Daniel Fisher: Why — or why not —  “survival of the fittest”?  Perpetual evolution and diversification from ecological feedback

 

Corina Tarnita – ‘More is different’: the origin of major evolutionary transitions

In the enormous catalog of innovation that is life—past and present—there are some events that mark major turning points. Examples include the evolution of multicellular organisms from a world dominated by single cells; the evolution of insect sociality culminating in the intricate complexity of ant colonies; and the evolution of social mammals, with human culture at the pinnacle. These rare events, known as major evolutionary transitions, have had a disproportionate impact on the history of life. The last 15 years have seen dramatic advances from comparative genomics, experimental evolution, and theoretical modeling. I will give an overview of our current knowledge and outline the potential for a comparative lens across major transitions to lead to a paradigmatic shift in our understanding of rare but transformative evolutionary events.

 

Microorganisms offer a remarkable window into the processes that drive evolutionary change in the environment. In this lecture, I will argue that one effective way to interpret the patterns of genome evolution in ecological context is to focus on what organisms eat —and the challenges they face in securing those resources.

 

I will begin by briefly introducing the oligotroph/copiotroph divide, a major axis of resource partitioning in the biosphere. I’ll then delve more deeply into the evolution of resource preferences among copiotrophic bacteria—the subset of microbes that specialize in decomposing and remineralizing complex macromolecules such as proteins and polysaccharides. These organisms evolve largely through the remodeling of their gene content via horizontal gene transfer, as well as through the acquisition of behavioral strategies that allow them to exploit transient nutrient patches in their environment—behaviors that regulate motility, attachment, growth, and dispersal.

 

We’ll then explore how expansions and contractions of gene repertoires drive substrate preferences, to the point where gene content alone can be used to predict growth rates on specific classes of substrates. I’ll discuss two possible selective pressures behind these repertoire expansions:
(i) a pressure to cope with enzymatic trade-offs by occupying a Pareto front of enzyme performance traits, and
(ii) an evolutionary arms race between the producers of complex macromolecules (e.g. plants) and the microbes that break them down.

 

Finally, I’ll present a recent example from the human gut microbiome, where certain bacteria have evolved the ability to distinguish macromolecules by their molecular weight. This capacity hinges on the evolutionary plasticity of transporter proteins, which can switch size preferences by acquiring or shedding specific domains.

 

Together, these findings illustrate how evolution in nature proceeds by remodeling gene and domain repertoires in ways that are tightly coupled to physiology, metabolism, and ecological strategy.

 

Homologous proteins share a common ancestry, a similar three-dimensional structure and a similar function. Their sequences carry the footprint of natural selection for structure and function. Statistical patterns in alignments of homologous protein sequences can thus reveal structural and functional properties of these proteins. Statistical-physics inspired methods and protein language models can both be used to infer them.  I will focus in particular on the problem of determining which proteins interact together among the members of two protein families, which we have addressed using both types of methods.

 

 

Polarity establishment plays an important role throughout the tree of life. For example, the yeast S. cerevisiae divides by budding. In order to establish a unique bud site, a cell needs to break its internal symmetry, a process known as polarization. The complex protein network responsible for cell polarity has been mapped and modelled in great detail. However, many existing models fail to capture the robustness of the network under molecular variation, which is clearly observed in both perturbation experiments as well as evolutionary history. Here I will present our efforts to quantify and describe evolution of cell polarity as a collective adaptive process.

 

 

Recent breakthroughs in computational protein design and structure prediction are revolutionizing the way we tailor biology, enabling bioengineering with unprecedented accuracy. In our group, we combine computational protein design with directed evolution to create new-to-nature enzymes for solar-energy conversion.

To accelerate enzyme engineering, we recently developed AI.zymes, an integrative platform that unites traditional and AI-based protein design within a computational evolution framework. AI.zymes leverages tools such as Rosetta, ESMFold, ProteinMPNN, and FieldTools in iterative design–selection cycles, enabling systematic improvement of activity, selectivity, and stability. Following the central idea of our work to integrate computational design with directed evolution, we have engineered photoenzymes by introducing high-affinity photosensitizer binding sites into existing enzyme scaffolds. Starting from heme-containing enzymes, we created light-driven catalysts for use in solar cells that convert sunlight into electricity. These methods address critical sustainability challenges by enabling the development of biocatalysts not only for photovoltaics but also for solar-powered carbon capture and nitrogen fixation.

In my group, we engineer synthetic gene regulatory networks producing spatiotemporal patterns such as oscillations and stripes. Synthetically recapitulating biological processes and interactions provides a bottom-up approach towards understanding the general principles of their mechanisms and evolution. By isolating gene regulatory networks from their native cellular context, we can systematically investigate how genetic mutations influence network function and phenotype.

In this talk, I will provide an overview of our previous work, where we employed synthetic circuits to explore how the structure and dynamics of gene regulatory networks shape evolutionary trajectories. In the second part of my talk, I will share our recent efforts to design gene regulatory networks beyond natural designs with potential applications in biocomputing.

 

Bacteria are excellent model systems to study evolution in the lab. Such studies are typically carried out on single species in isolation, for example to understand how they evolve resistance to drugs. In natural environments instead, bacteria are evolving within multi-species communities, which are likely to influence their evolutionary trajectories. In this talk, I will introduce a model system of small bacterial communities that degrade pollutants in industrial waste waters. I will talk about a long-term evolution experiment involving four bacterial species that can coexist over long time-scales. We compare how these species evolve when together versus when they are alone and how that affects their degradation ability. Second, I show how we can use the principles of group selection to breed new communities from scratch, leading to increased pollutant degradation over just a few weeks. Overall, this talk will illustrate how microbial communities can evolve through natural as well as directed evolution.

 

 

Microbial communities can carry out many functions relevant to biotechnology, such as food fermentation or biofuel synthesis in industrial biorefineries. Our ability to rationally design communities that optimize these functions is limited by the enormous complexity of microbial interactions. This work demonstrates that, despite this complexity, the function of a community is often predictable using very simple statistical models. These models mirror the patterns of global epistasis reported in quantitative genetics, which allow us to predict the fitness/phenotypic effect of a mutation despite the potential complexity of the interactions between genetic components. These results unify the task of predictively linking biological structure and function across scales, from molecules and organisms to entire communities, and illuminate a new path for the rational design of microbial consortia.

 

 

Our group engineers genetic systems that dramatically accelerate the speed of mutation and gene evolution in vivo so that we can drive the rapid evolution of new biomolecular functions and prospectively watch, analyze, and leverage the outcomes of long gene evolutionary processes on laboratory timescales. I will share a recent upgrading of our orthogonal DNA replication system (OrthoRep) to enable the experimental evolution of chosen genes at 1-million times the genomic error rate. This allowed us to extensively adapt and diverge an enzyme in a way that generates enormous diversity, from which we were able to detect new selective forces shaping how the gene evolved in vivo. I will also share work on the application of OrthoRep to antibody and enzyme engineering, with a focus on strategic data generation, as well as the evolution of arbitrary gene functions.

 

 

Abstract: Directed evolution experiments are typically carried out using in vitro systems, bacteria, or yeast—even when the goal is to probe or modulate mammalian biology for research purposes or in disease therapy. Performing directed evolution in systems that do not match the intended mammalian environment severely constrains the scope and functionality of the targets that can be evolved. This presentation will discuss motivation for this approach, present existing platforms, and describe ongoing method development that are now making it possible to use the mammalian cell itself as the setting for directed evolution. An overview of current accomplishments, frontier challenges, and high-impact targets for this approach will also be presented.

 

Frances Arnold – Innovation by Evolution: Bringing New Chemistry to Life

 

Chemistry encoded in DNA and optimized by evolution promises efficient, clean, sustainable routes to fuels, chemicals, materials, foods, medicines, and more. Evolution not only optimizes and tunes features such as activity or stereoselectivity—it also innovates. We are using evolution to create entirely new biocatalysts that catalyze reactions unknown in biology and sometimes unprecedented in human-invented chemistry. New-to-nature ‘carbene transferase’ and ‘nitrene transferase’ enzymes increase the scope of molecules and materials that can be made using biology’s remarkable chemical machinery. Such enzymes unlock chemicial transformations that were inaccessible to small-molecule catalysts. And, with modern machine learning and AI tools to aid discovery and optimization, we are closer than ever to encoding a vast array of chemical transformations in DNA.