Chemical Engineering Seminar

We are interested in building genetic systems that have extremely high mutation rates in order
to speed up the evolution of target proteins and enzymes in vivo as well as to record transient
information, such as lineage relationships or exposure to biological stimuli, as genetic
information in situ. I will present work on two synthetic genetic systems that we have recently
developed. First, I will describe a highly error-prone yeast orthogonal DNA replication
(OrthoRep) system that mutates user-selected genes at a base pair substitution (bps) rate of
1e-5 without any increase in the genomic mutation rate (1e-10 bps). This ~100,000-fold
mutational acceleration allows for the rapid continuous evolution of target biomolecules entirely
in vivo using a simple serial passaging process amenable to extensive parallelization. I will
summarize a large DNA polymerase engineering effort leading to these OrthoRep systems,
show that OrthoRep can stably cross the error threshold of haploid yeast, and discuss the use
of OrthoRep in a yeast model of Plasmodium falciparum DHFR where we evolved resistance
to the antimalarial drug, pyrimethamine, in 90 independent 0.5 mL cultures through serial
passaging alone. We find that a highly adaptive first-step mutation constrains path-choice
through sign epistasis, leading to convergence, but rare mutations direct trajectories to an
alternative smooth fitness peak, illustrating the balance between fate and chance in drug
resistance. OrthoRep may enable a paradigm of high-throughput directed evolution, where
scores of evolving lines are routinely used to sample adaptive landscapes, evolve multiple
functions in parallel, or exploit subpopulation structure in the search for desirable biomolecular
function and the study of molecular evolution. Second, I will describe new work on the
construction of a rapidly mutating "ticker-tape locus" in mammalian cells, where our ultimate
goal is to achieve single-cell resolution lineage tracing in developing animals.