About TSRI
Research & Faculty
News & Publications
Scientific Calendars
Scripps Florida
PhD Program
Campus Services
Work at TSRI
TSRI in the Community
Giving to TSRI
Site Map & Search

The Skaggs Institute
for Chemical Biology

Scientific Report 2007

Continuous In Vitro Evolution of RNA Enzymes

G.F. Joyce, T.A. Lincoln, B.M. Paegel, S.B. Voytek

We seek to understand the processes by which enzymes evolve. We have devised methods that enable us to evolve nucleic acid enzymes in the test tube. This work has led to the development of novel RNA and DNA enzymes, some of which have had applications in biomedicine. In addition, our studies have provided new insights into darwinian evolution at the molecular level.

Evolution by Serial Transfer

Most in vitro evolution studies involve a powerful but laborious process in which a population of molecules is first challenged to perform a biochemical task, segregated on the basis of whether or not they performed the task, and then amplified to produce progeny molecules that resemble but are not identical to the parent molecules. The progeny are treated similarly, and the entire process is repeated until the population adapts to the task at hand. Natural evolution does not require such intensive manipulation. Individuals that are well adapted to their environment replicate preferentially, and their progeny have the opportunity to do the same. Some organisms, such as bacteria, can be evolved in the laboratory by using a simple serial-transfer procedure. These organisms can be maintained in continuous culture if part of their growth mixture is transferred to a new vessel that contains a fresh supply of nutrients.

We have devised methods that enable us to culture enzymes in the same way bacteria are cultured. We use an RNA enzyme with RNA-joining activity and provide a system in which any RNA molecule that performs this reaction can be amplified to generate progeny molecules. The progeny have the opportunity to do the same, and so on, all within a common mixture. As the reaction components become consumed, a small part of the mixture is transferred to a new test tube that contains a fresh supply of reagents. With many such transfers, the population of RNA enzymes can evolve to meet imposed selection constraints.

Toward a Molecular Ecology

Until recently, all experiments involving the continuous evolution of RNA enzymes used descendants of the class I RNA ligase. This enzyme is one of fastest-reacting RNA enzymes ever described. During the past year, we obtained a second RNA enzyme, derived from the DSL RNA ligase developed by T. Inoue and colleagues at Kyoto University, Kyoto, Japan, that can undergo continuous evolution. This enzyme has a slow catalytic rate, especially under the conditions necessary for continuous evolution. However, by using stepwise in vitro evolution, we improved the rate by more than 10,000-fold. This improvement was achieved by adding an accessory domain of 35 nucleotides and using highly stringent selection pressure, with reaction times as short as 15 milliseconds, to drive the evolution of enhanced catalytic rate. The resulting fast-reacting enzymes can be propagated indefinitely by serial transfer.

The 2 distinct species of continuously evolving enzymes now are being made to evolve within a common environment in which they compete for limited resources. Initial experiments of this type led to the rapid extinction of one of the species. We currently are exploring conditions that will support sustained coexistence of the 2 species, for example, by modifying the 2 enzymes so that they behave cooperatively to support each other's function. The ability of continuous evolution to enable highly longitudinal studies that span many hundreds of successive generations will allow us to study ecologic processes at the molecular level.

Microchip-Based Evolution

Recently we devised a method that enables us to conduct evolution by using computer-controlled microfluidic chips. We monitor the increase in the population of evolving enzymes in real time and carry out precise dilutions and resupply of reagents, all without direct intervention by an experimenter. We have begun to conduct on-chip darwinian evolution with lineages of RNA enzymes that are subjected to various selection constraints.

In one experiment, we progressively reduced the concentration of substrate for the RNA-joining reaction, forcing the population of enzymes to adapt to these more stringent conditions. At the outset, the enzymes had a catalytic rate of 20 min-1 and recognized the substrate with a Km of 35 μM. The starting population was provided with only 1 μM substrate, causing the molecules to operate at only 3% of their maximum rate. The concentration of substrate then was reduced progressively during 500 logs of growth on the chip, eventually reaching just 0.05 μM (Fig. 1). At the final concentration of substrate, the starting enzyme cannot grow at all. During the course of evolution, however, the population acquired mutations that led to a 90-fold improvement in Km, while maintaining a catalytic rate of 21 min-1. We identified the specific mutations that led to this improvement, providing a molecular explanation for evolutionary adaptation.

Fig. 1. Microchip-based evolution of RNA enzymes with RNA-joining activity, carried out over 500 logs of growth. The population size was monitored in real time, and after each log of growth, a 10-fold dilution was performed by microfluidic manipulation. The substrate concentration was reduced periodically to maintain selective pressure on the evolving population.


Joyce, G.F. Forty years of in vitro evolution. Angew. Chem. Int. Ed. 46:6420, 2007.

Joyce, G.F. A glimpse of biology's first enzyme. Science 315:1507, 2007.

Voytek, S.B., Joyce, G.F. Emergence of a fast-reacting ribozyme that is capable of undergoing continuous evolution. Proc. Natl. Acad. Sci. U. S. A. 104:15288, 2007.


Gerald F. Joyce, M.D., Ph.D.
Dean, Faculty

Joyce Web Site