Loose Lips Sink Ships

In Broad Institute, Inc. et al. v. Regents for the University of California et al., 903 F.3d 1286 (Fed. Cir. 2018) [hereinafter ‘Broad’ and ‘UC’] the United States Court of Appeals for the Federal Circuit [hereinafter ‘the Federal Circuit’] concluded that Broad’s bioengineered molecule was not obvious in view of another bioengineered molecule created by UC. Both UC and Broad were aware that this bioengineered molecule possesses a superior ability to rapidly sever genetic molecules. If successful in eukaryotic cells (i.e., plant and animal cells) this molecule also possesses enormous commercial potential for medical applications. Because of this high-stake financial implication, Broad and UC each claimed first inventor status of the bioengineered molecule that functions within animal cells.

If either UC or Broad is the first inventor of a single invention, then that entity is the rightful owner of United States patents and patent applications for this bioengineered severing molecule in animal and plant cells. The Federal Circuit affirmed the United States Patent Trial and Appeal Board’s conclusion [hereinafter ‘the Board’] that Broad’s molecule was not obvious, or unduly similar, to UC’s molecule. More importantly for commercial purposes, the Federal Circuit also concluded that because of this non-obviousness (i) UC’s bioengineered molecule that exclusively operates within a cell-free environment, and (ii) Broad’s bioengineered molecule that operates within plant and animal cells (iii) are separate and distinct inventions which each deserve their own patent(s).

  1. Background

Broad is a research entity formed by Harvard University and the Massachusetts Institute of Technology, while UC comprises the state university system in California. Broad owns patents and patent applications for a method in which the bioengineered molecule [hereinafter ‘CRISPR-Cas9’] successfully and most efficiently severs genetic molecules in animal and plant cells.[1] UC’s patent application also requested protection for a method of severing genetic molecules with CRISPR-Cas9. In so doing UC implicitly represented that its own CRISPR-Case9 functioned within plant and animal cells as well as within cell-free environments. However, UC’s patent application only discloses technical information for CRISPR-Cas9 that is operable in a cell free environment (i.e., in vitro).  Similarly, an earlier publication by one of the UC CRISPR-Cas9 inventors only disclosed CRISPR-Cas9 that (i) solely comprised bacterial components and (ii) demonstrated severing activity exclusively in vitro [hereinafter ‘2012 Jinek’].[2]

This litigation arose at UC’s request for an interference between UC and Broad, and for the purpose of resolving which entity was the first to invent the bioengineered severing molecule [hereinafter ‘CRISPR-Cas9’] that functions within plant and animal cells.  An interference is an administrative adjudication by the U.S. Patent & Trademark Office [hereinafter ‘Patent Office] that determines the first inventor of subject matter

(i)disclosed in a patent or patent application, and

(ii)for which protection is requested, and

(iii) that comprises the same invention as the invention of another party’s patent application(s) or patent(s).

If each inventor claims the same invention as the other inventor, then the interference resolves which person is the first inventor. This first person, or the entity which employees this person, then exclusively owns the disputed invention claimed in the applications and/or patents.[3]

  1. Interference proceedings

On April 13, 2015, UC requested the Board to declare an interference between UC and Broad, because both parties requested protection for a method of using CRISPR-Cas9 that is operable in plant and animal cells. Thereafter Broad moved to terminate the interference.  In so moving, Broad contended that Broad’s method of using CRISPR-Cas9 in plant and animal cells was sufficiently different from UC’s CRISPR-Cas9 in vitro use to merit patent protection for a separate invention.

UC maintained that Broad’s method of CRISPR-Cas9 use in plant and animal cells was too unduly similar to UC’s CRISPR-Cas9 to qualify as a different invention. In this context ‘undue similarity’ is known as obviousness, and wherein obviousness is patent law term of art.  One question in an obviousness analysis is always: Whether one of ordinary skill in the same technology would conclude that there was a reasonable likelihood of success for predictably achieving the disputed invention in view of previously existing technology. The parties agreed that the activity of CRISPR-Cas9 for selectively severing genetic molecules, such as DNA, is a significant improvement over previous biochemical systems for this purpose.

In this case, UC contended that Broad’s claims[4] to operable CRISPER-Cas9 systems in plant and animal cells were obvious in view of UC’s claims, because

(i)  one of average skill in this technology could routinely develop

(ii)  operable plant and animal cell CRISPR-Cas9 systems with a reasonable expectation of success based upon in vitro systems disclosed in (a) UC’s patent application and (b) Jinek 2012.

Broad responded that soon after the 2012 Jinek publication UC’s own scientists stated that achievement of functional CRISPR-Cas9 systems in plant and animal cells was uncertain, because (a) one of ordinary skill in the technology would not have reasonably expected success in plant and animal cells (b) without significant modifications to UC’s CRISPR-Cas9 in vitro assembly.

The Board concluded that UC’s molecular assembly and Broad’s molecular assembly  sufficiently differed from each other (i.e., Broad’s CRISPR-Cas9 was non-obvious in view of UC’s CRISPR-Cas9) to each merit recognition as separate inventions. The Board then terminated the interference because there was now no need to resolve the first (earliest) inventor of a SINGLE invention.  The Board based its decision in large part upon the following:

(i)Neither party disputed that UC’s patent application exclusively disclosed CRISPR-Cas9 comprising bacterial (prokaryotic) components, and not plant or animal cell components.[5]  Furthermore, at the time of UC’s 2012 Jinek  disclosure of in vitro CRISPR-Case9, UC’s own  inventors expressed uncertainty that this in vitro CRISPR-Cas9 could function within plant or animal cells.  Although UC’s scientists testified at the interference hearing that there were no impediments to achieving Broad’s system, their earlier statements made soon after 2012 Jinek expressed doubt for achieving a successful animal cell model. The Board gave these earlier statements more weight than the scientists’ statements prepared for the interference hearing.

(ii) Even if several researchers reported success in plant and animal cells soon after the publication of 2012 Jinek, unpublished experiments do not necessarily indicate whether an ordinary skilled bioengineer would have expected success prior to actual experimental results. Instead, any rush to achieve operability in animal cells evidenced motivation, but motivation did not comprise evidence of a reasonable expectation of success in plant and animal cells.

(iii)Obviousness depends on the specific nature of previously existing subject matter which is closely related to the disputed invention. For unpredictable technologies, specific instructions for achieving similar or identical subject matter may indicate a reasonable expectation of success. However, the availability of only generalized instructions, along with evidence of failures with similar subject matter, generally indicates non-obviousness. In this instance, Broad’s evidence demonstrated that success with selected prokaryotic molecules among thousands of prokaryotic systems would not provide a reasonable expectation of success. Broad also provided evidence of specific failed attempts at transferring other prokaryotic systems into eukaryotic cells.[6] Development of these systems enjoyed limited success and required years of research and specific developed strategies. Furthermore, UC did not offer any commonality or analogy between these previous strategies and development of a CRISPR-Cas9 that functions in plant and animal cells.

(iv) UC’s examples of successful gene editing systems of other scientists were not sufficiently analogous, because they incorporated components which naturally infect eukaryotic cells, while UC’s CRISPR-Cas9 components do not. Furthermore, the Board concluded that these same examples from UC supported Broad’s position, because UC’s examples each required a unique developmental protocol for even limited success.[7] Other prokaryotic systems cited by UC as analogous examples did not provide a reasonable expectation of success, because they were smaller and less complex than UC’s CRISPR-Cas9 system.

UC’s Appeal to Federal Circuit

According to UC, the Board applied an incorrect legal standard that (i) required previous references to comprise specific instructions for achieving an invention (ii) to support a finding of obviousness.  UC maintained that the law of obviousness does not require that specific instructions be present in prior publications for an invention to be obvious.  UC also contended that the Board improperly relied upon statements of UC’s inventors and professional witness made shortly after the 2012 Jinek publication. According to UC, these prior statements actually established that an ordinary scientist in this technology would have reasonably expected UC’s CRISPR-Cas9 to function in plant and animal cells. Consequently, the Board should have credited more weight to actual success of scientists shortly after Jinek 2012 was released, because this success evidenced the routine predictable nature of Broad’s CRISPR-Cas9 in plant and animal cells.

UC further contended that success by other scientists supported a finding of simultaneous invention.   According to UC soon after the publication of 2012 Jinek six independent research groups implemented UC’s operable CRISPR-Cas9 system

(i) within animal cells, and

(ii) within a few months of U.C.’s initial publication about this invention, and

(iii) by implementing conventional biochemical techniques.

According to UC this simultaneous invention evidenced undue similarity of Broad’s invention (i.e., obviousness)to UC’s CRISPR-Cas9.

UC also contended that Broad did not establish that it was entitled to December 12, 2012 as its earliest patent application submission date.  In support of this contention, UC observed that Broad relied upon a provisional United States patent application with a filing date of December 12, 2012.[8] According to UC, Broad had never established that its provisional application was entitled to this early date for the purpose of excluding earlier references from the obviousness analysis.[9]

Broad’s Response

Broad replied that under federal administrative law, the Board’s decision was sufficiently supported by substantial factual evidence that was credible, reliable and relevant.[10] Under this standard the Board found that the statements of UC’s own inventors  in 2012 clearly established that there was no reasonable expectation of success for UC’s CRISPR-Cas9 operation in plant or animal cells. These earlier spontaneous statements were not made in anticipation of litigation, and so the Board properly found them to comprise the more credible and reliable evidence. Also, according to Broad, UC’s own evidence demonstrated that a successful transfer of a bacterial system into an animal cell required extensive development and surmounting numerous technical pitfalls.

As to UC’s simultaneous invention contention, Broad’s position was that UC had forfeited this argument because it was not raised before the Board. In any event, the Board had already addressed the merits of simultaneous invention under its reasonable expectation of success analysis. Under this analysis the Board properly concluded that these unpublished reports resulted from motivation and not from a reasonable expectation of success.  This motivation to further develop was separate and distinct from a reasonable expectation of success that UC’s in vitro prokaryotic CRISPR-Cas9 would operate in plant and animal cells. 

The Federal Circuit Decision

The Federal Circuit affirmed the Board’s decision to terminate the interference, because Broad’s CRISPR-Cas9 which was operable in plant and animal cells was not obvious in view of UC’s prokaryotic CRISPR-Cas9 that operated exclusively in vitro. The court also agreed that there were two patently distinct inventions, and that there was no reasonable expectation of success that one of ordinary skill in this technology could predictably develop Broad’s molecular system from that of UC. In reaching its conclusions the court held that substantial evidence supported the Board’s decision. The court also found that this evidence comprising expert testimony,  statements of the UC inventors themselves, the 2012 Jinek publication and examples of prior art failures was credible, reliable and relevant.

The court observed that Broad’s expert had testified that biological differences between bacteria(prokaryotes) and plant/animal cells (eukaryotes) result in unpredictability as to whether UC’s prokaryotic CRISPR-Cas9 would operate successfully in plant or animal cells. The court also observed that in September 2012, UC’s own expert witness had recognized that these differences could result in difficulties implementing UC’s CRISPR-Cas9 within eukaryotic cells. Furthermore, according to her statements contemporaneously with the 2012 Jinek publication,  one of UC’s inventors had experienced numerous frustrations in adapting UC’s CRISPR-Cas9 to animal cells and expressed her doubts for achieving this adaptation.

The court also looked at UC’s evidence for what UC characterized as analogous examples of predictable development of earlier gene editing systems. As did the Board, the court concluded that at least two gene editing molecular assemblies relied upon by UC were not relevant, because they comprised plant or animal cell components while UC’s CRISPR-Cas9 exclusively comprises bacterial (prokaryotic) components.  The court also confirmed that Broad effectively discounted UC’s examples of successful transfer of bacterial technologies for gene editing because of (i) their limited efficacy and (ii) requirement for specific innovations to successfully operate in animal cells. Based upon this evidence the court then concluded that Broad had credibly established that successful transfer of bacterial systems to plant or animal cells remained extremely limited.

The court then addressed UC’s contention that the Board erred in adopting a ‘specific instruction in the prior art’ test to reach a conclusion of non-obviousness in view of UC’s CRISPR-Cas9. The court found that the Board had acknowledged that certainty in the art is not necessary, and that the Board had applied the correct legal standard.  The court also found that the Board did not find non-obviousness solely on absence of specific instructions in the art for an operable CRISPR-Cas9 in plant or animal cells.  Instead, the Board looked at other factors such as industry failures for transfers of other gene editing systems, as well as other prokaryotic molecular systems, to plant and animal cells.

Next, the court agreed with the Board that the degree of predictability, and therefore a reasonable expectation of success, is determined on a technology by technology basis.  In this instance, the prior technology offered as evidence by both UC and Broad established that gene editing results were highly unpredictable. As a result, more specific instructions, and not merely generalized instructions that might contain prior laboratory techniques, were necessary for a reasonable expectation of success in this particular biotechnology.

Although the court did not consider all of UC’s remaining arguments, it did address UC’s evidence of simultaneous invention.  The court found that the Board had considered this evidence under the reasonable expectation of success approach, and that its relevance was merely one factor in the totality of circumstances. It also found that the Board correctly regarded simultaneous invention as exclusively evidencing motivation to achieve a successful CRISPR-Cas9 in plant or animal cells. Even if it existed this motivation did not establish that scientists experienced a reasonable expectation of success prior to attaining their goals.

Discussion and Conclusions

There is a constant tension in university research communities between scientific collaboration and the proprietary nature of research results. This case is an example of how statements of university inventors seriously damaged UC’s professional and financial interests in an unanticipated manner. These statements, although not the only factor in the court’s decision, nevertheless comprised significant evidence on which both the Board and Federal Circuit relied.

In addition to statements of UC’s inventors prior to the proceedings, even more damaging was the 2012 Jinek publication which supported these statements. According to UC, subsequent success in plant and animal cells, including that of Broad, relied upon the bacterial CRISPR-Cas9 technical information disclosed within this publication. The question then becomes whether this UC article should have even been published prior to UC’s own success in plant and animal cells.  After all, the possibility clearly existed that third parties could use the  publication information to develop another CRISPR-Cas9 system and then submit its own patent applications.[11] It  appears therefore, that without the publicized nature of UC’s work, there may never have been any statements which defeated UC in the Federal Circuit.[12]

However, the reality is that most university faculty and research professionals rely upon traditional research publications and the significance thereof, to augment resumes, employment opportunities, tenure and promotions. As a result, any solution for avoiding UC’s outcome inevitably rests with the university or other institution by which the researchers are employed. The solution for preventing UC’s outcome in entities with substantial commercially oriented research activities is best addressed by a risk/financial resources analysis. For example, how does a university affirmatively yet effectively manage risk of multi-million dollars lost in licensing revenue because of ‘information leaks,’ but in a cost-effective manner?

Some faculty and university sponsored startups already rely upon privately retained outside attorneys for advice on these matters. However, this legal advice is expensive on a long-term basis and becomes problematic depending upon the finances of each research group or individual. If the university assumes the risk mitigation task, then ideally whenever a research article is to be publicly distributed, there would be a (i) a prior review of the article by the legal department and (ii) a memorandum from the university legal department about risks of technical disclosures to outside research entities. However, there may be too many articles and research projects for a truly comprehensive process, and publication of time sensitive articles could be delayed to the detriment of the authors.

Furthermore, even if such a comprehensive approach were possible, a university legal department cannot realistically predict or prevent all possible confidentiality breaches. The university attorneys also most likely do not possess all the technical expertise to recognize potential financial injuries.  A potential solution is that the university retain outside counsel with specific technical expertise sets for all its research facilities, but the cost could be prohibitively high.

Therefore, the author proposes a pro-active approach with an oversight significantly smaller in scale.  For example, the university could identify which of its departments historically has provided the most commercially successful inventions.   In these departments a two-step procedure would then be implemented by the university legal department.  The first step would be a periodic meeting with an intellectual property attorney together with the relevant research staff and/faculty. The researchers would then present the status of their work, as well as the existence of communications about their work with colleagues outside the university.  If there is a potential for commercial success, the researcher should be encouraged to provide reasons that particular product or method would be commercial successful.

The second step of the proposed risk mitigation comprises review of articles or other publications from these selected research departments. If a researcher reports the preparation of a technical article for publication during a meeting, then the researcher should submit the article for legal review prior to its publication. However, this two-step process should be included in employment agreements for new faculty and researchers, as well as an amendment to employee agreements of those previously engaged by the university. The process should also be included in a university’s official employee handbook so everyone is aware and all persons are aware of these requirements.



[1]  Humans are within the larger classification of living organisms known as eukaryotes. There are three classifications of living organisms: bacteria, eukaryotes and archaea. Bacteria and archaea are each also referred to as ‘prokaryotes.’ CRISPR systems occur naturally only in prokaryotes such as some bacteria and most archaea. The invention at issue here addresses Type II CRISPR systems, and which are also referenced as CRISPR-Cas9 systems. CRISPR-Cas9 occurs naturally in prokaryotes, but not in eukaryotes (i.e., plant and animal cells)

[2] In vitro data originates from experiments conducted in a cell free environment. The article published in 2012 was written by Dr. Martin Jinek, Ph.D. who is a co-inventor of the UC CRISPR-Case9 molecular assembly. The 2012 Jinek article disclosed the UC’s inventors’ CRISPR-Cas9 in vitro results which were achieved in collaboration with other institutions. Jinek, Martin et al. “A Programmable Dual-RNA-guided DNA Endonuclease in Adaptive Bacterial Immunity” Science Vol. 337:816-821 (August 17, 2012); Public Affairs, “UC now holds largest CRISPR-Cas9 patent portfolio” Berkeley News October 1, 2019 at https://news.berkeley.edu/2019/10/01/uc-now-holds-largest crispr-ca…

[3] The America Invents Act does not authorize interferences, because under the Act the ‘first person to invent’ is not entitled to a patent. Instead, the person entitled to the patent is now presumptively the first person to apply to the United States Patent Office. See University of California et al.  v. Broad Institute, Inc. et. al., supra at 903 F. 3d at 1291 n.2. The case which is the subject of this article proceeded under the earlier U. S. patent statute which awards first to invent, and not first to file.

[4] Claims are sentences at the end of a United States patent application or patent that describe the invention in varying degrees of specificity. These sentences thereby designate the scope of patent protection that the applicant requests with varying degrees of specificity. Generally, if a claim sentence is (i) longer in text and (ii)describes numerous invention features, then the scope of protection accorded to the invention will be smaller.

[5] However, Broad contended that the CRISPR-Cas9 genetic component that UC disclosed in 2012 Jinek was biochemically distinct from the CRISPR-Cas9 genetic component of Broad’s (i) January 2012 National Institutes of Health Proposal, and (ii) disclosure in Cong, L. et al., “Multiplex genome engineering using CRISPR/Cas systems.” Science 2013 Feb. 15:339 (*121): 819-23. Appellees’ Brief filed October 25, 2018 pp. 24-25.

[6] These three prokaryotic systems that contained variants of genetic molecules (RNA) and operated in vitro were ribozymes, riboswitches and Group II introns.

[7]  The molecular systems contain ZFN and TALENS proteins, and these proteins comprise hybrids between prokaryotic and eukaryotic cell components.

[8]  Inventors may submit a U.S. provisional patent application for an initial semblance of an early filing date for a subsequent utility application. Provisional applications are not examined on the merits by the patent office, require fewer non-attorney patent office fees and never become full-fledged patents. For the risks of relying upon the filing date of a provisional application  see New Railhead Mfg. LLC v. Vermeer Mfg. Co., 298 F.3d 1290, 1294 (Fed. Cir. 2002).

[9] Under U.S. patent law, both a full-fledged U.S. utility patent application and a provisional patent application must comply with U.S. patent law and thereby

  • enabled a person of average skill in this technology to replicate the invention without undue experimentation, and
  • provides sufficient technical detail to establish that the Broad inventor(s) possessed all invention’s features and design at the time the provisional application was filed.


[10] Substantial evidence is the standard of review for a federal administrative agency’s factual findings. Substantial evidence is evidence that a reasonable mind might accept as adequate to support a conclusion. Substantial evidence has also been defined as that which a reasonable mind would accept as sufficient to support a particular conclusion and consists of more than a mere scintilla, but it may also comprise somewhat less than a preponderance.  Black H.C., Black’s Law Dictionary 1281 [5th Ed. West Publishing Co.: St. Paul Minn. 1979].

[11] Currently, under the AIA the right to a patent presumptively belongs to the first person to file the application for a particular invention.  As a result, a third party could benefit from another’s published work and then submit a patent application for a closely related invention, or even the same invention, prior the original true inventor.

[12] According to UC “all researchers’ experiments— including Broad’s—took UC’s disclosures in 2012 Jinek as their jumping off point.” Brief for Appellants filed July 25, 2018, page 41, first full paragraph.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: