Eric Libby bio photo

Eric Libby

Research fellow at the Sante Fe Institute studying the evolution of biological complexity

Email Google Scholar Twitter



Complex life has evolved through a series of major transitions in which existing individuals combine to become parts of a new kind of individual. One key transition was the evolution of multicellularity which made possible significant increases in organismal complexity. While multicellularity has evolved on dozens of independent occasions, these events occurred millions of years ago and are absent from the fossil record. Recent experiments, however, have made it possible to study this transition in the lab by using microorganisms to evolve primitive multicellularity de novo . An example of an experimental system that evolved multicellularity is the yeast Saccharomyces cerevisiae . Unicellular yeast were grown in liquid media and then subjected to selection for rapid settling by centrifugation. Cluster-forming “snowflake” yeast evolved readily and fixed in the population because they settled more quickly than their unicellular ancestors. Such experimental systems not only allow evolution to be viewed in action but provide excellent opportunities to test hypotheses about the conditions that govern major transitions and enable the evolution of additional forms of biological complexity. A major arc of my research is directed at understanding how multicellularity evolves and what happens following its initial appearance. To this end, I collaborate with excellent experimentalists to connect theory with empirical observations. In the case of the yeast snowflake system, I collaborate with William Ratcliff and his lab at Georgia Tech. Examples of the types of topics we consider include: the role of multicellular geometry in shaping evolution, mechanisms that prevent multicellularity from reverting back to unicellularity, and the feedback between group and cell selection on the evolution of complex traits. There are many other exciting avenues that I have not listed here– please do not hesitate to contact me if you would like to discuss further or even collaborate!

Snowflake yeast (image courtesy of William Ratcliff) and mathematical model published in Libby E., Kerr B., Ratcliff W.C., Travisano M. PLoS Comp Biol (2014) .

Phenotype switching

Many microorganisms have evolved the capacity to switch between at least two phenotypic states. This can allow them to survive harsh environmental fluctuations such as might occur during an antibiotic regimen. Indeed, in this context, phenotypic switching can often be seen as a form of bet hedging. Interestingly this same behavior can also be coopted to enact a multicellular life cycle. For instance, Pseudomonas fluorescens produces two phenotypes: “smooth” and “wrinkly”. They differ by a mutation that causes wrinkly types to secrete an extracellular glue and thus stick together in a mat. Smooth types grown in culture will eventually give rise to wrinkly types who in turn will eventually produce smooth types. The phenotypes cycle because each type modifies the environment to favor the other type. This coupling of environmental change to phenotypic change creates stable oscillations in phenotypes that select for a life-cycle between smooth (unicellular) and wrinkly (multicellular) states.

In collaboration with Paul Rainey’s group at NZIAS, we have investigated the evolution of switching frequencies in the smooth-wrinkly system as well as how the uracil utilization pathway can be coopted to give rise to bistable switching (published in PLoS Biol. and PLoS One ). Currently, we are collaborating with Peter Lind at Uppsala University to predict the likelihoods of specific genetic targets of mutation in the switch between smooth and wrinkly phenotypes. While there are many possible mutations that could give rise to a wrinkly, typically only a few pathways are used. Mutational hotspots and genetic architecture play an important role in determining the frequency of these targets. We are using a bayesian framework to predict the relative frequency of evolutionary targets and show how genetic architecture shapes the evolutionary path of phenotype switching.

Smooth/Wrinkly transitions. Photo on left from: PB Rainey and M Travisano. Nature, 1998. Model on right from: E Libby and PB Rainey. PLoS One, 2013.

Developmental programs

Many multicellular organisms grow and develop from a stage in which they are only a single cell. As cells reproduce they differentiate to produce a multicellular form composed of different cellular phenotypes often organized into specialized tissues. This process usually relies on genetically encoded developmental programs that regulate cell differentiation decisions. Although there is much research on how developmental programs are modified to give rise to new forms, little is known about how they originated. At a fundamental level developmental programs are phenotypic switching mechanisms tied to a signal or some easily characterized rule. Thus, the evolution of development requires both an informative signal and the ability to use that information to regulate switching. Emma Wolinsky and I explored how such developmental mechanisms of switching might evolve in the P. fluorescens system (published in Evol. Ecol. ). We compared switching mechanisms using either external information such as nutrient availability or internal information such as cellular state. The P. fluorescens system represents only one type of multicellularity. I am looking at other multicellular systems with different organizations to understand how developmental programs can arise in general.

Different ways of a regulating a decision as might evolve in a simple developmental program. Image from: E Wolinsky and E Libby. Evol Ecol, 2015.

Life cycles

The P. fluorescens system is an example of a simple life cycle that alternates between unicellular and multicellular phases. This differs from the yeast snowflake system which does not have a single-celled phase– yeast groups give rise to yeast groups. There are other life cycles, like those found in choanoflagellates, that switch between multiple multicellular and unicellular forms. And in some systems, life cycles can be very plastic and adopt different structures. For example, in the unicellular green algae Chlamydomonas reinhardtii , different environmental conditions can lead to the evolution of life cycles with or without a unicellular phase. I am interested in understanding the diversity of life cycle structure and producing a general theory that predicts which life cycles will evolve given certain environmental conditions. Justin Yeakel and I coordinated a workshop at SFI on Complex Life Investment Strategies to discuss these topics and identify common underlying themes among experimentalists and theoreticians in ecology and evolution. The results of this workshop indicate that this is a rich and fascinating field.

Figure of Dicrocoelium dendriticum life cycle from CDC website. Parasite life cycles are some of the most terrifying and amazing.