The Broad Institute is testing the limits of what ‘nonprofit’ means

When a federal patent court ruled that the nonprofit Broad Institute of MIT and Harvard could legally license its version of the CRISPR-Cas9 genome-editing system, it opened the door to millions of dollars of revenue for the institute. It also contributed to the seismic shift occurring in science whereby tax-exempt research institutes established under an emerging model of “free market philanthropy” can amass money to further their research and protect their commercial interests.

The Broad Institute, a tax-exempt organization established by contributions totaling $1.4 billion by Eli Broad and Ted Stanley, is hardly the only one of these nonprofit models. For instance, Napster cofounder and former Facebook president Sean Parker set up the nonprofit Parker Foundation to explore cancer immunotherapy and other public health goals while retaining rights to any patents that might emerge from it. Facebook founder Mark Zuckerberg and his wife, Priscilla Chan, are investing $3 billion to create the Chan Zuckerberg Biohub, which would retain the exclusive right to commercialize its inventions.

The tax-exempt Broad draws National Institutes of Health funding to subsidize dozens of basic research projects, many of them conducted with commercial partners. But it is the Broad’s handling of its own CRISPR business and partnerships that threaten to undermine its nonprofit mission. These relationships can challenge the concept of “public interest,” such as when the Broad sells CRISPR licenses to Monsanto for agricultural applications, since a sizable fraction of the public may not agree that genetically modified crops serve the public interest. Or it might not be in the public interest when a favored corporation reaps a financial benefit, such as when the Broad sells exclusive licenses for CRISPR to Editas Medicine, a company that was spun out of the institute.

These nonprofit-corporate partnerships raise questions about the extent to which taxpayers should subsidize tax-exempt biomedical research by nonprofits that retain the rights to license patents and appear willing to defend them in court.

Read more: Broad Institute prevails in heated dispute over CRISPR patents

The Broad is testing how close a nonprofit can be to a corporation. Soon after patents were issued to the Broad for CRISPR-Cas9, its director, Eric Lander, turned to Third Rock Ventures, a company to which he has close personal ties, for an early shot at investing in Editas. That startup was founded by Feng Zhang, a core researcher at the Broad. Editas obtained an exclusive license to CRISPR-Cas9 from patent holders including the Broad for cash and stock equity, and began funneling $34.1 million (by the end of 2016) to reimburse the Broad’s legal fees in its court battle for the rights to CRISPR-Cas9. As millions were flowing from the now publicly traded Editas to the Broad, Editas cofounder David Liu was installed as a core member at the Broad, further strengthening ties between the Broad and its commercial partner. Zhang and colleagues then patented an application of a related genome-editing technique called CRISPR-Cpf1 and, through the Broad and other institutions, granted Editas an exclusive license to use this technology for medical applications.

This kind of tight relationship with a for-profit corporation clearly puts the concept of “nonprofit” to the test. How many tens of millions should the Broad be allowed to take from Editas? How many of the founders of Editas should be installed into leadership positions at the Broad? Does granting of exclusive licenses constitute a special favor, a quid pro quo? Those are the kinds of questions that regulators should be asking, but aren’t.

The Broad, possibly aware that its nonprofit mission could be at risk, published at the end of 2016 a guide to its intellectual property licensing philosophy. It noted that “non-profit institutions [like the Broad] should, in general, favor non-exclusive licenses over exclusive licenses.” The document immediately walked back that principle, saying that investors “would need to make a large investment to turn [intellectual property] into a commercial product” and “could not recoup this investment without exclusive rights.” By the end of the statement, the Broad made a case that special deals were good for everyone as exclusivity “may be appropriate because there is a clear case that it will better serve the public good.”

In fact, a nonprofit may not confer a “private benefit” to a corporation. Private benefit is defined as “non-incidental benefits conferred on disinterested persons that serve private interests.” The law says that any private benefit must be relatively insignificant in size to the nonprofit’s overall revenue and a necessary side effect of achieving the nonprofit’s objectives. I asked the Internal Revenue Service and the office of the Massachusetts Attorney General about this, but both declined to comment on how financially entangled the Broad and Editas could legally become and what, if any, breaches could trigger an investigation.

The privatization of federally funded science could scarcely have been imagined when national funding was first proposed. During World War II, President Franklin D. Roosevelt directed his chief of military research and development, Vannevar Bush, to create a model for funding science after the war. In Bush’s poetic 1945 document, “Science the Endless Frontier,” he argued for robust funding through the NIH and the creation of the National Science Foundation. These and other efforts led to a briskly expanding scientific research base in the US. In the 1950s, physicist and MIT president Karl Compton noted of scientists in general, “I don’t know of any other group that has less interest in monetary gain.”

That’s changed. By 1980, a landmark ruling upheld early biotech patents. The Bayh-Dole Act enabled universities, small businesses, and nonprofit institutions to pursue ownership of inventions that arose from federally funded research. Today, scientists talk openly about “figuring out the model,” whereby public money can be turned into private wealth, and escalating biotech battles are a hot-button issue.

Science, long conceived as a public trust, is now characterized by an ownership culture. That is raising questions about how much taxpayers should pay for grants that feed into this culture and its incipient biotech wars. There is something intuitively wrong about a tax-exempt nonprofit organization such as the Broad being so financially aggressive. Taxpayers should not be paying for biomedical research and development that preferentially benefits the scientific elites who hold and legally defend their property ownership.

Michael Eisen, a University of California biologist who is now a candidate for US Senate, has argued that taxpayer-funded academic scientists should not patent seminal technologies such as CRISPR-Cas9. In a more nuanced opinion published in Science, patent lawyers Jacob S. Sherkow and Jorge L. Contreras argue that research institutions should limit their use of “surrogate licensors” — such as Editas — which have exclusive licenses to use a nonprofit’s technologies.

A clause in contracts between Editas and the Broad allows the company to permit licensing of CRISPR for single-gene targets it doesn’t plan to monetize to other “third parties.” But that clause may be moot, since Editas has the right to sublicense CRISPR and would and could monetize most anything within its grasp.

In fact, Editas has already struck up a $737 million partnership with Juno Therapeutics and a $90 million deal with Allergan. Sherkow and Contreras argue that, to be fair, a nonprofit such as the Broad should license CRISPR for only one application at a time — even to a partner like Editas. Sherkow told me that “to the extent they’re going to use exclusive licenses, they should do it narrowly, on a gene-by-gene basis.” In other words, a nonprofit should limit licensing to a specific target. In the CRISPR world, that might mean that the Broad could license to Editas the right to use its genome editing technology on the CEP290 gene to develop treatments for an inherited eye disease, but not hand over the whole basket of patent applications to this one friendly company.

Read more: Sean Parker’s put up big bucks for cancer. We’ve got questions

Under a 1980 law, the public could petition the government to make CRISPR licensing more accessible if the terms of the license are not reasonable — meaning that licensing terms should be affordable and not under the control of a single company. It’s clear the Broad and Editas do not want to allow that.

The public benefit of CRISPR is uncertain. An independent report issued in March showed that a dozen candidates vying to become the first gene therapy drug approved by the FDA (none of them developed using CRISPR technology) carry jaw-dropping costs, raising questions about how most folks could even pay for such drugs. CRISPR treatments, if and when they make it to the market, are likely to be equally unaffordable.

Competition in biomedical science will continue to ratchet up through the means of large nonprofit hubs and exclusive corporate partnerships. Regulators will need to step up and ensure more equitable fair play for licensing or begin to cut federal funding streams for these nonprofits.

As taxpayers, we need to support research by the Environmental Protection Agency, the Food and Drug Administration, and the NIH. But corporate welfare should be subject to the same rigorous standards as public welfare. Biomedical researchers who command surplus cash flows and convey private benefits should be subject to a means test to determine if they should continue to receive public assistance.

Writer Jim Kozubek is the author of “Modern Prometheus: Editing the Human Genome with Crispr-Cas9,” published by the Cambridge University Press. Between October 2013 and May 2016, he worked as a data scientist for Brigham and Women’s Hospital with an affiliation to the Broad Institute.