ABOVE: Ticks can acquire the bacteria that cause Lyme disease from infected white-footed mice, but genetically engineering Lyme-resistant mice could help stop the spread of the disease. Jimmy Day

Today, Massachusetts Institute of Technology biologist Kevin Esvelt is well known for his work on guided evolution technologies—creating systems for evolving biomolecules in the lab and developing techniques to shape the evolutionary trajectories of species in the wild—as well as forging new pathways to safeguard these technologies from misuse.1,2 

Esvelt’s entanglement with evolution began early. As a child, he visited the Galápagos, and was captivated by the islands’ stunning array of unique wildlife. “That sparked an interest in the evolution of creatures in the natural world,” said Esvelt. “It got me reading Darwin. And I started wondering—could we make things of comparable magnificence?” 

So, when he joined David Liu’s research group at Harvard University for his graduate studies in 2004, he jumped into exploring how to put evolutionary processes to work in the lab. “I love solving problems that I am not actually smart enough to solve. And to do that, you need access to something that is effectively smarter than you, or at least can execute search strategies that you can’t,” said Esvelt. “One of the reasons I love playing with different disciplines is they all give you different lenses on the world. And my favorite lens, evolution, is really about search strategies through a hyper astronomically vast space.” 

One of the reasons I love playing with different disciplines is they all give you different lenses on the world.

 —Kevin Esvelt,  Massachusetts Institute of Technology

Astronomically vast is no exaggeration; as Esvelt pointed out, there are more possible 100-amino-acid proteins than there are atoms in the universe.3 “So how do you search through that space for the ones that actually work? Obviously, evolution can do it because we’re all here.” Esvelt wanted to create a system that made it easier for scientists to harness these evolutionary strategies to develop new molecular tools. 

In some ways, humans have been using evolutionary strategies to develop new biological entities for millennia, selecting plants and animals with desired traits until they had lap dogs and an astonishing number of chicken breeds. As Esvelt noted, even before Darwin’s On the Origin of Species, veterinary scientist William Youatt, in the context of breeding sheep, wrote in the early 1800s that selection, “enables the agriculturist not only to modify the character of his flock, but to change it altogether—the magician’s wand, by means of which he may summon into life whatever form and mould he pleases.”4 By the 1970s, researchers began using this “magician’s wand” at the molecular level, directing the evolution of proteins in the lab.5

While these systems offered a promising new way to modify the stability or activity of proteins, they were also extremely time- and labor-intensive. Scientists generated large numbers of mutants and screened them for the desired functions. Then they used the best variants to generate another round of mutants, slowly and painstakingly accumulating beneficial mutations. Esvelt’s system, dubbed phage-assisted continuous evolution, or PACE, allowed scientists to hit fast forward, evolving a protein with a specific function in as little as a week.1 

The PACE system has two main parts: M13 phages and E. coli cells. The phage contains the gene of interest that scientists want to evolve to perform a specific function. Scientists then delete the gene encoding the phage’s pIII protein, which it needs to produce infectious progeny. Researchers insert this gene into a plasmid in the host E. coli and rig the cell so that its production of pIII depends on how well the infecting phage’s gene of interest can perform the desired function. In this way, the better the protein produced by the phage gene of interest performs a particular task, the more pIII the host cell produces, resulting in more infectious progeny. In each replication cycle, the best performing phage outcompetes the other phages, effectively automating the screening process that scientists previously had to do themselves. Throughout each phage generation, the gene continues to evolve for improved activity. 

This automated screening, coupled with the phage’s short life cycle, substantially accelerated the speed of directed evolution experiments. In the following years, researchers used PACE to examine the reproducibility of evolution, coax proteases into cleaving specific protein targets, and evolve improved gene editing tools.6–8

          Three grayish marine iguanas sun themselves on a rocky beach.
The unusual wildlife of the Galápagos, like the marine iguanas pictured here, helped spark Kevin Esvelt’s interest in evolution.
© istock.com, Maridav

First forays into CRISPR

Meanwhile, Esvelt moved to the Wyss Institute to work with geneticist George Church. He intended to continue working with the PACE system, but his work was repeatedly foiled, albeit unintentionally, by fellow Wyss Institute researchers.

“The DNA origami folks were doing eight-liter phage preps in the warm room,” Esvelt recalled, “which meant that the air was saturated with infectious wild type phage. I just could not maintain a PACE culture without it getting contaminated.”

Incidentally, this resulted in his first foray into the world of CRISPR. While CRISPR’s gene-editing applications were not yet widely recognized, the function of the CRISPR-Cas system as a bacterial immune system was well established.9 Esvelt wanted to use the system for its original purpose: to protect the bacteria in the PACE culture from infectious phages. But when University of California, Berkeley biochemist Jennifer Doudna and Umeå University molecular biologist Emmanuelle Charpentier published their seminal CRISPR research in 2012, it was clear that CRISPR would become a game-changing technology.10

Esvelt teamed up with Prashant Mali, a fellow Wyss Institute researcher who had been working on TALEN gene editing, to explore the possibilities of CRISPR. Only six months later, Mali, Esvelt, Church, and others at the Wyss Institute published one of the first papers on CRISPR gene editing in eukaryotic cells.11 

Even as other researchers swarmed into the nascent field of CRISPR research, “Prashant and I were getting kind of bored of it,” said Esvelt. “We wrote a review that basically said, ‘Here’s what everybody’s going to do for the next three years. This is blindingly obvious. We don’t want to do this, so we’re just going to point it out for everybody. So, here’s your roadmap. Goodbye. We’re going to do more interesting things.’ It was a silly thing to do from a career perspective.”

It was not really goodbye after all. Having laid out the “obvious” uses for CRISPR, Esvelt began mulling over some not-so-obvious applications. His work ultimately took CRISPR in a new direction, transforming it into a tool that could enable humans to precisely and purposefully alter the trajectory of evolution on a global scale. 

Into the wild: CRISPR gene drives

While it was clear that CRISPR could be used in cells, laboratory animals, and one day, humans, Esvelt began to toy with the idea of using it in a new context: in the wild. Although humans have been placing selective pressures on other organisms since time immemorial, intentionally evolving animals through domestication, for instance, this generally results in species that simply can’t compete in the wild. Any genetically edited wild organism would not only need to have the desired trait, but also need to out compete its wild type counterpart in order to survive and spread. 

With this technology, one person can decide to edit a lab organism that—if released—will change the entire species.

 —Kevin Esvelt,  Massachusetts Institute of Technology

Esvelt needed to tip the scales to favor the inheritance of the desired trait. In normal patterns of inheritance, if an organism with an edited gene mated with a wild type animal, on average, 50 percent of the offspring would inherit the edited version. These patterns, however, can be altered by a gene drive, which increases the probability that a particular version of a gene will be passed on to offspring, causing this gene to spread throughout the population even if it does not increase the organism’s fitness. 

          Esvelt walks through a field, dragging a piece of white fabric through the grass.
Esvelt uses a piece of white fabric to collect ticks as part of the Mice Against Ticks project.
Jimmy Day

In 2003, evolutionary geneticist Austin Burt of Imperial College London published a paper on the possibility of creating such a drive using selfish genetic elements called homing endonucleases.12 In the interim, research groups around the world worked on various strategies to engineer gene drives in insects with the eventual goal of eliminating malaria by modifying the parasite-carrying mosquitos.13 But a CRISPR gene drive would be different: It would be easier to engineer and more stable than the gene drives developed using the other contemporary gene editing strategies.2

To create a CRISPR gene drive, scientists engineer an organism’s DNA to contain the desired gene along with genes coding for guide RNA and Cas9 molecular scissors. When the edited organism mates with a wild type organism, they both contribute one copy of each chromosome. But the edited organism has the genetic instructions to cut out the wild type partner’s unwanted gene, and to insert the desired gene and the guide RNA and Cas9, guaranteeing that all of the offspring—and all their offspring, and so on—have the desired gene. 

“When I realized CRISPR was the answer, I was super excited,” recalled Esvelt. “I thought, ‘It’s not just malaria! If we get malaria working, then we can hit schistosomiasis, we can hit locusts, we can do all of these other things.’”

The excitement soon gave way to concern about the implications of such a powerful technology. “With this technology, one person can decide to edit a lab organism that, if released, will change the entire species. And I thought, ‘if you can engineer a mosquito to never carry a disease, I bet you can engineer a mosquito to always carry a disease.’”

“I didn’t tell anybody. I didn’t even tell George,” he confessed. “I love George, but I’m not totally sure he can keep a secret about a new idea if it’s sufficiently awesome. And this was a pretty awesome idea.”

However, after much thought, Esvelt concluded that this technology likely favored defense for three reasons. First, the drive would be fairly slow, only spreading from parents to offspring, not horizontally throughout the population. Second, the drive would be easily detectable: The DNA encoding the drive can’t hide from genetic sequencing. Third, a weaponized gene drive could be countered with another drive to overwrite any dangerous genetic material. 

We all know that evolution does really interesting things. So, if you put a selective pressure on something like a malaria parasite that’s been around since the dinosaurs, could you end up selecting for something that’s worse?

 —Omar Akbari, University of California, San Diego

In 2014, Esvelt and his colleagues became the first to publish the idea of CRISPR-based gene drives. Unbeknownst to him, a group at the University of California, San Diego had been independently working on a similar system and published their successful CRISPR gene drive in fruit flies the following year.14

Even after concluding that the technology favored defense, Esvelt has been a strong advocate for ensuring that it is used safely and ethically, citing the importance of public trust in determining the successful application of promising technologies. “I was afraid of a future in which gene drive was like GMO or [nuclear power].” 

See "Combating Mosquito-Borne Diseases with CRISPR"

At the time, said Omar Akbari, who studies mosquito genome engineering at the University of California, San Diego, Esvelt was one of the most vocal advocates for safety, endeavoring to draw attention to the risks and the possibility of unforeseen outcomes.

“We all know that evolution does really interesting things,” said Akbari. “If you put a selective pressure on something like a malaria parasite that’s been around since the dinosaurs, could you end up selecting for something that’s worse? How do you measure that risk? And how do you get people to accept such a thing?”

“These themes that Kevin illuminated were really important to discuss,” said Akbari. “In some ways, it helped me shape my own research direction….Before that, I wasn’t thinking about these kinds of things too much. But then when I started to listen to Kevin and talk to ethicists and social scientists, I understood that these fears are real. And they really haven’t gone away either. The kinds of things that people thought were scary back when gene drive was really becoming exciting back in 2014, they’re the same fears that still exist today.” 

In 2015, Akbari, Esvelt, and dozens of others in the field published a set of recommendations for physical, molecular, and ecological barriers to prevent the escape of organisms carrying gene drive constructs.15 In the following years, many researchers in the field shifted towards developing more easily confinable systems using CRISPR-based strategies that could affect local populations without the risk of global spread.16,17

Mice Against Ticks, and a community against Lyme disease 

While many scientists in the field are focused on malaria-carrying mosquitos, Esvelt landed on a project closer to home. The northeast United States is free from many vector-borne diseases, including heavy hitters like malaria, dengue, and schistosomiasis, but it does have an ever increasing burden of Lyme disease.

“I chose this not because I think Lyme disease is the worst problem in the world—although it is awful—but because I wanted to figure out how to develop ecotechnologies in a well-supported and ethical manner,” he said. For this project, he said, “We want community-guided technology development. We want to pioneer a new state of the world in which the people developing the technologies that will define the future actually go to the communities that they think would benefit from these technologies, and ask, ‘Do you want this at all?’” 

          Sam Telford III, Joanna Buchthal, and Kevin Esvelt stand outside on a dock on a sunny day.
Pictured from left to right, Sam Telford III, Joanna Buchthal, and Kevin Esvelt are working on community guided ecotechnology development.
Jimmy Day

The Mice Against Ticks project was born as a collaborative effort between Esvelt and Sam Telford III, a vector-borne disease researcher at Tufts University.18 Telford has been studying Lyme disease for more than three decades, and has been involved in numerous projects aimed at protecting the public: advocating for deer population reduction, promoting the use of personal protection strategies, and assisting with the clinical trials for LYMErix, a Lyme disease vaccine that was FDA approved in 1998. (This vaccine was voluntarily withdrawn after three years, potentially due to low uptake and largely unfounded fears about side effects.)19 Despite the efforts of Telford and many others in research and public health, Lyme disease cases only continued to rise.

“For the longest time, there had been nothing really new,” said Telford. “When Kevin contacted me and outlined what he was thinking of doing, it was the first truly new idea I had heard in my career of trying to prevent Lyme disease.”  

Esvelt, Telford, and Joanna Buchthal, a researcher in Esvelt’s group, worked closely with communities in Nantucket and Martha’s Vineyard to discuss and develop strategies for controlling Lyme disease through genetically editing wild populations. In the wild, ticks become infected with Lyme-disease-causing Borrelia bacteria when they feed on infected hosts, often white-footed mice. Researchers had known for years that mice could be immunized against Lyme disease by causing the mice to develop antibodies against OspA, a Borrelia membrane protein.20 With this immunity, mice would be unable to transmit the bacteria to ticks that fed on them, which in turn, would fail to infect any humans they fed on.

I chose this not because I think Lyme disease is the worst problem in the world—although it is awful—but because I wanted to figure out how to develop ecotechnologies in a well supported and ethical manner.

 —Kevin Esvelt,  Massachusetts Institute of Technology

Immunizing enormous numbers of wild mice, however, is prohibitively difficult. By using genetic engineering, researchers could create white-footed mice that produced these antibodies from birth and could pass this ability on to their offspring. But did the island residents want to live with genetically engineered mice?

The answer was perhaps, but with caveats. In consulting with communities on this technology development, researchers found that community members preferred a cisgenic approach: They wanted white-footed mice that were engineered with DNA only from other white-footed mice.18 This would make the project more difficult for the researchers, and meant that a CRISPR-based gene drive, even one with limited spread, could not be used, since no white-footed mouse naturally has this gene-editing system. However, said Esvelt, “It’s their environment, so it’s their call.” 

“We’re potentially causing an irreversible change to the environment,” said Telford. “We need to think about informed consent of the community as a proxy for informed consent of the environment. That’s been a real advance and something [that Esvelt] has pioneered—involving the communities from the very start.”

          a gloved hand holds a brown mouse with a white underside.
White-footed mice can carry the bacteria that cause Lyme disease.
Heidi Goethert

The weaponization of synthetic biology

While Esvelt’s anti-Lyme ecological engineering project is still in progress—his team published a preprint on the creation of transgenic white-footed mice using CRISPR in 2023—his research has also branched off in new directions.21 “[Esvelt’s] mind is just so expansive,” said Telford. “Mice Against Ticks is only one of many projects that he’s working on. He’s equally creative with all sorts of other issues that we’re faced with today.”

One of these issues is biodefense. While considering the dangers of CRISPR gene drives, Esvelt reasoned that this technology favored defense. “But then I thought, ‘What about the rest of biotechnology?’ It is not necessarily slow; it is not necessarily obvious; and it’s not necessarily easy to counter.”

He is particularly concerned about lab-created pathogens. “COVID has shown us that we do not have defenses in place against pandemics,” he said. But the SARS-CoV-2 virus may pale in comparison to the pathogens that humans could engineer.

“All of our defenses, our immune systems, evolved to counter the attacks that nature makes,” said Esvelt. “We have search strategies to find molecular tools and techniques and methods that nature cannot find, which is to say that every living thing on the planet is hardware insecure.”

[Esvelt] is on the forefront of thinking about what we need to do to reduce any risks associated with a very powerful and very useful technology.

 —Sam Telford III, Tufts University

“We will be able to make pandemics—arbitrary pandemics, infinite pandemics. We will be able to make them do nasty things to us that are unlike the way that natural [pathogens] harm us because nature isn’t trying to kill us. It’s just trying to replicate,” he said.

While vaccines and N95 respirators were helpful for reducing SARS-CoV-2 infections and deaths, Esvelt believes that they would be woefully inadequate at stopping a highly transmissible engineered pathogen, and that new strategies are needed to mitigate the risk of an engineered pandemic. 

“[Esvelt] is on the forefront of thinking about what we need to do to reduce any risks associated with a very powerful and very useful technology,” said Telford. He believes that Esvelt’s research is future driven, unlike many biodefense researchers who, he said, “are doing the same old stuff, focusing on the weapons from the Cold War days. We don’t need to worry about plague and tularemia and things like that. It’s really the engineered viruses that we need to worry about.” 

To this end, Esvelt is exploring several pathways as part of the nonprofit SecureBio, which he cofounded. The SecureDNA project aims to screen all DNA synthesis requests for pandemic-capable pathogens in order to prevent those who can design these pathogens—an ever-increasing number of scientists—from actually obtaining them. The Global Nucleic Acid Observatory would monitor wastewater for any organisms increasing at exponential rates, providing an early warning system for potential pandemic agents, even without researchers having to know what kind of agent to look for.22 His research team is also investigating the use of far-UVC light as a strategy for destroying many types of pathogens, both known and unknown.23

Once protective strategies against an engineered pathogen have been developed, Esvelt said, “then the incentives for people to actually build the thing go away because then they will know that we will see it. We will be able to respond… And we will survive.” 

  1. Esvelt KM et al. A system for the continuous directed evolution of biomolecules.Nature. 2011;472(7344):499-503.
  2. Esvelt KM et al. Concerning RNA-guided gene drives for the alteration of wild populations.Elife. 2014;3:e03401.
  3. Romero PA, Arnold FH. Exploring protein fitness landscapes by directed evolution.Nat Rev Mol Cell Biol. 2009;10(12):866-876.
  4. Abeles O. The agricultural figures of Darwin’s evolutionary rhetoric.Quarterly Journal of Speech. 2016;102(1):41-61.
  5. Yuan L et al. Laboratory-directed protein evolution. Microbiol Mol Biol Rev. 2005;69(3):373-392.
  6. Dickinson BC et al. Experimental interrogation of the path dependence and stochasticity of protein evolution using phage-assisted continuous evolution.Proc Natl Acad Sci U S A. 2013;110(22):9007-9012.
  7. Packer MS et al. Phage-assisted continuous evolution of proteases with altered substrate specificity. Nat Commun. 2017;8(1):956.
  8. Richter MF et al. Phage-assisted evolution of an adenine base editor with improved Cas domain compatibility and activity.Nat Biotechnol. 2020;38(7):883-891.
  9. Bhaya D et al. CRISPR-Cas Systems in Bacteria and Archaea: Versatile Small RNAs for Adaptive Defense and Regulation. Annu. Rev. Genet. 2011;45(1):273-297.
  10. Jinek M et al. A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity.Science. 2012;337(6096):816-821.
  11. Mali P et al. RNA-guided human genome engineering via Cas9. Science. 2013;339(6121):823-826.
  12. Burt A. Site-specific selfish genes as tools for the control and genetic engineering of natural populations.Proc. Royal Soc. B. 2003;270(1518):921.
  13. Akbari OS et al. A synthetic gene drive system for local, reversible modification and suppression of insect populations.Curr Biol. 2013;23(8):671-677.
  14. Gantz VM, Bier E. Genome editing. The mutagenic chain reaction: a method for converting heterozygous to homozygous mutations.Science. 2015;348(6233):442-444.
  15. Akbari OS et al. Safeguarding gene drive experiments in the laboratory.Science. 2015;349(6251):927-929.
  16. Noble C et al. Daisy-chain gene drives for the alteration of local populations. Proc Natl Acad Sci U S A. 2019;116(17):8275-8282.
  17. Smidler AL et al. A confinable female-lethal population suppression system in the malaria vector, Anopheles gambiae.Sci Adv. 2023;9(27):eade8903.
  18. Buchthal J et al. Mice Against Ticks: an experimental community-guided effort to prevent tick-borne disease by altering the shared environment.Philos Trans R Soc Lond B Biol Sci. 2019;374(1772):20180105.
  19. NIGROVIC LE, THOMPSON KM. The Lyme vaccine: a cautionary tale.Epidemiol Infect. 2007;135(1):1-8.
  20. Simon MM et al. Recombinant outer surface protein a from Borrelia burgdorferi induces antibodies protective against spirochetal infection in mice.J Infect Dis. 1991;164(1):123-132.
  21. Buchthal J et al. Low-cost camera-based estrous tracking enables transgenesis in Peromyscus leucopus, the primary reservoir for Lyme disease. Published online October 22, 2023:2023.10.20.563285. bioRxiv.org.
  22. Consortium TNAO. A Global Nucleic Acid Observatory for Biodefense and Planetary Health. arXiv.org.
  23. Görlitz M et al. Assessing the safety of new germicidal far-UVC technologies.Photochem Photobiol. Published online November 6, 2023.