The CRISPR Patent Interference Showdown Is On: How Did We Get Here and What Comes Next?

Jacob S. Sherkow

Without question, this year’s—and potentially, this century’s—biggest biotech story is the rise of CRISPR, arguably the most precise and flexible gene editing technology yet to be created. As others have reported at length, two research teams—Jennifer Doudna’s lab at UC Berkeley, and others, and Feng Zhang’s lab at the Broad Institute and MIT—are engaged in a patent dispute over fundamental aspects of the CRISPR technology. (And to be clear—it’s Doudna’s and Zhang’s attorneys running this show—not the scientists.) On December 21st—after months of waiting—that patent dispute heated up substantially. The primary examiner responsible for Doudna’s patent application, Michelle K. Joike, along with an interference specialist (likely Brandon Fetterolf) issued an Initial Interference Memo, a recommendation that the Patent Trial and Appeals Board (the PTAB) conduct a procedure known as an “interference proceeding” between Doudna’s and Zhang’s patent applications (no PDF currently available). This has the potential to decide who owns the core CRISPR intellectual property, possibly stripping Zhang of his near-dozen patents, and shaking up hundreds of millions dollars of investment in their respective companies. It also may see Doudna and Zhang deposed and cross-examined on the priority and scope of their inventions. Needless to say, this is a monumental event for patent attorneys, molecular biologists, the PTO, and the world.

The origins of the dispute began with Doudna’s original patent application (No. 13/842,859), filed on March 15, 2013—a day before the PTO’s first-to-file rules came into effect—but given a priority date of May 25, 2012. That original application contained 155 broad claims to the CRISPR technology, but described only “genetically modified cells that produce Cas9,” the enzyme critical to the CRISPR reaction, and “Cas9 transgenic non-human multicellular organisms.”

On October 15, 2013, Zhang filed his own patent application (No. 14/054,414)—months after the first-to-file rules came into effect—but claiming a December 12, 2012 priority date under the old first-to-invent rules. Concurrent with his application, Zhang also filed an Accelerated Examination Request (also known, somewhat comically, as a Petition to Make Special), a request that the PTO make an up or down decision on an inventor’s application so long as it’s short (no more than three independent claims), directed to a single invention, and under the condition that the inventor won’t argue the patentability (or lack thereof) of individual claims during prosecution (i.e., that the inventor agrees to an all-or-nothing decision on his application). Zhang’s patent application, unlike Doudna’s however, specifically contemplated adapting CRISPR in eukaryotic cells. The PTO accepted this accelerated timeline and issue Zhang his first CRISPR patent on April 15, 2014. Thus—even though Doudna had both an earlier invention date and filing date—Zhang won the initial patent race, and with specific claims covering profitable eukaryotic applications.

Meanwhile, Doudna’s patent application languished at the PTO. For most of 2014, Doudna’s patent attorneys resolved to correct technical deficiencies in the ownership of the patent application, complicated by Doudna’s co-inventor Emmanuelle Charpentier’s agreements with several European institutions and her initial representation by different attorneys. In addition, on September 5, 2014, a mysterious third-party submitted a 66 page report of prior art—including one of Zhang’s patents—in an effort to sink Doudna’s application.

Doudna’s attorneys then attempted to get the PTO to more aggressively prosecute her patent application. Over the next few months, they met with Examiner Joike and amended Doudna’s claims, on January 8, 2015. The amendment was striking—it cancelled all 155 of Doudna’s previous claims, and replaced them with only nine claims covering the broadest applications of using CRISPR as a gene editing tool—including, presumably, eukaryotic applications. But this, too, spurred a response from a mysterious third party, who submitted more literature—including UC Berkeley press releases—to sink Doudna’s application.

Doudna’s attorneys then went for broke. On April 13, 2015, they amended Doudna’s application once again, cancelling all previous claims—164 to date—and replacing them with 82 new claims directed to “a desired cell type”—an effort to include both prokaryotic and eukaryotic cells. But more importantly, Doudna’s attorneys filed a 114 page Suggestion of Interference, claiming that ten of Zhang’s issued patents—despite their later priority date—were interfering with their patent application. Two days later, Antonio Regalado of the Technology Review quoted Paul Goldsmith of the Broad Institute saying that the Institute had made “repeated efforts and trips since the beginning of 2013 to resolve this situation outside the legal system.”

We—the world, patent academics, technologists—have been waiting on the interference request since then. And, on December 21st, the examiner corps recommended to the PTAB that it initiate such an interference proceeding, and changed the status of Doudna’s patent application accordingly. Frustratingly—and almost suspiciously (almost)—the PTO then experienced, on the following day, a massive, crippling power outage that frustrated public attempts to access any application’s docket. While the decision to institute an interference proceeding is ultimately up to the PTAB, one former PTO employee I have spoken with claims that the decision is largely pro forma and follows the examiners’ recommendations. The fight is on.

So: What happens now? Well, over the next few months we can expect a final decision from the PTAB on the interference decision, and an assignment by the Chief Judge of the PTAB, Nathan Kelley, of a three patent judge panel to hear the case. (Given the spate of hires of patent judges by the PTAB, and their technical specialization, it’s highly likely that all three will have some life science technical experience.) And then we can commence with what could be the last, great interference proceeding—ever—in the PTO.

This will be an event unto itself for two reason: First, because it will likely—or at least discourage—the possibility of a settlement between the parties. The interference declaration is a Rubicon, of sorts, beyond which the parties must compete for territory. They may settle, but interference settlements must be approved by the PTO and cannot be anticompetitive. The reasons for this are arcane but sensible—we don’t want patentees to collude to divvy up patents rights to which they are not otherwise entitled.

Second, because interference proceedings are famous for their formalistic complexity—an almost Soviet-style detail to bureaucratic rules that gives even seasoned patent prosecutors the shudders. With that said, the proceedings look somewhat like an actual trial, with the interference divided into two phases: the interlocutory phase and the testimonial phase. At the interlocutory stage, the three judge panel will set a motions calendar—a timeline for when the two parties—Doudna, the “senior party,” and Zhang, the “junior party”—can present initial briefs. Those initial briefs will likely include preliminary substantive statements about the patentability of the inventions at-issue, whether there is truly an interference in-fact, or whether the PTAB should redefine the scope of the interference. The PTAB may also consider motions as to why each why each party is entitled to certain priority dates—the determining factor in who will be awarded their respective patents if there is an interference in-fact. Those motions often come down to inventor testimony—notarized notebooks and the like—as well as affidavits called Rule 131 Statements (based on their governance under 37 C.F.R. § 1.131). Because Rule 131 statements may decide the interference proceeding outright, they have long been notorious in the patent community for encouraging inventor duplicitousness—“Of course I invented that two years earlier than I originally claimed!” Whether Doudna and Zhang—two of the most well-liked and seemingly honest scientists in any field—will engage in this priority dispute through Rule 131 Statements will be interesting. (To be clear: I’m not at all saying that Doudna or Zhang are going to lie, or that if they do file Rule 131 Statements, they’re lying. I’m only saying that the procedure has a bad reputation among patent attorneys. It’ll be like watching Nobel Prize winners defend themselves in traffic court.) In addition, the parties are entitled to limited discovery during the interlocutory phase—mostly, document discovery concerning priority dates, but notably inventor depositions and cross-examinations. That’s right: Doudna and Zhang may be required to take the stand. (Or, at least testify under oath in a deposition.)

After the interlocutory phase, the three panel will rule on any outstanding motions. At that time, the panel may decide the issue outright—by determining, for example, that there is an interference in-fact and that Doudna wins because she has priority—or, given conflicting evidence, move the proceeding to the next, testimonial phase. In that case, the parties file substantive briefs on the issues, somewhat like Supreme Court merits briefs. The panel then hosts oral argument, typically 40 minutes total, 20 minutes per side. If the CRISPR patent dispute does come to this, the oral argument should—in theory—be open to the public.

After all that, the three judge panel decides the interference issue and awards one (or none) of the parties his or her respective patents. And, as elsewhere with the PTO, the losing party may appeal a negative decision to the U.S. Court of Appeals for the Federal Circuit, stringing out the final determination of priority for years to come. Time will tell. But the biotech patent dispute of this century may last well into this century.

Jacob S. Sherkow is a former CLB fellow and currently an Associate Professor at New York Law School

16 Responses to The CRISPR Patent Interference Showdown Is On: How Did We Get Here and What Comes Next?
  1. How much scrutiny does the PTO give settlements during interferences? It seems to me clearly in the interests of MIT and UC to settle and share the monopoly (though, of course, that’s been true for a couple of years now).

    Also, why exactly is this the LAST interference? Are they only about who invented first and thus gone when “first to file” takes over entirely? That sounds plausible to me, but I know just enough about patent law to know the correlation between “plausible” and “accurate” is far under 1.00.


  2. Scrutiny: It so rarely comes up, it’s hard to say. But settlements *have* been rejected by the PTO in the past, most famously in the Singer sewing machine cases. I do think that because they have to be filed with the PTO and reduced to writing–and because they can’t oppose the interlocutory evidence–it vastly complicates things. But hey–vast complication is what the U’s TTO’s pay their attorneys for.

    The Last Interference: I, of course, can’t tell whether this will really be the *last* interference. But man: we’re getting close. Even here, the only reason we’re in interference territory is because Zhang claimed a priority date *before* the AIA. His application was in fact filed *after* the AIA’s critical date. How many more applications are there out there that *haven’t* been issued that are like that? And how many are competing with other similar applications? And how many of *those* are filed by inventors who can pay their attorneys the vast sums it requires to complete an interference proceeding? It’s almost like Drake’s equation. Sure–*maybe* there’s another one in a sea of patent applications. But there can’t be many, even if there’s one.


  3. Great post, Jake. Several of the interferences in our case studies took up to a decade to fully resolve. The process should flesh out some of the history of this important technology, but with a very powerful bias in favor of each of the inventor teams. So weird that our legal system converts a multi-party innovation system into a winner-takes-all game. From your account, a first guess is that the UC lawyering was a couple steps down from the Broad’s. Does the whack-a-mole strategy UC used to finally converge on some claims factor in the interference proceedings?

    Will stay tuned with keen interest.



  4. Great post, Jacob! Wish I had talked about this in my IP class this past semester, but will discuss in my Transactional class in the Spring (at my position in Syracuse University College of Law). Will share your post.

    Hank, not much data on settlements and interferences that I am aware of. My hypothesis is that the parties prefer all to a part. I think Colleen Chien has an article (with Mark) on interferences. One observation from that article is that first to file tended to win interferences anyway.

    I looked a bit at settlements with IPR’s and there have been several. I wrote a bit about this in my Rutgers symposium piece on Actavis. My sense is that scrutiny was low-level–perhaps even deferential.


  5. Thanks a lot for this deep review on the USPTO battle for those pioneer CRISPR patents and the PTAB/interference insights, very useful! will all inventors be involved in such an interference or does PTAB typically only calls one representative to testify?
    We also noted that the mysterious third-party submission on Sep 2014 against the Doudna patent is not anonymous as it’s not possible at the USPTO (as opposed to EPO and WIPO), but the identity of the law firm, Aka Chan LLP in California, does not help identifying where it comes from, as this firm seems to be primarily active in the ICT/electronics fields… i.e. looks like a straw man opposition filing. Very mysterious indeed, and on the international side it’s no clearer – e.g. 9 oppositions have been filed by November 2015 in Europe against the first granted Zhang patent, by CRISPR Therapeutics, Novozymes A/S, Boxall Intellectual Property Management Ltd., Sarittarius Intellectual Property, Cabinet Regimbeau, Mr. George Schlich, Dr. Martin Grund, Mr. Harvey Adams and Dr. Ulrich Storz.


  6. Glad for the support, Bob and Shubha.

    Regarding your questions: “Does the whack-a-mole strategy UC used to finally converge on some claims factor in the interference proceedings?” Yes–essentially, it’s *those* claims that will be at-issue in the interference proceeding. I don’t think the PTAB will look at the history of claim amendments to determine *whether* to institute an interference. They *may* play a role in how the PTAB looks at the scope of the claims, but that’s a very, nitty-gritty question that someone with more interference expertise is probably better suited to answer.

    And, WRT to settlements in IPR’s: Yes, the PTO has to approve those as well, and the PTAB has been deferential. But because we’re dealing a single issued patent in IPR–as opposed to potentially two competing patents or an application and an issued patent–there’s less opportunity for collusion. In IPR, the worst case scenario, from a collusion perspective, is that the petitioner drops its challenge, which it never needed to bring. It simply leaves us with the status quo. For interferences, however, there’s the concern that two competitors will divvy-up rights to which neither was entitled. As a result, settlements, once an interference has been instituted, tend to be given more scrutiny.


  7. Thanks @ipstudies for the comments. WRT who gets to testify, the interference proceeding is only between the inventors of the patents or applications in the declared interferences. Other inventors who claim to have a dog in the fight don’t have a right to intervene. So, it’ll be Doudna v. Zhang, mano y mano.

    WRT to the third-party submission, I meant that the “real party in interest”–not the patent attorney–has chosen to remain anonymous. (Which is, indeed, the case.) Oftentimes, other inventors, or competitors, seek unrelated, outside counsel to shield their identity during third party challenges. So, I wouldn’t read too much into the patent attorney listed on the TPS. A good analogy, perhaps, is the situation involving BioGatekeeper and the Yamanaka patents, excellently detailed by Paul Knoeplfer here:

    Also, if you take a look at the signature page of the Sept. 5, 2014 challenge, it’s by an attorney named Rodney C. LeRoy, who appears to be a solo practitioner–not someone from Aka Chan LLP. Are you referring to a different challenge?


    1. Thank you for your detailed answer and extra link.
      Rodney LeRoy lists himself as a patent attorney at Aka Chan LLP on his linkedin profile so that seemed consistent to me, but indeed that’s certainly just an outside consel shield trick for anonymous filing…
      Do we have an idea of the timeframe for those interference proceedings? I’ve just started diving into the Editas IPO document which also lists another interference by Toolgen. I was also wondering whether Emmanuelle Charpentier would be asked to testify as well as a co-inventor and co-applicant of the Doudna et al. patent, as she is also listed on the Editas IPO S1 Sec filing list wrt the USPTO interference. What a landscape!
      Overall, this patent and licensing uncertainty makes CRISPR as complex to deploy on the business side as it seems simple to implement (or at least experiment) on the tech side… I’m curious how this will evolve in 2016.

  8. “Rodney LeRoy lists himself as a patent attorney at Aka Chan LLP on his linkedin profile”

    Whoa–good detective work. I just don’t know about him–or his firm.

    “Do we have an idea of the timeframe for those interference proceedings?”

    My understanding is that we first wait for the PTAB to make a final decision on institution–which should be somewhere between three and four months. (Yes; it’s slow.) Then, the proceedings themselves can take over a year. So it’s going to be a bit.

    “Overall, this patent and licensing uncertainty makes CRISPR as complex to deploy on the business side as it seems simple to implement (or at least experiment) on the tech side… I’m curious how this will evolve in 2016.”

    I agree. And I do wonder, especially with Zhang’s discovery of Cfp1, whether this is just going to encourage work-arounds. That’ll be incredibly interesting.


    1. Our genomics expert just completed the January 2016 CRISPR patent landscape analysis (489 published patent families worldwide) and yes, we can confirm it’s incredibly interesting! Beyond the initial interference, the research directions and IP developed by Zhang (36 families) and Doudna (10 families) seem more complementary than conflicting. And there are an increasing number of players securing their positions, either individually or through strategic collaborations. Would a patent pool make sense to gather essential IP on the “standard/core tech” part of CRISPR operation?

  9. On work-arounds, I wonder how relevant they will be. The current version, so widely adopted, is CRISPR/Cas9, a combination of a CRISPR component and “CRISPR associated protein 9”. There are a LOT of various proteins that, in various microbial species, work with CRISPR. My understanding is that Cfp1 is just one more of them, more or less modified from its natural state. It still needs CRISPR to work so any work-around using CRISPR/Cfp1 would still need a license from whoever has the base patent on CRISPR. No?

    Also, a different question – are the CRISPR patent claims (in Zhang – have the Doudna claims become public?) claims for uses of CRISPR/Cas9 only or are there also composition of matter claims? And, if the latter, will they stand up to Myriad? Maybe as a combination of two molecules found in nature? Maybe because either the CRISPR or the Cas9 is significantly (as significantly as cDNA from gDNA, so not necessarily much) modified? Thoughts?


  10. “It still needs CRISPR to work so any work-around using CRISPR/Cfp1 would still need a license from whoever has the base patent on CRISPR. No?”

    My reading of Doudna’s application claims is that they almost all require Cas9. So assuming you can get the CRISPR system to work without Cas9, then you wouldn’t infringe.

    “Also, a different question – are the CRISPR patent claims (in Zhang – have the Doudna claims become public?) claims for uses of CRISPR/Cas9 only or are there also composition of matter claims? And, if the latter, will they stand up to Myriad? Maybe as a combination of two molecules found in nature? Maybe because either the CRISPR or the Cas9 is significantly (as significantly as cDNA from gDNA, so not necessarily much) modified? Thoughts?”

    The Doudna application claims are public, and they’re linked here in this blog post. They are almost all method and system claims. But, claim 223 in the current Doudna application, for example, is a composition claim directed to an RNA. I think that claim withstands Myriad–in its myriad forms. It claims a hybrid RNA–the targeted and activator components important to CRISPR–and also specific limits itself to RNAs that have “one or more sequence modifications” compared to wild type. Good drafting, I think, by Doudna’s attorneys.


    1. No Clustered regularly-interspaced short palindromic repeats (Crispr) are involved in Cas9 mediated genome editing in eukaryotic cells. Cpf1 is new matter ( different enzyme and works differently compared to Cas9. Guide RNA is designed always specifically thus different in sequence from anything naturally occurring, accept it is composed of 4 nucleotides that in different combination make up all RNA.

    2. In the IPStudies landscape review, our expert identified that after initial R&D positioning on the CRISPR-Cas system development in the pioneering landscape and some design arounds of Cas9 (such as Cpf1), several players are now focusing their IP on gene-specific gRNA. This certainly goes in the “different from anything naturally occurring” post-myriad direction indeed…

  11. There is one count of interference against all claims of 11 patents. Would at least some of the claims be dismissed from the interference as not corresponding to the count? When will it be decided (if yes)?

    Could their be a situation where both parties have claims and neither can practice without a license?


  12. Great blog entry and comments exchange. I have to agree with one of the remarks above that, so far, MIT has out-lawyered UCB.


Leave a Comment