Editing Humanity: The CRISPR Revolution and the New Era of Genome Editing

CHAPTER 13 PATENT PENDING

“A few years ago with my colleague Emmanuelle Charpentier, I invented a new technology for editing genomes. It’s called CRISPR-Cas9.”¹ Doudna raised a few eyebrows with that offhand remark during a TED talk in London in 2015, which made light of a billion of years of evolution, not to mention the competing efforts of a few other investigators. But it is fairly ingrained in the popular culture. In November 2019, Alex Trebeck read a question on Jeopardy:

JENNIFER DOUDNA & EMMANUELLE CHARPENTIER ARE CO-INVENTORS OF THE REVOLUTIONARY TOOL CRISPR TO EDIT THESE IN THE BODY^I

Lest we forget, bacteria clearly invented CRISPR many hundreds of millions of years ago. But the legal question of who invented CRISPR genome editing technology is very real—and really important. At stake are commercial rights to a technology potentially worth billions of dollars. Law professor Jacob Sherkow calls it “the most monumental biotech patent dispute in decades.”² At the heart of the dispute is the Broad Institute, the base of Feng Zhang, versus the University of California (UC), home of Doudna at Berkeley, in collaboration with Charpentier and her host institution.

So who “invented” the revolutionary CRISPR gene-editing technology? The transatlantic team that devised the single-guide RNA and showed that Cas9 could be programmed to cut DNA on demand in June 2012? Or the group that first demonstrated editing of human DNA using CRISPR-Cas9 seven months later? What about other less-celebrated contenders, such as Virgis Šikšnys, who filed a patent application in March 2012? In his small office at the New York Law School in Tribeca, Sherkow enthusiastically guides me through the landmarks of the saga.

One might assume that, because the Doudna-Charpentier team published the first account of CRISPR-Cas gene targeting, the invention should belong to them. Had the CRISPR saga taken place twelve months later, that might be the end of the story. In April 2013, the criterion for awarding American patents changed to a “first to file” basis. Before then, however, applications were reviewed on a “first to invent” basis. Eight years after the first key publication, we still don’t have a definitive answer. Doudna and Charpentier, together with Jínek and Chyliński, filed their provisional US patent application in May 2012, shortly before submitting their manuscript to Science.³ The three patent co-owners were UC, the University of Vienna, and in an interesting twist, Charpentier herself. In Sweden, where Charpentier was based, academics enjoy “Professor’s privilege,” which grants them full rights to their own inventions. (Thus, Charpentier is both a co-inventor of her technology and a co-owner of it.) The triumvirate is known as CVC.

This was not the first patent application to mention CRISPR: in April 2004, the Danisco group of Barrangou and Horvath described a method for sequencing CRISPR regions in a dairy sample to fingerprint variants of Lactobacillus acidophilus. “The origins of the CRISPR patent estate, far from being forged in combat, were gently cultured in distant tuns,” observes Sherkow.⁴ Another interesting application was one from Erik Sontheimer and Luciano Marraffini. In his office at the University of Massachusetts Medical School, Sontheimer rifles through his computer files searching for what he calls “the best grant I’d ever written,” which mirrored his patent application.

In January 2009, Sontheimer submitted a five-year, $1.8 million grant application to the NIH, entitled “RNA-Directed DNA Targeting in Eukaryotic Cells.” “CRISPR interference could provide unique capabilities, if it can be ported to eukaryotic cells,” Sontheimer wrote. “It could be easily programmed (and, when desired, reprogrammed) by the introduction of small RNA molecules.” The goal was to manipulate the structure and activity of genomes of higher organisms using RNA that recognizes its target DNA using the well understood rules of Watson-Crick base pairing. The method “would have transformative and disruptive potential in biotechnology and medicine.”⁵ There were parallels to RNA interference, the technology codiscovered by the Nobel laureate next door, Craig Mello. But “CRISPR interference” was exciting because of the sequence specificity conferred by the twenty-four- to forty-eight-base spacers. Sontheimer and Marraffini’s discovery of “a natural, highly specific, RNA-directed, DNA-targeting machinery in bacteria” suggested a route toward a “reprogrammable genome targeting system in eukaryotic cells.”

Sontheimer’s plan was to port CRISPR into eukaryotic cells and test its ability to target specific genes. He would start small, working in yeast, then progress to Drosophila, and eventually mammalian cells in the final two years of the funding by 2012–13. RNA-directed genome targeting “has the potential to transform biomedical research, biotechnology, genetic medicine, and stem cell therapeutics,” he wrote. In retrospect, it was too much too soon. The proposal “hit a brick wall” and was rejected, and his companion patent application was abandoned. “The vision and idea were out there, but we hadn’t reduced it to practice,” said Sontheimer wistfully.⁶

Six months after CVC filed their patent application (’772) just prior to publication of their Science paper, the Broad Institute followed the same playbook, as Zhang filed an application covering the use of CRISPR editing in eukaryotic cells in December 2012. But in a nifty legal maneuver, the Broad lawyers paid the princely sum of $70—a round of cocktails at the local Meadhall gastropub—for an expedited “fast track” review, thereby leaping ahead of UC’s application. In January 2014, the United States Patent and Trademark Office (PTO) provisionally denied the application, citing the “prior art” of the CVC application. Zhang quickly filed a forty-page “personal declaration” rebuttal. It paid off: three months later, in April, the PTO awarded patent number 8,697,359 to the Broad. The decision sent shockwaves across the biotech communities. Doudna resigned as an advisor to Editas, the all-star company she had cofounded with Zhang and others. CVC filed for an interference proceeding, seeking a PTO ruling to establish who truly invented the technology.

The Broad responded by filing a “priority statement” with the PTO in May 2016. Exhibit A was an internal Broad document—a “Confidential Memorandum of Invention”—that Zhang had signed in February 2011. In this document, Zhang declared he had the initial idea for CRISPR gene editing on February 4, 2011—shortly after hearing Gilmore’s lecture—and first documented those ideas four days later.⁷ Zhang wrote:

The key concept of this invention is based on the CRISPR [repeats] found in many microbial organisms. Enzymes associated with the CRISPR complex use short RNA sequences to recognize specific target sites on the host genome and performs site-specific cleavage. The key novel feature of this invention is that it does not rely on the design of site-specific DNA binding proteins (i.e., zinc finger or TAL effector) and can be easily targeted to multiple sites through the use of multiple sequence-specific CRISPR spacer elements.

The Broad mounted an effective campaign to sway public opinion. “CRISPR itself cannot be patented,” explained communications director Lee McGuire. “Cas9 is a naturally occurring protein and part of a naturally-occurring bacterial process, but this process, on its own, does not work in mammalian cells. What [Zhang] has patented are engineered components and compositions specifically altered from their naturally-occurring form to be useful in methods for editing the genomes of living mammalian cells.”⁸ There were shortcomings in both the CVC application as well the earlier Šikšnys application. These applications, the Broad argued, merely showed that “purified Cas9 protein and a certain purified RNA could cut a short piece of DNA in a solution in a test tube.” And then the Broad dropped the hammer. Both applications “contained no disclosure of work in cells, no genomes, and no editing.”⁹

No cells, no genomes, no editing. It was a brutal rebuttal to the CVC appeal. But would it work in court?

Showtime came on a rainy Tuesday morning in December 2016. Legal scholars including Sherkow and Robert Cook-Deegan, attorneys and journalists formed an orderly queue more than an hour before the doors opened at the PTO’s headquarters in Alexandria, Virginia. They were there to witness history in the making: such courtroom dramas would likely become a thing of the past under the “first to file” rule.¹⁰ The Patent Trial and Appeal Board (PTAB) was unprepared for the crush of spectators who rapidly filled the courtroom and two overflow rooms. “It was probably the most well-attended interference proceeding in USPTO history,” said Sherkow.

The purpose of the hearing was not to establish who invented CRISPR genome editing. Rather, the three judges on the PTAB were trying to determine “what the what is,” as Sherkow put it. If the PTAB concluded that there were indeed two separate inventions with two distinct timelines, then the question of who came first was redundant. But if the dispute was over the same invention, then all bets were off. There is a reason, says Sherkow, that for example Bell Laboratories required all its engineers to sign and date every page of their notebooks before they left work each day to ensure there was documentary evidence in the event of an interference proceeding.

The panel of three judges, led by Judge Deborah Katz, who has a doctorate in molecular biology, fired their sharpest questions at the CVC lawyer, who insisted that Doudna’s discovery, restricted to DNA in a test tube, could be extrapolated to all organisms including humans.¹¹ Among hundreds of research papers, emails, PDFs, patents, and other documents, UC submitted statements from a pair of expert witnesses. One was from Nobel laureate Carol Greider, a Berkeley grad herself, who discovered telomeres (the DNA aglets that protect the ends of chromosomes from fraying). The other was from gene-editing pioneer Dana Carroll, who was compensated $500/hour to compile a brief totaling nearly two hundred pages.

Much of the deliberation centered on the contemporaneous commentaries of Carroll and other experts on the initial prospects of translating CRISPR gene targeting to human cells. Recall that in his September 2012 commentary, Carroll mulled over the prospects for CRISPR genome editing in human cells. In conclusion he wrote: “Whether the CRISPR system will provide the next-next generation of targetable cleavage reagents remains to be seen, but it is clearly well worth a try. Stay tuned.”¹²

Table: Key Dates in the CRISPR-Cas9 Patent Dispute

YEAR

DATE

EVENT

2012

March 20

Šikšnys files patent (’739)

May 25

CVC files patent application (’772)

December 6

MilliporeSigma files patent

December 12

Broad files first patent (’527)

2013

March 15

CVC files patent application (’859)

2014

April 15

Broad awarded ’527 patent

July 7

Rockefeller files CRISPR patent

2015

April 13

CVC files “suggestion of interference”

2016

January 11

PTO declares an interference

December 6

PTAB interference oral argument

2017

February 15

PTAB interference decision favors Broad

March 23

EPO grants CVC patent

April 12

CVC appeals Interference in federal court

2018

January 18

EPO revokes Broad patent in Europe (’468)

April 30

Appeal oral argument

June 19

PTO awards CVC its first patent (’772)

Sept 10

Court upholds PTAB appeal decision

2019

February 8

PTO awards CVC ’859 patent

June 24

PTO declares interference

2020

January 16

EPO dismisses Broad appeal of ’468

February 7

EPO upholds CVC

May 18

PTAB hearing

* CVC, University of California/University of Vienna/Charpentier; EPO, European Patent Office; PTAB, Patent Trials and Appeals Board; PTO, U.S. Patent and Trademark Office.

Incredibly, the PTAB judges tried to unpack just what Carroll meant by “Stay tuned.” Was Carroll skeptical that CRISPR would work in human cells, or suggesting it was only a matter of time? The judges dissected another line from Carroll’s article where he said: “there is no guarantee that Cas9 will work effectively on a chromatin target.” The judges summed it up: “We fail to see how ‘no guarantee’ indicates an expectation of success”—especially for someone having ordinary skill in the art. They concluded sternly that Carroll “did not have a reasonable expectation that the system would work.” And if it had been a formality, then one would have to question why the editors of Science wanted to publish the Zhang and Church reports six months after they published the Doudna-Charpentier masterpiece.

Among the hundreds of documents and exhibits made public in the Broad-CVC saga, was a bizarre email that Doudna received in February 2015. It was sent by Lin Shuailiang, a Chinese student who had spent nine months at the Broad from October 2011 through June 2012 and was a coauthor on Zhang’s Science article. Lin raised allegations about the provenance of the CRISPR work in Zhang’s lab. Lin called the PTO decision “a joke” and claimed, “Feng is not only unfair to me but also the science history.” He continued:

I began the CRISPR project solely the first day I came to Feng Zhang’s lab in 2011 as a side project for a visiting student. At that time, the whole lab except me focused on TALEN projects. I was working on the CRISPR project until I went back to China [in] 2012 June for my PhD… After seeing your [Science] paper, Feng Zhang and Le Cong quickly jumped to the project without letting me know. My lab notebooks, emails and other files… recorded every step of the lab’s failure process… We did not work it out before seeing your paper, it’s really a pity. But I think we should be responsible for the truth. That’s science.¹³

The Broad dismissed Lin’s assertions, noting that his visa was about to expire and he was out of a job at the Broad. Lin ended up joining a San Francisco nanomedicine biotech company called Ligandal. On the company’s website, he was described as “the very first scientist in the world to work with CRISPR.”

On February 15, 2017, the PTAB ruled decisively in favor of Zhang and the Broad Institute, finding no “interference-in-fact”—the Broad’s patent did not overlap significantly with the CVC application.¹⁴ Zhang’s invention of CRISPR genome editing was “not obvious” because “one of ordinary skill in the art would not have reasonably expected a CRISPR-Cas9 system to be successful in a eukaryotic environment.” As the rival patent applications did not overlap, the Broad’s patents could stand. “Even though it seemed like both parties were there at the same time and they seemed to be working in relatively the same area, their patent applications, at the very least, did not interfere with one another,” Sherkow summarized.¹⁵ Editas’s stock surged 30 percent, pushing its market cap above $1 billion.

Three months later, I was in Vilnius, meeting Šikšnys for the first time. On his desk was a newly arrived FedEx envelope. Before I left, he agreed to open it. Inside was an official copy of his newly approved U.S. patent, filed back in March 2012: “RNA-directed DNA cleavage by the Cas9-crRNA complex.” DowDuPont (now Corteva) wasted no time in licensing the rights to Šikšnys’ IP.

CVC might have lost the Battle of Alexandria, but it was still determined to win the war. Doudna put her own spin on the verdict: “They have a patent on green tennis balls. We will have a patent on all tennis balls.”¹⁶ The players regrouped and dug in for another nail-biting set.

Two months later, CVC filed an appeal in federal court, confident it would establish definitively that the Doudna-Charpentier partnership “was the first to engineer CRISPR-Cas9 for use in all types of environments” including human cells.¹⁷ It was an uphill battle—the federal court’s task was not to retry the case but merely to assess whether the PTAB had committed a legal error. At the appeal hearing in April 2018, UC’s lawyer Donald Verrilli, a former U.S. solicitor general, argued that if Doudna and colleagues hadn’t demonstrated the feasibility of genome editing in higher organisms, the multiple publications in early 2013 would have used different techniques. “This was like the California Gold Rush,” Verrilli said. The competing researchers were all trying to demonstrate genome editing first. “If they thought making changes was necessary, they would have.”¹⁸

But the three judges didn’t bite. “You start with the common technology,” declared Judge Kimberly Moore. She highlighted statements by Doudna and others expressing frustration about getting CRISPR to work in human cells. For example, in a profile for a Berkeley College of Chemistry magazine, Doudna openly admitted: “Our 2012 paper was a big success, but there was a problem. We weren’t sure if CRISPR-Cas9 would work in eukaryotes.”¹⁹ In the end, the flood of reports in early 2013 showed that it was fairly easy to get genome editing to work in human cells based on the UC methods. But the key question according to Moore wasn’t “Did it turn out to be easy to do?” It was: “What was the perception of the skilled artisan at the time?”

On September 10, 2018, the Court ruled in favor of the PTAB’s earlier judgment. Sherkow said that while legally correct, he wasn’t sure the decision meshed with the way biomedical research is actually practiced. “The PTAB’s decision and Šikšnys’s patent sit as the Scylla and Charybdis for UC’s continued prosecution of its application,” Sherkow wrote.²⁰ Some reporters thought it was game over: NPR declared “East Coast scientists” the victors.²¹

But not so fast: in another surprising twist in February 2019, the PTO assigned the CVC ’859 patent—in limbo for six years—to a new examiner, who promptly allowed it with minimal comment.²² By the end of 2019, CVC had been awarded some twenty CRISPR-related patents, gleefully punctuating each award with a press release. But questions lingered: was the key CVC patent overly broad (no pun intended)? And did it meet the enablement clause? Did it provide sufficiently detailed instructions to allow someone to edit a genome?

In June 2019, the PTO declared its own interference, hoping to resolve the dispute. In one corner, a dozen patents owned by the Broad (the senior party); in the other, ten awarded to CVC.²³ As the junior party, CVC still has the uphill task. The core of the case, Sherkow explains, is who can claim priority for the single-guide RNA?

Many have wondered why UC and the Broad have spent so many millions of dollars on this dragged-out patent fight? The bills are actually paid by the biotech companies; Editas has paid tens of millions of dollars on legal fees on behalf of the Broad Institute. The thought that the Nobel committee is unlikely to award the inevitable CRISPR prize to the loser in the patent drama “could be one factor in the Broad’s litigation strategy,” says Hank Greely.²⁴ The Broad insists it has been open to a settlement for years, but UC hasn’t shown any appetite for compromise. Indeed, CVC has filed motions alleging that Zhang did not include tracrRNA until after Doudna’s paper was published, also questioning authorship discrepancies in some of Broad’s patents in the United States and Europe.²⁵

In the run-up to the latest PTAB hearing in May 2020, the Broad filed an expert brief by Chad Mirkin, a chemistry professor at Northwestern University. Mirkin, who runs a massive lab of seventy people, charged his standard consulting rate of an eye-watering $1,600/hour in support of the Broad’s claims. He argued that the hypothetical “person of ordinary skill in the art” would not have successfully achieved genome editing in mammalian cells based on the CVC. During a deposition in early 2020, UC lawyers laid into Mirkin’s CRISPR credentials. He was forced to admit that his lab had not actually performed a CRISPR experiment until 2016.

Whether this legal sparring will have any material bearing on the case is unclear. At the PTAB encore hearing in May 2020, the judges did not tip their hand. Whoever wins the next round, the other side is still likely to come out swinging.

In Europe, CVC has fared better, thanks in part to a curious event that happened early on in the legal drama. The original Broad application in 2012 listed four co-inventors: Zhang plus Le Cong, Ann Ran, and Marraffini, all coauthors on the 2013 Science paper. But in the later ’359 patent application, Zhang was flying solo.^II In July 2014, the Rockefeller University filed its own patent, essentially a carbon copy of the Broad’s application, only with Marraffini’s name restored.²⁶ It was an unorthodox gambit that violated PTO rules because Zhang’s name had been included in this later filing without his notification or permission.

The dispute went to binding arbitration behind closed doors. On January 15, 2018, the Broad announced that it had prevailed. This meant that Marraffini, the microbiologist who had collaborated with Zhang throughout 2012, would not be listed as a co-inventor on the Broad’s major patents regarding eukaryotic editing.²⁷ The Broad’s press release stating that the two sides had “settled their disagreement” gave the appearance of an amicable resolution. But when I even broached the subject with Marraffini in his office months later, he shot me the universal my-lips-are-sealed gesture.

But the Broad’s celebrations lasted barely seventy-two hours, as it was dealt a major setback in Europe. In March 2017, the European Patent Office (EPO) had awarded CVC a wide-ranging patent across all organisms—microbes, plants, and animals.²⁸ Just a few days after the Broad-Rockefeller arbitration results, EPO announced that it was revoking the Broad patents ironically because of the Marraffini dispute. As the list of inventors of the Broad applications had changed from the earliest application in December 2012, EPO deemed the earliest application to be invalid. With that off the table, the Broad’s applications did not predate CVC. The EPO upheld that decision in January 2020 and reaffirmed the CVC patent three weeks later.²⁹

The dispute has driven a wedge between the two biggest names and ex-industry colleagues in CRISPR in the United States. It’s not a subject Doudna cares to discuss, saying only it is “very disappointing for me, as a scientist and as a person.”³⁰ Lurking in the wings is another patent application, originally filed on December 6, 2012, by MilliporeSigma researchers, who claim they were the first to perform CRISPR editing in eukaryotes. Ironically, the PTO judged that claim to be obvious in light of the CVC work, in contrast to the history of the Broad application. MilliporeSigma has petitioned the PTO to declare another interference with CVC.³¹

Hopefully this saga won’t be repeated as more CRISPR technologies are developed. In more recent iterations like base editing and prime editing (see chapter 22), the patent situation looks much cleaner. And the first to file rule is less contentious than first to invent. As patent attorney Joe Stanganelli says: “If there is any universally good news to take away from the CRISPR patent saga, it is this: It is unlikely to happen again.”³²

What hinges on these patent disputes? In the coming years, the CRISPR biotech companies will likely have several gene-editing therapies on the market, all using garden-variety Cas9. “There will be a lot of money changing hands to make sure those drugs get into patients’ veins without infringement,” says Sherkow, who is now a law professor at the University of Illinois. What about 2040? “We’ll have a million CRISPR therapeutics on the market, every nuclease, RNA editors, a complete riot of options. Will we look back and say, ‘What a godawful waste of money’? Absolutely. It’s a joke.”³³

Joke or not, the patent wars hold an “inside baseball” fascination that will at some point result in some momentous decisions about who will be paying licensing fees to whom. But in 2018, a new danger flared up. Some scientists and observers considered it nothing short of an existential threat to the fledgling CRISPR industry as it took its first nervous steps into the clinic.

I. The correct answer is “genes.”

II. The U.S. PTO is usually quite forgiving about such adjustments; less so in Europe.