In a lab pushing the boundaries of biology, an embedded ethicist keeps scientists in check

The young scientists had a question. They were working with mouse embryos from which all living cells had been chemically dissolved away.

So far, so good, thought the bioethicist, as she listened to the presentation at a Harvard Medical School lab meeting.

The scientists were seeding the mouse scaffolds with human stem cells. Those cells were expected to turn into human liver cells and perhaps a mini human liver; and human kidney cells and perhaps mini human kidneys; and human heart and brain cells and …

Wait.

Jeantine Lunshof insists she is not the “ethics police.” It says so on the door to her closet-sized office at Harvard. She doesn’t find reasons to reflexively shut down experiments. She doesn’t snoop around for deviations from ethical guidelines. But when scientists discuss their research in the twice-weekly lab meetings she attends, “I will say, hmm, that raises some good questions,” Lunshof said.

There is no shortage of “good questions” for Lunshof, who for the last three years has been embedded in the synthetic biology lab of George Church, the visionary whose projects include trying to resurrect the wooly mammoth and to “write” a human genome from scratch. Church is also famous for arguing that it is ethically acceptable to edit the genomes of human embryos if doing so will safely alleviate suffering, and for encouraging people to make their full genome sequence public, privacy be damned.

In the Church lab, Lunshof told STAT, “you have incredibly interesting conversations.”

Rapid advances in genomics and stem cell biology are forcing researchers to regularly confront ethical quandaries that seem straight out of science fiction. The power to create organisms with cells, tissue, and even organs from different species, called a chimera, raises thorny questions: What is the moral status of a primordial human brain nourished with a rudimentary heart and circulatory system, all inside a mouse scaffold? Can it feel pain? Should it not be created in the first place? Genome-editing presents other challenges: Where does therapy end and enhancement begin? Could genome-editing to prevent dwarfism, for example, go a little further and create a future NBA star? How should society balance competing values such as autonomy, like the freedom of parents to do everything they can for their children, and justice, as in not creating classes of genetic haves and have nots?

“George is far ahead of everyone else” in the kinds of experiments he undertakes, said John Aach, a senior scientist in Church’s lab who works closely with Lunshof. “She performs a service in making them slow down to where the rest of the world is. Otherwise George might stumble. It doesn’t take much to stumble and make a mess of things. Jeantine keeps things moving on the bioethics side as the science is moving ahead.”

Lunshof’s role is unusual if not unique. Genetics researchers “will tap a bioethicist to join a grant or consult on a project, but it is rarely if ever the case that a genetics lab has a full-time bioethicist,” said Brendan Parent, a bioethicist at New York University. He and others are unaware of any other such embeds. Instead, bioethicists and biologists tend to interact when they serve on committees convened by universities, scientific organizations, or government.

In contrast, Lunshof not only coauthors papers with Church and his colleagues, but also helps draft protocols for some of the cutting-edge science the lab conducts. By being present at the creation, she is able to flag ethical minefields before the lab finds itself bumbling across one.

“She provides me with a comfort zone,” Church said. “I think much more about societal concerns” that the lab’s research might raise. “She’s here while we’re just starting to think about experiments,” he added, and because of her “we talk about [bioethics] earlier than most groups do. Jeantine is fearless in what she tackles.”

The benefits of this collaboration extend beyond Church and his lab. Watching new biology emerge in real time has enabled Lunshof to develop much-needed new ways of thinking about bioethics, giving her field and the world outside the lab a fighting chance to keep up.

‘Don’t go down that road’

It’s at the lab meetings every Monday and Thursday afternoon that Lunshof typically learns what might next land on her to-do list. The 50 or so scientists in attendance update Church on their research and others offer comments. The rows of chairs are generally all filled. Lunshof, in typically casual lab attire, rarely asks questions, instead taking notes and keeping track of who she needs to follow up with.

This week, researchers discussed plans to do “cognitive testing” on participants in a project centered on having their genomes sequenced. Lunshof’s ears pricked up.

The combination of genetics and intelligence has long been a danger zone, largely because measurements of intelligence are imprecise and shaped by the dominant culture, as decades of debate about IQ tests have shown. The tests do not measure “cognition,” let alone intelligence, Lunshof said during the meeting, arguing for “staying away from linking the genome to cognition or IQ.” She urged the scientists to be more precise in describing what the tests measured: memory and mental processing speed. “Correcting things later by saying, ‘No, we are not measuring IQ, really we’re not,’ is very difficult,” she said.

“When I feel that something is a problem, I feel completely free to say, ‘Don’t go down that road,’” Lunshof said in an interview. She is not paid by Harvard. Born and raised in the Netherlands, she is an assistant professor there, at University Medical Center Groningen, and she was awarded a Marie Curie fellowship to move to Boston and support her work in Church’s lab.

“No one in the lab ever puts pressure on me to legitimize anything or to agree with what they’re doing,” she said. “I am always on the alert for things that could get into delicate areas.”

Lunshof’s collaboration with Church began in 2006. It was the start of the Personal Genome Project, an effort to sequence people’s full genomes and mine the data to link genetics to health. Church was causing consternation by proposing that people make their genome and their health history publicly available.

“My first reaction was, this is totally crazy,” Lunshof recalled. “Anonymity and confidentiality were central to everything we do in biomedical ethics.”

But then she thought, what if Church is right? He had argued that it’s impossible to guarantee that a DNA sample would remain anonymous. (He would be proved right in 2013.) So why not do away with that charade at the outset, and instead of making empty promises of anonymity, tell volunteers from the get-go that anyone could know who they were?

Lunshof had studied philosophy and Tibetan language and culture as an undergraduate, then had written a doctoral thesis on ethical issues in genomics. She also had earned a nursing degree, and worked at the Netherlands Cancer Institute in Amsterdam. In 2006, she stumbled on the PGP website and sent an email expressing her interest in it. Church replied within hours and their partnership was born.

Together, they developed a new form of patient consent for the Personal Genome Project. Called “open consent,” it was founded on principles new to the bioethics of genetic research. It tells participants they won’t have privacy and confidentiality. Instead, consent is based on values such as reciprocity (scientists and volunteers interact as equals) and veracity. Lunshof is also a big believer in the ethical concept of citizenry, including allowing one’s genetic data to be accessed by all qualified scientists to help advance medical progress and alleviate human suffering.

“Because of advances in genetics and genomics, it made sense to abandon the traditional idea of medical confidentiality,” Lunshof said, “or at least not make it central.”

That was a minority opinion. The National Institutes of Health, a main funder of Church’s lab, “wasn’t ready to embrace the idea of genetic privacy being violable,” said Aach. “It and the genetics community went in the other direction, saying we have to take steps to protect privacy,” a huge and costly undertaking.

With the development of open consent, the Personal Genome Project took off, and now has more than 5,000 participants in the United States alone.

The ethics debate around genomics intensified with publication of a breakthrough 2012 paper on CRISPR, the revolutionary new genome-editing technology. After Church and his team got CRISPR to edit the genomes of human cells, later that year, they and others quickly faced two quandaries: Should CRISPR ever be used to “enhance” people’s genetic inheritance? Should it be used to edit the genomes of human eggs, sperm, or early embryos, producing changes that could be inherited by offspring and, maybe, generations of designer babies?

For many scientists and ethicists, the line-in-the-sand position on such germline editing and genetic enhancement has long been no. Lunshof had other ideas.

“From the bioethics standpoint,” she told STAT one afternoon at a Harvard Medical School cafe, “it is not clear why altering genes [for enhancement] is by definition unethical. Some philosophers have consistently argued that there is a duty” to at least consider genetic enhancement.

In the real world, “Prospective parents decide to use — or not to use — reproductive technologies,” Lunshof argued, and that could one day include germline genome editing.

That reflects the balancing act she brings to the ethical puzzles she tries to unravel. Sometimes two core values are in conflict. In the case of germline editing and enhancement, parental autonomy (to make reproductive choices) might clash with the idea that all children are entitled to an equal start in the world. But the latter is honored in the breach more than the observance, Lunshof says, and so should not be allowed to trump parental autonomy.

Last week, a report from the National Academy of Sciences and the National Academy of Medicine opened the door to germline editing. It opposed enhancement, but called the line between enhancement and therapy blurry. Lunshof beat them to it: “The criteria for what is therapy and what is ‘enhancement’ are fluid,” she wrote two years ago.

For all the passions that germline editing incites, its effects would be small: It requires in vitro fertilization, so few parents would use it (unless reproductive sex goes the way of flip phones). Other applications of CRISPR could be more consequential. One could alter ecosystems. Called “gene drive,” it is a technology for editing the genomes of an organism in a way that causes the change to be inherited by every offspring, contrary to usual inheritance patterns.

As scientists in Church’s lab and elsewhere involved the public in conversations about testing gene drives in wild populations of mice or mosquitoes, Lunshof recently raised a novel bioethics question: If a “bioneer” community says yes to gene drive, “it sets a precedent” and could lead people in other places to allow it, too, she said. “How much would this community be held morally accountable for genetic interventions elsewhere that go wrong?”

No rules applied

The ethical minefield created by the possibility of seeding mouse embryo scaffolds with human stem cells, and possibly growing a functional, if mini, human brain, has been trickier to navigate. “You’d grow human organs,” Lunshof said. “My question was, what if this worked?”

There didn’t seem to be any government or other rules against it. Scientists using stem cells from embryos are supposed to clear experiments with an Embryonic Stem Cell Research Oversight (ESCRO) Committee, which many research universities have established. But Church’s lab proposed to use stem cells produced by reverting adult cells back to an embryo-like state. And although there are rules against creating human chimeras, it wasn’t clear whether this thing would be a chimera: It wouldn’t be a single living entity, though it might have living human cells or even organs. It seemed there was no bureaucracy to stop the experiment.

Lunshof spent hours with the two scientists who were planning the experiment. “She primes the lab to be sensitive to ethical issues even when they don’t know what to be sensitive about,” Aach said. She proposed asking the ESCRO committee. Church agreed. It decided that the experiment did not violate any known guideline but asked him to keep the committee informed as the experiments progressed.

As it happens, the experiments didn’t work and the lab moved on — smack into another ethical conundrum.

This time, postdoctoral fellow Eswar Iyer was using a process called micropatterning to create special surfaces on glass slides. Placed on them, human stem cells formed a precisely shaped little colony that differentiated into one or another organ.

Iyer described this work at a 2015 lab meeting. Two phrases made Lunshof sit up: “embryo-like features” and “generation of cerebral organoids.”

It was the de-cellularized mouse dilemma all over again, but with glass slides instead of mouse scaffolding, and, again, no rules seemed to apply. There are federal prohibitions against allowing an embryo to develop past the point where it forms a structure called a primitive streak, which happens on the 15th day after fertilization. At this point the embryo can no longer split (into twins) and is therefore widely regarded as a “morally significant individual.” But human cells or tissues developing on the micropatterned surfaces never form a primitive streak; only whole embryos do.

The question, Lunshof said, was, “What is the threshold where a synthetic entity is enough of an embryo that the same moral questions” must be considered?”

That question loomed even larger with those cerebral organoids, primordial mini-brains that are even more realistic and “much more embryo-like,” Church said.

Cerebral organoids, too, fall through the cracks of the rules on embryo research, Lunshof said, “but we know we’re doing things that involve the same ethical issues that inspired the rules,” such as when human life begins, when something has a moral status, and whether — since this is brain tissue — the thing is sentient.

After the lab meeting, Iyer dropped by her office. The 10-minute visit he expected lasted two-and-a-half hours. Lunshof not only asked him to explain every detail of every slide he had shown. Their conversation also ranged into Western and Eastern philosophy (Iyer is a Hindu), especially views on when life begins. They agreed to keep talking.

Lunshof gave Church a rundown of the discussion, began looking for scholarly papers that might shed light on the ethically-uncharted territory, and figured out what rules are applicable. She also took the helm of a working group on the ethics of embryo-like entities.

One result is a paper to be published in eLife, an online biology journal. In it, Lunshof, Aach, Iyer, and Church propose “that research limits for these entities be based as directly as possible on the generation of morally concerning features.” (The “entities” are called SHEEFs: Synthetic Human Entities with Embryo-like Features.) For instance: How human are the cerebral organoids? Do they feel pain? How could you tell?

Just because the thing cannot develop into a baby is not a valid reason to green-light the experiments, Lunshof said. She believes that if human cells are highly organized and display “functional interactivity” — as a blood supply in a cerebral organoid would — then one must at least consider the possibility that the SHEEF has “moral status.”

Lunshof also initiated a discussion of SHEEFs with Harvard’s stem-cell oversight committee, which led to a meeting last November at Harvard Law School. There, Church explained that it is possible to get blood vessels to infuse cerebral organoids, which “allows us to go to larger and larger organoids.” So far, he said, “we can see beautiful structures very similar to advanced cerebral [tissue]. … There is essentially no limit to the technology, so we need to focus on the ethics and the humanity” as guides to how far to take the science.

Which means Lunshof is unlikely to run out of “good” questions.