University of Groningen On the origin and function of phenotypic variation in bacteria Moreno Gamez, Stefany

University of Groningen

On the origin and function of phenotypic variation in bacteria

Moreno Gamez, Stefany

DOI:

10.33612/diss.146787466

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from

it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date:

2020

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Moreno Gamez, S. (2020). On the origin and function of phenotypic variation in bacteria. University of

Groningen. https://doi.org/10.33612/diss.146787466

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

7

A F T E RT H O U G H T S

In this thesis I have studied phenotypic variation in different species of bacteria. I have shown how this variation results from the dynamics of resource uptake and metabolism during starvation (Chapter 3) and from both the effects of the environ-ment and of direct cell-cell interactions on signal transduction pathways (Chapter 4). I have also studied the functional consequences of this phenotypic variation. I have shown that it might allow bacteria to cope with fundamental trade-offs between rapid growth and survival (Chapter 2), to respond to changes in their environment (Chapter 5) and to expand their spatial range and rapidly adapt to antibiotics (Chap-ter 6).

An important motivation for my work that I offered in the introduction of this thesis, is that phenotypic variation within populations provides the raw material for the origin of biodiversity at the species level and above. Therefore, it is appro-priate to finish this thesis with some thoughts on the evolutionary consequences of phenotypic variation. To prepare the ground for this discussion, I will emphasize a common thread in my work, namely that phenotypic variation needs to be stud-ied at different levels of biological organization, along the continuum from genes to phenotypes, individuals and populations. I will argue that taking such a systems perspective is particularly important for understanding the functional relevance of phenotypic variation and its consequences for evolution.

���������� ��������� ���� � ������� �����������

As already pointed out in the introduction, an essential lesson from studying the molecular basis of phenotypic traits is that the phenotype of an individual is not merely determined by what each of its genes does in isolation. Instead, phenotypic traits can be seen as emergent properties of molecular interactions that are at the basis of a complex, hierarchical mapping from genotype to phenotype.

These molecular interactions occur in networks consisting of various components, including proteins, metabolites and genes themselves, that control gene expression by regulating the processes of transcription and translation. By interacting in a net-work, even a small set of components can give rise to a large range of emergent behaviors that are complex enough to support the different ways in which organ-isms sense and respond to their environment (Guet, Elowitz, Hsing, & Leibler, 2002). In fact, as illustrated in this thesis and by the work of many others (Pfennig & Ehren-reich, 2014; Promislow, 2005), molecular networks underlie the ability of organisms to modify their phenotype through the regulation of gene expression in response to environmental cues. Hence, by regulating phenotypic plasticity, they play a pivotal role in shaping phenotypic variation.

The regulation of pneumococcal competence studied in Chapter 4 is a good ex-ample of how phenotypic variation depends on emergent features of a molecular network. First, the competence regulatory network exhibits bistability (i.e. the

pres-ence of two stable states). By modeling this network we showed that this property is achieved by the coupling of a positive feedback loop with a highly non-linear dynamics of transcription; the latter resulting from the interaction of the response regulator of the two-component system controlling competence with the promoters of the competence operons. As shown by our model and experiments, bistability is responsible for the switch-like development of competence that allows bacteria to transiently engage in the competence program. Second, the molecular architecture of competence regulation leads to hysteresis. This is a property of many molecu-lar networks that makes them dependent on their history, introducing short-term phenotypic memory of previous environmental conditions. Although I did not test the adaptive benefit of hysteresis in the context of competence regulation, I specu-late that history-dependence might generate a short-term memory that speeds up collective decision making and confers some degree of robustness to noise.

Interactions within molecular networks are so essential to determine emergent phenotypic traits that a major way in which organisms cope with changes in their environment is by dynamically rewiring their gene regulatory networks. An ex-ample of such a plastic nature of signal transduction is the response of yeast to DNA damage. By assessing approximately 80000 pairwise interactions among more than 400 genes involved in core functions like transcription, and cell cycle control,

Bandyopadhyay et al. (2010) showed that there are considerable differences between the networks of gene interactions in yeast in the presence and absence of a DNA-damaging agent. For instance, more than 70% of the positive interactions that exist in the presence of the DNA-damaging agent and that are associated with higher cell viability are not detected in its absence, indicating that yeast can cope with stress as-sociated to DNA damage by changing their phenotype through a massive rewiring of their gene regulatory networks.

Molecular networks are not only integral to how organisms sense and respond to their environment but also to the propagation of expression noise that leads to phenotypic heterogeneity as shown in Chapters 2 and 3. On the one hand, gene regulatory networks can amplify stochastic fluctuations in gene expression by the presence of positive feedback loops that lead to distinct stable states (Chalancon et al., 2012; Veening, Smits, & Kuipers, 2008b). For instance, in B. subtilis the decision to become competent depends on the expression of the transcriptional activator ComK (Maamar et al., 2007). Noise in the basal expression level of ComK can be amplified by positive autoregulation leading to stochastic switching to the competent state. As a result, clonal populations of B. subtilis in a constant environment might contain ⇠ 10-20% of competent cells. On the other hand, the architecture of gene regulatory networks can dampen expression noise reducing phenotypic variation. This can happen by the presence of negative feedback loops or by functional redundancy (Chalancon et al., 2012).

Finally, understanding the molecular architecture of phenotypic traits is essential to determine how genetic mutations will influence phenotypic variation. For in-stance, while genetic circuits with different regulatory interactions can produce the same phenotype, they might respond differently to the action of mutation

constrain-ing in distinct ways the outcome of evolution. This has been shown in E. coli usconstrain-ing two simple synthetic networks made of a few genes. Although these circuits can produce the same spatiotemporal pattern of gene expression, Schaerli et al. (2018) found that upon mutagenesis they produce distinct sets of gene expression profiles due to their different topologies. Chapter 2 of this thesis illustrates a similar prin-ciple; in particular, it shows that the evolutionary outcome of selection for rapid growth resumption after starvation strongly depends on the molecular mechanisms underlying lag time because these mechanisms determine how tight is the relation-ship between the mean and the variance of this trait.

From individuals to populations

Individuals in nature very rarely live in isolation. This is true in particular for bacte-ria, which tend to grow in highly diverse ecological communities, where individuals interact with members of their own and many other species. These interactions may be direct, or mediated by feedbacks with the abiotic environment that individuals may respond to and/or modify. To understand the functional value of phenotypic variation, it is necessary to study how variation in a phenotypic trait affects eco-logical and evolutionary dynamics at the level of a population, community, or even higher levels of biological organisation.

The functionality of phenotypic variation in a clonal population depends on how individual behavior affects population-level performance. As discussed in this thesis, phenotypic variation in clonal populations can affect population-level performance when environments fluctuate. In this scenario, the performance of a population is determined by its geometric mean fitness across the different environments, which in turn can be influenced by the presence of phenotypic variation within the popula-tion. In particular, when the mapping between the phenotype of an individual and its performance is convex, phenotypic minorities can have a large effect on overall population dynamics. In Chapter 2, I showed that this property of the mapping between individual and population behavior determines growth and survival of bac-terial populations coming out of starvation: While population growth is dominated by the individuals with the shortest lag times, tolerance to antibiotics is provided by the individuals with longest lags. As a result, clonal populations of bacteria can break a trade-off between rapidly resuming growth from starvation and sur-viving antibiotic exposure by evolving phenotypic heterogeneity in lag. Nonlinear mappings between phenotype and performance have been found in other biological systems (Pickett, Thomson, Li, & Xing, 2015; Waite et al., 2016). As in the context of feast-and-famine dynamics in bacteria, in these scenarios the shape of a phenotypic distribution becomes as important as its mean in determining population perfor-mance because few individuals can have disproportionately large effects on overall population dynamics.

The consequences of phenotypic variation in the dynamics of clonal populations also heavily rely on interactions between individuals expressing different phenotypes. In particular, phenotypic heterogeneity is a way in which individuals in a clonal

population can divide tasks that they would otherwise not be able to perform indi-vidually due to the potential conflicts between them. For example, competition for resources such as intracellular space, or access to the transcriptional machinery can prevent a cell from synthesising enzymes to degrade a large variety of substrates (Zhou, Rivas, & Minton, 2008). Also, intermediate products from one metabolic re-action can inhibit a second metabolic rere-action, preventing a single individual from carrying both reactions simultaneously at a high rate (Costa, Pérez, & Kreft, 2006). In these scenarios, biochemical and biophysical constraints can lead to metabolic specialization of different subpopulations by avoiding that a cell performs every function by itself (Johnson et al., 2012). Individuals in clonal populations can then compensate for the missing functions by interacting with metabolically distinct indi-viduals carrying out complimentary functions for instance by means of cross-feeding (Ackermann, 2015; Pfeiffer & Bonhoeffer, 2004).

Emergent dynamics resulting from interactions between phenotypically distinct in-dividuals becomes particularly relevant in the context of genotype and multi-species communities (D’Souza, 2020; DalCo et al., 2020). In fact, collective properties emerging from such interactions can allow individuals to cope with environmental challenges that they could not otherwise withstand in isolation. One of the main in-stances of collective behavior in bacteria are biofilms. These multicellular aggregates often made of cells with different genotypes and even from different species are one of the most common forms in which bacteria live in nature (Claessen et al., 2014;

Nadell, Xavier, & Foster, 2009; Stoodley, Sauer, Davies, & Costerton, 2002). Biofilms constitute a hallmark of cellular differentiation in microbes: due to the emergent me-chanical, biochemical and physical properties of these collectives resulting from in-teractions between cells with distinct phenotypes, individuals in a biofilm can better tolerate various kinds of biotic and abiotic stressors and benefit from cross-feeding interactions with other cells. Additional examples of such bacterial multicellularity, where collective behavior results from the interactions of individuals with distinct phenotypes, include the formation of fruiting bodies, cellular filaments and migra-tion collectives (Claessen et al., 2014; Julien, Kaiser, & Garza, 2000; van Gestel, Vlamakis, & Kolter, 2015).

Since microbial communities are primarily shaped by interactions between their in-dividual members, these interactions are essential to understand the origin and func-tionality of phenotypic variation in microbes (D’Souza, 2020). Current approaches to characterize microbial communities are mainly based on high-throughput tech-niques aimed at finding which species and genes are present in a given commu-nity. Although this information certainly has a descriptive value, it does not pro-vide a mechanistic understanding of community dynamics (Konopka, Lindemann, & Fredrickson, 2014; Widder et al., 2016), which would require mapping the in-teractions between the different community members over time and determining how they ultimately lead to emergent community-level properties and functions. As shown in this thesis, phenotypic variation resulting from phenotypic heterogeneity and plasticity is inherent to bacterial behavior and as a result has a major effect on determining how individuals interact with one another. For this reason, future

ap-proaches to understand how microbes live in nature should aim at integrating these sources of phenotypic variation at the individual level within the broader community context.

���������� ��������� ��� ���������

In the previous section, I illustrated how emergent behavior of systems at different levels of organization affects the origin and functional consequences of phenotypic variation. Here I will discuss how phenotypic variation can in turn influence the course of evolution. Naturally, I will focus on phenotypic variation with a non-genetic origin since by definition evolution is driven by non-genetic variation. First, I will argue that phenotypic variation in molecular networks can control the rate at which genetic mutations arise and determine their probability of fixation (i.e., their evolu-tionary fate). Next, I will focus on the cumulative effect of these processes over longer periods of evolutionary time, and review how the complexity of molecular networks underlies both the robustness and the evolvability of biological systems. Finally, I will speculate about the processes that shape the genotype-phenotype map, and ask whether they are capable of producing not only adaptive phenotypic variants but also adaptive patterns of phenotypic variation.

How phenotypic variation controls the rate of mutations

Genetic mutation is often thought as one of the major forces driving evolutionary change along with selection, recombination, drift and migration. In the classical population genetics framework, mutation is a force that is independent of the other forces and that is inherently random: a gene has a fixed probability of mutation that is independent of its identity and that is constant across environments. This is a very useful abstraction to understand the origin and evolution of genetic diversity as shown by many population genetic models including the one presented in Chapter 6 of this thesis. However, the study of the molecular basis of phenotypic traits has shown that this abstraction is far from true.

Below I will highlight some recent findings that indicate that the rate of mutation is in itself a phenotypic trait. In fact, it is subject to variation by the same three mechanisms that I have discussed throughout this thesis: it can be hard-wired and differentially encoded in the genome, it can change plastically and be regulated by environmental stress and cell-cell communication; and it can vary among genetically identical individuals in a constant environment as a result of stochastic fluctuations in gene expression.

The notion that the rate of mutations can be genetically encoded is supported by several findings. For example, the basal pace at which genetic mutations are accumu-lated along different regions of the genome is not random, and this variability has been exploited by organisms to optimise their genome organisation. Highly

struc-tured mutational patterns are the result of how DNA sequences affect the chance of mistakes during DNA replication (Martincorena & Luscombe, 2012). For in-stance, tandem repeats tend to be hypermutable and can be used by organisms to increase mutation rates in a targeted manner (Vinces, Legendre, Caldara, Hagi-hara, & Verstrepen, 2009); this is the case of contingency loci in bacteria that often mediate pathogenic interactions and have promoter regions made up of tandem re-peats (Moxon, Bayliss, & Hood, 2006; Zhou, Aertsen, & Michiels, 2014). Moreover, variation in the mutation rates across genes can result from the differential action of repair mechanisms that correct replication errors. These mechanisms are energet-ically costly for a cell so they are not used equally for every gene (Martincorena & Luscombe, 2012; Svejstrup, 2002).

Mutation rates have also been shown to be variable in time or across environments in a plastic manner. The reason is that mutation rates are ultimately determined by the machinery used to replicate DNA and to correct mistakes from the process of replication (Chatterjee & Walker, 2017; Goodman, 2002). This machinery is made of proteins that are susceptible to environmental stress and can mutate. Moreover, since this machinery is encoded by genes, its expression can be regulated to induce mutagenesis in response to stressful environmental conditions. For instance, bacteria use stress-induced mutagenesis to cope with starvation or antibiotics by upregulat-ing the expression of error-prone polymerases or downregulatupregulat-ing the expression of enzymes in charge of DNA repair (Foster, 2007).

In addition to mistakes in DNA replication, a major source of variation in the ge-netic makeup of many species is horizontal gene transfer (HGT). HGT is a main evo-lutionary driver in prokaryotes and eukaryotes that, although less recognized in the latter, has led to key innovations like the evolution of mitochondria and chloroplasts (Keeling & Palmer, 2008). In Chapters 4 and 5 of this thesis I showed that HGT by genetic competence is a highly plastic phenotype that is subject to regulation by both the biotic and abiotic environment of a cell in a way that can be potentially adaptive for S. pneumoniae. First, the uptake of DNA during the competent state is induced by antibiotics and has been directly linked to the evolution and spread of antibiotic resistance in the pneumococcus (Hsieh et al., 2006; Reinert, 2009). In fact, environ-mental stress is thought to be a main driver of pneumococcal genome evolution by means of competence induction (Straume, Stamsås, & Håvarstein, 2015). Second, competence illustrates how the rate of mutation is a phenotype that can also be di-rectly regulated by interactions among individuals through communication signals; a regulation that is potentially adaptive: As I argue in this thesis, cell-cell communi-cation can be instrumental to the decision of when to upregulate DNA uptake as it might allow cells to collectively sense the environment and decide the optimal time to become competent.

Finally, a recent study has shown that, as any other phenotype, the rate of muta-tion can also vary among isogenic individuals in constant environments as a result of stochastic gene expression. Phenotypic heterogeneity in mutation rates was recently documented in E. coli where stochastic fluctuations in the Ada protein involved in DNA repair can result in cells containing either one or zero copies of Ada (Uphoff

et al., 2016). This difference is enough to change the fate of cells upon DNA-damage leading to a subpopulation of hypermutators. A high rate of mutation is a ‘high risk-high gain’ strategy. It increases the chance of gaining an adaptive mutation at the cost of very possibly introducing deleterious changes. Although phenotypic het-erogeneity in the Ada protein in E. coli likely results from the toxicity of this protein at high expression levels, the finding that mutation rates can also vary stochastically opens the question of whether organisms could bet-hedge on this trait as they do with other phenotypes.

How phenotypic variation controls the fate of mutations

Given that most genes exert their phenotypic effects through interactions with other genes, proteins or environmental factors, the fitness effect of most mutations is con-tingent on other sources of phenotypic variation. This often involves variation in the genetic background where a mutation occurs, which is a widespread phenomenon known as genetic epistasis. However, it may also involve phenotypic variation with a non-genetic basis as I will argue below.

The action of selection on a new mutation can be influenced both indirectly and directly by phenotypic variation with a non-genetic origin. Indirect effects are medi-ated by the biotic and abiotic environment of an organism, which is what ultimately sets the selective pressures acting on new mutations. First, plasticity and phenotypic heterogeneity determine to a major extent where organisms live and with whom they interact. As discussed before, they allow individuals to thrive in environments that fluctuate over time and to survive in conditions where certain phenotypic functions are incompatible by enabling clonal populations to divide these functions among different cells (Ackermann, 2015; Pfennig et al., 2010). Second, although not exten-sively discussed in this thesis, phenotypic plasticity is also instrumental on deter-mining where an organism can potentially live; this is, on the colonisation of novel environments. Once an individual has colonized a novel environment as a result of the expression of a plastic phenotypic trait, the forces of selection operating on new mutations change (Lande, 2009; Pfennig et al., 2010). On the one hand, selection on the plastic trait itself might favor genetic mutations that replace plastic phenotype determination due to the inherent limits and costs of phenotypic plasticity ( Calla-han, MaugCalla-han, & Steiner, 2008; Murren et al., 2015; Snell-Rood, 2013). On the other hand, the dynamics of selection on other traits can also change because their fitness relevance in the newly colonized environment changes (Pfennig et al., 2010). This can in turn modify the direction and strength of selection operating on these traits. For example, deleterious mutations in genes encoding a trait that is not longer needed in a new environment might go from being purged by selection in the old environment to having a neutral or even an adaptive value.

In Box 1 I argue that plasticity and phenotypic heterogeneity can also directly influence evolution because epistatic interactions do not only occur among genes but also among genes and phenotypes. In other words, the fitness effect of a mutation

does not only depend on the genetic but also on the phenotypic background where the mutation occurs, a phenomenon that I define as phenotypic epistasis.

BOX 1. Phenotypic epistasis

The fate of a mutation can be influenced by the phenotypic background where it occurs if two prerequisites are held. First, phenotypic states must be inherited. Second, the fitness consequences of a mutation must vary depending on the phenotypic background where the mutation occurs. Note that by phenotypic state I refer to phenotypic variation that does not have a genetic origin; otherwise, interactions among genes and phenotypes could be reduced to the framework of genetic epistasis.

Evidence for the first prerequisite is extensive. Non-genetic inheritance can occur through the transmission of epigenetic marks, cytoplasmic content, be-haviors and culture (Bonduriansky & Day, 2009). In this thesis for example, I showed that the inheritance of cellular components upon cell division under-lies phenotypic differences lasting for multiple generations once bacteria resume growth from starvation (Chapter 3) and can explain history-dependence in the context of quorum sensing responses (Chapter 4). Importantly, the inheritance of phenotypic states can last from various generations with examples of traits inherited for tens (Rechavi, 2014) or even hundreds of generations as it could be the case of certain human behaviors.

Regarding the second prerequisite, I will first discuss how the fitness effect of a mutation can depend on the phenotypic background where it occurs by in-teractions that are mediated by the environment. To start, the phenotype of an individual can change its environment and as a result the selective pressures that a new mutation encounters. This has been largely discussed in the context of the debate on the evolutionary significance of phenotypic plasticity (Pfennig et al., 2010) which I summarized before. In addition to these indirect interactions, plas-tic phenotypic traits can directly modulate the effect of geneplas-tic mutations. For example, the fitness effect of a mutation in a plastic trait will only be visible to selection in environments where the trait is expressed, whereas in environments where the trait is not expressed this mutation will most likely be neutral. This explains the loss of sporulation in B. subtilis evolving in nutrient-rich environ-ments, which do not induce sporulation making deleterious mutations in this trait invisible to selection (Maughan, Masel, Birky, & Nicholson, 2007). A less straightforward example are molecular chaperones, which are proteins that can mask the effect of genetic mutations by assisting with protein folding (Taipale, Jarosz, & Lindquist, 2010). Under environmental stress, the expression of chap-erones like Hsp90 is upregulated which in turn has a direct effect on the phe-notypic consequences of mutations appearing during stress. For instance, such increased Hsp90 buffering capacity has been found to facilitate the evolution of drug resistance in yeast and the evolution of oncogenic cancer lines (Cowen, 2005; Falsone, Leptihn, Osterauer, Haslbeck, & Buchner, 2004).

The effect of a mutation can also depend on phenotypes that are not plastically determined but instead result from stochastic gene expression. In this scenario, there can be an interaction between phenotypic variation and genetic mutations that is not mediated by the environment. The simplest of these interactions can occur between a mutation and the expression level of the gene where it occurs, a trait that can be prone to high levels of phenotypic noise. One can envision at least two types of mutations that would be susceptible to this effect. First, mu-tations that reduce protein stability. These mumu-tations, which often confer new functionality to a protein, give rise to proteins with an unstable structure. When highly expressed, unstable proteins can be particularly harmful for a cell because they form aggregates that considerably perturb cellular functioning and can lead to disease (Dobson, 2006; Lee, Lim, Masliah, & Lee, 2011; Schröder, Schäfer, & Friedl, 2002). As a result, the detrimental fitness effect of a destabilizing muta-tion can be highly contingent on stochastic fluctuamuta-tions in the expression level of the gene where this mutation occurs. Other type of mutations whose effect is susceptible to expression level are the ones affecting enzymatic efficiency. This is often the case of drug resistance mutations that make enzymes less suscep-tible to drugs by changing their structural properties (Schrag, Perrot, & Levin, 1997). The consequences of mutations reducing enzyme efficiency will depend on the expression level of the enzyme - they can be masked when expression level is high or further revealed if an enzyme is expressed at low levels. A recent theoretical study suggested that these interactions between expression level and enzyme function could speed up the evolution of antibiotic resistance by allow-ing bacteria to temporally reduce the cost of mutations that lower susceptibility to a drug while they are compensated genetically (Tadrowski, Evans, & Waclaw, 2018). Finally, the fitness of a mutation can also depend on stochastic fluctuations in the expression of a different gene. Examples include stochastic expression of chaperone proteins which would lead to similar effects as the ones discussed be-fore (Burga, Casanueva, & Lehner, 2011; Holmstrom, Tolker-Nielsen, & Molin, 1999) and stochastic switching of a translation termination factor in yeast from a normal to a prion-based form (Avery, 2006). The latter form would change se-lective forces operating on genetic mutations occurring in regions of the genome right after stop codons.

Overall, the previous examples show that there is plenty of evidence for both prerequisites of phenotypic epistasis across various biological systems. In fact, many of the interactions that I describe have been largely discussed before in the context of the evolutionary consequences of epigenetics and phenotypic plastic-ity (Pfennig et al., 2010). One could argue that the term phenotypic epistasis points out to the common factor underlying all of them: Individuals do not only inherit genes but also the propensity to express certain phenotypes.

Complexity as a basis for robustness and evolvability

In the previous section, I illustrated how phenotypic plasticity and heterogeneity can influence evolution because of various features of the molecular networks underlying phenotypic expression. Here I will briefly discuss how these features are not only relevant to determine the direct fitness effects of mutations but underlie more general evolutionary properties of biological systems like robustness and evolvability. These two concepts have attracted much attention in evolutionary biology because they are pivotal for understanding how biological systems evolve innovations which is at the root of many processes of diversification (Wagner, 2008b).

The study of genotype-phenotype maps has provided major insights on the ques-tion of how biological systems can be robust to genetic mutaques-tion while remaining evolvable. In particular, research on various systems like secondary RNA structures, transcription factor binding sites and metabolic networks has shown that two prop-erties of the genotype-phenotype map facilitate the search for novel adaptive pheno-types (Matias Rodrigues & Wagner, 2009; Payne & Wagner, 2014; Pigliucci, 2010;

Wagner, 2008a): First, the mapping is redundant or degenerate in the sense that the same phenotype can be produced by very different genotypes. Second, small changes in the genotypic space can translate into large changes in the phenotypic space such that a single mutation in a genotype can result in a very different phe-notype. These two properties allow a population to search for novel phenotypes by moving through neutral genotype networks. Since these neutral networks are large, a population can explore very distant regions of the genotypic space and thus explore a large variety of reachable alternative phenotypes. In this way, biological systems can have high evolvability while being robust to genetic mutations.

The previous properties of the genotype-phenotype map rely to a large extent on phenotypic variation with a non-genetic origin, which as discussed previously has a major role on determining the fate of genetic mutations (Box 1). First, mech-anisms that grant phenotypic plasticity can increase robustness to genetic change. For instance, robustness to mutation increases with the complexity of regulatory and metabolic networks (Siegal & Leu, 2014; van Gestel & Weissing, 2016), which is often a result of higher phenotypic plasticity. Another example are chaperones, proteins that are essential to respond to environmental stress and that constitute one of the main mechanisms that biological systems have to maintain mutational robust-ness (Taipale et al., 2010). Second, phenotypic variation with a non-genetic origin can increase evolvability by allowing organisms to further explore the phenotypic effects of a genetic mutation. Not only does this variation make the search for adap-tive genotypes more efficient (Espinosa-Soto, Martin, & Wagner, 2011), but it can also become a major source of evolutionary innovations by allowing organisms to change their environment (Wagner, 2017). These changes, which are often facilitated by phenotypic plasticity and heterogeneity as discussed before, lead to the release of cryptic genetic variation that can often become adaptive in new environments.

How is phenotypic variation shaped by evolution?

I argued that the architecture of genotype-phenotype map shapes the expression of phenotypic variation in populations, in a way that promotes robustness and evolv-ability. However, it remains unclear whether this relationship can itself be seen as a product of evolution. In other words, should we think of evolvability as an adapta-tion or as a fortunate by-product; and can we expect patterns of phenotypic variaadapta-tion to be adaptive in themselves, or do they rather reflect an indiscriminate shotgun ap-proach to producing an adapted variant? In this final section, I will discuss how molecular networks can be substrates for the evolution of the genotype-phenotype map and of adaptive patterns of phenotypic variation.

Possibly the most direct way of studying and testing the role of selection on shap-ing genotype-phenotype maps is by usshap-ing experimental evolution. This implies designing regimes that could both favor or hinder the evolution of genetic deter-mination, phenotypic plasticity or phenotypic heterogeneity and evolving popula-tions through repeated cycles of selection to study adaption from the genetic and subsequent regulatory changes in the genotype-phenotype map. There are various examples of evolution experiments showing transitions between genetic and plastic phenotype determination in response to different levels of environmental variability (Ketola et al., 2013; Listmann, LeRoch, Schlüter, Thomas, & Reusch, 2016; Murren et al., 2015; Suzuki & Nijhout, 2006; Torres-Dowdall, Handelsman, Reznick, & Gha-lambor, 2012). For instance,Suzuki and Nijhout (2006) showed that both plastic and non-plastic responses to heat stress can be selected in a moth species. In this system, selection can act in opposite directions on the evolution of plasticity as a result of genetic mutations in a sex-linked gene and various regulatory genes that modify the heat sensitivity of a hormonal pathway determining coloration (Suzuki & Nijhout, 2007).

Evolving phenotypic heterogeneity experimentally has been more elusive. One of the few examples is the evolution of stochastic colony switching in Pseudomonas fluorescens (Beaumont, Gallie, Kost, Ferguson, & Rainey, 2006). In this experiment, re-peated rounds of selection in a regime selecting for variability in colony morphology led to the evolution of a bacterial strain that switches stochastically between two phe-notypes. A single mutation in the carB locus was found responsible for the synthesis of a cellular capsule that led to the two observed morphologies. This mutation low-ers the levels of precursors in the pyrimidine byosinthetic pathway and subsequently exposes a decision point where cells either invest in nucleotide metabolism and do not form a capsule or stop cell division and synthesise a capsule (Gallie et al., 2015). In Chapter 2, I presented an evolution experiment that acts in the opposite direction by selecting for a reduction in phenotypic heterogeneity in lag time upon growth resumption from starvation. Using a mathematical model I showed that this reduc-tion can be adaptive in the regime that bacteria encountered during the experiment, which strongly favored rapid growth resumption when resources appeared.

The previous evolution experiments indicate that natural selection can shape molec-ular networks over few hundreds of generations to favor or hinder particmolec-ular modes

of phenotype determination. One can examine how this process occurs over longer time scales by studying the genetic basis of molecular networks. One of the main findings from studying such genetic basis is that DNA sequences involved in tran-scription regulation like promoters, enhancers and sequences involved in chromatin organization or transcription factor expression have diverged much more than cod-ing sequences across many species over evolutionary time (Levine & Tjian, 2003;

Villar et al., 2015; Wagner & Lynch, 2008; Wittkopp & Kalay, 2011; Yue et al., 2014;

Zheng, Gianoulis, Karczewski, Zhao, & Snyder, 2011). This is consistent and explains to some extent one of the key ideas I have presented so far: much of the observed biological diversity is not a result of differences in genes per se but of differences in their regulation and in how they are used by organisms to sense and respond to their environment. Perhaps the most famous example of a major process of diversification originating from changes in genome regulation is the evolution of morphology and body plans in animals, which has mostly occurred through changes in the expression and regulation of a basic genetic toolkit (Carroll, 2008).

Other important realizations arising from studying the genetic basis of molecular networks are that the regulatory portion of the genome is in many species larger than the coding part and that the phenotypic complexity of an organism correlates with the amount of genes devoted to regulation (Ashby, 2004; Berezikov, 2011;

Lang et al., 2010; Ponting & Hardison, 2011). The latter is well illustrated in bacteria by the tight correlation between the number of one and two-component systems of a species and both its genome size and the complexity of its natural environment (Capra & Laub, 2012; Galperin, Nikolskaya, & Koonin, 2001; Ulrich et al., 2005). Whereas bacteria that live in stable environments with few fluctuations (e.g. obli-gate intracellular parasites) have a handful of signal transduction systems, bacteria living in highly dynamic environments can encode hundreds of one-component sys-tems as well as histidine kinases and response regulators. Examples of these versatile bacteria include highly motile species and species that use a large variety of electron donors and acceptors.

Overall, the above findings illustrate that (i) selection can and has actively shaped the molecular basis of phenotypic traits and that (ii) much of the molecular complex-ity underlying phenotype determination has evolved as a way to generate adaptive genetic and phenotypic variation.

���������� �������

Understanding why organisms vary in their phenotypes and what are the conse-quences of this variation are two of the most fundamental questions in biology. In this thesis, I showed how phenotypic variation in bacteria arises from the interplay between molecular networks, the environment and interactions among individuals and manifests in phenotypic differences that exist even between genetically identi-cal cells encountering the same environment. I also showed that this variation can

be functional in various contexts. It allows bacteria to break a fundamental life history trade-off arising with the appearance of new resources, to improve their esti-mates of the state of the environment and to expand their spatial range and rapidly evolve drug resistance. Furthermore, the work presented in this thesis illustrates that studying the origin and function of phenotypic variation benefits from understand-ing emergent behavior at different levels of organization: From genes interactunderstand-ing within molecular networks to individuals interacting within diverse communities. Finally, most of the work presented here deals with the functional consequences of phenotypic variation with a non-genetic origin. I closed this thesis by showing how, despite not arising from genetic differences, both phenotypic heterogeneity and phe-notypic plasticity can have a major influence on the rate and direction of evolutionary processes.

B I B L I O G R A P H Y

Abel zur Wiesch, P., Kouyos, R., Abel, S., Viechtbauer, W., & Bonhoeffer, S. (2014). Cycling empirical antibiotic therapy in hospitals: Meta-analysis and models. PLoS Pathog, 10(6), e1004225.

Abramson, J., Iwata, S., & Kaback, H. R. (2004). Lactose permease as a paradigm for membrane transport proteins (review). Molecular Membrane Biology, 21(4), 227– 236.

Ackermann, M. (2015). A functional perspective on phenotypic heterogeneity in mi-croorganisms. Nature Reviews Microbiology, 13(8), 497–508.

Ackermann, M., Stecher, B., Freed, N., Songhet, P., Hardt, W., & Doebeli, M. (2008). Self-destructive cooperation mediated by phenotypic noise. Nature, 454. Adams, M. H., & Roe, A. S. (1945). A partially defined medium for cultivation of

Pneumococcus. J. Bacteriol. 49, 401–409.

Adler, M. (2010). Monte carlo methods applied to the ising model. Norderstedt, Germany: Diplomica Verlag.

Ahn, S. J., Wen, Z. T., & Burne, R. A. (2006). Multilevel control of competence devel-opment and stress tolerance in Streptococcus mutans UA159. Infect. Immun. 74(3), 1631–1642.

Alberghini, S., Polone, E., Corich, V., Carlot, M., Seno, F., Trovato, A., & Squartini, A. (2009). Consequences of relative cellular positioning on quorum sensing and bacterial cell-to-cell communication. FEMS Microbiol. Lett. 292(2), 149–161. Albertson, R. C. (2003). Genetic basis of adaptive shape differences in the cichlid

head. Journal of Heredity, 94(4), 291–301.

Alexander, H. K., & Bonhoeffer, S. (2012). Pexistence and emergence of drug re-sistance in a generalized model of intra-host viral dynamics. Epidemics, 4(4), 187–202.

Allen, G. E. (2003). Mendel and modern genetics: The legacy for today. Endeavour, 27(2), 63–68.

Althaus, C. L., & Bonhoeffer, S. (2005). Stochastic interplay between mutation and recombination during the acquisition of drug resistance mutations in Human Immunodeficiency Virus Type 1. Journal of Virology, 79(21), 13572–13578. Alvarado, S., Rajakumar, R., Abouheif, E., & Szyf, M. (2015). Epigenetic variation in

the Egfr gene generates quantitative variation in a complex trait in ants. Nature Communications, 6.

Anetzberger, C., Pirch, T., & Jung, K. (2009). Heterogeneity in quorum sensing-regulated bioluminescence of Vibrio harveyi. Molecular Microbiology, 73(2), 267–277.

Ankomah, P., Johnson, P. J. T., & Levin, B. R. (2013). The Pharmaco –, Population and Evolutionary Dynamics of Multi-drug Therapy: Experiments with S. aureus and E. coli and Computer Simulations. PLoS Pathog, 9(4), e1003300.

Ankomah, P., & Levin, B. R. (2012). Two-Drug Antimicrobial Chemotherapy: A Math-ematical Model and Experiments with Mycobacterium marinum. PLoS Pathog, 8(1), e1002487.

Antinori, A., Perno, C., Giancola, M., Forbici, F., Ippolito, G., Hoetelmans, R., & Piscitelli, S. (2005). Efficacy of cerebrospinal fluid (CSF)-penetrating antiretro-viral drugs against HIV in the neurological compartment: Different patterns of phenotypic resistance in CSF and plasma. Clinical Infectious Diseases, 41(8), 1787–1793.

Aprianto, R., Slager, J., Holsappel, S., & Veening, J. W. (2016). Time-resolved dual RNA-seq reveals extensive rewiring of lung epithelial and pneumococcal tran-scriptomes during early infection. Genome Biol. 17(1), 198.

Ashby, M. K. (2004). Survey of the number of two-component response regulator genes in the complete and annotated genome sequences of prokaryotes. FEMS Microbiology Letters, 231(2), 277–281.

Avery, O. T., Macleod, C. M., & McCarty, M. (1944). Studies on the chemical nature of the substance inducing transformation of pneumococcal types: Induction of transformation by a desoxyribonucleic acid fraction isolated from pneumococ-cus type III. J. Exp. Med. 79(2), 137–158.

Avery, S. (2006). Microbial cell individuality and the underlying sources of hetero-geneity. Nature Reviews Microbiology, 4(8), 577–587.

Ayala-del-Rio, H. L., Chain, P. S., Grzymski, J. J., Ponder, M. A., Ivanova, N., Bergholz, P. W., . . . Tiedje, J. M. (2010). The genome sequence of Psychrobacter arcticus 273-4, a psychroactive Siberian permafrost bacterium, reveals mechanisms for adaptation to low-temperature growth. Applied and Environmental Microbiology, 76(7), 2304–2312.

Bacheler, L. T., Anton, E. D., Kudish, P., Baker, D., Bunville, J., Krakowski, K., . . . Abremski, K. (2000). Human immunodeficiency virus type 1 mutations se-lected in patients failing efavirenz combination therapy. Antimicrobial Agents and Chemotherapy, 44(9), 2475–2484.

Baker, M. D., Wolanin, P. M., & Stock, J. B. (2005). Signal transduction in bacterial chemotaxis. BioEssays, 28(1), 9–22.

Bakkeren, E., Diard, M., & Hardt, W.-D. (2020). Evolutionary causes and conse-quences of bacterial antibiotic persistence. Nature Reviews Microbiology. Balaban, N. Q., Helaine, S., Lewis, K., Ackermann, M., Aldridge, B., Andersson, D. I.,

. . . Zinkernagel, A. (2019). Definitions and guidelines for research on antibiotic persistence. Nature Reviews Microbiology, 17(7), 441–448.

Balaban, N. Q., Merrin, J., Chait, R., Kowalik, L., & Leibler, S. (2004). Bacterial persis-tence as a phenotypic switch. Science, 305(5690), 1622–1625.

Baldomá, L., Badía, J., Sweet, G., & Aguilar, J. (1990). Cloning, mapping and gene product identification of rhaT from Escherichia coli K12. FEMS Microbiology Let-ters, 72(1-2), 103–107.

Bandyopadhyay, S., Mehta, M., Kuo, D., Sung, M.-K., Chuang, R., Jaehnig, E. J., . . . Ideker, T. (2010). Rewiring of genetic networks in response to DNA damage. Science, 330(6009), 1385–1389.

Bangsberg, D. R., Kroetz, D. L., & Deeks, S. G. (2007). Adherence-resistance rela-tionships to combination HIV antiretroviral therapy. Current HIV/AIDS Reports, 4(2), 65–72.

Baranyi, J. (2002). Stochastic modelling of bacterial lag phase. Int. J. Food Microbiol. 73, 203–206.

Bardwell, L. (2005). A walk-through of the yeast mating pheromone response path-way. Peptides, 26(2), 339–350.

Beaumont, H., Gallie, J., Kost, C., Ferguson, G., & Rainey, P. (2006). Evolution of a polyphenism by genetic accommodation. Science, 311.

Beckert, B., Turk, M., Czech, A., Berninghausen, O., Beckmann, R., Ignatova, Z., . . . Wilson, D. N. (2018). Structure of a hibernating 100s ribosome reveals an inac-tive conformation of the ribosomal protein s1. Nature Microbiology, 3(10), 1115– 1121.

Berdahl, A., Torney, C. J., Ioannou, C. C., Faria, J. J., & Couzin, I. D. (2013). Emergent sensing of complex environments by mobile animal groups. Science, 339(6119), 574–576.

Berezikov, E. (2011). Evolution of microRNA diversity and regulation in animals. Nature Reviews Genetics, 12(12), 846–860.

Bergkessel, M., Basta, D. W., & Newman, D. K. (2016). The physiology of growth arrest: Uniting molecular and environmental microbiology. Nature Reviews Mi-crobiology, 14(9), 549–562.

Bird, A. (2007). Perceptions of epigenetics. Nature, 447(7143), 396–398.

Bliss, C. I. (1939). The toxicity of poisons applied jointly. Ann. Appl. Biol. 26(3), 585– 615.

Boland, P. J. (1989). Majority systems and the condorcet jury theorem. The Statistician, 38(3), 181.

Bolger, A. M., Lohse, M., & Usadel, B. (2014). Trimmomatic: A flexible trimmer for illumina sequence data. Bioinformatics, 30(15), 2114–2120.

Bonduriansky, R., & Day, T. (2009). Nongenetic inheritance and its evolutionary im-plications. Annual Review of Ecology, Evolution, and Systematics, 40(1), 103–125. Bonhoeffer, S., & Nowak, M. A. (1997). Pre-existence and emergence of drug

resis-tance in HIV-1 infection. Proceedings of the Royal Society B: Biological Sciences, 264(1382), 631–637.

Bonhoeffer, S., Chappey, C., Parkin, N. T., Whitcomb, J. M., & Petropoulos, C. J. (2004). Evidence for positive epistasis in HIV-1. Science, 306(5701), 1547–1550.

Boos, W., Ehmann, U., Forkl, H., Klein, W., Rimmele, M., & Postma, P. (1990). Tre-halose transport and metabolism in Escherichia coli. Journal of Bacteriology, 172(6), 3450–3461.

Botero, C., Weissing, F., Wright, J., & D.R., R. (2015). Evolutionary tipping points in the capacity to adapt to environmental change. Proceedings of the National Academy of Sciences, 112(1).

Boulineau, S., Tostevin, F., Kiviet, D., Rein ten Wolde, P., Nghe, P., & Tans, S. (2013). Single-cell dynamics reveals sustained growth during diauxic shifts. PLoS ONE, 8.

Boyer, M., & Wisniewski-Dye, F. (2009). Cell-cell signalling in bacteria: not simply a matter of quorum. FEMS Microbiol. Ecol. 70(1), 1–19.

Bozic, I., Reiter, J. G., Allen, B., Antal, T., Chatterjee, K., Shah, P., . . . Nowak, M. A. (2013). Evolutionary dynamics of cancer in response to targeted combination therapy. eLife, 2.

Brauer, M. J., Yuan, J., Bennett, B. D., Lu, W., Kimball, E., Botstein, D., & Rabinowitz, J. D. (2006). Conservation of the metabolomic response to starvation across two divergent microbes. Proceedings of the National Academy of Sciences, 103(51), 19302–19307.

Brauner, A., Fridman, O., Gefen, O., & Balaban, N. Q. (2016). Distinguishing between resistance, tolerance and persistence to antibiotic treatment. Nature Reviews Mi-crobiology, 14(5), 320–330.

Bretscher, M. T., Althaus, C. L., Müller, V., & Bonhoeffer, S. (2004). Recombination in HIV and the evolution of drug resistance: For better or for worse? BioEssays, 26(2), 180–188.

Buchler, N., Gerland, U., & Hwa, T. (2005). Nonlinear protein degradation and the function of genetic circuits. Proc. Natl. Acad. Sci. U. S. A. 102, 9559–9564. Burga, A., Casanueva, M. O., & Lehner, B. (2011). Predicting mutation outcome from

early stochastic variation in genetic interaction partners. Nature, 480(7376), 250– 253.

Callahan, H. S., Maughan, H., & Steiner, U. K. (2008). Phenotypic plasticity, costs of phenotypes, and costs of plasticity. Annals of the New York Academy of Sciences, 1133(1), 44–66.

Capra, E. J., & Laub, M. T. (2012). Evolution of two-component signal transduction systems. Annual Review of Microbiology, 66(1), 325–347.

Carey, L. B., van Dijk, D., Sloot, P., Kaandorp, J., & Segal, E. (2013). Promoter se-quence determines the relationship between expression level and noise. PLoS Biology, 11(4).

Carnes, E. C., Lopez, D. M., Donegan, N. P., Cheung, A., Gresham, H., Timmins, G. S., & Brinker, C. J. (2010). Confinement-induced quorum sensing of individual Staphylococcus aureus bacteria. Nature Chemical Biology, 6(1), 41–45.

Carroll, S. (2008). Evo-devo and an expanding evolutionary synthesis: A genetic the-ory of morphological evolution. Cell, 134.

Carvajal-Rodríguez, A., Crandall, K. A., & Posada, D. (2007). Recombination favors the evolution of drug resistance in HIV-1 during antiretroviral therapy. Infection, Genetics and Evolution, 7(4), 476–483.

Chait, R., Craney, A., & Kishony, R. (2007). Antibiotic interactions that select against resistance. Nature, 446(7136), 668–671.

Chalancon, G., Ravarani, C. N., Balaji, S., Martinez-Arias, A., Aravind, L., Jothi, R., & Babu, M. M. (2012). Interplay between gene expression noise and regulatory network architecture. Trends in Genetics, 28(5), 221–232.

Chatterjee, N., & Walker, G. C. (2017). Mechanisms of DNA damage, repair, and mutagenesis. Environmental and Molecular Mutagenesis, 58(5), 235–263.

Chen, J. D., & Morrison, D. A. (1987). Modulation of competence for genetic trans-formation in Streptococcus pneumoniae. J. Gen. Microbiol. 133(7), 1959–1967. Chen, M., Chory, J., & Fankhauser, C. (2004). Light signal transduction in higher

plants. Annual Review of Genetics, 38(1), 87–117.

Chernish, R. N., & Aaron, S. D. (2003). Approach to resistant gram-negative bac-terial pulmonary infections in patients with cystic fibrosis. Current Opinion in Pulmonary Medicine, 9(6), 509–515.

Choi, P. J., Cai, L., Frieda, K., & Xie, X. S. (2008). A stochastic single-molecule event triggers phenotype switching of a bacterial cell. Science, 322(5900), 442–446. Claessen, D., Rozen, D. E., Kuipers, O. P., Søgaard-Andersen, L., & van Wezel, G. P.

(2014). Bacterial solutions to multicellularity: A tale of biofilms, filaments and fruiting bodies. Nature Reviews Microbiology, 12(2), 115–124.

Claverys, J. P., Prudhomme, M., & Martin, B. (2006). Induction of competence regu-lons as a general response to stress in gram-positive bacteria. Annu. Rev. Micro-biol. 60, 451–475.

Claverys, J.-P., & Havarstein, L. S. (2002). Extracellular-peptide control of competence for genetic transformation in Streptococcus pneumoniae. Front. Biosci. J. Virtual Libr. 7, d1798–d1814.

Clifton, M. C., Simon, M. J., Erramilli, S. K., Zhang, H., Zaitseva, J., Hermodson, M. A., & Stauffacher, C. V. (2014). In vitro reassembly of the ribose ATP-binding cassette transporter reveals a distinct set of transport complexes. Journal of Bio-logical Chemistry, 290(9), 5555–5565.

Conrad, T. M., Joyce, A. R., Applebee, M. K., Barrett, C. L., Xie, B., Gao, Y., & Pals-son, B. Ø. (2009). Whole-genome resequencing of Escherichia coli K-12 MG1655 undergoing short-term laboratory evolution in lactate minimal media reveals flexible selection of adaptive mutations. Genome Biology, 10(10), R118.

Cornforth, D. M., Popat, R., McNally, L., Gurney, J., Scott-Phillips, T. C., Ivens, A., . . . Brown, S. P. (2014). Combinatorial quorum sensing allows bacteria to resolve their social and physical environment. Proceedings of the National Academy of Sciences, 111(11), 4280–4284.

Cornick, J. E., & Bentley, S. D. (2012). STREPTOCOCCUS PNEUMONIAE: THE EVO-LUTION OF ANTIMICROBIAL RESISTANCE TO BETA-LACTAMS, FLUORO-QUINOLONES AND MACROLIDES. Microbes Infect. 14(7-8), 573–583.

Costa, E., Pérez, J., & Kreft, J.-U. (2006). Why is metabolic labour divided in nitrifica-tion? Trends in Microbiology, 14(5), 213–219.

Cowen, L. E. (2005). Hsp90 potentiates the rapid evolution of new traits: Drug resis-tance in diverse fungi. Science, 309(5744), 2185–2189.

Cox, M. M., & Battista, J. R. (2005). Deinococcus radiodurans — the consummate sur-vivor. Nature Reviews Microbiology, 3(11), 882–892.

Croucher, N. J., Mostowy, R., Wymant, C., Turner, P., Bentley, S. D., & Fraser, C. (2016). Horizontal DNA Transfer Mechanisms of Bacteria as Weapons of Intragenomic Conflict. PLoS Biol. 14(3), e1002394.

Cvikel, N., Egert Berg, K., Levin, E., Hurme, E., Borissov, I., Boonman, A., . . . Yovel, Y. (2015). Bats aggregate to improve prey search but might be impaired when their density becomes too high. Current Biology, 25(2), 206–211.

D’Souza, G. G. (2020). Phenotypic variation in spatially structured microbial commu-nities: Ecological origins and consequences. Current Opinion in Biotechnology, 62, 220–227.

Dadiani, M., van Dijk, D., Segal, B., Field, Y., Ben-Artzi, G., Raveh-Sadka, T., . . . Segal, E. (2013). Two DNA-encoded strategies for increasing expression with oppos-ing effects on promoter dynamics and transcriptional noise. Genome Research, 23(6), 966–976.

DalCo, A., van Vliet, S., Kiviet, D., Schlegel, S., & Ackermann, M. (2020). Short-range interactions govern the dynamics and functions of microbial communities. Na-ture Ecology & Evolution, 4(3), 366–375.

Dalmay, T. (2006). Short rnas in environmental adaptation. Proceedings of the Royal Society of London B: Biological Sciences, 273.

Darch, S. E., West, S. A., Winzer, K., & Diggle, S. P. (2012). Density-dependent fitness benefits in quorum-sensing bacterial populations. Proceedings of the National Academy of Sciences, 109(21), 8259–8263.

Dartois, V. (2014). The path of anti-tuberculosis drugs: From blood to lesions to my-cobacterial cells. Nature Reviews Microbiology, 12(3), 159–167.

Daruwalla, K. R., Paxton, A. T., & Henderson, P. J. (1981). Energization of the trans-port systems for arabinose and comparison with galactose transtrans-port in Es-cherichia coli. Biochemical Journal, 200(3), 611–627.

Dean, A. M. (1995). A molecular investigation of genotype by environment interac-tions. Genetics, 139(1), 19–33.

Deatherage, D. E., & Barrick, J. E. (2014). Identification of mutations in laboratory-evolved microbes from next-generation sequencing data using breseq. In Meth-ods in molecular biology (pp. 165–188). Springer New York.

Decho, A. W., Visscher, P. T., Ferry, J., Kawaguchi, T., He, L., Przekop, K. M., . . . Reid, R. P. (2009). Autoinducers extracted from microbial mats reveal a surprising diversity of N-acylhomoserine lactones (AHLs) and abundance changes that may relate to diel pH. Environ. Microbiol. 11(2), 409–420.

Decho, A. W., Norman, R. S., & Visscher, P. T. (2010). Quorum sensing in natural en-vironments: Emerging views from microbial mats. Trends in Microbiology, 18(2), 73–80.

Deresinski, S. (2009). Vancomycin in combination with other antibiotics for the treat-ment of serious methicillin-resistant staphylococcus aureus infection. Clinical Infectious Diseases, 49(7), 1072–1079.

Dobson, C. (2006). Protein aggregation and its consequences for human disease. Pro-tein & Peptide Letters, 13(3), 219–227.

Duan, K., & Surette, M. G. (2007). Environmental regulation of Pseudomonas aerugi-nosa PAO1 Las and Rhl quorum-sensing systems. Journal of Bacteriology, 189(13), 4827–4836.

Dubnau, D., & Losick, R. (2006). Bistability in bacteria. Molecular Microbiology, 61(3), 564–572.

Dunlap, P. V., & Kuo, A. (1992). Cell density-dependent modulation of the Vibrio fischeri luminescence system in the absence of autoinducer and LuxR protein. J. Bacteriol. 174(8), 2440–2448.

Dutta, T., & Srivastava, S. (2018). Small RNA-mediated regulation in bacteria: A grow-ing palette of diverse mechanisms. Gene, 656, 60–72.

Dykhuizen, D., & Hartl, D. (1978). Transport by the lactose permease of Escherichia coli as the basis of lactose killing. Journal of Bacteriology, 135(3), 876–882. Eames, M., & Kortemme, T. (2012). Cost-benefit tradeoffs in engineered lac operons.

Science, 336(6083), 911–915.

Echenique, J. R., Chapuy-Regaud, S., & Trombe, M. C. (2000). Competence regulation by oxygen in Streptococcus pneumoniae: involvement of ciaRH and comCDE. Mol. Microbiol. 36, 688–696.

Edén, A., Fuchs, D., Hagberg, L., Nilsson, S., Spudich, S., Svennerholm, B., . . . Giss-lén, M. (2010). HIV-1 viral escape in cerebrospinal fluid of subjects on sup-pressive antiretroviral treatment. The Journal of Infectious Diseases, 202(12), 1819– 1825.

Eldon, B., & Wakeley, J. (2006). Coalescent processes when the distribution of off-spring number among individuals is highly skewed. Genetics, 172(4), 2621– 2633.

Else, L., Taylor, S., Back, D. J., & Khoo, S. H. (2011). Pharmacokinetics of antiretroviral drugs in anatomical sanctuary sites: The male and female genital tract. Antiviral Therapy, 16, 1149–1167.

Engelmoer, D. J., & Rozen, D. E. (2011). Competence increases survival during stress in Streptococcus pneumoniae. Evolution, 65(12), 3475–3485.

Espinosa-Soto, C., Martin, O. C., & Wagner, A. (2011). Phenotypic plasticity can facil-itate adaptive evolution in gene regulatory circuits. BMC Evolutionary Biology, 11(1).

Esser, D. S., Leveau, J. H., Meyer, K. M., & Wiegand, K. (2015). Spatial scales of interactions among bacteria and between bacteria and the leaf surface. FEMS Microbiology Ecology, 91(3).

Falsone, S. F., Leptihn, S., Osterauer, A., Haslbeck, M., & Buchner, J. (2004). Onco-genic mutations reduce the stability of src kinase. Journal of Molecular Biology, 344(1), 281–291.

Ferrell, J. E. (2002). Self-perpetuating states in signal transduction: positive feedback, double-negative feedback and bistability. Curr. Opin. Cell Biol. 14, 140–148. Figueira, R., Brown, D. R., Ferreira, D., Eldridge, M. J. G., Burchell, L., Pan, Z., . . .

Wigneshweraraj, S. (2015). Adaptation to sustained nitrogen starvation by Es-cherichia coli requires the eukaryote-like serine/threonine kinase YeaG. Scientific Reports, 5(1).

Fisher, R. A. (1919). The correlation between relatives on the supposition of mendelian inheritance. Transactions of the Royal Society of Edinburgh, 52(2), 399–433.

Fletcher, C. V., Staskus, K., Wietgrefe, S., Rothenberger, M., Reilly, C., Chipman, J. G., . . . Schacker, T. W. (2014). Persistent HIV-1 replication is associated with lower antiretroviral drug concentrations in lymphatic tissues. Proceedings of the Na-tional Academy of Sciences, 111(6), 2307–2312.

Folmes, C., Dzeja, P., Nelson, T., & Terzic, A. (2012). Metabolic plasticity in stem cell homeostasis and differentiation. Cell Stem Cell, 11(5), 596–606.

Foster, P. L. (2007). Stress-induced mutagenesis in bacteria. Critical Reviews in Bio-chemistry and Molecular Biology, 42(5), 373–397.

Fridman, O., Goldberg, A., Ronin, I., Shoresh, N., & Balaban, N. Q. (2014). Optimiza-tion of lag time underlies antibiotic tolerance in evolved bacterial populaOptimiza-tions. Nature, 513(7518), 418–421.

Fu, F., Nowak, M. A., & Bonhoeffer, S. (2015). Spatial Heterogeneity in Drug Concen-trations Can Facilitate the Emergence of Resistance to Cancer Therapy. PLoS Comput Biol, 11(3), e1004142.

Fujita, M. (2005). Evidence that entry into sporulation in Bacillus subtilis is governed by a gradual increase in the level and activity of the master regulator Spo0A. Genes & Development, 19(18), 2236–2244.

Fuqua, W. C., Winans, S. C., & Greenberg, E. P. (1994). Quorum sensing in bacte-ria: The LuxR-LuxI family of cell density-responsive transcriptional regulators. Journal of Bacteriology, 176(2), 269–275.

Fux, C., Costerton, J., Stewart, P., & Stoodley, P. (2005). Survival strategies of infec-tious biofilms. Trends in Microbiology, 13(1), 34–40.

Gallie, J., Libby, E., Bertels, F., Remigi, P., Jendresen, C., Ferguson, G., . . . Rainey, P. (2015). Bistability in a metabolic network underpins the de novo evolution of colony switching in Pseudomonas fluorescens. PLoS Biology, 13.

Galperin, M. Y., Nikolskaya, A. N., & Koonin, E. V. (2001). Novel domains of the prokaryotic two-component signal transduction systems. FEMS Microbiology Letters, 203(1), 11–21.

Gardan, R., Besset, C., Gitton, C., Guillot, A., Fontaine, L., Hols, P., & Monnet, V. (2013). Extracellular life cycle of ComS, the competence-stimulating peptide of Streptococcus thermophilus. J. Bacteriol. 195(8), 1845–1855.

Garmyn, D., Gal, L., Briandet, R., Guilbaud, M., Lemaıtre, J.-P., Hartmann, A., & Piveteau, P. (2011). Evidence of autoinduction heterogeneity via expression of the Agr system of Listeria monocytogenes at the single-cell level. Applied and Environmental Microbiology, 77(17), 6286–6289.

Gefen, O., Fridman, O., Ronin, I., & Balaban, N. Q. (2014). Direct observation of single stationary-phase bacteria reveals a surprisingly long period of constant protein production activity. Proceedings of the National Academy of Sciences, 111(1), 556– 561.

Gefen, O., Gabay, C., Mumcuoglu, M., Engel, G., & Balaban, N. Q. (2008). Single-cell protein induction dynamics reveals a period of vulnerability to antibiotics in persister bacteria. Proceedings of the National Academy of Sciences, 105(16), 6145– 6149.

Geisel, N., Vilar, J. M. G., & Rubi, J. M. (2011). Optimal resting-growth strategies of microbial populations in fluctuating environments. PLoS ONE, 6(4), e18622. Gillespie, D. (1977). Exact stochastic simulation of coupled chemical reactions. J. Phys.

Chem. 81, 2340–2361.

Gohara, D. W., & Yap, M.-N. F. (2018). Survival of the drowsiest: The hibernating 100s ribosome in bacterial stress management. Current Genetics, 64(4), 753–760. Goldkorn, T., Rimon, G., & Kaback, H. R. (1983). Topology of the lac carrier protein in the membrane of Escherichia coli. Proceedings of the National Academy of Sciences, 80(11), 3322–3326.

Golub, B., & Jackson, M. O. (2010). Naïve learning in social networks and the wisdom of crowds. American Economic Journal: Microeconomics, 2(1), 112–49.

Goodman, M. F. (2002). Error-prone repair DNA polymerases in prokaryotes and eukaryotes. Annual Review of Biochemistry, 71(1), 17–50.

Gottesman, S., & Storz, G. (2010). Bacterial small RNA regulators: Versatile roles and rapidly evolving variations. Cold Spring Harbor Perspectives in Biology, 3(12), a003798–a003798.

Graveley, B. R. (2001). Alternative splicing: Increasing diversity in the proteomic world. Trends in Genetics, 17(2), 100–107.

Greco, W. R., Bravo, G., & Parsons, J. C. (1995). The search for synergy: A critical review from a response surface perspective. Pharmacological Reviews, 47(2), 331– 385.

Greulich, P., Waclaw, B., & Allen, R. (2012). Mutational pathway determines whether drug gradients accelerate evolution of drug-resistant cells. Physical Review Let-ters, 109, 088101.

Grimbergen, A. J., Siebring, J., Solopova, A., & Kuipers, O. P. (2015). Microbial bet-hedging: The power of being different. Current Opinion in Microbiology, 25, 67– 72.

Grote, J., Krysciak, D., & Streit, W. R. (2015). Phenotypic heterogeneity, a phenomenon that may explain why quorum sensing does not always result in truly homoge-nous cell behavior. Applied and Environmental Microbiology, 81(16), 5280–5289. Grube, M., Dimanta, I., Gavare, M., Strazdina, I., Liepins, J., Juhna, T., & Kalnenieks,

U. (2014). Hydrogen-producing Escherichia coli strains overexpressing lactose permease: FT-IR analysis of the lactose-induced stress. Biotechnology and Applied Biochemistry, 61(2), 111–117.

Guet, C. C., Elowitz, M. B., Hsing, W., & Leibler, S. (2002). Combinatorial synthesis of genetic networks. Science, 296(5572), 1466–1470.

Guiral, S., Henard, V., Granadel, C., Martin, B., & Claverys, J. P. (2006). Inhibition of competence development in Streptococcus pneumoniae by increased basal-level expression of the ComDE two-component regulatory system. Microbiology (Reading, Engl.) 152(Pt 2), 323–331.

Guiral, S., Mitchell, T. J., Martin, B., & Claverys, J. P. (2005). Competence-programmed predation of noncompetent cells in the human pathogen Streptococcus pneumo-niae: genetic requirements. Proc. Natl. Acad. Sci. U.S.A. 102(24), 8710–8715.

Gulick, R. M., Mellors, J. W., Havlir, D., Eron, J. J., Gonzalez, C., McMahon, D., . . . Schleif, W. A., et al. (1998). Simultaneous vs sequential initiation of ther-apy with indinavir, zidovudine, and lamivudine for HIV-1 infection: 100-week follow-up. Journal of the American Medical Association, 280(1), 35–41.

Guo, M., & Gross, C. (2014). Stress-induced remodeling of the bacterial proteome. Current Biology, 24(10), R424–R434.

Gupta, R. K., Jordan, M. R., Sultan, B. J., Hill, A., Davis, D. H., Gregson, J., . . . Bertag-nolio, S. (2012). Global trends in antiretroviral resistance in treatment-naive in-dividuals with HIV after rollout of antiretroviral treatment in resource-limited settings: A global collaborative study and meta-regression analysis. The Lancet, 380(9849), 1250–1258.

Gustafsson, E., Nilsson, P., Karlsson, S., & Arvidson, S. (2004). Characterizing the dy-namics of the quorum-sensing system in Staphylococcus aureus. Journal of Molec-ular Microbiology and Biotechnology, 8(4), 232–242. doi:10.1159/000086704

Gutowski, S. J., & Rosenberg, H. (1975). Succinate uptake and related proton move-ments in Escherichia coli K12. Biochemical Journal, 152(3), 647–654.

Haeno, H., & Iwasa, Y. (2007). Probability of resistance evolution for exponentially growing virus in the host. Journal of Theoretical Biology, 246(2), 323–331. Halfmann, A., Kovacs, M., Hakenbeck, R., & Bruckner, R. (2007). Identification of the

genes directly controlled by the response regulator CiaR in Streptococcus pneu-moniae: five out of 15 promoters drive expression of small non-coding RNAs. Mol. Microbiol. 66(1), 110–126.

Harrigan, P. R., Hogg, R. S., Dong, W. W. Y., Yip, B., Wynhoven, B., Woodward, J., . . . Montaner, J. S. G. (2005). Predictors of HIV drug-resistance mutations in a large antiretroviral-naive cohort initiating triple antiretroviral therapy. Jour-nal of Infectious Diseases, 191(3), 339–347. 11th Conference on Retroviruses and Opportunistic Infections FEB 08-11, 2004 San Francisco, CA.

Hastings, I., Whatkins, W., & White, N. (2002). The evolution of drug-resistant malaria: The role of drug elimination half-life. Philosophical Transactions of the Royal Soci-ety B, 357(1420), 505–519.

Hattori, D., Millard, S. S., Wojtowicz, W. M., & Zipursky, S. L. (2008). Dscam-mediated cell recognition regulates neural circuit formation. Annual Review of Cell and De-velopmental Biology, 24(1), 597–620.

Hausmann, B., Pelikan, C., Rattei, T., Loy, A., & Pester, M. (2019). Long-term tran-scriptional activity at zero growth of a cosmopolitan rare biosphere member. mBio, 10(1).

Havarstein, L. S., Coomaraswamy, G., & Morrison, D. A. (1995). An unmodified heptadecapeptide pheromone induces competence for genetic transformation in Streptococcus pneumoniae. Proc. Natl. Acad. Sci. U.S.A. 92(24), 11140–11144. Havarstein, L., & Morrison, D. (1999). in Cell-Cell Signaling in Bacteria. ASM Press,

9–26.

Hedrick, T., Schulman, A., McElearney, S., Smith, R., Swenson, B., Evans, H., . . . Sawyer, R. (2008). Outbreak of resistant Pseudomonas aeruginosa infections