Sculpting Evolution
Exploring evolutionary and ecological engineering.

Evolved systems are very different from technologies designed by humans. They're harder to predict and sometimes evolve away from intended purposes, but can also achieve outcomes we could never have rationally designed. The Sculpting Evolution group invents new ways of engineering self-replicating systems, including but not limited to gene drive elements capable of altering wild populations. By ensuring that candidate gene drives are developed transparently and responsively from project inception, we hope to catalyze a shift towards community-guided research that will change the relationship between science, society, and the natural world.

Research Projects

  • Computer-Assisted Transgenesis

    Kevin Esvelt, Erika Alden DeBenedictis, Cody Gilleland and Jianghong Min

    This is a new platform to automate experiments in genetic engineering and bring large-scale moonshot projects within reach. Too often, lab experiments are limited in scale by human fatigue and the costs associated with manual labor. In particular, the process of delivering genetic materials via manual microinjection remains a long-standing bottleneck. We are developing a computer-assisted microinjection platform to streamline the production of transgenic organisms. Briefly, organisms are immobilized in a gel and microinjections are performed using precision robotics using computer vision algorithms. This platform demonstrated high-throughput gene editing in an animal model (C. elegans) for the first time. We will be using this technology to refine and create safeguards for our gene drive technology.

  • Engineering Microbial Ecosystems

    Kevin Esvelt, Erika Alden DeBenedictis, Jianghong Min and Devora Najjar

    We are developing methods of controlling the genetic and cellular composition of microbial communities in the gut. Stably colonized microbes could be engineered to sense disease, resist pathogen invasion, and release appropriate therapeutics in situ.

  • Preventing Lyme Disease by Permanently Immunizing Mice

    Kevin Esvelt, Devora Najjar and Joanna Buchthal

    Lyme disease is the most common vector-borne infection in North America. People are infected when bitten by ticks; ticks are typically infected when they bite white-footed mice, the primary "reservoir" of the disease. We are exploring the possibility of permanently immunizing mouse populations to block transmission by making and releasing mice that produce protective mouse antibodies from birth and pass immunity on to their pups. The project has been guided by representatives in offshore island communities from inception. Communities will choose which type of antibodies, pick uninhabited islands to serve as field trial sites, select independent monitors, and ultimately decide whether to volunteer their own islands for the next stage. If successful, prevention could be expanded to the mainland using local or global gene drive systems. Whether or not communities decide to proceed, we hope the process will become a model for responsive science worldwide.

  • Reducing Suffering in Laboratory Animals

    Kevin Esvelt and Devora Najjar

    The world uses an estimated 20 million mice in laboratory research experiments each year. These experiments are monitored and regulated to protect animal welfare whenever possible. However, analgesics cannot completely eliminate suffering, while many studies cannot use opiates or anti-inflammatory drugs because they would interfere with the biological process being studied. The benefits of animal research may outweigh the cost in animal suffering, but it would be better to perform these experiments without animal suffering. This project seeks to develop strains of mice that experience far less pain and suffering than current animals but are equally suited to laboratory and medical research. If successful, widespread adoption of these mice could drastically reduce animal suffering in laboratories worldwide.

  • Studying the Evolution of Gene Drive Systems

    Kevin Esvelt, Cody Gilleland and Jianghong Min

    How will gene drive systems evolve once released into the wild? Can they be reliably overwritten and blocked by immunizing reversal drives? Might they spread into related species? These are difficult questions because wild populations are so much larger than laboratory colonies, meaning critical evolutionary events would never be observed in the lab. We seek to develop nematode worms as a model system to help answer these questions. Nematodes are genetically tractable, reproduce twice each week, and are readily grown in populations numbering in the billions. This allows us to study drive systems intended for other organisms in nematodes. Synthetic site targeting, split drives, and ecological confinement will prevent spread into wild nematodes. Because nematodes are easy to culture and count using Foldscope microscopes, we intend to work with educators to enable students, museum-goers, and citizen scientists to participate in gene drive research.

  • Understanding Molecular Evolution

    Kevin Esvelt and Erika Alden DeBenedictis

    Humanity has harnessed evolution to sculpt domesticated animals, crops, and molecules, but the process remains a black box. Which combinations of evolutionary parameters will enable us to discover the best solutions? We plan to answer this question by performing massively parallel directed evolution experiments. Our system will use phage-assisted continuous evolution (PACE), a method of building synthetic ecosystems in which billions of fast-replicating viruses compete to optimize a molecular function of our choice. We are developing methods of running many experiments in parallel, each with real-time fitness monitoring and customized evolutionary conditions such as mutation rate, selection stringency, and evolutionary goal-switching. We will use these methods to systematically characterize the relationship between evolutionary parameters and outcomes.