The evolution of new functions has not been fully explained through standard population genetic theory. One alternative hypothesis is based on relaxing the assumption of strict genotype-to-phenotype maps; perhaps populations explore novelty through evolving phenotypic plasticity, allowing new forms and functions. Destabilizing mutations in bacteriophage λ’s receptor binding protein (RBP) cause the protein to fold into a range of confirmations, with differing receptor function. These observations led to a hypothesis for how viruses evolve to use new receptors and shift hosts (Receptor Expansion via Multimorphic Intermediates; REMI). Here we test REMI through the evolution of a new virus, bacteriophage 21 (Φ21), as it evolved to use two new receptors. Multiple lines of evidence were found: First, Φ21 evolves biphasic decay where an isogenic culture of Φ21 particles decays at different rates indicating phenotypic heterogeneity in particle stability. Second, this heterogeneity was found to be associated with the evolution of new functions since particles with novel function decay fastest. Third, we found evidence that this heterogeneity arises during protein folding by manipulating chaperone expression during Φ21 genesis and found that increased chaperones decreased the fraction of particles able to use the new receptors. This last test was not performed on λ, representing a novel test of REMI and is the first to directly link protein folding to the generation of heterogeneity in receptor use. Altogether we provide evidence for REMI in a new virus, lending support to the generalizability of REMI, and we provide a qualitatively new test of the hypothesis.

Coevolution between species can drive complex evolutionary and ecological processes of the living world. However, evolution of multiple mutations and constantly changing selective pressures during coevolution creates a tangled web of interactions between species that makes it hard to investigate processes underlying the rich observations. Here we combine next-generation sequencing, quantitative experimental assays, and computational analyses to understand the true complexity coevolution causes by studying coevolution between bacteriophage lambda and its host Escherichia coli.

In Chapter 2, we test the mechanistic underpinnings of the arms race dynamics model in host-parasite coevolution. By coupling large-scale phenotypic assays with whole genome sequencing, we showed that although lambda and E. coli engaged in arms race at the phenotypic level, multiple lineages of both species coexisted and shifted in dominance during the coevolution, a feature of the complementary fluctuation selection dynamics model.

Chapter 3 uses fitness landscapes to test the role and extent of coevolution in promoting evolution of a new function. During coevolution, lambda evolves to target E. coli through a new membrane receptor, OmpF, by fixing multiple mutations in its host-recognition protein J. By measuring lambda’s fitness landscapes at two stages of coevolution: before and after E. coli evolved resistance, we show that evolution of resistance in E. coli deformed lambda’s landscape such that it opened adaptive pathways for lambda to evolve the new function. Multiple replays of the coevolution experiment with different starting conditions and a detailed examination of a coevolving laboratory population further confirmed that lambda required both ancestral and resistant hosts from coevolution to evolve the OmpF-function.

Lastly, we studied how perturbations in the host-parasite interactome could lead to adaptions at the population level. When we deleted E. coli’s dnaJ gene that lambda requires to initiate DNA replication, lambda evolved mutations in genes unrelated to dnaJ. Lambda adapted by improving its adsorption rate and lysis timing, and evolving intracellular mutualism.

Altogether, this dissertation offers insights into mechanisms that structure ecological networks, demonstrates that coevolution of multiple species can promote innovation, and shows how parasites can adapt in unintuitive ways to counter genetic deficiencies in their host.

When a gene experiences new selective pressures from its environment, adaptative mutations may arise that alter its amino acid sequence that produce a protein better suited to the new pressure, or through mutations that tune the gene’s expression to adjust to the new environment. Here we report evolution experiments to understand the second mode of adaptation by studying the evolution of λ, a bacterial virus or bacteriophage that infects E. coli, under controlled laboratory settings. We pressured  to alter its lysis timing, which is controlled by a well-studied gene regulatory system, primarily driven by the S gene. For 10 days, we pressured λ towards shortened or lengthened lysis times and found that  evolved shorter lysis times repeatedly in both treatments. Lysis time evolved in most replicates without evolving cis-regulatory S mutations, suggesting more genes are involved in lysis timing circuit than previously reported. We found a single S regulatory mutation among 12 experimental populations. This mutation account for some of the fitness gains  evolved, and the mutation also increased the stochasticity of lysis timing. We hypothesize that the stochasticity helped increase phenotypic variance and allowed  to adapt to changing pressures. Altogether we found that  rapidly adapts to selective pressure on lysis timing, but in unintuitive ways. Clearly a more sophisticated understanding how gene regulation networks respond to evolutionary pressure is required to predict adaptation.

The number of multidrug resistant strains of bacteria is increasing rapidly, while the discovery of new antibiotics has stagnated. Subsequently, interest in bacteriophages as anti-bacterial therapeutics has surged, in part, because there is near limitless diversity of phages to harness. While this diversity provides opportunity, it also creates the dilemma of having to decide which criteria to use to select phages. Two traits previously proposed are thermostability and reproductive rate. Here we show that focusing on phage’s current abilities may be shortsighted if maximizing traits like stability and reproduction limits future evolution. We studied the ability of three phages to suppress bacteria. Each phage differed by only one amino acid at site 987 in their host-recognition protein (J), yet they varied significantly in their stability, growth rate, and evolvability. We uncovered a three-way tradeoff such that each phage maximized only two of the three traits. The most suppressive phage was an evolvable, fast-reproducing variant, supporting the importance of evolvability and reproduction rate. We tested whether these traits were interdependent but found that each had individual effects on suppression. Seeking to test the environmental contingency of our results, we altered the experiment to reflect more challenging conditions inside a patient’s body. The stable, fast-reproducing phage was most suppressive in the short term, but the fast-reproducing, evolvable phage was more suppressive in the long term. Our results highlight the importance of evolvability, an often-overlooked trait in phage therapy, and underscore the need to consider long-term dynamics when testing phage for therapeutic use.

For billions of years, bacteria and their viruses (termed phages) have engaged in never-ending cycles of coevolution, as bacteria evolve resistance and phages counter resistance. This process of reciprocal adaptation has fundamentally shaped the biology of bacteria and phages, as well as structured the microbial ecosystems of our planet’s oceans, soils, and organisms. In recent years, the evolution and spread of antibiotic resistant bacterial pathogens have renewed interest in phages as an alternative way to treat bacterial infections. Yet, only by understanding how phages interact and coevolve with their hosts can we devise effective treatment strategies that avoid recreating the past mistakes made with antibiotic drugs. In this dissertation, I take an applied perspective to study and better understand bacteria-phage coevolution. In the first chapter, I demonstrate that the therapeutic potential of phages can be enhanced by pre-adapting them to target pathogens via a process called coevolutionary training. In the second chapter, I evaluate how bacteria respond to trained phages compared to favored approaches in phage therapy, such as phage cocktails. In the third chapter, I study how bacteria-phage coevolution generates diversity and creates complex ecological networks. And in the fourth chapter, I use an in vivo mouse model to show that phages administered to the gut can modify the gut microbiome and affect host mouse body weight phenotypes. Altogether, this work reveals the importance of coevolution as a fundamental process that shapes the diversity of life and as a useful tool for applications such as phage therapy.

In order to carry out successful infections, bacteriophage lambda depends on many proteins in its host, the bacterium Escherichia coli. lambda proteins and lambda DNA interact with host molecules to facilitate infection and replication. These interactions form a molecular network known as the interactome. Viral evolution is therefore largely governed by changes that optimize the productivity of the interactome. Here, I present studies where we perturbed lambda’s interactome and allowed it to regain functionality through adaptive evolution. These studies provide insight into how viruses adapt, and more generally, how gene networks evolve. Previous studies have identified host genes used by lambda during infection. We used this knowledge in order to perturb the interactome by culturing lambda on a variety of host strains with one of these genes deleted. We proceeded to study lambda evolution to the 16 KOs and a wild type control to distinguish between specialized adaptations to specific KOs, or general adaptations to suboptimal laboratory conditions. Overall, most perturbations to the interactome could be solved with generic mutations that improve lambda growth, although, there are a few host gene deletions that require specialized mutations. Our study shows that viruses can easily adapt to perturbations in their interactomes through different adaptive solutions. While mutations with general benefits aid the correction of node deletions by providing disproportionate benefits for especially destructive deletions, some host gene deletions require specialized mutations.

For decades, scientists have been fascinated by the ease with which viruses, seemingly simple life forms, evolve new feats of innovation. One viral innovation relevant to humans is gaining infectivity on a new host type. Although numerous instances of viral host-range expansion have been documented, we still lack the ability to predict them reliably. In part, this is because many factors affect whether a host shift will occur, ranging from molecular interactions to host behavior. To increase the tractability of this problem, scientists are beginning to ask whether viruses might vary in their innate evolvability, or capacity for adaptive evolution. I focused on the role of stability, here defined as thermodynamic stability, an intrinsic trait of viral proteins thought to enhance evolvability. Using the well-studied host-range expansion of bacteriophage λ, I first showed how mutations that confer expanded host range destabilize the receptor binding protein and allow it to assume alternative conformations with new binding activity. Then, I showed that among λ genotypes varying in stability, the most evolvable tended to be the most unstable, and the stable genotypes that did evolve gained destabilizing mutations. Instability promoted the evolution of new host range, in contrast to the widely cited consensus that stability enhances evolvability. I discovered one λ genotype that exhibited high stability and evolvability, but it grew poorly, suggesting a three-way tradeoff between stability, evolvability, and reproduction. This result led to another question: if traits that affect current fitness, like stability and reproductive rate, trade off with future evolutionary capacity, then which traits most influence the outcome of coevolutionary arms races between viruses and their hosts? I examined which traits influence λ’s ability to overcome host resistance and maintain infectivity on its coevolving host bacteria. The fast-reproducing, evolvable, but unstable genotype emerged most successful, suggesting that lineages that initially appear poorly adapted may give rise to progeny that persist due to their capacity to evolve. This work suggests that a positive, linear relationship between stability and evolvability does not hold in all scenarios, with important implications for predicting viral emergence and selecting genotypes for phage therapy.

The advent of genomic editing technology offers the tantalizing promise of precisely treating a variety of disease states at the genetic level. CRISPR-Cas has democratized this capability in laboratory settings across the world. However, developing these exciting tools into therapies appropriate for clinical use involves several challenges including the potential recognition of foreign components such as the Cas effector proteins by the adaptive immune system which may limit the effectiveness of gene therapy. Here I approach this problem by first characterizing the immune response to AAV-delivered CRISPR-Cas in a mouse model targeting the PCSK9 protein, known to play a major role in atherosclerotic disease. Next, I propose and test a method to enable redosing by leveraging immune orthogonal components of AAV-CRISPR-Cas therapeutics, systematically defining the capability of this approach to avoid immune-mediated inhibition of therapeutic efficacy. Finally, in an effort towards building a single deimmunized Cas9 protein usable across diverse contexts while circumventing pre- existing immunity, I develop a scalable and portable long-range multiplexed protein engineering platform to progressively de-immunize target proteins by abolishing the most immunogenic MHC-restricted epitopes without disrupting protein function. By applying this technique to Cas9, I identify a Cas9 variant with 7 simultaneously deimmunized epitopes that retains near wild-type functionality for both direct editing and gene activation/repression. Taken together, this work represents a meaningful step towards unlocking the potential to precisely edit genomes and gene expression in the clinical setting.

The Severe Acute Respiratory Syndrome Coronavirus-2 (SARS2) is a novel epidemic strain of Betacoronaviruses responsible for the Covid-19 pandemic. This virus, and its corresponding disease has claimed millions of lives and caused major obstructions to social and economic platforms around the world. Other notable Betacoronaviruses include the 2003 SARS1 which caused a brief epidemic in China, and the 2011 Middle East Respiratory Syndrome Virus (MERS) which affects northern Africa and central Asia to this day. Despite issuing multiple FDA emergency-use authorized vaccines for the virus, symptomatic breakthrough cases are still possible largely due to already established variants of the SARS2 virus, in particular the delta-variant. With such a large spread, and propensity to evolve, the Betacoronaviruses are likely to remain a challenge to civilization along with the seasonal flu. In any case, preparation against novel Betacoronaviruses is a necessity. In this review, we investigate the early replication stages of Coronaviruses, focusing on SARS1/2. We discuss the proteins associated with genome replication, and the formation of the coronaviral replication organelle, and the discovery of a novel putative coronaviral nuclear pore protein complex. Despite major differences in amino acid sequences, the proteins associated with early replication share strong structural homology. Additionally, stages of membrane rearrangements are shared amongst all the coronaviruses. Understanding the shared mechanisms of replication amongst these viruses is the first step in discovering potential drug targets for coronavirus antivirals.

Bacteriophages are the most abundant and genetically diverse life forms, even greater than their numerous bacterial prey. This observation seems to violate fundamental ecological theory that predicts higher trophic levels should be less diverse than lower trophic levels, leading to the question of what mechanisms could support greater parasite diversity than host diversity. Here we provide unexpected results from a bacteriophage λ evolution experiment where λ was continually supplied a single isogenic host, yet repeatedly evolved a diversity of genotypes that diverged in their receptor use. Measurements of negative frequency-dependent selection between the receptor specialists revealed that they evolved because of diversifying selection. Such selection in viruses usually stems from adaptations to specialize on different hosts, but only a single host was provided and replenished every eight hours. A series of experiments conducted over the 8-hour incubation period revealed that selection for the different specialist phages oscillated because gene expression of the two receptors oscillated (E. coli outer membrane proteins LamB and OmpF). The cause of the receptor expression changes was in part attributable to rapid evolution in the bacteria within the eight-hour period that was documented through both phenotypic resistance assays and population genome sequencing. Altogether this suggests that rapid host evolution on the timescales of hours can drive the emergence of multiple host genotypes and their oscillations can promote the evolution of phage diversity. This result suggests that rapid host evolution may play a role in driving viral diversity and may promote diversity in other non-microbial systems too.