“Nature Is the Best Inventor”: A Conversation With Feng Zhang

Jason Pontin: What was your experience of moving from China to Iowa at the age of 11?

Feng Zhang: I was born in China, and I went to elementary school in China before immigrating to the U.S. with my parents. I feel very lucky to have immigrated to the U.S. The school system in China is very much about rote memorization, but in the U.S. the teachers encouraged us to pursue things that we found interesting. I liked to take things apart, and the school had plenty of classes and outside activities that encouraged kids to experiment and learn how to build robots or interesting devices.

JP: Can you recall what most startled you about America?

FZ: Of course China has many, many more people than the U.S. When I first moved to Iowa, I’d hardly ever see any people on the street. In China, I was used to seeing crowds of people walking or crossing intersections, and in the U.S., it was very different.

JP: If you were taking apart robots and interested in mechanical things, what led you to biology and your current research at the Broad [Institute]?

FZ: I went to a public school in Des Moines, Iowa. One of the programs was a Saturday enrichment program. They had Saturday classes for kids who wanted to just learn something different. One of the classes was on molecular biology. I hadn’t really known what molecular biology was before that, and in fact hadn’t enjoyed the biology classes I took in school. It had mostly been about memorizing things and dissecting smelly frogs, but going to that class, we did experiments like extracting DNA from strawberries. We ran gels to see what DNA looks like and we watched movies like Jurassic Park. It was that combination of learning about the fundamental principles of DNA and biology and watching a movie that showed how these concepts were not too unreal that got me really excited.

JP: It’s a miraculous moment—at least it was for me—when you realize how molecular biology is at the one and the same time underpinned by extraordinarily simple principles and yet possesses an endlessly fascinating, blooming complexity.

FZ: Exactly. And seeing how there is a code for biology and how you can make sense of that code and use the code to do things that would help people—creating new genetic medicine, engineering plants that can give higher yield or be resistant to drought—that just seemed really amazing.

JP: What is current focus of your lab, one of the largest and most active at the Broad Institute?

FZ: One of the main areas that we focus on has been to develop useful biotechnology for improving human health, and the approach that we take is by learning from nature. So far, I’ve been fortunate to work on a few problems that have borrowed intricate and also powerful mechanisms that nature has evolved over billions of years, and turning those into tools. Those experiences taught me that nature is really the best inventor out there in terms of biotechnology, and it’s had a lot more time and a lot more experiments to help it develop its inventions.

We can learn a lot from nature. We are using computational approaches combined with experimentation to discover new mechanisms from nature—new ways that bacteria or cells defend themselves against infection, or things that allow them to survive in interesting or very diverse environments, and also things that allow them to rapidly evolve, to be able to adapt to changes in the environment. And a lot of those mechanisms are some of the most robust mechanisms in life, because they are things that decide whether a cell survives or dies. Under that kind of strong selection, the molecular mechanisms that are developed are very robust. We have benefited from some of them—restriction enzymes for recombinant DNA, microbial defense mechanisms like CRISPR. They all came from these kinds of systems involved in providing survival mechanisms. We’re trying to find more, and some of these we’re hoping to also turn into new biotechnology to improve people’s lives.

JP: You are most associated with CRISPR-Cas9. What is it, and why was it such a breakthrough over previous gene-editing techniques like TALENs or zinc fingers?

FZ: Since the sequencing of the human genome, one of the things that scientists wanted to do is to be able to go into the DNA of cells and be able to change those DNA letters. That way we can ask questions like What does one DNA sequence do? or What is the mutation that leads to disease? If we can figure out what causes disease, then the hope is to also correct the mutation, to undo the disease, to improve health. And for that particular problem, several iterations of technologies have been developed. But most of them were very difficult to use. It’s very challenging to have to engineer a complicated protein to be able to home in on a specific mutation in the genome. CRISPR, on the other hand, is a much simpler mechanism. Rather than engineering a protein, we can simply specify the mutation that we want to change in the DNA by synthesizing a new piece of RNA. This is nowadays something that you can go online and order from a company. Once you type in the sequence, you get a FedEx envelope a couple days later. And it’s that simplicity that makes the tool significantly more accessible, and also that simplicity made it possible to develop it for many other applications that we couldn’t do with previous technologies. For example, we can now target every single gene in the human genome in one experiment and ask questions like Which gene is involved in the metastasis in cancer? or Which genes are involved in driving increased growth rate or transformation and differentiation into a different type of cell? And that kind of scale and accessibility opens up new ways to study biology.

JP: What is CRISPR best at?

FZ: I think CRISPR is a very broadly useful technology. Scientists are applying CRISPR to study anything from the basic biology of how cells divide and replicate all the way through to understanding diseases like cancer, psychiatric disease, and diabetes. Because CRISPR is so much easier to use and more scalable than anything before, it is making science much faster. Just to give you one comparison, scientists use mice as a model to study disease, and oftentimes they would make a transgenic mouse by modifying a gene in the mouse to see what happens. Using previous approaches, it would take maybe a year, or a couple of years, to make a mouse. Now with CRISPR you can do this in three weeks.

JP: What are the limitations of CRISPR as a technology, and how does your work address those limits?

FZ: CRISPR-Cas9 is like a pair of molecular scissors: It can cut DNA at a specific place. Once you cut DNA, you can disrupt that DNA. So it’s good for inactivating a gene. But there are even more sophisticated changes we want to do. We may want to swap out something, or we may want to insert a large piece of DNA. CRISPR-Cas9 by itself is not so effective at doing those other things. Scientists are working on developing new ways to be able to make more complicated changes. So, for example, just earlier this year we reported a new mechanism: It's called a CRISPR-associated transposase, and it provides a new way to be able to insert large fragments of DNA precisely into the genome. And that will open up opportunities for engineering cells for therapeutic applications or for research applications. The field is still moving quite rapidly: I think there are more natural mechanisms we’ll discover, and some will help us further advance genome editing.

JP: Does CRISPR have off-target effects, and are they predictable?

FZ: CRISPR is known to be able to not only target where you program it to target, but also go elsewhere in the genome—that’s called off-targeting. And depending on the specific CRISPR system that the scientist uses—nowadays there are many more systems besides Cas9; there’s also CRISPR-Cas12, Cas12a, Cas12b, and so forth—some are more specific than others, especially for therapeutic applications. This is a major area that needs to be further developed, because we want to be sure that we can be very precise at changing only what we want to change.

JP: Can these gene-editing techniques be used for neurons to treat neurodegenerative diseases like Parkinson's and Alzheimer's?

FZ: I think that’s one of the really exciting potential applications. Using the standard CRISPR-Cas9 system, it’s possible to disrupt genes. So if a specific gene is mediating disease, we can try to get rid of that gene and then treat the disease—for example, maybe a gene that’s involved in the glycogen storage disorder. And then as we develop more systems, we may be able to fix mutations in neurons, maybe treat things like Huntington’s disease or replace big pieces of DNA for different forms of genetic neurological disease.

JP: Somewhat unbelievably, you are still in your late thirties. Where do you think these biological controls will be in, say, 25 years?

FZ: The more we know, the more we can do, and the more we can help treat disease. But it’s hard to predict what will happen 25 years from now. It’s easier to see what may happen in five to 10 years. In the next 10 years the whole area of genetic medicine or cellular medicine is going to take off. Our ability to modify genes or cells is providing a new way to be able to get at the root cause of disease. So rather than just treating the symptom, we can correct the underlying cause of the mutation, so that the person gets one treatment and is cured. At the same time, the ability to engineer cells—especially cells that we can transplant into the body, whether they’re immune cells or stem cells—provides new ways to be able to treat illnesses by restoring function in the body. So, for example, if someone has a liver disease, we may be able to put back into the person liver cells that have been repaired, and they can regenerate liver tissue. Or if someone has cancer, there’s already great progress in immuno-oncology, engineering immune cells to get rid of cancer cells.

JP: Will we need new computational methods to understand the irreducible complexity of the cell? The cell has emergent properties that are difficult for our ape brains to understand.

FZ: Computation already plays a huge role in biological research and in the future will play an even greater role in the development of medicines. We only know what we don’t know. There are probably many things that we don’t even know about yet that we don’t know. As we study biology and begin to uncover new things, it will open up many new questions: the use of technology to profile biology, to measure things all the way from DNA to RNA to protein and then to the chemical molecules that cells and organisms are producing. It is going to be generating massive amounts of data, and that will require not only powerful computational capabilities but new algorithms to allow us to make sense of that information and to use that information to inform new treatments. So I see the synergy between computation and the development of higher-throughput experimentation to be critical for moving biology forward.

JP: What don’t you know now that you still yearn to understand?

FZ: I think it’s amazing how much diversity there is in nature. There is an enormous amount of effort going on right now to profile that diversity. People are sequencing organisms from many different diverse ecological environments. But that’s not enough, because a lot of natural diversity is becoming extinct. Species are going away. The Amazon rain forest is being burned. We’re not capturing that snapshot of nature’s creations quickly enough. I think we need to do more about that and make sense of the information we’re collecting about nature’s diversity—both preserving and understanding it.

In the long run something that I got really excited about recently was reading about the discovery of a new planet within the habitable zone of a galaxy, more than a hundred million light years away from us. Is there intelligent life out there? What is the form of this intelligent life? What is their biological substrate? Do they use DNA? Do they have something completely different from us that allows them to replicate and transmit information from generation to generation?

JP: Do you have an intuition? Do you believe that complex polymers would inevitably develop into macro-molecules like our own—or do you think it could be something absolutely dissimilar?

FZ: I think we have to be humble, because nature has had a lot more time to invent new things. We should be open-minded that there could be many other ways that nature could have made things work.

JP: Do you think we are close at all to understanding the biological code of the brain?

FZ: I think that’s one of the most mysterious questions. I don’t think we’re quite there yet, but scientists are making good progress in dissecting circuits in the brain and understanding what one brain cell is sending into another brain cell to mediate things like memory or emotion.

JP: Do you think that any biological, mechanistic explanation could in principle unpack the hard problem of consciousness?

FZ: Consciousness is a difficult thing to think about because in order to answer a question we will have to frame the question in a way that defines what consciousness is. It’s kind of hard to frame what we are trying to figure out. There are really great thinkers who have begun to put a framework around the biological substrate for consciousness. They ask, How do cells organize themselves or send information to each other to mediate what may be conscious thoughts? We need to put a framework together to allow us to ask the right questions and do the right experiments. But I think that will also take some time.

JP: You believe there should be a moratorium on the application of CRISPR to so-called germline interventions—no permanent changes to the human gene pool—until we know more about the risks. What would have to change for you to feel comfortable about altering our common genetic inheritance?

FZ: For the foreseeable future, it’s very difficult to imagine a situation where we absolutely have to use germline editing. That’s because there are alternatives to germline editing for the purpose of treating disease or for preventing disease. Couples may elect to use in vitro fertilization followed by a genetic screening to identify embryos that don’t carry specific mutations. That would prevent disease without having to use gene editing.

The even bigger ethical issue is really around the use of gene editing for enhancement: making designer babies and so forth. But biology is complicated, and we don’t really know when you change one thing how it could impact something else in the body. We have seen examples of nature’s own mutations. A mutation in a gene called CCR5 can prevent HIV infection, but it also increases susceptibility for West Nile virus or influenza. Because of these complicated interactions, I think it’s difficult to think that we know enough to do these enhancement experiments. But that’s something that I think society will have to think about as we begin to accumulate much more knowledge.

JP: Genes are mostly conserved for good reasons, and we could knock out one gene that is deleterious late in life but has important functions in early development, or confers some other advantage.

FZ: I agree. It’s important to be humble.

JP: To what do you attribute your success as a researcher? You’ve been very original in the way you think.

FZ: I think there are a number of things that contribute to what I do. I think first and foremost, having great parents who encouraged me to pursue things that I like and found the right opportunities for me. Also immigrating to the U.S. so that I could get a different type of education, and of course family being very supportive. And then, also the mentors who are there to help me discover what I enjoy and what I intellectually find exciting. That together with my co-workers who work with me on a day-to-day basis to solve some of these challenges. But the other thing that I try to think about whenever I am pursuing a new project is Will this be useful? Can we somehow see if there is a path to turning it into something that can benefit people? And try to follow that path.

JP: You are one of that small cadre of scientists who has attracted a public reputation outside science. What do you think it is about your work that fascinates the public?

FZ: I think it’s probably a combination of the work having captured people’s imagination and also the way it touches on many different areas of society. On one hand, there has been a lot of science fiction about changing DNA and controlling our own biology. But then at the same time, because of the potential impact of gene editing on health care, on agriculture, on many other ways of using biology for bioproduction, people are excited about the continued development of this technology to solve the world’s problems.

Editor's note: This conversation has been edited for length and clarity.