacteria. The sequences are derived from the DNA of viruses that have previously infected the bacterium and act as genetic memory. If the bacterium is infected by the same virus again, the CRISPR system will destroy any DNA matching the CRISPR sequence, thus protecting the cell from viral attack.
This leads to the question of how the viral DNA is destroyed? Cas9 (or ‘CRISPR-associated protein 9’) endonuclease cleaves the viral DNA at a specific site. Cas9 is guided to the cleavage site by crRNA which is transcribed from the CRISPR DNA sequence(s) and binds to a complementary strand of viral DNA. tracrRNA, which binds to the crRNA and forms an active complex, is also required. For cleavage to occur a protospacer adjacent motif (PAM), NGG, must be present in the viral DNA. Cas9 endonuclease cleaves the DNA 3 base pairs upstream of the PAM.
The CRISPR-Cas9 system has been utilised as an efficient and highly selective gene editing tool. Genomic modification using the technology can be used to reliably create a targeted break at a specific location. The DNA break can be ‘repaired’ using endogenous non-homology end joining (NHEJ), resulting in gene deletion or silencing, or by homology directed repair (HDR), in which donor DNA can be inserted.
Since its discovery there has been a great deal of activity in the area of CRISPR-Cas9-mediated gene editing. Indeed, the CRISPR-Cas9 gene editing market is very large and is estimated to be worth $10 billion by 2025.
IP Background
Perhaps unsurprisingly given the nature of the technology, a large number of applications have been filed in relation to CRISPR-Cas9 gene editing and there are numerous applicants. For example, over 600 applications are pending before the European Patent Office (EPO). Furthermore, over 40 European patents have been granted and over 20 of these have been opposed (often with multiple opponents).
Some of the applicants include Vilnius University (who filed first but have an invalid priority claim), the Broad Institute/MIT/Harvard (‘Broad’), the University of California/University of Vienna/Charpentier (‘UC’), Sigma and Toolgen. For the purpose of this article, focus will be on Broad and UC and their on-going patent issues in Europe and the US.
It is also worth noting, as will become clear later on, that in 2012 a paper was published by Jinek et al (“A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity”, Science, 337 (6096): 816–821) which described for the first time the use of a ‘single guide’ RNA, in which the crRNA and tracrRNA are fused together (i.e. as single strand of RNA), in prokaryotes. Other authors notably included are Dounda from UC and Charpentier.
European Patent Office
Broad
Although UC had the earliest priority claim, Broad adopted an aggressive filing strategy and used expedited prosecution such that their patent, EP 2,771,468 (‘EP’468’), was granted first.
EP‘468 was granted with claims directed to CRISPR-Cas9 gene editing using single guide RNA in eukaryotic cells and was opposed by nine parties. The patent was revoked by the Opposition Division and the decision was upheld on appeal; but why?
The issue came down to priority (and whether novelty destroying prior art was therefore citable). Broad’s first and second priority documents were filed as US provisional applications in the name of eight inventors, including Marraffini from the Rockefeller University. However, neither Marraffini nor his successor in title, i.e. Rockefeller University, were named on the priority-claiming PCT application.
Priority is dealt with under Article 87(1) EPC, which states that:
“Any person who as duly filed, in or for any state party to the Paris Convention… an application… or his successor in title shall enjoy, for the purpose of filing a European patent application in respect of the same invention, a right of priority during a twelve month period from the date of filing of the first application”
There is also a large body of case law relating to invalid priority claims, including T788/45 which states that:
“the priority right belongs simultaneously and jointly to the two applicants, who thus constitute a legal unity unless one of them decides to transfer his right to the other applicant, who then becomes his successor in title and this is before the filing date of the application”
Therefore, it would appear that all of the applicants named on the priority document should have been named on the PCT application in order for the priority claim to be valid. This is because the priority right belongs jointly to all the original applicants based on the established case law.
Although arguably facing an uphill, Broad appealed the first instance decision of the Opposition Division on three grounds:
- the EPO should have no power to asses legal entitlement to the right of priority;
- in case of joint/multiple applicants in a first application, the meaning of the term ‘any person’ under Article 87 EPC should be interpreted to mean ‘one or some indiscriminately’ of the co-applicants; and
- the meaning of ‘any person who has duly filed’ should be interpreted according to national law, in this case US law.
It came as a surprise to the opponents that the Board of Appeal considered referring questions related to Broad’s arguments to the Enlarged Board of Appeal. However, after hearing further arguments from both sides, the Board ultimately decided against a referral and were able to come to a decision on lack of entitlement to priority. As a result, the earliest priority claims were found to be invalid, the claims were found to lack novelty and the appeal was dismissed.
As such, EP‘468, stands revoked. However, there are numerous other patents and applications owned by Broad related to various aspects of CRISPR technology. Therefore, the story is not over yet.
University of California
UC’s patent, EP 2,800,811 (EP’811), was granted with claims to CRISPR-Cas9 gene editing using single guide RNA. However, in contrast to the Broad patent, the claims were not restricted to a particular cell type (and, as such, covered CRISPR-Cas9 mediated gene editing in all cells including both prokaryotes and eukaryotes). The patent was opposed by seven parties.
In terms of the arguments, several issues were discussed.
Firstly, the issue of priority was raised. The opponents argued that the earliest priority document did not render the invention as claimed plausible because the protospacer adjacent motif (PAM) was not disclosed. In fact, PAM was only added in the second priority document. It was also argued that use of CRISPR-Cas9 gene editing in eukaryotic cells was not enabled in the first priority document.
The Opposition Division decided that the priority claim was valid on the basis that the requirement for PAM is well known based on the common general knowledge (CGK) of the skilled person. Indeed, the Opposition Division concluded that PAM was implicitly disclosed, stating that “[the first priority document] implicitly discloses a PAM consistent with its role according to the CGK”. It was also decided that the earliest priority document enabled Cas9-mediated and sgRNA-directed cleavage of DNA in both prokaryotic and eukaryotic cells, despite examples only being provided in prokaryotic systems.
Secondly, the claims were found to be novel and inventive. On the issue of inventive step, the Opposition Division decided that the closest prior art was in the field of transcription activator-like effector nucleases (TALENs), i.e. general gene editing, and not CRISPR. Initially this may appear surprising. However, the EPO Guidelines require that in determining the closest prior art, the first consideration is that it should aim at a similar purpose or effect (GL G-VII 5.1). Taking this into account, the Opposition Division stated that “the invention underlying the opposed patent is concerned with a site specific cleavage of DNA using genome editing…the other documents deal with different aspects of the CRISPR system”. Therefore, the position appears consistent with the Guidelines in that it is the purpose of the claimed subject matter, and not necessarily the detail of the claims, that determines the closest prior art.
Thirdly, sufficiency was discussed. The opponents argued that no examples were provided in eukaryotic cells. However, the Opposition Division found the patent to be sufficient, stating that “the examples render the achievement of said DNA cleavage in eukaryotes at least plausible. The numerous post-published documents on file that confirm this cleavage in eukaryotes back-up this finding”. The post-published evidence referred to by the Opposition Division showed that CRISPR-Cas9 gene editing could be used successfully in eukaryotic cells.
In conclusion, the Opposition Division decided to maintain the patent. However, it was maintained in amended form. The claims were amended to include the feature that “said method is not a method of modifying the germ line identity of a human being” in view of the exceptions to patentability set out in Article 53(a) and Rule 28(b).
Whilst this is a victory for UC, the decision has subsequently been appealed. Therefore, we may not have heard the end of the dispute yet.
US Interference Action
A first interference action was brought before the United States Patent and Trademark Office (USPTO) in respect of a UC application and 11 patents/applications owned by Broad. The claims are analogous to those discussed in Europe, with UC having claims to the use of CRISPR-Cas9 in all cells and Broad’s claims being limited to its use in eukaryotic cells. Having the earlier filing date, UC were the senior party.
The Patent Trail and Appeal Board (PTAB) of the USPTO considered whether the claims of the two parties were patentably distinct. The question asked was: if one took the subject matter of the claims of the UC application, would it anticipate or render obvious Broad’s claims? The PTAB found that “one of ordinary skill in the art would not have reasonably expected a CRISPR-Cas9 system to be successful in eukaryotes”. This was partly due to the publication of Jinek et al. Thus, the PTAB found in Broad’s favour concluding that there was no interference in fact.
The decision was appealed to the Courts of Appeal of the Federal Circuit (CAFC); where it was upheld.
Interestingly, both the PTAB and CAFC gave significant weight to contemporaneous statements made by inventors from UC after the earliest priority date. For example, in a publication of January 2013, which referenced Jinek et al, it was stated that “it was not known whether such bacterial systems would function in eukaryotic cells”. Also, in 2014 an article was published in Catalyst Magazine where is was stated that “our 2012 paper [Jinek et al] was a big success but there was a problem. We weren’t sure if CRISPR-Cas9 would work in eukaryotes”. Taking these statements into account, the CAFC came to the decision that “if the inventors themselves were uncertain, it seems that ordinary skilled artisans would have been even more uncertain”.
So where are we now? The USPTO has issued a second interference action. The issue has arisen because UC amended their claims in a series of continuation applications to be restricted to eukaryotic systems. As such, there appears to be a clearer overlap between the claims of the two parties. In response, UC have argued that Broad are not entitled to priority. If this is found to be the case, it would mean that that the Broad applications fall under the first to file, rather than the first to invent, system (hence, there would be no requirement for an interference action at all). Their second argument is that Broad lied about inventorship. This is based on a statement made by Broad during proceedings at the EPO that one of the inventors, Cong, “contributed to the invention of CRISPR-Cas9 in eukaryotes”. However, Cong is only mentioned on 8 out of the 14 applications. Lastly, UC have accused Broad of inequitable conduct.
As for Broad, they have argued that UC should be estopped because the issue has already been decided in the first interference action. It will be interesting to see if this argument is successful given that the facts of the case appear to have changed compared to the first interference action.
A further complication is that both Sigma and Toolgen petitioned the USPTO to ask for declarations of interference in respect of their own applications. Each of these parties have filed applications directed to the use of CRISPR-Cas9 in eukaryotes. However, these have been refused by the USPTO over the original UC application. This appears to be in direct contradiction to the rulings of the PTAB and CAFC, which found that the UC application was not anticipatory for eukaryotic cells.
As with the situation in Europe, there is no conclusion to the matter yet.
Summary and Conclusions
It appears that Broad have the upper hand at the USPTO, having won the first interference action. However, the reverse appears to be the case in Europe, where Broad’s EP’468 patent has been revoked. In contrast, UC’s EP’811 patent was maintained with only relatively minor amendments at first instance. However, as already mentioned, each of the parties, as well as others, have numerous other patents in opposition proceedings as well as applications going through prosecution at the EPO and, of course, the decision of the Opposition Division in respect of EP’811 has been appealed. Further developments are also to be expected in the US.
It is worth mentioning that there appears to be inconsistency between the USPTO and EPO, with the USPTO finding that UC’s claims, which are not limited to use of CRIPSR-Cas9 in a particular cell, are patentably distinct from Broad’s claims to the use of CRISPR-Cas9 in eukaryotic cells and the EPO finding that UC’s patent is ‘plausible’ for all cells. Indeed, the EPO have acknowledged this, with the Opposition Division stating that “whilst the question in said proceedings related to obviousness (under US law!) of carrying out the method in eukaryotic cells, the question in present proceedings in whether it is plausible that said method can be carried out”. It would be interesting to see if the decisions are more in line with one another if the validity of the claims are put to the test in the US.
So, are there lessons that can be learned?
Firstly, it is important that the Applicant has a clear right to priority at the filing date of the PCT (or other priority-claiming application). Ownership and/or entitlement can be perfected at a later date, but the right to priority exists at the date of filing the priority-claiming application.
Secondly, care must be taken regarding inventorship. This is particularly the case if the application is being filed in the name of the inventor(s), as is common practice with US provisional applications. Accordingly, any necessary assignments should be executed in the priority year.
Lastly, care must be taken when making public statements even after the priority date! Statements questioning the viability of CRISPR-Cas9 gene editing in eukaryotic cells made by UC inventors after the earliest priority date were pivotal to the decision of both the PATB and CAFC in the first US interference action. It is perhaps counterintuitive to take care what you say after an application is filed. However, in this case, statements were made that clearly discussed what was not present in the application (i.e. evidence that CRISPR-Cas9 would work in eukaryotes). As such, and perhaps more intuitively, statements as to what is not shown or not disclosed in the application as filed should be avoided.
All in all, it is clear that we have not yet heard the end of the CRISPR patent saga.
