THE E FFECTS OF GENETIC BACKGROUND ON THE EVOLUTION OF ANTIBIOTIC RESISTANCE AND ITS FITNESS COSTS By Kyle J oseph Card A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Microbiology and Molecular Genetics Doctor of Philosophy Ecology, Evolutionary Biology and Behavior Dual Major 2020 ABSTRACT THE EFFECTS OF GENETIC BACKGROUND ON THE EVOLUTION OF ANTIBIOTIC RESISTANCE AND ITS FITNESS COSTS By Kyle Joseph Card A ntibiot ic resistance is a growing public - health concern. Efforts to control the emergence and spread of resistance would benefit from an improved ability to forecast when and how it will evolve. To predict the evolution of resistance with accuracy, we must unders tand and integrate centers on the effects of genetic background on the evolution of phenotypic resistance, its genetic basis, and its fitness costs. To a ddress these issues, I used Escherichia coli strains from the long - term evolution experiment (LTEE) that independently evolved for multiple decades in an environment without antibiotics. First, I examined how readily these LTEE strains could overcome prior losses of intrinsic resistance through subsequent evolution when challenged with antibiotics. Second, I investigated whether lineages founded from different genotypes take parallel or divergent mutational paths to achieve increased resistanc e. Third, I tested wh ether fitness costs of resistance mutations are constant across different genetic backgrounds. In these studies, I focused attention on the interplay between repeatability and contingency in the evolutionary process. My findings demons trate that genetic ba ckground can influence both the phenotypic and genotypic evolution of resistance and its associated fitness costs. I conclude this dissertation with a broader discussion about these and other factors that can influence the evolution of antibiotic resistanc e, and their clinical and public - health implications. Copyright by KYLE JOSEPH CARD 2020 iv This dissertation is dedicated to my family, and to that little boy who dreamed of becoming a scientist one day. v ACK NOWLEDGEMENTS There are many individuals who have encouraged, guided, and inspired me on this journey. First and foremost, I want to thank my Ph.D. advisor, Dr. Richard Lenski. I had the fortune of meeting Rich while I was participating in undergraduate re search with Dr . Frank Dazzo. I had become interested in evolutionary biology during this time, particularly as it related to my research work. I therefore applied to the BioMolecular Science Gateway program with aspirations of studying lab; I had rea The Greatest Show on Earth and was intrigued by the multitude of interesting questions one could address with such a simple experiment. E ven though my knowledge of evolution was limited to the few boo ks I had read, I decided to take a chance and reach out to Rich about a rotation. The rest is history. I could scarcely understand it at the time, but that decision would prove profound on my life. I had set forth on a path of personal and professional dis covery, and Ri ch facilitated the journey. Rich is both incredibly sincere and kind. I learned this fact early on as a rotation student when he invited me and several other lab members to join him and his family for Thanksgiving dinner. Or on the several oc casions he wou I thank my committee members Dr. Christoph Adami, Dr. Frances Downes, and Dr. Christopher Waters for their guidance and helpful suggestions over these past several yea rs. I also wan t to thank the Department of Microbiology and Molecular Genetics (MMG) ; the Ecology, Evolutionary Biology, and Behavior Program; and the BEACON Center communities for encouraging and inspiring me. I also acknowledge the Howard Hughes Medical Institute Gill iam Fellowship and MMG for funding. In addition, I must thank my two undergraduate mentees, Jasper Gomez and Jalin Jordan. It has been a pleasure to work alongside them and bear witness to their vi academic and professional successes. Their futu re is bright, and I am immeasurably proud of them both. They are terrific scientists and friends. I made some wonderful friends while in graduate school. First, I want to thank Dr. Zachary Blount. My first interaction with Zack was when I was a rotation st udent. I had o pened one of the lab bench drawers and there was a cockroach staring up at me. I was startled, but quickly realized that it was fake. Zack had put it there to mess with the new recruit. My life is enriched because of our friendship. I am also delighted for our conversations on a great many topics. Zack, if you are to walk many more trails with you, my friend. Second, I want to thank Dr. Thomas LaB ar and Josh Fr been wonderful to get to know them. I will cherish the many movie and board game nights with them and Nkrumah Grant, Devin Lake, Elizabeth Baird, Jasper Gomez, Tanush Jagdish, and John Baltusis. Their friendships mean the world to me and they were always there to uplift me during the difficult times of graduate school. I cannot thank them enough. Lastly, I must thank my family. I had a wonderful up bringing with a sister, mother, and father who believed in me, unwaveringly. My sister, Kaci, is the most caring person I know, and I am so proud of the strong and confident woman she has become. One of the happiest days of my life was watching her walk do wn the aisle. Kaci is a wonderful mother to Rylee and my best friend. My mother, Lisa, is my stalwart advocate. She gave me the opportunities to follow my own interests and never gave up on me when I failed. She taught me to always keep fighting. My father , Dennis, died shortly after I began graduate school. The most profound lesson I have learned in graduate school and life in general is that d eath ends a life, not a relationship . I continue to learn from my father every day , and I know I inherited his strength of w ill . I am his monument. vii TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ .................... ix LIST OF FIGURES ................................ ................................ ................................ .................... x KEY TO ABBREVIATIONS ................................ ................................ ................................ .... xi CHAPTER 1: Historical contingency in the evolution of antibiotic resistance after decades of relaxed selection ................................ ................................ ................................ .......................... 1 Abstract ................................ ................................ ................................ ................................ ... 2 Introduction ................................ ................................ ................................ ............................. 3 Materials and Methods ................................ ................................ ................................ ............ 8 Bacterial strains ................................ ................................ ................................ ................... 8 Culture conditions and measurements of resistance and evolvability ................................ .... 9 Experimental design and da ta analyses ................................ ................................ ............... 11 Results ................................ ................................ ................................ ................................ .. 13 Antibiotic susceptibility profiles of the LTEE ancestral and derived clones ........................ 13 Changes in susceptibility under relaxed selection during the LTEE ................................ .... 14 Evolvability profiles of the ancestor and deri ved clones ................................ ..................... 14 Effects of genetic background on the evolvability of resistance ................................ .......... 15 Multiple factors can contribute to differences in evolvability ................................ ............. 20 Discussion ................................ ................................ ................................ ............................. 22 CHAPTER 2: Genomic evolution of antibiotic res istance is contingent on genetic background following a long - term experiment with Escherichia coli ................................ ............................ 28 Abstract ................................ ................................ ................................ ................................ . 29 Introduction ................................ ................................ ................................ ........................... 30 Materials and Methods ................................ ................................ ................................ .......... 33 Evolution experiments and bacterial strains ................................ ................................ ........ 33 Library preparation and genomic sequencing ................................ ................................ ..... 34 Mutation identification ................................ ................................ ................................ ....... 34 Statistical methods ................................ ................................ ................................ ............. 36 Results ................................ ................................ ................................ ................................ .. 38 Genomic evolution of strains evolved under four different antibiotic treatments ................. 38 Genomic parallelism at the functional level ................................ ................................ ........ 41 Specificity of genomic evolution in the different antibiotic environments ........................... 44 Specificity of genomic evolution with respect to genetic background ................................ . 48 Discussion ................................ ................................ ................................ ............................. 51 CHAPTER 3: Idiosyncratic variation in the fitness costs of tetracycline - resistance mutations in Escherichia coli ................................ ................................ ................................ ........................ 56 Abstract ................................ ................................ ................................ ................................ . 57 Introduction ................................ ................................ ................................ ........................... 58 viii Materials and Methods ................................ ................................ ................................ .......... 63 Experimental conditions and bacterial strains ................................ ................................ ..... 63 Fitness assays ................................ ................................ ................................ ..................... 64 Results ................................ ................................ ................................ ................................ .. 66 Tetracycline - resistant lines have reduced fitness in the absence of the antibiotic ................ 66 Cost of resistance varies among resistant mutants ................................ .............................. 66 Possible reversions of unstable mutations do not explain the variation in fitness cost ......... 67 Level of phenotypic resistance does not explain the variation in fitness cost ....................... 67 Genetic background does not explain the variation in fitness cost ................................ ....... 68 Idiosyncratic differences between mutant lin es in the cost of resistance ............................. 68 Discussion ................................ ................................ ................................ ............................. 71 CHAPTER 4: Conclusions and future outlook ................................ ................................ ........... 77 Summary ................................ ................................ ................................ ............................... 78 The multifactorial nature of antibiotic - resistance evolution ................................ ................... 79 Mutational supply rate ................................ ................................ ................................ ....... 80 Future directions ................................ ................................ ................................ ................ 83 Population bottlenecks ................................ ................................ ................................ ....... 85 Conclu ding remarks ................................ ................................ ................................ .............. 87 APPENDICES ................................ ................................ ................................ .......................... 88 APPENDIX A: Supplementary Ta bles ................................ ................................ .................. 8 9 APPENDIX B: Supplementary Figures ................................ ................................ ................. 96 REFERENCES ................................ ................................ ................................ ......................... 99 ix LIST OF TABLES Table 1.1 . Statistical analyses of declines in intrinsic resistance during relaxed selectio n of clones sampled at generation 50,000 of the LTEE. ................................ ................................ ............... 14 Table 1.2 . Statistical ana lyses of diminishing - returns trends in resistance evolvability of clones sampled at generation 50,000 of the LTEE. ................................ ................................ ............... 17 Table 1.3 . Statistical analyses of idiosyncrat ic patterns in resistance evolvability of clones sampled at generation 50,000 of the LTEE. ................................ ................................ ............................. 18 Table 1.4 . Statistical analyses comparing tetracycline resistance evolvability of c lones isolated from the Ara+5 population at different generations to the LTEE ancestor. ................................ . 20 Table 2.1 . Gene targets that contributed to antibiotic - treatment specificity . . .............................. 48 Table 2. 2 . Candidate gen es that may contribute to genetic - background specificity. ................... 51 Table 3.1 . ANOVA on the log - transformed fitness estimates of 16 tetracycline - resistant l ines, each measured relative to its sensitive parent. ................................ ................................ .................... 66 Table A. 1 . Bacterial strains used in Chapter 1 . . ................................ ................................ ......... 90 Table A. 2 . Statistical significance for changes in intrinsic resistance of the individual clones sampled at generation 50,000 of the LTEE for the fo ur an tibiotic treatments. ............................ 91 Table A. 3 . Statistical significance for trends of diminishing - returns resistance evolvability of the individual clones sampled at generation 50,000 of the LTEE for the four antibiotic treat ments. 92 Table A. 4 . Bacterial strains sequenced in Chapter 2. ................................ ................................ . 93 Table A. 5 . Bacterial strains used in Chapter 3 . . ................................ ................................ ......... 95 x LIST OF FIGURES Figure 1.1 . Schematic illustration of the evolvability of antibiotic resistance under three scenarios. ................................ ................................ ................................ ................................ ................... 6 Figure 1.2 . Intrinsic resistance usually declined over time in the absence of drug exposure. ...... 13 Figure 1.3 . Genetic background affects the evolvability of LTEE lines exposed to antibiotics. .. 16 Figu re 1.4 . Capacity to evolve tetracycline resistance was diminished early in one LTEE lineage. ................................ ................................ ................................ ................................ ................. 19 Figure 2.1 . Numbers and types of mutations in evolved genomes. ................................ ............. 40 Figure 2.2 . Distribution of independent mutations in different functional categories. ................. 43 Figure 2.3 . Specific ity of genomic evolution with respect to antibiotic treatment. ...................... 45 Figure 2.4 . Identity of mutated genes in antibiotic - resistant lines. ................................ .............. 47 Figure 2.5 . Specificity of g enomic evolution with respect to genetic background. ..................... 49 Figure 3.1 . Schematic illustration of fitness effects of antibiotic resistance mutations under two scenarios . ................................ ................................ ................................ ................................ .. 62 Figure 3.2 . Schematic illustration showing the derivation of the strains used in this study and the methods employed to measure the fitness of resistant lines r elative to their sensitive parents. .... 65 Figure 3.3 . Fitnesses of 14 tetracycline - resistant mutants relative to their parental st rains. ......... 70 Figure B. 1 . Schematic illustration of the LTEE and evolvability study design. .......................... 97 Figure B. 2 . Experimental plates. ................................ ................................ ............................... 98 xi KEY TO ABBREVIATION S AMP a mpicillin ANOVA analysis of variance bp base pair CF cystic fibrosis CIP ciprofloxacin CRO ceftriaxone df degrees of freedom DM Davis Mingioli DNA deoxyribonucleic acid dsDNA double - stranded deoxyribonucleic acid ESBL extended - spectrum beta - lactamase indel insertion or deletion IS insertion sequence LB Luria Bertani LTEE long - term evolution exp eriment MH Mueller Hinton MIC minimum inhibitory concentration MSW mutant selection window NCBI National Center for Biote chnology Information rpm rotations per minute TA tetrazolium arabinose xii TET tetracycline tmRNA transfer - messenger ribonucleic acid tRNA transfer ribonucleic acid 1 CHAPTER 1: Historical contingency in the evolution of antibiotic resistance after d ecades of relaxed selection Work presented in this chapter has been published as Card, K. J., T. LaBar, J. B. Gomez, and R. E. Lenski. 2019. Historical contingency in the evolution of antibiotic resistance after decades of relaxed selection. PLoS Biol . 17:e3000397. Reproduced with permission from the publisher: https://journals.plos.org/plosbiology/s/licenses - and - copyright 2 Abstract Populations often encounter changed environments that remove selection for the maintenance of particular phenoty pic traits. The resulting genetic decay of those traits under relaxed selection alters the subsequent evolvability of a population upon restoration of selection for a previously diminished trait is not well understood. We addressed this question using Escherichia coli strains from the long - term evolution experiment (LTEE) that independently evolved for multiple decades in the absence of antibiotics. We first confirme d t hat these derived strains are typically more sensitive to various antibiotics than their common ancestor. We then subjected the ancestral and derived strains to various concentrations of these drugs to examine their potential to evolve increased resista nce . We found that evolvability was idiosyncratic with respect to initial genotype; o e volve increased resistance was constrained in some backgrounds, implying that evolvability depended upon prior mutations in a historically contingent fashion. We further subjected a time series of clones from one LTEE population to tetracycline and dete rmi ned that an evolutionary constraint arose early in that population, corroborating the role of contingency. In summary, relaxed selection not only can drive populations to increased antibiotic susceptibility, but it can also affect the subsequent evolvab ili ty of antibiotic resistance in an unpredictable manner. This conclusion has potential implications for public health, and it underscores the need to consider the genetic context of pathogens when designing drug - treatment strategies. 3 Introduction A pop ula tion may encounter an environmental change that removes or reduces a selective pressure that was previously important for the maintenance of a trait (Darwin 1859; Lahti et al. 2009) . Adaptation to the itn ess in its prior environment. These correlated responses may lead to the functional decay of unused traits over time or, conversely, their maintenance despite relaxed selection (Lahti et al. 2009) . However, the evolutionary processes driving these respo nse s are often hard to disentangle because one must rely on retrospective studies and historical inference. By contrast, evolution experiments with microorganisms provide a powerful approach to study correlated responses. Microbes often have large populati on sizes and fast generations, and they are amenable to freezing and revival. One can therefore observe evolution in action, directly compare ancestral and derived forms, and simultaneously assess adaptation to one environment and quantify correlated fitne ss responses in another. Accordingly, numerous studies with bacteria (Chao et al. 1977; Lenski 1988; Reboud and Bell 1997; Cooper and Lenski 2000; Cooper et al. 2001; Ellis and Cooper 2010; Leiby and Marx 2014) , viruses (Turner and Elena 2000; Duffy et al. 20 06; Agudelo - Romero et al. 2008; Coffey and Vignuzzi 2011; Wasik et al. 2015; Meyer et al. 2016) , and yeast (Wenger et al. 2011; Ratcliff et al. 2012; Koschwanez et al. 2013) have found that fitness trade - offs between environments are common. Trade - offs are often caused by antagonistic pleiotropy, which occurs when a mutation that is beneficial in one environment is deleterious in another. This process can have important public - health consequences when antibiotic - resistance mutations or acquired resistanc e g enes impose costs on bacterial growth and competitiveness relative to their sensitive counterparts in the absence of drugs (Lenski 1997; Andersson and Hughes 2010) . Previous studies have shown that pleiotropic 4 fitness costs are widespread among resistan ce determinants to diverse drug classes (Nguyen et al. 1989; Schrag et al. 1997; Rozen et al. 2007; Han et al. 2009) , although their magnitudes are variable and may also depend on the genetic background (Lenski et al. 1994; Andersson and Hughes 2010; Melny k e t al. 2015; Palmer et al. 2018) . Given that antibiotic - resistance mutations and genes commonly impose fitness costs, one would expect that resistance should decline over time in the absence of antibiotic exposure. However, compensatory evolution often r edu ces or eliminates these trade - offs (Bouma and Lenski 1988; Schrag et al. 1997; Reynolds 2000; Rozen et al. 2007) . Adaptive trends during compensatory evolution have been studied using a number of E. coli mutants resistant to the drug rifampicin (Barrick et al. 2010) . That study found that the mutants were generally less fit than their sensitive progenitors in a permissive antibiotic - free environment; moreover, the compensatory effects of subsequent beneficial mutations were greater when the resistance wa s m ore costly. Thus, compensation exhibited a pattern of diminishing - returns adaptation in that study. Even when bacteria have no known history of exposure to antibiotics, they may have low level resistance to some drugs because of intrinsic structural or functional features, including their cell envelope and efflux pumps (Cox and Wright 2013). As a consequ ence, intrinsic resistance may decline in the absence of drug exposure if relevant genes accumulate mutations either by selection or drift in permissive environments (Cooper and Lenski 2000). A recent study used the E. coli long - term evolution experiment ( LTEE), and antibiotic stresses when it evolves for a lo ng period in the absence of those stresses (Lamrabet et al. 2019). In the LTEE, 12 replicate population s were founded from a common ancestor and have been propagated daily for over 30 years in a medium without antibiotics (Lenski et al. 1991; Tenaillon 5 et al. 2016). In particular, Lamrabet and colleagues measured changes in mostly low - level intrinsic resist ance between ancestral and derived strains isolated from each population after generations 2,000 and 50,000. They found that derived strains were usually more susceptible to most antibiotics than their ancestor, and from multiple lines of evidence they inf erred that these losses of intrinsic resistance resulted primarily from pleiotropic side effects of beneficial mutations that arose during the LTEE. Alth ough the lineage leading to the LTEE ancestor has no known history of exposure to industrially manufact ured antibiotics (except streptomycin), it might nevertheless have a history of ex p osure to similar compounds produced by competitors and to host bile sa lts. Adaptations that provide resistance to these other stressors, such as those involving the cell env elope and efflux pumps (Thanassi et al. 1997) , often confer intrinsic resistance to antibiotics (Cox and Wright 2013) . Thus, even the low - level resistanc e of the ancestor might reflect this prior natural history, and selection was relaxed on these traits in the LTEE environment. Taken together, the experimental evolution studies described above have two contrasting implications relevant for medicine and pu blic health. First, resistance to antibiotics (including even low - level intrinsic resistance) m ay decline in the absence of drug exposure. Second, evolution can often compensate for deleterious side effects of mutations, thereby facilitating the maintenanc e of evolved resistance. The question then arises how readily bacteria can overcome losses of a ntibiotic resistance that arose during periods of relaxed selection through subsequent evolution in the presence of drugs. In this study, we address this fundame ntal question by using the LTEE ancestor and derived strains isolated from four populations aft er 50,000 generations to examine how evolution in the absence of when drugs are intro duced. In so doing, we examine the role that genetic background plays in 6 resistance evolvabilit y (Fig ure 1 .1 ). Does resistance evolution tend to follow a general trend of diminishing returns (Travisano et al. 1995; Barrick et al. 2010; Wiser et al. 2013; K ryazhimskiy et al. 2014) , such that derived strains that are initially more susceptible to a dr ug can increase their resistance disproportionately relative to their ancestor (Fig ure 1 .1 B)? Or is evolvability idiosyncratic with respect to prior evolutionary history (Blount et al. 2008, 2018) , such that the relative gains in resistance are independent of ure 1 .1 C)? Figure 1.1 . Schematic illustration of the evolvability of antibiotic resistance under three scenarios. A s from an ini tially susceptible genotype during one round of drug selection. (A) Null model, with no effect of genetic background on evolvability. (B) Diminishing - returns mod el, such that backgrounds with low initial resistance are more evolvable than backgrounds that are initially more resistant. (C) Idiosyncratic - effects model, in which evolvability varies among genetic backgrounds but is uncorrelated with their initial leve l of resistance. Throughout this chapter , we discuss how differences in genetic background may affect the evolvability of antibiotic resistance. This focus brings to mind the concept of epistasis, whereby the marginal effect of a particular mutation on som e phenotype of interest depends on its interaction with another mutation or, more generally, the set of mutations that distinguish genetic backgrounds (Levin et al. 2000; Moore et al. 2000; Trindade et al. 2009; MacLean et al. 2010; 7 Chou et al. 2011; Khan et al. 2011; Kryazhimskiy et al. 2014; Wong 2017) . In the context of our study, ep istasis could arise in at least two ways. First, mutations in the same target gene may confer different levels of resistance depending on other mutations that differ between backgrounds. Second, the physiological mechanisms and associated loci underlying r esistance may differ across genetic backgrounds. It might seem unexpected that the genetic basis of resistance would differ among closely related backgrounds. However, as we will show, the initial levels of resistance vary among backgrounds, and the physio logical mechanisms that allow cells to resist drugs may depend on their concentrations, such that the mechanisms used may also differ across backgrounds. Without a more preci se mechanistic understanding at this stage of our work, we cannot distinguish betw een these forms of epistasis. More generally, we will use the term evolvability (rather than epistasis) because it emphasizes the consequences of these effects for antibiotic resistance. We confirmed the finding of Lamrabet and colleagues (2019) that the L TEE - derived strains had typically become more susceptible to antibiotics during relaxed selection. However, contrary to our expectation based on a diminishing - returns model, we discovered that these derived strains were usually no more evolvable (and somet imes less evolvable) than their ancestor when exposed to various antibiotics. Instead, idiosyncratic responses dominated over any diminishing - returns tendency, such that the capacity to evolve resistance was hampered on some LTEE - derived genetic background s. These results indicate that evolution and diversification of a single bacterial species in a permissive environment can lead to unpredictable changes in the potential to e volve antibiotic resistance. Our work suggests that methods for predicting, at the evolutionary potential should be developed in light of the global threat of antibiotic resistance. If successful, such methods could become an int egral aspect of resistance surveillance and patient treatment. 8 Materials and Metho ds Bacterial strains All of the strains used in this study are from the E. coli LTEE. In the LTEE, 12 replicate populations were founded from a common ancestral strain called REL606 (Lenski et al. 1991) . These populations have been propagated for over 3 2 y ears by daily 1:100 transfers in glucose - supplemented Davis minimal (DM) medium without any antibiotics (Lenski et al. 1991) , resulting in >7 3 ,000 cell generations to date. S In this study, we qu antified the intrinsic antibiotic resistance and evolvability of the ancestor and derived clones isolated from four populations (designated Ara 5, Ara 6, Ara+4, and Ara+5) af ter 50,000 generations of the LTEE. We chose these strains for two reasons. First, the source populations of these derived clones retained the low ancestral mutation rate, and therefore they accumulated many fewer mutations than their counterparts from sev eral populations that evolved hypermutability (Tenaillon et al. 2016) . This charac teristic should increase the tractability of identifying candidate alleles affecting resistance evolvability, which we hope to achieve in future work. Second, generation 50,0 00 is the latest point at which whole - genome sequence data are available for the c lonal samples (Tenaillon et al. 2016) . We also examined when the Ara+5 population evolved a diminished capacity to increase its tetracycline resistance (as described in the R esults) by testing two strains isolated from this population at several earlier ti me points (generations 500, 1,000, 1,500, 2,000, 5,000, and 10,000). All of the strains used in this study are listed in Appendix A Table 1 . 9 Culture conditions and measurem ents of resistance and evolvability ial strains were revived from frozen stocks by overnight growth in Luria Bertani (LB) medium. Cells from these cultures were then streaked onto DM agar plates supplemented wi th 4 mg/mL glucose. We randomly picked single isolated colonies from these plates to start multiple replicate populations in LB. Final population sizes in the LB 9 cells/mL. When an initially susceptible cell expands i nto a colony and then a population, new mutations spontaneously occur and increase in number during growth (Luria and Delbrück 1943) . The evolution of antibiotic - resistant mutants will therefore originate by independent mutational events in each replicate population (Luria and Delbrück 1943; Kassen and Bataillon 2006) . We define a initially susceptible genotype during one round of drug selection. Evolvability experiments were per formed using Mueller - Hinton (MH) agar (Acumedia, Lansing, MI) supplemented wi th 1 mg/mL glucose, 0.1 mg/mL magnesium sulfate, 0.01 mg/mL thiamine, and a series of 2 - fold dilutions of an antibiotic. We used MH agar because a previous study used this medium to quantify the susceptibilities of LTEE clones to various antibiotics (Lamra bet et al. 2019) . Our study, in part, sought to replicate these findings. We chose the four antibiotics in our study because they have diverse cellular targets: ampicillin and cef triaxone inhibit cell - wall synthesis, ciprofloxacin inhibits DNA replication, and tetracycline inhibits protein synthesis. We prepared stock solutions of One - millil iter samples of each population were centrifuged at 8,000 rpm for 2 minutes a nd 1 2 × 10 8 cells) from each suspension onto the antibiotic - amended MH agar plates, an d MICs were 10 evaluated after 48 hours of incubation. For this study, we operat ionally define a pair of MICs for each series of antibiotic - amended plates as the lowest concentration that prevents either confluent growth or isolated colonies. According to thi s approach, confluence indicates growth by the evolvability was calculated from the difference in MIC between these two genotypes. For each experimental blo ck, putative resistant mutants were confirmed by streaking one randomly chose n colony per strain onto fresh antibiotic - amended MH plates. All clones regrew at the corresponding concentration. This approach indicated that a selected clone was indeed a resis tant mutant with a stably inherited increase in its MIC, as opposed to a so - c - than - average phenotypic tolerance relative to genetically identical cells (Balaban et al. 2019) . Cultures of mutant clones were then frozen glycerol as a cryoprotectant. Ap pendix B Figure 1 provides a schematic representation of our methods for measuring the MICs of sensitive parental strains and their resistant daughter derivatives. In Appendix B F igure 2 , we show an image of the resulting plates for one replicate series ac ross a 256 - fold (= 28) range of ciprofloxacin concentrations for the LTEE ancestral clone. In this image, one sees confluent growth on the first 3 plates, isolated colonies on the next 2 plates, and no evident growth on the 4 plates with the highest concen trations. Based on these plates, we scored the MIC of the sensitive parental strain as the lowest concentration that inhibited confluent growth, which was mple. We scored the MIC of the resistant daughter derivative as the lowest co ncentration where even isolated colonies were absent, 2 - transformed difference between these values (i.e., log 2 0.01/0.0025 = 2 in this example) provides one estimate of the evolvability of t he LTEE ancestral strain with res pect to ciprofloxacin. We 11 obtained 32 independent esti mates of t hese MICs and the associated evolvabilities for the ancestral strain against each of the four antibiotics used in our study. We similarly obtained eight independent estimates of t he MICs and a ssociated evolvabilities for each of the four 50,000 - g eneration strains used in our study against each of the same antibiotics. Photographs of all of the replicate plate series used to estimate these values have been archived on the Dryad Digital Repository: https://datadryad.org/stash/dataset/doi:10.5061/dry ad.g41hg96 (Card 2019) . Experimental design and data analyses All MIC values were transformed by taking their base - 2 logarithm because the antibiotic concentrations were tested across a series of 2 - fold dilutions. For each experimental block, an independe ntly isolated LTEE ancestral clone was paired with each derived clone. We had two p redictions when we began this study: (i) the derived bacteria would be more susceptible to antibiotics (lower MICs) than their common ancestor as a consequence of the relaxe d selection they experienced in the permissive LTEE environment and (ii) the derive d bacteria would be more evolvable than their ancestor when challenged with antibiotics, following a general trend of diminishing - returns adaptation. Statistical tests that rely on normally distributed data were deemed inappropriate for this study owing to the discrete, lumpy nature of the measurements. Instead, we used nonparametric methods. There were also numerous instances in which the derived clones were equal both in MI C and evolvability to the paired assays for the ancestor, and these ties introduced additional complications. Therefore, we used trinomial tests to examine changes in the direction of our expectations relative to the null hypothesis that changes are equall y frequent in either direction (Bian et al. 2011). We performed these analyses by i ndividually comparing the four derived clones 12 with their paired ancestors across each antibiotic. Probabilities were then combined from these independent significance tests freedom (i.e., df = 2 k ; where k is the numb er of comparisons) (Fisher 1934; Sokal and Rohlf 1994). As explained previously (Fig ure 1 .1 C), evolvability might be idiosyncratic and therefore not correlated with the init ial level of resistance. To assess this possibility, we performed a Kruskal - Wallis one - way ANOVA to test for heterogeneity in evolvability among the LTEE lines. Datasets and the details of our statistical analyses are provided in an R Notebook on the Dryad Digital Repository: https://datadryad.org/stash/ dataset/doi:10.5061/dryad.g41hg96 (Card 2019) . We performed fluctuation tests to estimate effective mutation rates (Luria and Delbrück 1943) r a derived clone from the two strains were taken from the freezer and grown overnight in LB broth. Each culture was then serially diluted in saline solution, and fewer than 1,000 cells were transfer red into each well of a 96 - well plate; each well contained 0.1 mL of LB broth. Afte r 24 hours, we removed the entire volume from each of 84 wells and or REL606, respectively. The other 12 wells were sampled to enumerate the bacteria using a Cou lter counter (Multisizer 4e, Beckman) with a 30 - distinguish cells from background debris and subtracted counts from a sterile LB - on ly negative control. We incubated the antibiotic - amended plates for 48 hours and we then scored each plate for the absence or presence of one or more colonies, as required for the p 0 method to estimate mutation rates. 13 Results Antibiotic susceptibility pr ofiles of the LTEE ancestral and derived clones Antibiotic susceptibility measureme nts were generally quite repeatable (Fig ure 1.2 ). For each antibiotic, all 32 independent ancestral replicate minimum inhibitory concentration (MIC) measurements were identi cal. Among the 16 sets of derived - clone replicates (4 clones × 4 antibiotics), the 8 replicate assays gave identical MICs in 2 cases (12.5%), they deviated minimally by a factor of 2 in 12 cases (75%), and in only 2 cases they deviated by a factor of 4 (12 .5%). These results provide strong support for the use of our plate - based approach, as described in the Materials and Methods and shown in Appendix B Figure 1 , to quantify antibiotic susceptibility profiles. Figure 1.2 . Intrinsic resistance usually decli ned over time in the absence of drug exposure. Comparison of t he LTEE ancestor and four independently derived clones sampled after 50,000 generations for their susceptibilities to ampicillin, ceftriaxone, ciprofloxacin, and tetracycline (A D). MICs are sho wn on a log 2 - transformed scale to reflect the fact that antibi otic concentrations were tested across a series of 2 - fold dilutions. In each panel, points show values obtained from 32 and 8 replicate assays for the ancestor and derived strains, respectively. Horizontal bars show the median of the log 2 - transformed MIC v alues for each strain on each antibiotic. The absolute values of the concentrations shown on the y - axis differ among the four antibiotics, but the range is the same in each panel. 14 Changes in sus ceptibility under relaxed selection during the LTEE For each a ntibiotic, we made 32 comparisons between the MICs of derived clones (4 clones × 8 replicates) against their paired and independently isolated ancestral clones. On balance, we observed increased susceptibility of the strains that evolved under relaxed sele ction (i.e., in the absence of antibiotics) during the LTEE, consistent with recently published results (Lamrabet et al. 2019) (Fig ure 1.2 ). All four derived strains have increased sensitivity t o ampicillin (Fig ure 1. 2A), ceftriaxone (Fig ure 1. 2B), and tet racycline (Fig ure 1. 2D) relative to their common ancestor, and two of the derived strains were more susceptible to ciprofloxacin (Fig ure 1. 2C). These trends toward lower resistance are well supp orted by trinomial tests, as described in the Materials and M e thods , and as shown in Table 1 .1 and Appendix A Table 2 . Table 1.1 . Statistical analyses of declines in intrinsic resistance during relaxed selection of clones sampled at generation 50,000 of th e LTEE. Antibiotic 2 p Ampicillin 40.06 < 0.0001 Ceftriaxone 45.33 < 0.0001 Ciprofloxacin 27.15 0.0007 Tetracycline 41.40 < 0.0001 independent hypothesis test s of the same hypothesis, with an underlying trinomial distribution for the null . Evolvability profiles of the ancestor and derived clones Next, we examined how the prior history of relaxed selection affected the evolvability of antibiotic resistance in th e different genetic backgrounds. To address this question, we selected mutants of the ancestral and LTEE - derived strains that survived and grew sufficiently to form colonies at higher concentrations of the four antibiotics than their corresponding parental strains (Appendix B 15 Figure 1). As described in the Materials and Methods, we operationally def ine evolvability as the maximum observed increase in antibiotic resistance from an initially susceptible genotype during one round of drug selection (Figure 1.1) . Evolvability measurements tended to be more variable than the MIC measurements. We examined t he evolvability of 128 independent ancestral clones across the four antibiotics. There were 73 cases (57%) in which these measurements corresponded to the median for that antibiotic, 49 cases (38.3%) in which they differed by a factor of 2, and 6 cases (4. 7%) in which they differed by a factor of 4. Likewise, among the 16 sets of replicates for the LTEE - derived clones, the 8 assays varied by a factor of 2 in 12 ca ses (75%) and by a factor of 4 in 4 other cases (25%). The greater variation in evolvability me asurements in comparison with MIC values among replicate assays is expected given the stochastic appearance of mutations in replicate cultures (Luria and Delbrüc k 1943) . Also, increased resistance can occur through multiple mutational paths (Toprak et al. 2012; Baym et al. 2016) , and those mutations affecting one mechanism might confer greater resistance evolvability relative to mutations affecting some other mech anism. Effects of genetic background on the evolvability of resistance We examined the possibi lity of two broad patterns of genetic - background effects with respect to resistance evolvability in our study. First, we asked whether evolvability followed a tr end of diminishing returns, such that the more susceptible LTEE - derived genetic backgrounds gen erally produced mutants with proportionally greater grains in resistance than the ancestor. Both the ancestral and derived strains evolved resistance to varying degrees (Fig ure 1. 3). The evolutionary potential of two of the four derived clones (Ara 5 and A ra 6) was noticeably greater relative to 16 their ancestor in the ampicillin environment (Fig ure 1. 3A), but there were no clear instances of similar trends in the t hree other drug environments (Fig ure 1. 3B D). Figure 1.3 . Genetic background affects the evolvability of LTEE lines exposed to antibiotics. Lines joining susceptible parental strains with their daughter mutants show the increases in resistance during one round of selection with ampicillin, ceftriaxone, ciprofloxa cin, and tetracycline (A D). If the slope of a derived strain is greater than that of the ancestor, then it has greater evolvability; and vice versa. Median MICs are shown on a log 2 - transformed sc ale to reflect the fact that antibiotic concentrations were tested across a series of 2 - fold dilutions. The y - axis ranges for the four drugs have been scaled to the ceftriaxone environment, which had the largest gains in resistance between the susceptible parental cells and the resistant daughter cells. Overall, th ere was no statistical support for the diminishing - returns trend, despite the visual with its pair ed ancestor in the four drug environments. We used trinomial tests to quantify the tested against the null hypothesis of equally frequent changes in either d irection, after taking into account the many numerical ties (Bian et al. 2011) . Although the capacity of the derived Ara 5 clone to evolve increased resistance was significantly greater than its ancestor when considered in isolation ( Appendix A Table 3 ), i t was marginally nonsignificant when we examined overall tre nds 17 for each antibiotic (Table 1. 2) using a meta - analysis approach (Fisher 1934; Sokal and Rohlf 1994) . Table 1.2 . Statistical analyses of diminishing - returns trends in resistance evolvability of clones sampled at generation 50,000 of the LTEE. Antibiotic 2 p Ampicillin 14.63 0.0668 Ceftriaxone 0.18 1 Ciprofloxacin 7.88 0.4456 Tetracycline 5.97 0.6511 independent tests of the same hypothesis, with an underlying tr inomial distribution for the null hypothesis. We then asked whether the proportional resistance gains when exposed to the antibiotics were idiosyncratic among LTEE lines. For example, the capacity to evolve ceftriaxone resistance appeared to b e reduced amo ng three LTEE - derived backgrounds (Ara+5, Ara 6, and especially Ara+4) relative to their common ancestor (Fig ure 1. 3B). Similarly, the evolvability of the Ara+5 background with respect to tetracycline appears to be constrained (Fig ure 1. 3D). I ndeed, this l atter case was the only one in which the mutants of a strain systematically achieved a lower level of resistance than did the mutants of other strains that were initially more susceptible (indicated by the crossing lines in Fig ure 1. 3D). These idiosyncrati c tendencies are statistically well supported by Kruskal - Wallis tests. For both ceftriaxone and tetracycline, these tests reject the null hypothesis of homogeneity in proportional resistance increases across the different genetic backgrounds ( Table 1. 3). 18 Table 1.3 . Statistical analyses of idiosyncratic patterns in resistance evolvability of clones sampled at generation 50,000 of the LTEE. Antibiotic 2 p Ampicillin 9.19 0.0566 Ceftriaxone 23.45 0.0001 Ciprofloxacin 7.59 0.1077 Tetracycline 10.18 0. 0376 Analyses were performed using a Kruskal - Wallis one - way nonparametric ANOVA (df = 4). Given these idiosyncratic effects of genetic background, we chose to examine one of the cases in greater detail. In particular, we asked when the evolvability of the Ara+5 background declined with respect to tetracycline. To address this question, we examined clones isolated during resistance during a single exposure evolvability. As shown in Fig ure 1. 4, the reduced evolvability was evident in all of the clones isolated from generation 2,000 onward as well as in one of two clones isolated at generation 1,500. With one exception, all of the LTEE - derived parental backgrounds across this time series had the same MIC value as the ancestor (Fig ure 1. 4A). However, the daughter mutants from the later - generation Ara+5 genetic backgrounds had progressively lower levels of t etracycline resistance (Fig ure 1. 4B), which when coupled with the same initial resistance level indicates they had become less evolvable in this respect (Fig ure 1. 4C). A Kruskal - Wallis test decisively rejects the null hypothesis of equal evolvabilities acr oss the entire set of 2 = 67.89, df= 12, p < 0.0001), each derived clone from LTEE population Ara+5 with that of the ancestor support the temporal break point described above (Table 1. 4). 19 Figure 1.4 . Capacity to evo lve tetracycline resistance was diminished early in one LTEE lineage. (A) Comparison of the intrinsic tetracycline resistance of the ancestor and a time series of derived strains isolated from the Ara+5 population. Strains are ordered by their time of isol ation. The strain identifiers begin with a number corresponding to the generation (in thousands) of their isolation, followed by an arbitrary letter; strains 2A and 2B, for example, are two clones isolated at generation 2,000 of the LTEE. (B) Comparison of the evolved resistance levels after one round of drug selection, based on the MICs of the mutant daughter cells derived from the corresponding parental strains. MICs are shown on a log 2 - transform ed scale to reflect the fact that the concentrations of anti biotics were tested across a series of 2 - fold dilutions. (C) Evolvability is quantified for each strain as the difference in the log 2 - transformed MICs of the parental strain and its corresponding daughter mutant. Points show 10 independent replicates per s train. Horizontal bars show the median log 2 - transformed MICs (A, B) for 10 replicate assays and the evolvability (C) based on the corresponding 10 paired differences. 20 Table 1.4 . St atistical analyses comparing tetracycline resistance evolvability of clone s isolated from the Ara+5 population at different generations to the LTEE ancestor. Strain Difference Lower 95% CI Upper 95% CI p 0.5A - 0.3 - 0.9 0.3 0.6921 0.5B - 0.3 - 0.9 0.3 0.6921 1A 0.0 - 0.6 0.6 1 1B - 0.2 - 0.8 0.4 0.9607 1.5B - 0.3 - 0.9 0.3 0.6921 1.5A - 0.6 - 1.2 0.0 0.0412 2A - 1.0 - 1.6 - 0.4 < 0.0001 2B - 0.8 - 1.4 - 0.2 0.0021 5A - 0.7 - 1.3 - 0.1 0.0101 5B - 1.2 - 1.8 - 0.6 < 0.0001 10A - 1.2 - 1.8 - 0.6 < 0.0001 10B - 1.3 - 1.9 - 0.7 < 0.0001 ers begin with a number that corresponds to the generation (in thousands) of their isolation, followed by an arbitrary letter; strains 2A and 2B, for example, are two clones isolated at generation 2,000 of the LTEE . Multiple factors can contribute to diffe rences in evolvability The observed changes in evolvability of tetracycline resistance in the Ara+5 population could, in principle, reflect several factors, including differences among genotypes in cell density, mutation rate, and the number of potential m utations that confer sufficient resistance to allow growth at a given drug concentration (i.e., the effective mutational target size). To examine these facto rs, we performed Luria - Delbrück fluctuation tests (Luria and Delbrück 1943; Pope et al. 2008; Lang 2018) with the LTEE ancestor and 2,000 - generation clone 2A at tetracycline concentrations of 4 iotic concentrations because, in our evolvability assays, the 2,000 - generation clone had never prod uced any mutants that formed ure 1. 4). We grew 96 replicate 0.1 - mL cultures of each strain, starting from small population sizes to e nsure mutational independence ( Luria and Delbrück 1943 ; Pope et al. 2008) . Because antibiotic - resis tant mutants often grow more slowly than their progenitors 21 (Schrag et al. 1997; Rozen et al. 2007; Han et al. 2009; Andersson and Hughes 2010; Barrick et al. 0 meth od is insensitive to possible differences in growth rate between parent strains and daughter mutants, as it uses only the fraction of the replicate assays th at do not yield any mutants ( Luria and Delbrück 1943 ; Pope et al. 2008) . For each strain, we used 1 2 cultures to estimate the population size and 84 cultures to test for resistant mutants at the relevant concentration. The 2,000 - generation clone yielded on ly about half the cell density as the ancestral strain (8.4 × 10 7 versus 1.8 × 10 8 cells per 0.1 - mL culture, p < 0.0001, based on a two - t test), which indicates one factor that would contribute to its lower evolvability. For the ancestral cult ures yielded none. For the derived strain, 26 test cu ltures yielded one or more mutants resistant 0 method, the estimated effective mutation rate for the ancestral strain is 7.7 × 10 10 per cell generation [ approximate 95% confidence limits of 4.2 × 10 10 9 based on the uncertainty in p 0 only (Agresti and Coull 1998)] , and for clone 2A the estimated rate is 4.4 × 10 9 (approximate 95% confidence limits of 9 ). At first glance, it may seem counterintuitive that the estimated mutat ion rate was higher for the derived clone than for the ancestor, but recall that we tested this clone at a lower antibiotic concentration. Hence, the two rates are not directly comparable. Moreover, published genome sequences and analyses (Tenaillon et al. 2016) indicate that the underlying point - mutation rate in clone 2A is the same as the ancestral rate. The effective mutation rate that is estimated using a fluctuation test depends on the product of the underlying mutation rate and the effective mutationa l target size (i.e., the number of mutations that would allow a colony to grow on the antibiotic test 22 plate). Given the same underlying mutation rate, the difference in the estimated mutation rates implies that the derived clone has a larger mutational tar tetracycline than the that antibiotic. These analyses indicate that differences in evolvability between genotypes may reflect differences in several factors cell density, underlying mutation rate, and ef fective target size that depend not only on the genetic background but also reflect complex interactions between the genetic background, potential resistance mutations, and the selective environment. Discussion In this study, we addressed a fundamental qu estion about how relaxed selection on a particular set of organismal traits affects their evolvability in situations in which those traits again become advantageous. The traits we studied are resistances to several antibiotics, and the question of how chan ges in genetic background that occur during relaxed selection affect the subsequent evolvability of resistance has potentially important implications for public health. To address these issues, we examined the capacity of E. coli strains to evolve increase d resistance to four different antibiotics after they had evolved in a drug - free environment for 50,000 generations as part of the LTEE. We confirmed that intrinsic resistance tended to decay among the LTEE - derived clones to all four antibiotics we tested (Fig ure 1. 2, Table 1 .1 ). Our results are consistent with a recent study that examined losses of intrinsic resistance in all 12 LTEE lines at generations 2,000 and 50,000 (Lamrabet et al. 2019) . Unlike that previous work, however, we then also examined whet her and how the LTEE - derived bacteria had changed in their evolvability, specifically their potential to evolve resistance when challenged across a range of concentrations of the same four antibiotics. 23 We examined two alternativ e hypotheses that might bea r on resistance evolvability. The first is called diminishing returns (Figure 1.1B), and it often characterizes the course of adaptive evolution (Moore et al. 2000; Orr 2005; Barrick et al. 2010; Khan et al. 2011; Wiser et al. 20 13; Kryazhimskiy et al. 201 4; Passagem - Santos et al. 2018) . For example, one study used rifampicin - resistant mutants to examine the relation between their initial fitness costs in the absence of this drug and their ability to reduce or eliminate those cost s during subsequent evoluti on, again in a drug - free environment (Barrick et al. 2010) . They first isolated eight rpoB mutants after a single round of antibiotic selection and showed that the mutants varied in their fitness defects. The authors then propaga ted these mutants and detec ted the first beneficial mutations to sweep to high frequency in those populations. They found that the lower - fitness backgrounds gave rise to mutations that conferred greater advantages than did the backgrounds that initially ha d higher fitness, in accord ance with a diminishing - returns model. If the evolution of antibiotic resistance after a period of decay under relaxed selection conformed to the diminishing - returns model, then we would expect the more susceptible LTEE - derived backgrounds to be more evol vable than their common ancestor. Howev er, we found little statistical support for diminishing returns in our study (Figure 1.3, Table 1.2). There was one instance in which an individual LTEE - derived clone was significantly more evolvable than the ancestor in the ampicillin environment, and two other clones trended in this direction (Figure 1.3A, Appendix A Table 3 ). However, the statistical support, even for ampicillin, was marginal at best when the evolvabilities of the four clones were analyzed together to account for multiple tests of the sa me hypothesis (Table 1.2). In any case, diminishing returns was not typical across the entire set of experiments. 24 The absence of an overall trend toward diminishing returns might be attributable in part to two methodo logical issues. First, it might point t o a limitation of our plate - based approach, and conventional MIC assays in general, to discern subtle differences in MICs and hence in evolvabilities based on differences in MIC values. That is, slight differences in evolvability may be obscured by the dis crete resolution of the assays using 2 - fold increasing concentrations of an antibiotic. Consistent with this possibility, the range of initial susceptibilities was greatest in the ampicillin environment, where the tre nd toward diminishing returns was most evident (Figure 1.3A). An alternative approach that might better capture subtle trends would be to use a continuous culture device that dynamically adjusts drug concentration in the growth medium to match the ongoing adaptive dynamics of the population und er study (Toprak et al. 2012). Second, we might have had insufficient statistical power to resolve diminishing - returns trends in evolvability. We tested four LTEE - derived lines and their ancestor, whereas some other s tudies that show diminishing returns in other contexts have used as many as hundreds of lines (Kryazhimskiy et al. 2014) . There is a third factor one that is biological, rather than methodological that could also obscure any tendency toward diminishing ret urns, and that is idiosyncratic heterog eneity among genetic backgrounds in their evolvability (Figure 1.1C). This pattern occurs when particular future evolut ionary potential with respect to a give n selective pressure. We found that the capacity to evolve ceftriaxone resistance tended to be lower for the derived clones than for the ancestor (Figure 1.3B, Table 1.3). That tendency is in contrast to the ampicilli n environment, and it is unexpected giv - lactams that target cell - wall synthesis. In a similar vein, we also demonstrated that the Ara+5 lineage had become significantly constrained in its ability to 25 evolve tetracycline resistance re lative to both its ancestor and the oth er LTEE - derived lineages (Figure 1.3D) . We conclude, therefore, that historical contingency has played an important role in the capacity of the LTEE - derived populations to respond evolutionarily to changed environments, in particular when challenged with a ntibiotics. That is, different lineages accumulated genetic differences even in replicate populations that evolved in the same environment that influence their ability to evolve and adapt in new directions. Several other microbial evolution studies have al so documented cases of historically contingent outcomes. For example, the mutations that accumulated in one LTEE population potentiated the subsequent evolution of a novel metabolic capacity that arose in only that one population, despite comparable time a nd opportunity in the 11 other replicate populations (Blount et al. 2008; Quandt et al. 2015; Leon et al. 2018) . Similarly, an experiment with Pseudomonas aeruginosa showed that high - level colistin resistance was potentiated by prior mutations in transcrip tional regulators phoQ and pmrB (Jochumsen et al. 2016) . However, the consequences of contingency in these two cases were the opposite of what we saw in our study: namely, evolvability was potentiated in these studies, whereas it became more constrained in ours. Still other studies have found little evidence for historical contingencies affecting evolvability. For example, Travisano and colleagues (1995) isolated a clone from each LTEE population after 2,000 generations in the glucose - limited medium. They t hen founded three replicate populations from each clone and let them evolve for 1,000 generations in the same environment, except with maltose replacin g glucose as the limiting resource. The founding clones had independent histories in the glucose environm ent, and they varied greatly in their initial fitness in the maltose environment. Despite this initial heterogeneity, however, the populations rapidly 26 converged on similar fitnesses in the new maltose environment. Thus, adaptation dominated over contingenc y in that experiment. Returning to our study, we ask the following questions: When did the evolvability with respect to tetracycline exposure decline i n the Ara+5 population? And why, in molecular - genetic terms, did it decline? To answer the first question , we tested clones from throughout this the same degree as the ancestral strain. We found that this constraint was already present in one of two clones s ampled at 1,500 generations of the LTEE, and it was evident in all of the clones we tested from generation 2,000 and onward (Table 1.4) despite these b ackgrounds having retained the same initial resistance level as the ancestor (Figure 1.4). These data thu s confirm the idiosyncratic effects of genetic background on the evolvability of resistance. By performing fluctuation tests to measure effective mutat ion rates, we also showed that evolvability can depend in complex ways on several factors, including not the cell density it achieves and the effective mutational target size for a given antibiotic concentr ation. With respect to the second question, we do not yet know the answer, but the early timing of the ch ange in evolvability allows us to narrow substantially the genetic possibilities. By combining our phenotypic results with previously obtained genomic data (Tenaillon et al. 2016), we have identified three candidate mutations that alone or in combination c ould explain this reduced evolvability. These mutations arose in the following genes: mreB , which encodes a protein involved in cell - wall structuring; pykF , which encodes pyruvate kinase that catalyzes the last step of glycolysis; and trkH , which encodes a potassium ion transporter. Interestingly, recent studies have discovered that mutations in trkH can cause increased susceptibility to tetracycline thr ough 27 changes to the proton - motive force, and this relationship may depend upon the genetic background (Lá zár et al. 2013; Apjok et al. 2019) . In future work, we hope to make genetic constructs that will allow us to investigate the genetic basis for the low evolutionary potential of this background when exposed to tetracycline. We will also sequence some of th e antibiotic - resistant mutants that evolved in our experiments to test whether there are any systematic differences among the various strains in the ge netic targets of the resistant mutations and whether such differences correlate with the history of relax ed selection and concomitant increased susceptibility. In summary, we have shown that bacterial evolution in the absence of antibiotic exposure can lea d not only to increased susceptibility but also to genetic background - dependent changes in resistance evo lvability when cells are exposed to those drugs. The evolution of resistance can thus depend upon previously accumulated mutations in a historically co ntingent fashion. These findings could have important health implications if evolution in an antibiotic - f ree environment sometimes therefore suggest that strategic antibiotic management may benefit not only from surveillance of current resistance levels in pa thogens but also from analyses of their potential to evolve increased resistance. This approach could be valuable both on the scale of the individual p atient, where effective treatment is paramount, and on a community - wide scale, where judicious efforts to control the spread of drug resistance become critical. We hope that our evolvability - based approach and extensions thereof prove useful in achieving t hese objectives. 28 C HAPTER 2: Genomic evolution of antibiotic resistance is contingent on genetic backgro und following a long - term experiment with Escherichia coli 29 Abstract Antibiotic resistance is a growing health concern. Efforts to control resistance w ould benefit from an improved ability to forecast when and how it will evolve. Epistatic interactions bet ween mutations can promote divergent evolutionary trajectories, which complicates our ability to predict evolution. W e recently showed that differences between genetic backgrounds can lead to idiosyncratic responses in the evolvability of phenotypic resist ance, even among closely related Escherichia coli strains. In this study, we examined whether a strain's genetic background also influences the genotyp ic evolution of resistance. Do lineages founded by different genotypes take parallel or divergent mutatio nal paths to achieve their evolved resistance states? We addressed this question by sequencing the complete genomes of antibiotic - resistant clones that evolved from several different genetic starting points during our earlier experiments. We first validate d our statistical approach by quantifying the specificity of genomic evolution with respect to antibiotic treatment. As expected, mutations in particul ar genes were strongly associated with each drug. Then, we determined that replicate lines evolved from t he same founding genotypes had more parallel mutations at the gene level than lines evolved from different founding genotypes, although these effects w ere more subtle than those showing antibiotic specificity. Taken together with our previous work, we conc lude that historical contingency can alter both genotypic and phenotypic pathways to antibiotic resistance. 30 Introduction Convergent evolution is comm on in nature. The independent emergence of winged flight in insects and mammals, and of camera - like eyes in vertebrates and cephalopod mollusks, are familiar but striking examples of how evolution can drive distantly related lineages to similar phenotypic outcomes (Conway Morris 2003). For over a century, biologists have sought to understand the processes und erlying these patterns and quantify the extent of convergent evolution in the natural world. However, quantifying convergence in nature is difficult fo r at least two reasons. First, one typically observes only a biased sample of possible outcomes. For exam ple, extinct lineages that evolved different, but ultimately unsuccessful, adaptations usually go undetected, causing one to overestimate the extent of convergence (Woods et al. 2006; Lenski 2017). Second, comparative studies generally cannot account for s light differences in environments or lineages' prior evolutionary histories as causes of divergent adaptation, leading to an underestimation of converg ence (Woods et al. 2006; Achaz et al. 2014; Lenski 2017). Controlled and replicated evolution experiments with microbes allow one to overcome these limitations and more precisely quantify the extents of parallel and convergent evolution (Blount et al. 2018 ). Populations of bacteria, yeast, and viruses reproduce quickly and grow to large numbers, permitting on e to observe evolution over timescales of days to years. One can also propagate many independent lines under identical conditions and characterize evol utionary repeatability at both the phenotypic and genotypic levels. Evolution experiments with bacteria c an also shed new light on the growing crisis of antibiotic resistance. It has been estimated that at least 700,000 people die each year from resistant infections, and the mortality rate has been projected to rise to 10 million by 2050, outpacing deaths fro 31 especially given the slowing pace at which new drug classes are being developed. Drug discovery through large - scale chemical screen s or isolation of natural products is one approach to addressing the problem of resistance (Wohlleben et al. 2016). However, bacteria will likely evolve resistance to these new compounds, as they have to previous antibiotics, diminishing the effectiveness of drug discovery over the long - term. Alternatively, it might be possible to extend the usefulness of existing therapeutics by channeling evolution toward drug susceptible states with an improved understanding of the factors that shape evolutionary traject ories (Nichol et al. 2015; Baker et a l. 2017; Hughes and Andersson 2017; Iram et al. 2019) . To that end, the phenotypic and genetic repeatability of resistance evolution has motivated several studies (Toprak et al. 2012; Imamovic and Sommer 2013; Lázár et al. 2013; Baym et al. 2016; Yen and P apin 2017; Card et al. 2019; Maltas and Wood 2019; Nichol et al. 2019). In particular, two landmark studies evaluated the reproducibility of Escherichia coli populations evolving in and adapting to increasing antibiotic concentrations in spatially homogene ous (Toprak et al. 2012) and structured environments (Baym et al. 2016). They found strong signatures of genomic parallelism; that is, replicate lines tended to evolve high - level resistance through mutations in a limite d set of genes. However, the replicat e lines in these experiments were all founded from genetically identical cells, so it is unknown whether selection would target the same genes if the founding genetic background had been different. Epistatic interaction s between mutations (including betwee n resistance mutations and the genetic background in which they occur) can alter adaptive trajectories. Thus, they complicate our ability to predict resistance evolution and design effective treatment strategies. In the ory, one can leverage collateral drug responses where the evolution of resistance to one drug increases a 32 al. 2014; Chevereau et al. 2015; Yen and Papin 2017; Ima movic et al. 2018) to forestall or ev en reverse antibiotic resistance in pathogens. Yet differences in genetic background can make this approach difficult in practice. For example, replicate E. coli and Enterococcus faecalis populations can take different mutational paths to increased resista nce that change their collateral responses to second - line antibiotics (Maltas and Wood 2019; Nichol et al. 2019). Despite stochasticity at the level of individual mutant strains, one can still exploit statistical patter ns in resistance profiles across many replicates to optimize drug - cycling protocols (Maltas and Wood 2019). Recent work showed that such an approach could drive populations to eventual long - term susceptibility through intermediate states of high - level resi stance (Maltas and Wood 2019). Replic ate populations also accumulate genetic differences as they evolve in permissive environments, and these differences can change their ability to adapt when challenged with antibiotics. We recently used several strains from the E. coli long - term evolu tion e xperiment (LTEE) to examine the role that genetic background plays in the evolution of drug resistance (Card et al. 2019). In the LTEE, 12 replicate populations were founded from a common ancestral strain and have been propagated daily for over 30 ye ars in an environment without antibiotics (Lenski et al. 1991; Tenaillon et al. 2016). Clones from several populations evolved increased antibiotic sensitivity compared to their ancestor (Card et al. 2019; Lamrabet et al. 2019). We asked whether these stra ins co uld compensate for their increased sensitivity through subsequent evolution under drug selection. W e found that their evolutionary potential was idiosyncratic with respect to their initial genotype, such that resistance was constrained in some backgr ounds but not in others, indicating the role of historical contingency in this process (Blount et al. 2018; Card et al. 2019). In this study, we sequenced the complete genomes of antibiotic - resistant clones that evolved from several different founding stra ins du ring our earlier experiments and used this 33 information to examine how genetic background affects the genomic evolution of antibiotic resistance. First, we validated our statistical approach by demonstrating that mutations in particular target genes w ere as sociated with each of the four antibiotic treatments. We then showed that evolution was also contingent, albeit more subtly, on differences in genetic background, such that resistant lines that evolved from the same genotype had, on average, slightly more mutations in common. These results, taken together with our previous work, indicate that even slight differences of antibiotic resistance evolution. Mate rials and Methods Evolution experiments and bacterial strains The LTEE has been described in detail elsewhere (Lenski et al. 1991; Lenski 2017). Briefly, 12 replicate populations of E. coli were founded from a common ancestral strain called REL606. These p opulat ions have been propagated for more than 73,000 bacterial generations by daily 1:100 transfers in a glucose - limited Davis minimal (DM) medium without antibiotics. In a previous study (Card et al. 2019), we measured the intrinsic resistance of the LTEE ances tor and derived clones isolated from four populations (designated Ara 5, Ara 6, Ara+4, and Ara+5) at generation 50,000 to the antibiotics ampicillin (AMP), ceftriaxone (CRO), ciprofloxacin (CIP), and tetracycline (TET). We also quantified these strai pacities for evolving resistance by challenging them across a range of concentrations to these same drugs during one round of selection. In this study, we sequenced the complete genomes of a subset of the resistant mutants that evolved during these e xperim ents, and we examined whether the genetic targets of the resistance mutations systematically differed between the four antibiotics and five genetic 34 backgrounds. Specifically, for each antibiotic treatment we sequenced 4 mutants that independently evo lved f rom the ancestral background, and 3 mutants from each derived background, for a grand total of 64 sequenced mutants (16 mutants × 4 antibiotics) ( Appendix A Table 4 ). Library preparation and genomic sequencing Glycerol stocks of frozen samples were grown overnight in 3 mL of Luria Bertani (LB) medium at 37 o C with shaking at 250 rpm. Overnight cultures were harvested by centrifugation and genomic DNA was extracted using the E.Z.N.A. Bacterial DNA kit (Omega Bio - tek). DNA was quantified using the Quant iFluor dsDNA system (Promega). Then, 250 ng of purified genomic DNA was used from each sample for library preparation using the Nextera DNA Flex Library Prep Kit - cycle V3 MiSeq reagent cartridge and sequenced on an Illumina MiSeq at North Carolina A&T State University. The resultant FASTQ files of 300 - base paired - end reads were deposited to the NCBI Sequence Read Archive. Mutation identification We trimmed sequenc ing re ads to remove low - quality bases using Trimmomatic v0.38 (Bolger et al. 2014) on the Galaxy web platform (Afgan et al. 2018). A sliding - window approach was used; reads were clipped when the average quality score was < 20 in a 4 - bp window and to a mini mum le ngth of 36 - bp. Mutations were then identified in the genomes using breseq v0.35 (Barrick et al. 2014) with default parameters. We used a version of the REL606 reference genome with updated gene annotations for variant calling (Jeong et al. 2009; Tena illon et al. 2016). 35 Each population evolved unique substitutions in pykF during the LTEE that distinguish them from one another (Tenaillon et al. 2016). Therefore, we first compared this locus for each resistant clone against its corresponding parental strain to test for possible external and cross - contamination. Strains KJC184 and KJC217 from the CIP and CRO treatments, respectively, were s upposed to derive from the Ara+4 and Ara+5 parental backgrounds, respectively, but they had pykF alleles corresponding to ot her backgrounds used in this study. Likewise, strain KJC152 from the CIP treatment was supposed to derive from the Ara+4 background, but its genome was identical to a resistant mutant derived from the ancestral clone. We discarded these three cross - contami nants from our study. The breseq results for each of the other 61 sequenced resistant clones gave information on both its genetic b ackground and the mutations that evolved during our previous antibiotic - selection experiments. We manually curated the resul ts by removing the background - specific mutations (i.e., those that arose during the LTEE), which we did by comparing each resistant clone to its parental strain. We also excluded expansions and contractions of hypermutable short sequence repeats that are u nlikely to contribute to stably inherited resistance, and mutations within multi - copy elements (e.g., ribosomal RNA operons and inse rtion sequences) that may result from gene conversions but cannot be fully resolved using the short - read sequencing data. In addition, we resolved numerous structural variants by manually examining the depth of read coverage across the genome and predictio ns of new sequence junctions from split - read mapping for each clone (Barrick et al. 2014). To verify the predicted mutations , we applied the genomic changes in each parental background to the REL606 reference genome and reran breseq . In total, we identifie d mutations in 59 of the 61 antibiotic - resistant clones. Two clones (KJC65 and KJC66) had no clear genetic changes despite e xhibiting increased phenotypic 36 resistance in our earlier study (Card et al. 2019). This discrepancy suggests at least two possible e xplanations. First, these resistant lines might have mutations that could not be adequately resolved by short - read sequencin g, including amplifications of genes or chromosomal regions and inversions bounded by identical sequences (e.g., multi - copy IS eleme nts) (Raeside et al. 2014). Second, these lines might have had unstable genetic changes, including copy number changes in ho mopolymeric tracts and amplifications, which are often unstable and can lead to hypermutability, phase variation, and other complica tions (Moxon et al. 1994; Sandegren and Andersson 2009; Blount et al. 2012; Jiang et al. 2019). To look for amplifications t hat might have been missed by the breseq pipeline, we also used another pipeline (Blount et al. 2020) to examine the two genomes wit hout identifiable mutations for evidence of regions with above - average read coverage. However, this analysis did not reveal any amplifications in these two clones. More details can be found in the R Notebook provided on GitHub ( https://github.com/KyleCard/ LTEE - ABR - mutant - sequencing ). Statistical methods We quantified the specificity of genomic evolution with respect to antibio tic treatment by comparing the gene - level similarity of mutations between independent resistant lines that evolved under the same tr eatment versus different treatments. Similarly, for each antibiotic we quantified the specificity of genomic evolution with respect to genetic background by comparing the gene - level similarity of mutations between lines derived from the same parental genot ype versus different parental genotypes. Following Deatherage et al. (2017), we included in our comparisons nonsynonymous po int mutations, small indels, and IS element insertions in genes or within 150 - bp upstream of the start of a gene. However, we modifi ed their approach to also include large 37 deletions and amplifications if at least one of the affected genes was also found to be mutated in another clone or if there were parallel changes across lines. We excluded from these analyses synonymous mutations, t he two clones with no identified genetic changes, and a third clone with only a large amplification that was unique and coul d not be assigned to any particular gene. A total of 71 mutations qualified based on these criteria. We then calculated Dice's coefficient of similarity, S , for each pair of evolved clones, where . Here, |X| and |Y| represent the number of genes with qualifying mutations in each clone, and is the number of mutated genes in com mon between them. S therefore ranges from 0, when the pair of clones have no mutated genes in common, to 1, when both have mutations in exactly the same set of genes (Sokal and Rohlf 1994; Deatherage et al. 2017; Blount et al. 2020). We then calculated the average of these coefficients for all pairs of clones evolved within the same treatment or f rom the same parental genotype, S s , and for all pairs of clones evolved across different treatments or different genotypes, S d . The difference between these two va lues serves as a test statistic for the specificity of genomic evolution. The observed outcom e can be seen as one of many possible but equally likely outcomes that could have arisen by chance. One can therefore perform a randomization test to evaluate the significance of the test statistic associated with the observed outcome (Sokal and Rohlf 1994 ). To do so, we repeatedly rearranged the clones associated with each antibiotic treatment, or the clones within each treatment when testing for background specifi city, while maintaining the number and identity of the mutations in any clone (Deatherage et al. 2017). For example, if mutations A and B were found together in the same sequenced clone, we retained their association throughout the procedure but randomly a ssigned the set to a different clone label. We calculated the specificity test statistic for each of 10,000 permutations of the clone labels. This procedure yields the expected 38 distribution of the test statistic under the null hypothesis that the similarit y among lines is independent of the antibiotic treatment or founding genotype. We then calcul ated an approximate p - value for rejecting this null hypothesis from the proportion of permutations in the expected distribution with a specificity statistic value greater than or equal to the observed value. To quantify the specificity of genomic evolution with respect to genetic background, we performed an independent randomization test for each of the four antibiotics. Because these tests address the same null hyp othesis, we combined the resulting p - 2 k = 8 degrees of free dom, where k is the number of comparisons (Fisher 1934; Sokal and Rohlf 1994). We provide the datasets and details of our statistical analyses in an R Notebook on GitHub ( https://github.com/KyleCard/LTEE - ABR - mutant - sequencing ). Results Genomic evolution o f strains evolved under four different antibiotic treatments In our previous study (Card et al. 2019), we isolated antibiotic - resistant mutants that evolved from f ive different parental genotypes: the LTEE ancestor and four derived clones isolated from the LTEE at generation 50,000. The experiment was performed over one round of drug selection. Here, we sequenced the complete genomes of 64 of the resistant clones, b ut we discarded 3 that were identified as cross - contaminants (Materials and Methods) ( Appendi x A Table 4 ). The 61 remaining resistant clones had a total of 76 mutations. Forty - five genomes had a single mutation, 11 others had two mutations, and 3 genomes had three mutations (Figure 2. 1). Two other clones (both in the tetracycline treatment) had no identifiable mutations; they might have had unstable genetic changes or types of mutations that could not be resolved by our analyses of the short - read sequencing data (see Materials and Methods). In any case, we have excluded 39 these two clones from the an alyses that fo llow. Twenty - seven of the 76 mutations (35.5%) were single - base substitutions; of these, 22 were either nonsynonymous or nonsense mutations that - acid sequence, 1 was synonymous, and 4 occurred in alaT , whic h encodes a tR NA rather than a protein. Five other mutations (6.6%) were in intergenic regions within 150 - bp upstream of a gene, which suggests that they affect regulation. 40 Figure 2.1 . Numbers and types of mutations in evolved genomes. Summary of the 76 mutations obse rved in 61 antibiotic - resistant clones after selection in ampicillin, ceftriaxone, ciprofloxacin, or tetracycline. Mutations are color - coded by the type of genetic change, according to the legend mutations in the non - protein - coding gene alaT . Evolved genomes are labeled according to their parental genetic background and replicate. Two tetracycline - selected clones (Ara 5 - 1 and Ara+5 - 1) had no identifiable mutations (see Materials and Methods). 41 The largest propor tion of mutations in the resistant lines were structural variants. They comprise 44 (57.9%) of the observed changes (Figure 2. 1). Eleven of these were IS - element insertions in protein - coding genes; 8 were small insertions and deletions (indel s) of less tha n 50 bp, 13 were large deletions, and 12 were large amplifications. Twelve of the 13 large deletions and 11 of the 12 large amplifications were found in lines derived from the generation - 50,000 backgrounds (Figure 2. 1). However, 45 of the 61 clones (73.77% ) belong to that group, and neither observed distribution deviates significantly from that null expectation (binomial tests, p = 0.1077 and p = 0.1368 for large deletions and large amplifications, respectively). Genomic parallelism at the fu nctional level Antibiotic resistance can arise through mutations that change gene regulation and expression, cell permeability and efflux, and metabolism (Blair et al. 2015). To determine how drug selection acted on these functions in our experiment, we qu antified the e xtent of genomic parallelism in the resistant lines at the functional level i.e., sets of genes that share broadly defined functions. We used the curated descriptions of cellular processes in EcoCyc (Keseler et al. 2017) to match each mutat ed gene to an associated function. We excluded large deletions and amplifications when the affected genes do not share a common function. About 37% of the 57 mutations that fit the criteria for inclusion occurred in regulatory genes, ~26% in metabolic gene s, ~21% in gen es that encode transporters, and ~11% in genes involved in transcription or translation (Figure 2. 2). More mutations in some of these functional categories than in others suggest a pattern of parallel evolution. However, more E. coli genes ar e involved in some functions than in others, and therefore a random mutation is more likely to occur in those categories that constitute a larger proportion of the genome. To examine whether the 42 observed number of mutations in each category occurred more f requently than expected, we modeled the data using the Poisson cumulative expectation , where is (total number of mutations) [(combined length of all genes in functional category) / (total genome length)] (Tenaillon et al. 2012). The deriv ed parental backgrounds evolved slightly smaller genomes during the LTEE (Tenaillon et al. 2016), so we used their average genome size (4,602,572 bp) when calculating . 43 Figure 2.2 . Distribution of independent mutations in different functional categories. The observed and expected distributions are shown as shaded regions and outlines, respectively. See text for statistical analysis. Regulatory genes accrued mutations about 5 times m ore often than expected from the Poisson distribution (Figure 2. 2), and this difference is highly significant ( p < 0.0001). Genes involved in t ransport functions had about 1.6 times more mutations than expected by chance, but this difference was marginally non - significant ( p = 0.0723). Genes involved in transcription or translation had about 1.8 times as many mutations as expected, but this diffe rence was also not statistically significant given the small number of mutations in these targets ( p = 0.1245). I t should be emphasized that this analysis is conservative because it lumps together all of the genes in each functional category. However, muta tions in only a subset of these genes are likely to cause resistance. Therefore, the effective mutational target size and the resulting expected number of mutations is presumably much smaller. 44 Specificity of genomic evolution in the different antibiotic e nvironment s We compared the gene - level similarity of mutations between independent lines that evolved in the same antibiotic treatment and across the four different treatments to evaluate the effect of the selective environment on the genetic paths to incr eased antibiotic resistance. As described in the ity for each pair of clones using the 71 qualifying mutations that can be assigned to a particular gene. The average within - treatment similarit y was 0.089 and the average between - treatment similarity was 0.032 (Figure 2. 3). In other words, two clones that independently evolved under the same antibiotic selection had on average 8.9% of their mutated genes in common, whereas those that evolved unde r different antibiotics shared on average only 3.2% of their mutated genes. A randomization test shows that the 5 .7% difference in similarity is highly significant ( p < 0.0001). Thus, genomic evolution was demonstrably specific with respect to the antibiot ic treatment. 45 Figure 2.3 . Specificity of genomic evolution with respect to antibiotic treatment. Treatments a similarity for all clone pairs evolved in the sam e antibiotic ( S s ) and in different antibiotics ( S d ), respectively. Only the 71 qualifying mutations (see Material s and Methods) were included in the calculations. The weighted averages of S s and S d are shown in the grey box. The difference between these tw o values indicates the extent to which genome evolution was specific to the antibiotic treatment. The resulting p - value was calculated using a randomization test. T he similarity analysis does not r eveal the specific genes that contribute to the antibiotic - an excess of qualifying mutations in the replicate lines evolved under the four treatments (Fig ure 2. ontributed significantly to antibiotic specificity (Figure 2. 4 , Table 2.1 ). The alaT gene encodes a tRNA; it was mutated in 4 of the 14 CIP - resistant lines, but in none of the other 44 lines with q ualifying mutations (Figure 2. 4). The ompR gene is part of the two - component system that regulates the production of outer - membrane 46 proteins; it was mutated in 6/14 TET - resistant lines as well as in 4/44 lines that evolved resistance to other drugs. The ot her gene in this two - component system, envZ , was mutated in 2 of the 10 TET - resistant lines that did not have an ompR mutation. Two genes, ompF and hns , were associated with resistance to ceftriaxone (Table 2. 1). The former encodes an outer - membrane porin and was mutated in 6/15 CRO - resistant lines along with 3/43 other lines; the latter encodes a histone - like global regulator and acquired mutations in 3/15 CRO - resistant lines and 1 of the 43 lines that became resistant to another antibiotic (Figure 2. 4). F inally, a large deletion was found in 3 of the 15 AMP - resis tant lines but not in any of the other 43 lines (Table 2. 1); this deletion affects multiple genes including phoE , which encodes an outer membrane porin (Figure 2. 4). 47 Figure 2.4 . Identity of mut ated genes in antibiotic - resistant lines. A total of 61 lin es (labels at left) evolved from 5 different genetic backgrounds in ampicillin, ceftriaxone, ciprofloxacin, or tetracycline environments. Two (TET Ara 5 - 1, TET Ara+5 - 1) had no identifiable mutation s; a third (AMP Ara+4 - 2) had no qualifying mutation that co uld be assigned to a specific gene (see Materials and Methods). These three lines are not shown. Filled cells identify the 71 qualifying 48 Fig ure 2.4 . (con t d) mutations by the affected genes (shown along the bottom and listed in order of the number of mutations). The darkly sha ded cells identify the signature genes, in which mutations are significantly associated with one antibiotic treatment (Table 2.1). A deletion or amplification spanning a given genomic region is ind icated when two gene names are shown. If a gene name is sho those genes are deleted; otherwise, they are amplified. Part of the ompF gene is deleted in the CRO Ara 5 - 1 line. T able 2.1 . Gene targets that contributed to ant ibiotic - treatment specificity. Antibiotic Gene target p Ampicillin [ECB_00212] [phoE] 0.0147 Ceftriaxone ompF 0.0064 hns 0.0493 Ciprofloxacin alaT 0.0024 Tetracycline ompR 0.0087 the association between antibiotic treatment and genetic targets. Specificity of genomic evolution with respect to genetic background evolution with respect t as an example (Figure 2. 5A), the clones that evolved independently from the same founding genetic background and from different backgrounds had, on average, 14.8% and 3.1% of their mutated genes in common , respectively, indicating a difference of 11.7%. This trend of greater similarity for clones derived fro m the same genetic background also occurred in the three other antibiotics (Figure 2. 5B D). 49 Figure 2.5 . Specificity of genom ic evolution with respect to genetic background. Five different s coefficient scores that show the average similarity for all clone pairs evolved from the same genetic background ( S s ) and from different backgrounds ( S d ), respectively, in ampicillin (A), ceftriaxone (B), ciprofloxacin (C), and tetracycline (D). The diff erence between S s and S d indicates the extent to which genome evolution was specific to the genetic background. Two of the three replicates derived from the Ara+4 background in ciprofloxacin were excluded owing to cross - contamination, and S s cannot be calc ulated in that case (*). See Figure 2.3 for additional details. We performed separate randomization tests for the clones in each antibiotic treatment to evaluate whether the effects of genetic background were significant. The associations between genomic e volution and the identity of the parental strain were significant for the lines that evolved 50 in the ampicillin and ceftriaxone e nvironments (Figure 2. 5A and 2. 5B), and they were marginally nonsignificant for the lines in the ciprofloxacin and tetracycline environments (Figure 2. 5C and 2. 5D). When we combined the probabilities from these four independent tests of the hypothesis that differences in genetic background influence the genetic basis of antibiotic resistance using l and Rohlf 1994), the overall trend toward greater similarity (gene - level parallelism) of lines evolved from the same founding genotype was highly significant ( 2 = 24.67, df = 8, p = 0.0018). In our analysis of genome specificity with respect to antibiot ic treatment, we identified several signature genes that contributed significantly to that specificity (Figure 2. 4 , Table 2. 1 ). We have much less statistical power to identify particular genes that contribute to specificity with respect to genetic background, because only 3 or 4 replicate lines derive from any given background in each antibiotic treatment. Nonetheless, we can ident ify candidate loci that may contribute to that specificity, whic h might be further studied in the future. Table 2. 2 shows all of those genes that fulfilled both of the following criteria for a given antibiotic treatment: (i) two or more lines derived from the same background had mutations affecting the same gene; and ( ii) that background produced at least as many mutations affecting that gene as did the other four backgrounds combined. In the case of each antibiotic treatment, at least two genetic backgroun ds have candidate signature genes that fulfill these criteria. 51 Table 2.2 . Candidate genes that may contribute to genetic - background specificity. Antibiotic Genetic background Gene target Ampicillin Ancestor ompR Ampicillin Ara 6 yfiH Ampicillin Ara+5 [ECB_00212] [phoE] Ceftriaxone Ara 5 [insA - 26] [fimB] Ceftriaxone Ara 6 [yeaS] [menC] Ceftriaxone Ara+4 hns Ciprofloxacin Ancestor alaT Ciprofloxacin Ara 5 alaT Ciprofloxacin Ara+5 [yedV] [cbl/nac] Tetracycline Ancestor ompR Tetracycline Ara 6 [yfc TU] [yfiH] Discussion How does genetic background affect the evolution of antibiotic resistance? We previously addressed this question by examining the resistance potential of the E. coli ancestor of a long - term experiment and derived clones isolated fro m four populations after generation 50,000. We challenged these strains using a series of drug concentrations, and we found that several strains had a reduced capacity to evolve resistance relative to their ancestor, implicating the role of historical cont ingency in this process (Blount et al. 2008, 2018; Card et al. 2019). In this study, we asked whether and how genetic background influences the evolution of resistance through chromosomal mutations. We sequenced the complete genomes of 61 resistant lineage s that evolved in our ea rlier experiment to identify the mutations that conferred resistance. We then determined if there were particular signatures of (i) the antibiotic treatment and (ii) the initial background evident from the identities of the mutated genes. The populations t hat evolved resistance to the four different drugs in our study exhibited divergent underlying genetic changes (Figure 2. 3). This result was expected given that bacteria generally evolve resistance through mutations specific to a dr (Toprak et 52 al. 2012; Chevereau et al. 2015; Baym et al. 2016). This specificity was driven by parallel mutations in several genes (Table 2. 1). Overall, ompR and ompF had more mutations than any other genes (Figure 2. 4). OmpR is a t ranscriptional regulator involved in responses to osmotic and acid stress; mutations to OmpR also contribute to antibiotic resistance by altering the expression of the OmpF major porin (Aiba and Mizuno 1990; Chakraborty and Kenney 2018; Choi and Lee 2019). A recent study showed t hat ompF - lactams (e.g., ampicillin and ceftriaxone) across the outer membrane, thus increasing resistance (Choi and Lee 2019). We found that ompF mutations were strongly associated with ceftria xone - resistant lines, co nsistent with this prior study. However, the evolution of ampicillin resistance occurred through more diverse mutational paths in our experiment. Although there were some mutations in ompF and ompR in the ampicillin treatment, we sa w a significant association of that treatment with deletion of a different outer - membrane porin, PhoE. The down - regulation of this porin also In the tetracycline treatment, muta tions were more common in ompR than in ompF , which suggests that altering the expression of other genes in this regulon also contributes to resistance. Mutations in hns were associated with ceftriaxone - resistant lines. Ni shino and Yamaguchi (2004) showed t hat deletion of this global transcriptional regulator increases resistance to multiple drugs because it causes overexpression of several efflux pumps. In contrast to the results for the other three antibiotics, the evolut ion of ciprofloxacin resistance was not associated with mutations in genes related to outer - membrane proteins. Instead, mutations in alaT , which encodes an alanine tRNA, were a signature of this treatment. The mechanism behind this resistance is unknown. O ne possibility is that these mutati ons modulate interactions that have been reported between this tRNA and tmRNA, which rescues stalled ribosomes after aberrant translation (Gillet and Felden 2001). Enhanced rescue 53 might directly promote survival or indire ctly affect the expression of other vital genes, when cells are treated with this antibiotic. The signature genes that we observed for each treatment are not the canonical resistance genes for the respective antibiotics (Blair et al. 2015). Ampicillin and ceftriaxone irreversibly bind to tr anspeptidases and disrupt cell - wall synthesis; ciprofloxacin targets topoisomerase and inhibits DNA replication; and tetracycline targets the ribosome and hinders protein synthesis. Drug resistance often arises through mo difications to these targets, yet t hese changes rarely occur in our study. This discrepancy may reflect two factors: one environmental and the other genetic. First, altering a drug target often confers high - level resistance, but at the expense of bacterial growth rate (Andersson and Hughes 2010; Hughes and Andersson 2017). We used moderate drug concentrations to select for mutants (Card et al. 2019), and the observed resistance rarely reached levels thought to be clinically relevant (EUCAST 2020). This mode rate environment should favor mutat ions that provide sufficient resistance at a low fitness cost, because they will leave more descendants during population growth before treatment, and consequently they will be seen more often after the antibiotic challen ge. Second, the E. coli used in our experiments are all derived from a B strain that differs in important ways from the K - 12 strains that are more widely used in studies of antibiotic resistance (Studier et al. 2009). In particular, E. coli K - 12 has two ma jor porins, OmpC and OmpF, whereas E. coli B expresses only OmpF (Pugsley and Rosenbusch 1983; Schneider et al. 2002). Thus, the use of the E. coli B background may well have influenced which genes could mutate to yield resistance in our experiments. We al so found genomic signatures of adap tive divergence associated with differences in genetic background, and these differences are far smaller than those between E. coli B and K - 12. We sequenced and analyzed resistant lines that evolved from five genetic back grounds that were 54 separated in time by only a few decades, and which differed only in the mutations that had accumulated in the antibiotic - free environment of the LTEE (Figure 2. 1). Three or four resistant lines independently evolved from each parental bac kground for each of the four antibi otics studied, allowing us to assess the genomic specificity of resistance with respect to the genetic background (Figure 2. 5). Although these background effects were more subtle than those showing antibiotic specificity, they are compelling when taken tog ether. Various factors might contribute to the genetic background specificity. Most broadly, epistatic interactions can cause the same mutation to have different effects on resistance, or on its fitness costs, in differen t genetic backgrounds. The rates at which particular resistance mutations arise may also vary between different genetic backgrounds. Imagine that the same mutation arises in separate populations founded from two distinct backgrounds. If the mutation confer s less resistance in one background than the other, then it may go undetected when those populations are challenged at a high drug concentration. This type of epistasis could therefore generate a signature of genomic specificity of resistance mutations wit h respect to the genetic background . It is also possible that different genetic backgrounds affect the evolution of resistance by changing the likelihood of certain genomic amplifications or deletions. These types of structural mutations often occur by hom ologous recombination between IS el ements, and they can confer resistance by altering the number of membrane transporters or drug targets (Sandegren and Andersson 2009). Such mutations can also occur spontaneously at very high rates (Cooper et al. 2001; Sa ndegren and Andersson 2009). In our study, many of the resistance mutations were mediated by insertion sequences, including new copies in the derived backgrounds that previously arose during the LTEE (Tenaillon et al. 2016). The evolution of 55 resistance in these cases is therefore influenced , at least in part, by changes in the rates at which certain types of mutations arise in the derived backgrounds. Antibiotic resistance is a growing public - health concern. If the likely routes to resistance can be predict ed, at least to some extent, then t here exists a potential opportunity to control the emergence of resistance through rational treatment strategies. However, to predict the evolution of resistance with accuracy, we must understand and integrate information about many factors, including a ba future evolution. The results from this study, together with our previous findings, demonstrate the importance of historical continge ncy in the evolution of drug resist ance at both the phenotypic and genotypic levels. This contingency underscores the importance of accounting for stochasticity, in the past as well as at present, when designing evolutionarily informed treatment strategies . 56 CHAPTER 3: Idiosyncratic variat ion in the fitness costs of tetracycline - resistance mutations in Escherichia coli 57 Abstract plays a key role in its ecological success and clinical impact. In this study, we examine whether tetracycline - resistant mutants are less fit in the absence of the drug than their sensitive parents, and whether the fitness cost of resistance is constant or variable across independently derived lines. Tetracycline - resi stant lines suffered, on average, a reduction in fitness of almost 8%. There was substantial among - line variation in the fitness cost. This variation was not associated with the level of phenotypic resistance conferred by the mutations, nor did it vary sig nificantly across several different genetic backgrounds. The two resistant lines with the mos t extreme fitness costs involved functionally unrelated mutations on different genetic backgrounds. However, there was also significant variation in the fitness co sts for mutations affecting the same pathway and even different alleles of the same gene. Our findings demonstrate that the fitness costs of antibiotic resistance do not always correlate with the phenotypic level of resistance or the underlying genetic cha nges. Instead, these costs reflect the idiosyncratic effects of particular resistance mutatio ns and the genetic backgrounds in which they occur. 58 Introduction Antibiotics are an essential component of modern medicine. Although they have dramatically reduc ed the morbidity and mortality caused by severe bacterial infections, their benefits have dim inished in recent years because of their overuse in the clinic and in agriculture, which has led to the evolution and proliferation of antibiotic - resistant pathoge ns. As a result, many infections have become more difficult to treat with mainline drug thera pies, and in some severe cases, some pathogenic strains have become resistant to all available drugs. An understanding of the forces underlying and shaping antibio tic resistance is therefore critical to the future health of the human population. Bacteria c an evolve resistance by either spontaneous mutations or horizontal acquisition of resistance genes. Spontaneous mutations commonly confer resistance by altering th e cellular target of the antibiotic or increasing its efflux (Blair et al. 2015). Mechanisms associated with horizontal gene transfer include target modification, drug detoxification, and the acquisition of novel efflux pumps (Blair et al. 2015). In either case, resistant variants have a clear advantage over their sensitive counterparts when expos ed to the corresponding antibiotic. However, these resistant types often suffer fitness costs because they disrupt the normal functioning of metabolic pathways and physiological processes or increase the energetic burden on the cell (Lenski and Bouma 1987; Nguyen et al. 1989; Andersson and Hughes 2010; Vogwill and MacLean 2015). Resistant types should therefore have lower growth rates than, and be outcompeted by, th eir sensitive counterparts in the absence of drugs. ss, both in the presence and absence of a drug, is an important factor that contributes to its ecological success and thus its clinical impact (Lenski 1997; Vogwil l and MacLean 2015; Hughes and Andersson 2017). For example, the fitness of a 59 resistance muta tion determines its likelihood of persisting in a bacterial population prior to drug exposure, its maintenance in a population at a particular drug concentration, and its reversibility when the antibiotic is reduced or removed from the environment (Hughes and Andersson 2017; Santos - Lopez et al. 2019). The expected time required to reduce the frequency of a resistant mutant in a bacterial population following the ces sation of antibiotic use is inversely proportional to the fitness cost of the resistance muta tion (Lenski 1997). Although mathematical models can predict the rate of these frequency declines (Levin et al. 1997), these theoretical expectations often are not met under real - world scenarios for at least two reasons. First, some resistance mechanisms a re inherently cost free, at least in certain environments. Several mutations in the gene rpsL confer resistance to streptomycin, but they have little or no fitness cost in both Escherichia coli and Salmonella typhimurium (Tubulekas and Hughes 1993), and they even confer a competitive advantage over wild - type strains in some animal infection models (Björkman et al. 1998; Enne et al. 2005). These cost - free rpsL mutati ons are also found in streptomycin - resistant Mycobacterium tuberculosis populations, where they may facilitate the long - term maintenance of this resistant type (Böttger et al. 1998; Ander sson and Hughes 2010). Similarly, treatment of Helicobacter pylori in fections with clarithromycin has been found to select for highly resistant commensal Enterococcus species that persist for years after drug treatment (Sjölund et al. 2003). This last outc ome demonstrates a troubling side - effect of antibiotic use, in which the microbiome can act as both a reservoir for resistance genes and as a conduit for their horizontal transfer to pathogens (Sommer et al. 2010). Second, pleiotropic costs associated wit h chromosomal - or plasmid - mediated resistance can often be reduced or even eliminated through subsequent compensatory evolution (Bouma and Lenski 1988; Schrag et al. 1997; Kugelberg et al. 2005; Nilsson et al. 2006; Andersson and Hughes 60 2010; Barrick et al . 2010). For example, clinically relevant levels of fluoroquinolone r esistance occur through the sequential substitution of mutations in several genes (Lindgren et al. 2003). Early genetic changes in the mutational pathway exact a cost on bacterial growth in both laboratory media and mouse models, but the cost can be amelio rated through later resistance mutations in the absence of drug selection while simulta neously preserving resistance in the event of future exposure to anti biotics. Moreover, compensatory evolution can sometimes drive multidrug resistance; this outcome has been seen when a genetic change simultaneously provides resistance to a newly imposed drug while reducing the fitness cost associated with resistance to a previous antibiotic (Trindade et al. 2009). Compensatory evolution shows how pleiotropic effects of one mutation can set the stage for epistatic interactions with subsequent mutations. In costs of antibiotic resistance. For example, Vogwill and colleagues (2016) examined the costs of rifampicin - resistance mutations in the gene rpoB across several Pseudomonas species. They found that some mutations vary in their fitness effects across backgr ounds, and these costs correlate with transcriptional efficiency. Thus, the same rpoB mutation can differentially affect transcriptional efficiency depending on the genetic background, an d these idiosyncratic effects in turn lead to heterogeneity in costs. This work evaluated genetic - background effects across a fairly broad phylogenetic scale, while focusing on mutations in a single gene. One can also ask whether genetic background affects the fitness cost of resistance even among recently diverged clones o f a single species, and for resistance that has evolved through more diverse mutational pathways. To address these issues, we evaluated the competitive fitness in the absence of drugs of tetracycline - resistant clones that evolved from several different E. coli backgrounds, which 61 previously diverged during a long - term evolution experiment (LTEE). We ask several questions. First, is there a fitness cost to resistance? Second, is the cost gre ater for mutants that evolved higher levels of resistance (Figure 3. 1 A)? Third, do fitness costs vary in an idiosyncratic manner that does not depend on the level of resistance achieved (Figure 3. 1B)? Fourth, if there is indeed idiosyncratic variation amon g lines in the cost of resistance, what factors contribute to that va riability? On balance, we found that the resistant lines are indeed less fit than their sensitive counterparts. These fitness costs do not correlate with the level of resistance achieved, nor do they vary among the several genetic backgrounds that we exami ned (Card et al. 2019). Some variation in cost of resistance occurs even among different mutations in the same gene, on the same genetic background, and conferring the same phenotypic res istance. In any case, further research on the fitness effects of anti biotic resistance should be pursued because of its potential implications for public health and patient treatment. 62 Figure 3.1 . Schematic illustration of fitness effects of antibiotic resistance mutations under two scenarios. ( A ) Tradeoff model, in whic h the fitness of a resistant line, when measured in the absence of drugs, is negatively correlated with the level of resistance conferred by its mutations. ( B ) Idiosyncratic model, in whi ch the fitness of resistant lines varies for reasons unrelated to the level of resistance. This idiosyncratic variation might, in principle, reflect differences between genetic backgrounds, mutations in different target genes, different alleles in the same target gene, secondary mutations, and epistatic interactions between the resistance mutations and their genetic backgrounds. The fitness of each resistant line is expressed relative to its sensitive counterpart. A log - transformed relative fitness of 0 ind icates no fitness cost associated with resistance, while values below and above 0 represent fitness costs and benefits, respectively. 63 Materials and Methods Experimental conditions and bacterial strains The LTEE has been described in detail elsewhere (Len ski et al. 1991; Lenski 2017). In brief, 12 replicate populations of E. coli were founded from a common ancestral strain, called REL606 (Daegelen et al. 2009). These populations have been propagated for over 32 years and 73,000 generations by daily 100 - fol d dilutions in Davis Mingioli minimal medium supplemented with 25 mL glucose (DM25). In this study, we examined the competitive fitness of tetracycline - resistant mutants that evolved from the LTEE ancestor and clones sampled from four LTEE populations ( denoted Ara 5, Ara 6, Ara+4, and Ara+5) after 50,000 generations. Specifically, we analyzed 4 mutants that independently evolved from the ancestral background, and 3 mutants that evolved from each derived background, for a total of 16 mutants (Appendix A T able 5) . We also used three clones as common competitors: REL607, REL10948, and REL11638. REL607 is a spontaneous Ara + mutant of REL606, the LTEE ancestor (Lenski et al. 1991). REL10948 is an Ara clone isolated from the Ara 5 population at 40,000 generati ons, and REL11638 is a spontaneous Ara + mutant of that clone (Wiser et al. 2013; Lenski et al. 2015). The Ara marker i s selectively neutral in the glucose - limited medium; it serves to differentiate competitors during fitness assays because the Ara and Ara + cells form red and white colonies, respectively, on tetrazolium - arabinose (TA) agar. We used REL607 as the common co mpetitor for REL606 and the four tetracycline - resistant clones derived from it. The 40,000 - generation clones served as common competitors for the four 50,000 - generation parental clones and twelve resistant mutants that evolved from them; using these common competitors ensured that the differences in fitness were not so large that their densities would fall below the detection limit during th e fitness assays. 64 Fitness assays Assays were performed in the absence of antibiotics to assess the relative fitness of drug - resistant mutants and their susceptible counterparts. Fitness was measured in an environment identical to that of the LTEE, except t cose (DM250). Resistant mutants and their sensitive parents each competed, in paired assays, against the same common competitor with the opposite Ara - marker state (Figure 3. 2). To set up each competition assay, the competit ors were revived from frozen stoc ks, and they were separately acclimated to the culture medium and other conditions over two days. The competitors were then each diluted 1:200 into fresh medium, and a sample was immediately plated on TA agar to assess thei r initial densities based on colo ny counts. The competition cultures were then propagated for 3 days, with 1:100 dilutions each day final densities. We quantifi ed the realized growth rate of ea ch competitor based on its initial and final density and the net dilutions imposed (Lenski et al. 1991). We then calculated relative fitness as the ratio of the realized growth rate of the clone of interest (either a resist ant clone or its sensitive parent ) to that of the common competitor. Lastly, the fitness of a resistant mutant in each assay was normalized by dividing it by the relative fitness of the paired assay obtained for its parental strain. We performed a total of 80 pairs of fitness assays (160 competitions in total) to produce 5 replicate estimates of the fitness of each of the 16 tetracycline - resistant mutants relative to its sensitive parent. The relative fitness values were log e - transformed before the statisti cal analyses reported in the Resu lts below. 65 Figure 3.2 . Schematic illustration showing the derivation of the strains used in this study and the methods employed to measure the fitness of resistant lines relative to their sensitive parents. Twelve initi ally identical E. coli population s were founded from the same ancestral strain to start the LTEE. A genetic marker distinguishes two sets of six populations each. These populations have evolved for >73,000 generations with daily transfers in a minimal gluc ose medium. In paired assays, we examined the fitness of tetracycline - resistant mutants (shown in red) that evolved either from the LTEE ancestor or one of four clones sampled at generation 50,000 by competing them against marked susceptible competitors (s hown in yellow). We used REL607 a s the common competitor for the LTEE ancestor and resistant lines evolved from it, and two 40,000 - generation clones (see Materials and Methods) as common competitors for the derived parental strains and their evolved resist ant lines. After acclimation to t he culture conditions, competitors were mixed at an equal volumetric ratio in a common medium. These cultures were propagated for three days in the absence of tetracycline by serial 1:100 transfers. We quantified each compe the initial and final densities after plating on TA agar, taking into account the net dilution over the three days. These realized growth rates were then used to calculate the fitness of a resistant line relative to its se nsitive parent (see Materials and Methods). 66 Results Tetracycline - resistant lines have reduced fitness in the absence of the antibiotic We ask first whether tetracycline resistance is costly, on average, in the absence of the drug. The grand mean of the l og e - transformed fitness of the 16 resistant lines relative to their paired parental strains is - 0.0771, indicating that the resistant mutants grow ~7.7% more slowly than their sensitive counterparts during head - to - head competitions with a common competitor . This value differs significantl y from the null hypothesis that the resistant lines and their sensitive parents are equally fit ( t s = 2.9973, df = 15, one - tailed p = 0.0045). Cost of resistance varies among resistant mutants We measured the relative fitn ess of each resistant line with 5 - fold replication. This replication allows us to test whether the variation in fitness among the 16 tetracycline - resistant lines is simply measurement noise or, alternatively, reflects genetic v ariation in the cost of resis tance. Table 3. 1 shows the analysis of variance (ANOVA). The variation among the 16 lines is about 10 - fold greater than expected from the variation between replicate assays performed on the same line ( F 15,64 = 10.34, p << 0.000 1). Table 3.1 . ANOVA on the l og - transformed fitness estimates of 16 tetracycline - resistant lines, each measured relative to its sensitive parent. Source SS df MS F p Line 0.7948 15 0.0530 10.3384 << 0.0001 Error 0.3280 64 0.0051 Total 1.1228 79 67 There are many possible reaso ns why the cost of resistance might vary including mutations in different genes, different alleles even in the same gene, different genetic backgrounds, epistatic interactions between mutations and genetic backgrounds, and so on. In the sections that follo w, we examine various possibilities. Possible reversions of unstable mutations do not explain the variation in fitness cost We previously sequenced the complete genomes of the 16 resistant lines, and we compared them to their parental strains to identify the mutations responsible for their resistance (Chapter 2). Two lines had no identifiable mutations (Ara 5 - 1 and Ara+5 - 1), even though they had increased phenotypic resistance relative to their respective parent strains (Card et al. 2019). This discrepancy suggested that these two resistant lines may have had unstable genetic changes , which might have reverted prior to the genomic analysis and our fitness assays. Potentially unstable mutations include changes in the copy number of oligonucleotide repeats an d gene amplifications . We repeated the ANOVA, except excluding the two resistant lines without identifiable mutations. The variation in the cost of resistance remains highly significant in the 14 lines with known, stable mutat ions ( F 13,56 = 10.15, p << 0.0 001 ). Level of phenotypic resistance does not explain the variation in fitness cost All of the resistant lines evolved during a single round of exposure to tetracycline. However, they vary in the resulting minimum inhibitory concentration (MIC) that they achieved. They also vary in the magnitude of the increase in their MICs relative to their p arental strains, which also varied in their MICs. It is possible that mutations that provide greater resistance have higher fitness costs (Fig. 3.1A). To test that p ossibility, we examined the correlation between the log - transformed 68 fitness values of the 1 4 resistant lines and their log - transformed MICs, as previously reported (Card et al. 2019). However, the correlation is not significant; in fact, it is weakly posit ive ( r = 0.1682, two - tailed p = 0.5655). We also computed the correlation between the log - t ransformed fitnesses and log - transformed fold - increases in resistance, but again the correlation is weakly positive and not significant ( r = 0.1002, two - tailed p = 0 .7332). Thus, we find no evidence that the variation in the fitness cost of tetracycline re sistance is related to the level of resistance conferred by the underlying mutations. Genetic background does not explain the variation in fitness cost The 14 tetra cycline - resistant mutants with identifiable mutations evolved on five different genetic bac kgrounds. We asked whether the average cost of resistance differed between the backgrounds. In this case, the ANOVA tests whether the variance in the average cost of resistance for mutants derived from different backgrounds is greater than expected given t he variance in the average cost for mutants derived from the same background. This analysis indicates no significant effect of the genetic background on the cost of resistance ( F 4,9 = 0.47, p = 0.7570 ). Idiosyncratic differences between mutant lines in th e cost of resistance Neither the level of phenotypic resistance conferred by mutations nor the genetic background in which they arose explains the substantial variat ion in the fitness effects of tetracycline resistance. Instead, it appears there are idios yncratic differences in the fitness costs associated with different resistance mutations (Fig. 3. 1B). These idiosyncratic effects could, in principle, reflect mutati ons in different genes, different mutations in the same target gene, secondary mutations th at might have hitchhiked with the mutations conferring resistance, or epistatic interactions between any of 69 these new mutations and the existing mutations that disti nguished the different parental strains. Without a much larger number of resistant lines, i t is not possible to rigorously disentangle these various sources of idiosyncratic fitness costs. However, by examining and contrasting specific cases, we are able t o shed light on some of the sources of these differences. Two resistant clones, Ara+4 - 3 and Ara+5 - 2, have fitness costs that are very similar to one another, but more than double the cost of any of the other 12 resistant mutants (Figure 3 .3 ). Yet these two cases occurred on different genetic backgrounds and have different mutations. A ra+4 - 3 has mutations in hns , which encodes a histone - like global regulator, and lpcA , which encodes a phosphoheptose isomerase; Ara+5 - 2 has a single mutation in ompF , which enc odes an outer - membrane porin ( Chapter 2 ). We asked whether these two extreme cases are solely responsible for the heterogeneity in fitness costs by performing an ANOVA that excludes them. The variation in fitness costs among the other 12 clones is reduced, but it nonetheless remains highly sign ificant ( F 11,48 = 4.44, p = 0.0001). 70 Figure 3.3 . Fitnesses of 14 tetracycline - resistant mutants relative to their parental strains. The mutants are arranged from lowest to highest fitness. Each symbol shows the me an log e - transformed fitness, based on 5 - fold replication of paired fitness assays. Error bars show 95% confidence limits calculated using the t - distribution with 4 df and the pooled standard deviation estimated from the ANOVA (Table 1). Letters above the e rror bars identify mutants with relativ e test for multiple comparisons. Nine of the 14 resistant clones have a single mutation each, while four of them (Ara 5 - 2, Ara 6 - 2, Ara+4 - 3, Ara+5 - 3) have tw o mutations, and another (Ancestor - 2) has three mutations ( Chapter 2 ). It is reasonable to imagine that in each clone one mutation confers the drug resistance, while the others merely hitchhiked with the resistance mu tation. Such hitchhikers might include deleterious mutations that reduce fitness. Therefore, we compared the fitness costs for the resistant clones with and without secondary mutations. The average fitness cost of the clones with multiple mutations is high er (13.8%) than the average of those wi th single mutations (5.5%), but the 71 difference is only marginally nonsignificant given the small number of clones in each group and t - test, t s = 1.4751, 9.3 df, one - tailed p = 0.0866). It is also interesting to compare the four resistant clones derived from the ancestral LTEE background. All four resistant clones evolved the same level of phenotypic resistance, with MICs that are 4 - fold higher than their parental strain (Ca rd et al. 2019). Moreover, all four hav e mutations affecting the same two - component system that regulates the synthesis of outer - membrane proteins: one clone (Ancestor - 1) has a 11 - bp deletion in envZ , which encodes the sensory histidine kinase; the others (Ancestor - 2, Ancestor - 3, Ancestor - 4) have nonsynonymous mutations in ompR , which encodes the DNA - binding response regulator. Even with these striking phenotypic and genetic similarities, an ANOVA sho ws significant heterogeneity in the fitness of these clon es ( F 3,16 = 4.50, p = 0.0180). We can also compare only Ancestor - 3 and Ancestor - 4 (each having a single mutation in ompR and no other mutation), and the variation in fitness remains significant ( F 1,8 = 5.71, p = 0.0439). These results show that different m utations in the same target pathway, and even different alleles in the same gene, can lead to different fitness costs of drug resistance. Discussion In previous work, we examined the role that genet ic background plays in both the phenotypic and genotypic evolution of antibiotic resistance. First, we examined the potential of several different LTEE backgrounds to evolve increased resistance to several antibiotics. We found that evolvability was idiosy ncratic with respect to the parental genotype, such that resistance was more constrained in some backgrounds than in others (Card et al., 2019). Genetic differences will accumulate between populations, even if they evolve in the same permissive environment . These differences can unpredictably alter their ability to respond evolutionarily when challenged with antibiotics. 72 Second, we sequenced the complete genomes of some of these resistant mutants and assessed whether the different initial genotypes took sim ilar or divergent mutational paths to increased resistanc e ( Chapter 2 ). Again, we found that the initial genetic background is important. On average, the replicate lines that evolved from the same founding genotypes had more gene - level mutations in common than lines derived from different founding genotypes. The aim of this study was to examine whether and how the genetic background influences the fitness effects of resistance mutations in the absence of antibiotic. In particular, we examined the fitness co sts of tetracycline resistance in 16 lines that evolved f rom five sensitive parental backgrounds. We found that the resistant lines are, on average, less fit than their sensitive counterparts in the absence of the antibiotic. This result is not surprising, given that resistance mutations often disrupt the normal function of metabolic or physiological processes, or impose energetic demands that reduce growth and competitiveness (Andersson and Hughes 2010). We also observed highly significant variation among the resistant lines in their fitness costs (Table 3. 1). T his variation remained substantial (Figure 3 .1 ) even after we excluded two strains without identified mutations ( Chapter 2 ). These two strains exhibited phenotypic resistance in our earlier work (Car d et al. 2019), but that resistance might have been confe rred by unstable genomic changes, such as gene amplifications or frameshift mutations in homopolymeric tracts that can have reverted prior to the genomic analysis and the comp etition assays that we performed. We then addressed two broad possibilities regarding the variation in fitness cost between the 14 lines with known, stable mutations. First, we asked whether there is a relation between a s fitness cost, such that mutations that confer greater resistance are more costly (Figure 3. 1A). A meta - analysis of fitness costs across several species and drug 73 classes by Melnyk and colleagues (20 15) supported this association, and the authors suggested it could be understood from evolutionary and mechanistic perspectives. Imagine a population that is well - adapted to one environment and hence near a local fitness optimum. If the environment changes , such as with the addition of an antibiotic, then the po pulation may evolve toward a different optimum through the substitution of new mutations. Mutations of large effect will bring the population closer to this new optimum than mutations of small effect . However, if the environment later reverts to its previo us state, then populations that substituted the large - effect mutations will be further from their previous optimum than those populations that acquired small - effect mutations. From a mechanistic stan dpoint, the increased expression of efflux pumps or drug targets diverts resources from other cellular processes. Also, resistance mutations that change evolutionarily conserved proteins are more likely to disrupt their functions than improve them. In our study, however, there was no significant association betw een fitness costs and the level of resistance conferred by mutations, whether on an absolute basis or relative to the parent strain. The second broad possibility is that the fitness costs of resistan ce can vary for reasons unrelated to the level of resista nce conferred (Figure 3. 1B). There are several potential reasons for such idiosyncratic variation. One possibility is that the same resistance mutation may have different fitness costs in different g enetic backgrounds. In Campylobacter jejuni , for example, a C257T mutation in the gene gyrA confers fluroquinolone resistance. When fluroquinolone - resistant and - susceptible strains were inoculated separately into chickens, they colonized equally well and each persisted even in the absence of drug exposure (Luo et al. 2005). However, when resistant and sensitive strains were co - inoculated, the resistant variants often prevailed. Further work indicated that this particular gyrA mutation was beneficial in som e genetic backgrounds, even in the absence of antibiotic, and costly in others (Luo et al. 2005). In our study, by contrast, 74 the variation in fitness costs among strains was not explained by genetic - background effects, but instead involved several other fa ctors. One such factor is that resistance mutations can o ccur in different genes, which can lead to different fitness costs. In this study, the relative fitnesses of clones Ara+4 - 3 and Ara+5 - 2 were significantly lower than the other 12 strains. Ara+4 - 3 is the only line with mutations in either lpcA or hns . Mutat ions in the former gene have been shown to confer tigecycline resistance in E. coli through modifications to the lipopolysaccharide biosynthesis pathway, and these mutations have moderate fitness cos ts in vitro (Linkevicius et al. 2013, 2016). The latter g ene encodes the global transcriptional regulator H - NS, and mutations in it affect acid resistance (Giangrossi et al. 2005), the modulation of osmotic stress (Lucht et al. 1994), and several other imp ortant cellular processes. ructure and function might therefore have large fitness costs via widespread pleiotropic effects. The Ara+5 - 2 clone evolved a 9 - bp insertion in ompF , which encodes the sole major porin in the LTEE an cestral strain (Crozat et al. 2011); this mutation presum 2005; Phan and Ferenci 2017). Thus, resistance mutations that affect different cellular pathways and func tions can have variable fitness costs, a finding that is consistent with many other studies (Vogwill and MacLean 2015). Another factor is that mutations in different genes that are part of the same physiological pathway may confer similar resistance levels but have different fitness costs. In our study, four tet racycline - resistant lines derive from the same LTEE ancestor: one had a mutation in envZ , while the other three had mutations in ompR . These genes encode proteins that comprise a two - component regula tory system that regulates cellular responses to osmotic stress, and which affects antibiotic resistance through altered expression of the major porin OmpF (Chakraborty and 75 Kenney 2018; Choi and Lee 2019). We observed significant heterogeneity in fitness e ven among these lines, implying that different changes wi thin this one pathway can impose unique burdens. The evolution of carbapenem resistance in E. coli K12 can also occur by mutations in this same two - component system, again with variable fitness costs (Adler et al. 2013). In their study, Adler and colleague s (2013) found that envZ mutants had no measurable loss of fitness in the absence of antibiotic, whereas ompR mutations suffered a large cost. By contrast, in our study the envZ mutation was more costly, which may reflect differences between the E. coli K1 2 and B strain backgrounds or the use of different culture media. Yet another factor is that different mutations in the same gene can have different costs. The evolution of rifam picin resistance, for example, typically occurs via mutations in several canon ical regions of rpoB al. 2002; Barrick et al. 2010; MacLean et al. 2010; Hall and MacLean 2011). Diffe rent alleles have widely varying costs that impact their competitive ability a nd, moreover, affect the dynamics of subsequent compensatory evolution (Barrick et al. 2010). In our study, two clones derived from the same parent had different nonsynonymous mu tations in ompR . Both conferred the same level of resistance to tetracycline, but they had different fitness costs in the absence of the drug. Such differences can have important public - competitive fitness in the absence of antibiotics is critically important for its long - term persi stence in a heterogeneous environment. More generally, we argue that further studies of the fitness costs of antibiotic resistance are needed, because this phenomenon can inform treatment strategies. Standard clinical practice calls for aggressive treatmen t to eliminate an infecting pathogen before it has time to evolve resistance (Craig 2001; Drlica 2003; Mehrotra et al. 2004; Abdul - Aziz et al. 2015; Hansen et al. 2020). This 76 app roach is likely beneficial if the population is composed of only drug - suscepti ble cells. However, if the pathogen population already contains drug - resistant cells, then aggressive treatment may promote the proliferation of the resistant population by elimi nating susceptible competitors. To address this problem, an alternative treatm ent strategy was recently proposed (Day and Read 2016; Hansen et al. 2020). Given that resistance often imposes a cost, resistant variants might be at a competitive disadvantage relative to their sensitive counterparts at low antibiotic concentrations that nonetheless reduce the growth rate of both types. If so, the resulting to clear the infection. Both mathematical models (Hansen et al. 2017) and experiments wit h the LTEE ancestor (Hansen et al. 2020) have shown that competition between susceptible and resistant populations, mediated in part by fitness costs, can indeed slow the time to treatment failure. However, these expectations are complicated by (i) the pot ential for higher mutation rates, and (ii) idiosyncratic fitness costs that depend on the specific resistance mutation and its interaction with the genetic background in which it occurs. Regarding the first complication, Hansen and colleagues (2020) used a strain with a low mutation rate (Sniegowski et al. 1997). However, six LTEE populations evolved hypermutability by generation 50,000 (Tenaillon et al. 2016), and mutation rates vary in some pathogens by orders of magnitude (Hughes and Andersson 2017). Wit h respect to the second complication, the competitive release of a resistant population should occur faster when fitness costs are lower. Given that the cost may depend on the pa rticular mutation and its genetic background, the time to treatment failure is harder to predict. We think that these issues and their relevance for treatment options are important avenues for future research. 77 CHAPTER 4: Conclusions and future outlook 78 Summary Chapter 1 examined how readily bacteria could overcome prior losses of intrinsic resistance through subsequent evolution when challenged with antibiotics. My co - authors and I focused on the role that genetic background played in this process, with particular attention to the interplay between repeatability and contingency in the evolutionary process (Card et al. 2019). We addressed these questions using Escherichia coli strains from the long - term evolution experiment (LTEE) that independently evolv ed for multiple decades in an environment without antibiotics. We first confirm ed that these LTEE - derived strains had typically become more susceptible to various drugs during this period of relaxed selection. We then subjected the strains to a range of co ncentrations of these same drugs. We found that evolvability was idiosyncratic with respect to initial genotype, such that resistance was more constrained in some backgrounds than in others. These results show that replicate populations accumulate genetic differences, even as they evolve in permissive environments, that affect their ability to adapt when challenged with antibiotics. In that study we focused on empirical trends in the evolution of phenotypic resistance and the effects of genetic background o n those trends. Chapter 2 asked also influences t he genetic basis of resistance. Do lineages founded by different genotypes take parallel or divergent mutational paths to increased resistance? My co - authors and I addressed thi s question by sequencing the complete genomes of antibiotic - resistant clones th at evolved from several parental genotypes during the earlier experiment described above. Using a statistical approach developed to compare mutational targets in populations tha t evolved under different thermal regimes (Deatherage et al. 2017), we demonstr ated that genomic evolution was specific to antibiotic treatment and that particular gene - level mutations were associated with each drug. These results were expected given that antibiotic resistance tends to evolve through ordered mutational 79 pathways in a limited set of genes (Toprak et al. 2012; Baym et al. 2016), but they served to validate our statistical approach. Next, we examined the specificity of genomic evolution with re spect to genetic background. We found that, on average, replicate lines evolved from the same founding genotypes had more mutations in common at the gene level than did lines evolved from different founding genotypes. Taken together, the results from Chapt ers 2 and 3 demonstrate that s phenotypic and genotypic paths to antibiotic resistance. Chapter 3 examined whether the fitness cost of tetracycline resistance is conserved across different LTEE backgroun ds. Accordingly, my co - authors and I quantified costs by comparing resistant mu tants to their sensitive parents in the absence of this drug. On average, the resistant lines had reduced fitness, and there was significant among - line heterogeneity in fitness costs. However, this variation did not correlate with the level of resistance c onferred by the mutations, nor did the cost vary significantly across the different genetic backgrounds tested. Instead, these idiosyncratic differences can be explained by muta tions in different genetic targets, mutations in different genes that comprise the same physiological pathway, and even different alleles of the same gene. We conclude in this Chapter that fitness costs often depend on idiosyncratic effects of particular r esistance mutations and the genetic backgrounds in which they arise. The multi factorial nature of antibiotic - resistance evolution Antibiotic resistance is a growing public - health crisis. Efforts to control resistance would benefit from an improved ability to forecast when and how it will evolve. However, to effectively predict evolu evolutionary history. My dissertation centers on the effects of different genetic backgrounds on 80 the evolution of antibiotic resistance and its fitness costs. Other factors , including the mutation supply rate and population bottlenecks, can also affect evolutionary trajectories and complicate our ability to predict resistance evolution. I discuss these additional factors and their medical implications in the following sectio ns. Mutational supply rate The adaptive potential of a population depends upon the number of potential evolutionary paths to improvement and the likelihood of finding each of t hose paths. Antibiotic resistance can arise through mutations that change gene regulation, cellular permeability and efflux, or the structure of the drug target (Blair et al. 2015). However, not all paths are equal phenotypically because these mechanisms c an confer different levels of drug resistance; mutations that reduce drug accum ulation in the cell, for example, tend to be less advantageous than those which alter drug targets (Hughes and Andersson 2017). Also, the rate at which these mutations occur dep ends on e sometimes called the The effects of population demography and mutation rates on adaptive potential are not well understood. This situation reflects cha llenges presented by the great variability of pathogen populations. Clinically relevant population sizes can span many orders of magnitude. Many newborn deaths worldwide have been caused by sepsis (United Nations Children's Fund 2009), and those infections might be driven by bacterial loads as low as 40 cells per milliliter (Stranier i et al. 2018); tuberculosis infections can reach 10 5 cells per milliliter of sputum (Palaci et al. 2007); and titers from meningitis patients can exceed 10 9 cells per millilite r (Bingen et al. 1990; Hughes and Andersson 2017). On balance, large population s should be more evolvable because they are 81 more likely to sample rare mutations that provide high levels of resistance, although this expectation is complicated by differences in mutation rates between populations. In vitro mutation rates are typically ch aracterized by the frequency at which detectable mutants arise in a population under selection with a given drug concentration. However, the rate of those mutations with suffici ent resistance is likely to vary with the drug concentration (see Chapter 1 Res ults and Discussion), making this parameter challenging to quantify. For example, the so - requi red to inhibit susceptible variants in the population up to that required to in hibit growth of resistant variants (Drlica 2003; Drlica and Zhao 2007). Moreover, increased antibiotic resistance can arise through mutations in different genes, and these mutat ions may provide different levels of resistance. At low antibiotic concentratio ns, mutations in any of these genes can effectively protect the bacteria. However, as the antibiotic concentration rises, the number of potential mutations that confer sufficien t resistance (i.e., the effective mutational target size) should decrease. One the MSW and underestimate this parameter at the upper boundary (Martínez and Baquero 2000). Bac terial populations can evolve elevated mutation rates, typically ranging from 1 0 - to 1,000 - fold higher than their wild - type counterparts (Macía et al. 2005; Wielgoss et al. 2013). Hypermutable mutants have a higher probability of discovering rare beneficia l alleles compared to nonmutator strains, and therefore these mutants may be fa vored, at least indirectly (Taddei et al. 1997; Tenaillon et al. 2001; Lenski 2004), in novel or rapidly changing environments (Sniegowski et al. 1997; Jayaraman 2011). For inst ance, P. aeruginosa infections are a major cause of morbidity and mortality in individuals with cystic fibrosis (CF) (Smith et al. 1996). Lung defenses against bacterial colonization primarily occur through ciliary action and neutrophilic 82 phagocytosis, both of which are limited by the viscous and hyperosmotic environment of the CF lu ng, resulting in long - term infection ( Oliver et al. 2000). This environment, along with recurrent and prolonged drug therapy, can select for P. aeruginosa lineages with elevated rates of mutation, in some cases 100 - fold higher than the ancestral wild - type (Govan and Deretic 1996; Oliver et al. can have mutation rates that are orders of magnitude higher than the opportunistic, short - term infections of the bloodstr eam, further suggesting that chronic i nfections of the lung environment impose repeated, strong selection on P. aeruginosa populations (Oliver et al. 2000), which generates second - order selection for higher mutation rates. Population size is another factor in the relation between mutation supp ly rate and resistance potential. For example, Adler and colleagues (2013) examined the effect of an extended - spectrum - lactamase (ESBL) - producing plasmid on the evolution of carbapenem resistance in E. coli . They found that the plasmid can increase the m utation rate to ertapenem specifically by 1800 - to 6000 - fold with an associated rise in resistance levels. They examined whether this observation could be explained by an increased overall mutation r ates to a separate antibiotic (i.e., rifampicin) were nearly equivalent between strains with and without the plasmid, indicating that overall mutation rates were not altered. Instead, the authors demonstrated that the ESBL - producing bacteria survived at la rger population sizes, and therefore were able to generate more resistance mutations, compared to cells that did not harbor the plasmid. The studies mentioned above underscore the importance of exploring the roles of population size and mutation rate in th e evolution of drug resistance. However, it is not well understood how genetic background interacts with population size and mutation rate to influence evolutionary trajectories to resistance. 83 Future directions In my future work, I plan to use E. coli stra ins from the LTEE to examine the contributions of genetic background, population size, and mutation rate to resistance evolvability. The LTEE growth medium contains glucose as the limiting nutrient, along with abundant citrate that is present as an iron - ch elating agent. E. coli cannot normally use citrate as an energy source under the oxygen - rich conditions of the experiment. However, after ~31,000 generations the population designated Ara 3 evolved the capacity to grow on this substrate (Cit + ) (Blount et a l. 2008). The Cit + trait caused a several - fold increase in population size. Additionally, the population had the low, ancestral mutation rate until a hypermutator phenotype evolved around 35,000 generations y includes changes in both population size and mutation rate, one could investigate the resistance potential of strains isolated from across this evolutionary time series. By generation 31,000, the Ara 3 population had evolved nonsynonymous point mutations in genes encoding DNA topoisomerase, efflux pump regulators, and multiple ribosomal proteins (Tenaillon et al. 2016). These mutations could conceivably influence resistance evolution to various drug classes. Nevertheless, rare pathways leading to high - lev el resistance may become more accessible due to the larger population size after Cit + evolution. One objective of this study could be to assess whether an increase in evolutionary potential due to the larger population size outweighs, or is outwei ghed by, any potentiating or constraining effects of changes in the genetic background (Card et al. 2019). Similarly, resistance potential should also increase when this population evolved hypermutability. However, this response becomes harder to predict f or later - generation strains that, owing to hypermutability, have accumulated hundreds of background mutations that may idiosyncratically interact with resistance mutations. 84 The fitness of a resistant bacterium, including relative to its sensitive counterpa rts in th e absence of antibiotics, can affect its evolutionary success over time (Chapter 3). For example, a mutation that confers less resistance may ultimately prevail over one that confers greater resistance if the latter carries a higher fitness cost t han the f ormer (Andersson and Hughes 2010; Melnyk et al. 2015). Discovery of more resistant variants in the short - term may not necessarily equate to their success over longer timescales. To examine this possibility, one could compete mutants against each o ther and their wild - type progenitors in continuous culture devices (i.e., morbidostats) ( Toprak et al. 2012; Gopalakrishnan et al. 2019; Kaznatcheev et al. 2019). Morbidostats steadily increase the drug concentration in a growth medium as the population un der study evolves resistance (Toprak et al. 2012; Gopalakrishnan et al. 2019) . The evolution of resistance will therefore oc cur over time with increasing antibiotic concentrations and by the sequential substitution of multiple muta tions. One could then sample the evolving populations periodically and perform metagenomic sequencing to examine how the frequency of competitors cha nge over time. One might also examine the individual effects of population size and mutation rate on evolva bility by using a variant of the LTEE ancestor. One could compare this hypermutable strain against its wild - type counterpart using the experime ntal approach as described. Population sizes could be directly controlled by altering the concentration of glucose in the medium. In any Predictions should i population size, and the rel ative fitness of resistance mutations. In this study, one could examine the interaction of these factors on resistance evolution using a well - charact erized experimental population of E. coli . 85 Population bottlenecks As discussed in this Chapter, and throughout this dissertation, evolutionary trajectories to antibiotic resistance can depend upon several factors including the supply of beneficial mutation s (dependent upon the mutation rate and population size under sele context; and the fitness of resistance mutations across varying drug concentrations. In this section, I discuss one additional factor with clinical relevance: t he size of the transmitted population that establishes new infecti ons (Hughes and Andersson 2017) . Population bottlenecks play important roles in disease dynamics and the evolution of antibiotic resistance in pathogen populations. The sexual transmission o f HIV - 1 can occur through stringent bottlenecks as low as one viri on particle (da Silva 2012) , and in the case of bacterial infections, clonal expansions can drive the spread of tetracycline - resistant Group B Streptococcus (Da Cunha et al. 2014) . Bottlenec ks also reduce genetic diversity, and they may reduce fitness duri ng transmission events if less - fit genotypes survive a bottleneck event and displace their more - fit counterparts (Moxon and Kussell 2017) . Given that genetic variation is the substrate that fuels natural selection, bottlenecks may also impose indirect sel ection that favors mechanisms that can rapidly regenerate diversity (Moxon and Kussell 2017) . Two important mechanisms include hypermutability (discussed above) and phase variation. Phase va riation occurs when gene expression is altered, in a reversible ma nner, by insertions and deletions in short genomic repeats, homologous or site - specific recombination, or epigenetic modifications ( Moxon et al. 1994; van der Woude and Bäumler 2004; Moxon a nd Kussell 2017) . Phase variable loci in pathogens can promote vir ulence by altering cell - surface molecules that modulate host - microbe interactions, and they may also facilitate the maintenance of some antibiotic - resistance determinants as well (Jiang et a l. 2019) . 86 Mutation rates, the relative fitness of antibiotic resis tance mutations, and population bottlenecks can simultaneously affect evolutionary trajectories of antibiotic resistance. For instance, high - level fluoroquinolone resistance in E. coli occur s through the sequential substitution of mutations in several diff erent genes (Lindgren et al. 2003) . The first mutation to be selected in this pathway generally occurs in gyrA , which encodes the DNA topoisomerase II drug target. Mutations in gyrA confer l ow levels of fluoroquinolone resistance with minimal associated fi tness costs (Huseby et al. 2017) . However, the second mutation in this sequence is more variable. Mutations in the regulator of the major efflux pump AcrAB TolC that confer resistance occur at a high rate of 10 6 per base pair per generation, whereas mutat ions in the secondary drug target topoisomerase IV, parC , occur at the comparatively low rate of 10 9 per base pair per generation. This difference in mutation rate can be understood in term s of different mutational target sizes (Hughes and Andersson 2017; Huseby et al. 2017) . Huseby and colleagues (2017) found that changes in efflux pump regulation predominate when fluoroquinolone resistance is selected for in vitro, in contrast to clinical strains in which second - step mutations often occur in parC . The au thors hypothesized that this discrepancy might be explained by differences in fitness costs, with parC mutations having smaller costs than mutations that upregulate drug efflux. To test thei r hypothesis, they challenged populations (already containing the first - step mutation in gyrA ) with increased levels of ciprofloxacin across several population bottleneck sizes. They found that when the population bottleneck was large enough for both varia nts to survive the bottleneck, the outcome matched that seen in cl inical isolates: most replicate lines had mutations in parC (Hughes and Andersson 2017) . Therefore, this work suggests that the bottleneck size during transmission of this pathogen even betw een hosts is sufficiently large to maintain the rarer, low - cost al leles compared to the more frequent, but costlier, 87 alternative solutions. More generally, more stringent bottlenecks might constrain the trajectories of resistance evolution by limiting gene tic diversity. Concluding remarks ng by itself, we find it hitched to everything else in the John Muir, My First Summer in the Sierra In closing, my dissertation research addressed questions about the importanc e of genetic background on the evolution of antibiotic resistance, including its phenotypic and genotypic patterns, and its associated fitness costs. Nonetheless, several additional factors can alter evolutionary trajectories and complicate our ability to forecast resistance evolution in clinical settings. Thus, an evolu tionary perspective might complement and benefit efforts to develop new strategies to limit the emergence and spread of antibiotic - resistant bacteria. And this work has set me on a trajecto ry of my own one that is both fascinating and important for our collective well - being. 88 APPENDICES 89 A PPENDIX A : Supplementary Tables 90 Table A. 1 . Bacterial strains used in Chapter 1. Strain Generation LTEE Population REL60 6 0 - REL772A 500 Ara+5 REL772B 500 Ara+5 REL962A 1,000 Ara+5 REL962B 1,000 Ara+5 REL1066A 1,500 Ara+5 REL1066B 1,500 Ara+5 REL1162A 2,000 Ara+5 REL1162B 2,000 Ara+5 REL2177A 5,000 Ara+5 REL2177B 5,000 Ara+5 REL4534A 10,000 Ara+5 REL4534B 10,00 0 Ara+5 REL11339 50,000 Ara 5 REL11389 50,000 Ara 6 REL11348 50,000 Ara+4 REL11367 50,000 Ara+5 All clones were derived from REL606, the ancestral strain of the LTEE. 91 Table A. 2 . Statistical significance for changes in intrinsic resistance o f the individual clones sampled at generation 50,000 of the LTEE for the four antibiotic treatments. Antibiotic Clone p Ampicillin Ara 5 0.0080 Ara 6 0.0039 Ara+4 0.0080 Ara+5 0.0080 Ceftriaxone Ara 5 0.0031 Ara 6 0.0031 Ara+4 0.0039 Ara+5 0. 0039 Ciprofloxacin Ara 5 0.2177 Ara 6 0.1004 Ara+4 0.0149 Ara+5 0.0039 Tetracycline Ara 5 0.0031 Ara 6 0.0031 Ara+4 0.0039 Ara+5 0.0278 Analyses were performed based on a trinomial distribution, which reflects the many ties in these datasets . The reported p - values are one - tailed, which reflects the expectation that resistance should decline under relaxed selection in the antibiotic - free LT EE environment. 92 Table A. 3 . Statistical significance for trends of diminishing - returns resistan ce evolvability of the individual clones sampled at generation 50,000 of the LTEE for the four antibiotic treatments. Antibiotic Clone p Ampicillin Ar a 5 0.0456 Ara 6 0.1094 Ara+4 0.1592 Ara+5 0.8408 Ceftriaxone Ara 5 0.9481 Ara 6 0.9969 Ara+4 0.9969 Ara+5 0.9722 Ciprofloxacin Ara 5 0.9250 Ara 6 0.8778 Ara+4 0.1964 Ara+5 0.1222 Tetracycline Ara 5 0.2895 Ara 6 0.1946 Ara+4 0.9250 Ara+5 0.9722 Analyses were performed based on a trinomial distribution, which reflects the many ties in these datasets. The reported p - values are one - tailed, which reflects the directional expectation implied by diminishing returns. 93 Table A. 4 . Bacterial stra ins sequenced in Chapter 2. Antibiotic* LTEE parent Replicate Strain AMP Ancestor 1 KJ C108 AMP Ancestor 2 KJC109 AMP Ancestor 3 KJC110 AMP Ancestor 4 KJC111 AMP Ara 5 1 KJC114 AMP Ara 5 2 KJC122 AMP Ara 5 3 KJC130 AMP Ara 6 1 KJC115 AMP Ara 6 2 KJC123 AMP Ara 6 3 KJC131 AMP Ara+4 1 KJC112 AMP Ara+4 2 KJC120 AMP Ara+4 3 KJC128 A MP Ara+5 1 KJC113 AMP Ara+5 2 KJC121 AMP Ara+5 3 KJC129 CRO Ancestor 1 KJC212 CRO Ancestor 2 KJC213 CRO Ancestor 3 KJC214 CRO Ancestor 4 KJC215 CRO Ara 5 1 KJC218 CRO Ara 5 2 KJC226 CRO Ara 5 3 KJC234 CRO Ara 6 1 KJC219 CRO Ara 6 2 KJC227 CRO A ra 6 3 KJC235 CRO Ara+4 1 KJC216 CRO Ara+4 2 KJC224 CRO Ara+4 3 KJC232 CRO Ara+5 1 KJC217 CRO Ara+5 2 KJC225 CRO Ara+5 3 KJC233 CIP Ancestor 1 KJC148 CIP Ancestor 2 KJC149 CIP Ancestor 3 KJC150 CIP Ancestor 4 KJC151 CIP Ara 5 1 KJC154 CIP Ara 5 2 KJC162 CIP Ara 5 3 KJC186 CIP Ara 6 1 KJC155 CIP Ara 6 2 KJC163 CIP Ara 6 3 KJC187 CIP Ara+4 1 KJC152 94 Table A. 4 CIP Ara+4 2 KJC160 CIP Ara+4 3 KJC184 CIP Ara+5 1 KJC153 CIP Ara+5 2 KJC161 CIP Ara+5 3 KJC185 TET Ancestor 1 KJC60 TET Ancestor 2 KJC61 TET Ancestor 3 KJC62 TET Ancestor 4 KJC63 TET Ara 5 1 KJC66 § TET Ara 5 2 KJC74 TET Ara 5 3 KJC82 TET Ara 6 1 KJC67 TET Ara 6 2 KJC75 TET Ara 6 3 KJC83 TET Ara+4 1 KJC64 TET Ara+4 2 KJC72 TET Ara+4 3 KJC80 TET Ara+5 1 KJC65 § TET Ara+5 2 KJC73 TET Ara+5 3 KJC81 *AMP, ampicillin; CRO, ceftriaxone; CIP, ciprofloxacin; TET, tetracycline. Ancestor, E. coli B strain REL606. All other parental strains are clones sampled at generation 50,000 from the indicated LTEE popula tion. Three strains (KJC152, KJC184, KJC217) are cross - contaminants, and they were discarded from all analys es. § Two strains (KJC65, KJC66) have no identifiable mutations. 95 Table A. 5 . Bacterial strains used in Chapter 3. Evolved tetracycline - res istant clones Strain name Derived from Freezer ID Ancestor - 1 REL606 KJC60 Ancestor - 2 REL606 KJC61 Ancestor - 3 REL606 KJC62 Ancestor - 4 REL606 KJC63 Ara 5 - 1 REL11339 KJC66 Ara 5 - 2 REL11339 KJC74 Ara 5 - 3 REL11339 KJC82 Ara 6 - 1 REL11389 KJC67 Ara 6 - 2 REL11389 KJC75 Ara 6 - 3 REL11389 KJC83 Ara+4 - 1 REL11348 KJC64 Ara+4 - 2 REL11348 KJC72 Ara+4 - 3 REL11348 KJC80 Ara+5 - 1 REL11367 KJC65 Ara+5 - 2 REL11367 KJC73 Ara+5 - 3 REL11367 KJC81 Tetracycline - sensitive parental strains LTEE population LTEE generati on Freezer ID Ancestor 0 REL606 Ara 5 50,000 REL11339 Ara 6 50,000 REL11389 Ara+4 50,000 REL11348 Ara+5 50,000 REL11367 Strains used as common competitors LTEE population LTEE generation Freezer ID Ancestor 0 REL607 Ara 5 40,000 REL10948 Ara 5 40 ,000 REL11638 96 A PPENDIX B : Supplementary Figures 97 Figure B. 1 . Schematic illustration of the LTEE and evolvability study design. (A) Twelve initially identical E. coli populations were founded from a common ancestor to start the LTEE. These pop ulations have evolved for >7 3 ,000 generations with daily serial transfers in a minima l medium without antibiotics. (B) In this study, antibiotic - susceptible ancestral or derived clones from generation 50,000 were inoculated into replicate cultures. A resis tance mutation may arise xpansion, resulting in two genetic variants: the susceptible parental cells and their descendent resistant daughters. (C) These whole populations were then spread onto aga r plates supplemented with two - fold increasing concentrations of an antibiotic (shown in red). Minimum inhibitory concentrations (MICs) of these two variants correspond to the lowest antibiotic concentration that inhibits confluent growth and that prevents even isolated colonies, respectively. Resistant clones were confirmed by streaking o nto fresh plates with relevant antibiotic concentrations. 98 Figure B. 2 . Experimental plates. Whole populations containing susceptible parental and resistant dau ghter cells were spread onto MH agar amended with two - fold increasing concentrations of ciprofloxacin (left to right, and down). Confluent lawns of bacterial growth (plates 1 3) consist largely of drug - susceptible cells. Isolated colonies (plates 4 5) are putatively resistant mutants. 99 REFERENCES 100 REFERENCES Abdul - Aziz, M. H., J. Lipman, J. W. Mouton, W. W. Hope, and J. A. Roberts. 2015. Applying pharmacokinetic/pharmacodynamic principles in critically ill patients: Optimizing effic acy and reducing resistance development. Semin. Respir. Crit. Care M ed. 36:136 153. AbdulWahab, A., K. Zahraldin, M. A. Sid Ahmed, S. A. Jarir, M. Muneer, S. F. Mohamed, J. M. Hamid, A. A. I. Hassan, and E. B. Ibrahim. 2017. The emergence of multidrug - res istant Pseudomonas aeruginosa in cystic fibrosis patients on inhaled antibiotics. Lung India 34:527 531. Achaz, G., A. Rodriguez - Verdugo, B. S. Gaut, and O. Tenaillon. 2014. The reproducibility of adaptation in the light of experimental evolution with who le genome sequencing. Pp. 211 231 in C. R. Landry and N. Aubin - Horth , eds. Ecological Genomics. Advances in Experimental Medicine and Biology. Springer, Dordrecht. Adler, M., M. Anjum, D. I. Andersson, and L. Sandegren - lacta mases on the evolution of spontaneous carbapenem resistance in Escherichia coli . J. Antimicrob. Chemother. 68:51 59. Cora or, B. A. Grüning, A. Guerler, J. Hil lman - Jackson, S. Hiltemann, V. Jalili, H. Rasche, N. Soranzo, J. Goecks, J. Taylor, A. Nekrutenko, and D. Blankenberg. 2018. The Galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2018 u pdate. Nucleic Acids Res. 46:W537 W54 4. binomial proportions. Am. Stat. 52:119 126. Agudelo - Romero, P, F. de la Iglesia, and S. F. Elena. 2008. The pleiotro pic cost of host - specialization in To bacco etch potyvirus . Infect. Genet. Evol. 8:806 814. Ahmad, S., E. Mokaddas, and E. Fares. 2002. Characterization of rpoB mutations in rifampin - resistant clinical Mycobacterium tuberculosis isolates from Kuwait and Du bai. Diagn. Microbiol. Infect. Dis. 4 4:245 252. Aiba, H., and T. Mizuno. 1990. Phosphorylation of a bacterial activator protein, OmpR, by a protein kinase, EnvZ, stimulates the transcription of the ompF and ompC genes in Escherichia coli . FEBS Lett. 261:1 9 22. Andersson, D. I., and D. Hughe s. 2010. Antibiotic resistance and its cost: is it possible to reverse resistance? Nat. Rev. Microbiol. 8:260 271. 101 Limited ev olutionary conservation of the phenot ypic effects of antibiotic resistance mutations. Mol. Biol. Evol. 36:1601 1611. Baker, S., N. Thomson, F. - X. Weill, and K. E. Holt. 2017. Genomic insights into the emergence and spread of antimicrobial - resistant bacterial pathogens. Science 360:733 738. Balaban, N. Q., S. Helaine, K. Lewis, M. Ackermann, B. Aldridge, D. I. Andersson, M. P. Brynildsen, D. Bumann, A. Camilli, J. J. Collins, C. Dehio, S. Fortune, J. M. Ghigo, W. - D. Hardt, A. Harms, M. Heinemann, D. T. Hung, U. Jenal, B. R. Levin, J. Michiels , G. Storz, M. - W. Tan, T. Tenson, L. Van Melderen, and A. Zinkernagel. 2019. Definitions and guidelines for research on antibiotic persistence. Nat. Rev. Microbiol. 17:441 448. Barrick, J. E., M. R. Kauth, C. C. Strelioff, and R. E. Lenski. 2010. Escheric hia coli rpoB mutants have increased evolvability in proportion to their fitness defects. Mol. Biol. Evol. 27:1338 1347. Barrick, J. E., G. Colburn, D. E. Deatherage, C. C. Traverse, M. D. Strand, J. J. Borges, D. B. Knoester, A. Reba, and A. G. Meyer. 20 14. Identifying structural variation in haploid microbial genomes from short - read resequencing data using breseq . BMC Genomics 15:1039. Baym, M., T. D. Lieberman, E. D. Kelsic, R. Chait, R . Gross, I. Yelin, and R. Kishony. 2016. Spatiotemporal microbial e volution on antibiotic landscapes. Science 353:1147 1151. Bian, G., M. McAleer, and W. - K. Wong. 2011. A trinomial test for paired data when there are many ties. Math. Comput. Simul. 81:115 3 1160. Bingen, E., N. Lambert - Zechovsky, P. Mariani - Kurkdjian, C. Doit, Y. Aujard, F. Fournerie, and H. Mathieu. 1990. Bacterial counts in cerebrospinal fluid of children with meningitis. Eur. J. Clin. Microbiol. Infect. Dis. 9:278 281. Björkman, J., D. Hughes, and D. I. Andersson. 1998. Virulence of antibiotic - resista nt Salmonella typhimurium . Proc. Natl. Acad. Sci. 95:3949 3953. Blair, J. M. A., M. A. Webber, A. J. Baylay, D. O. Ogbolu, and L. J. V. Piddock. 2015. Molecular mechanisms of antibiotic re sistance. Nat. Rev. Microbiol. 13:42 51. Blount, Z. D., C. Z. Borl and, and R. E. Lenski. 2008. Historical contingency and the evolution of a key innovation in an experimental population of Escherichia coli . Proc. Natl. Acad. Sci. 105:7899 7906. Blount, Z. D., J. E. Barrick, C. J. Davidson, and R . E. Lenski. 2012. Genomi c analysis of a key innovation in an experimental Escherichia coli population. Nature 489:513 518. Blount, Z. D., R. E. Lenski, and J. B. Losos. 2018. Contingency and determinism in evolution: aam5979. 102 Blount, Z. D., R. Maddamsetti, N. A. Grant, S. T. Ahmed, T. Jagdish, J. A. Baxter, B. A. Sommerfield, A. Tillman, J. Moore, J. L. Slonczewski, J. E. Barrick, and R. E. Lenski. 2020. Genomic and phenotypic evolution of Escherichia coli in a novel citrate - only resource env ironment. eLife, in press. Bolger, A. M., M. Lohse, and B. Usadel. 2014. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30:2114 2120. Böttger, E. C., B. Springer, M. Pletschette, and P. Sander. 1998. Fi tness of antibiotic - resis tant microorganisms and compensatory mutations. Nat. Med. 4:1343 1344. Bouma, J. E., and R. E. Lenski. 1988. Evolution of a bacteria/plasmid association. Nature 335:351 352. Card, K. J., T. LaBar, J. B. Gomez, and R. E. Lenski. 2 019. Historical contingen cy in the evolution of antibiotic resistance after decades of relaxed selection. PLoS Biol. 17:e3000397. Card, K. J. 2019. Data from: Historical contingency in the evolution of antibiotic resistance after decades of relaxed select ion. Dryad Digital Reposi tory. Openly available via https://datadryad.org/stash/dataset/ doi:10.5061/dryad.g41hg96. Chakraborty, S., and L. J. Kenney. 2018. A new role of OmpR in acid and osmotic stress in Salmonella and E. coli . Front. Microbiol. 9. Cha o, L, B. R. Levin, and F. M. Stewart. 1977. A complex community in a simple habitat: An experimental study with bacteria and phage. Ecology 58:369 378. Chevereau, G., M. Dravecká, T. Batur, A. Guvenek, D. H. Ayhan, E. Toprak, and T. Bollenbach . 2015. Quantifying the determinants of evolutionary dynamics leading to drug resistance. PLoS Biol. 13:e1002299. Choi, U., and C. - R. Lee. 2019. Distinct roles of outer membrane porins in antibiotic resistance and membrane integrity in Escherichia coli . F ront. Microbiol. 10. Chou, H. - H., H. - C. Chiu, N. F. Delaney, D. Segrè, and C. J. Marx. 2011. Diminishing returns epistasis among beneficial mutations decelerates adaptation. Science 332:1190 1192. Coffey, L. L., and M. Vignuzzi. 2011. Host alteration of Chikungunya virus increases fitness w hile restricting population diversity and adaptability to novel selective pressures. J. Virol. 85: 1025 1035. University Press, New York. Cooper, V. S., and R. E. Lenski. 2000. The population genetics of ecological specialization in evolving Escherichia coli populations. Nature 407:736 739. 103 Cooper, V. S., A. F. Bennett, and R. E. Lenski. 2001. Evolution of thermal dependence of growth rate of Escherichia coli popul ations during 20,000 generations in a constant environment. Evolution 55:889 896. Cooper, V. S., D. Schneider, M. Blot, and R. E. Lenski. 2001. Mechanisms causing rapid and parallel losses of ribose catabolism in evolv ing populations of E. coli B. J. Bact eriol 183:2834 2841. Cox, G., and G. D. Wright. 2013. Intrinsic antibiotic resistance: Mechanisms, origins, challenges and solutions. Int. J. Med. Microbiol. 303:287 292. Craig, W. A. 2001. Re - evaluating current antibiotic therapy. Respir. Med. 95:S12 S1 9. Crozat, E., T. Hindré, L. Kühn, J. Garin, R. E. Lenski, and D. Schneider. 2011. Altered regulation of the OmpF pori n by Fis in Escherichia coli during an evolution experiment and between B and K - 12 strains. J. Bacteriol. 193:429 440. Da Cunha, V., M. R. Davies, P. - E. Douarre, I. Rosinski - Chupin, I. Margarit, S. Spinali, T. Perkins, P. Lechat, N. Dmytruk, E. Sauvage, L. Ma, B. Romi, M. Tichit, M. J. Lopez - Sanchez, S. Descorps - Declere, E. Souche, C. Buchrieser, P. Trieu - Cuot, I. Moszer, D. Clermont, D. M aione, C. Bouchier, D. J. McMillan, J. Parkhill, J. L. Telford, G. Dougan, M. J. Walker, M. T. G. Holden, C. Poyart, P. Glaser, P. Melin, A. Decheva, B. Petrunov, P. Kriz, R. Berner, A. Buchele, M. Hufnagel, M. Kunze, R. Creti, L. Baldassarri, G. Orefici, A. Berardi, J. R. Granger, M. D. L. R. Fraile, B. Afshar, and A. Efstratiou. 2014. Streptococcus agalactiae c lones infecting humans were selected and fixed through the extensive use of tetracycline. Nat. Commun. 5:1 12. da Silva, J. 2012. The dynamics of HIV - 1 adaptation in early infection. Genetics 190:1087 1099. Daegelen, P., F. W. Studier, R. E. Lenski, S. C ure, and J. F. Kim. 2009. Tracing ancestors and relatives of Escherichia coli B, and the derivation of B strains REL606 and BL21(DE3). J. Mol. Biol . 394:634 643. Darwin, C. 1859. On the Origin of Species. John Murray, London. Day, T., and A. F. Read. 201 6. Does high - dose antimicrobial chemotherapy prevent the evolution of resistance? PLoS Comput. Biol. 12:e1004689. Deatherage, D. E., J. L. Kepner, A. F. Bennett, R. E. Lenski, and J. E. Barrick. 2017. Specificity of genome evolution in experimental popula tions of Escherichia coli evolved at different temperatures. Proc. Natl. Acad. Sci. 114:E1904 E1912. Drlica , K. 2003. The mutant selection window hypothesis and antimicrobial resistance. J. Antimicrob. Chemother. 52:11 17. Drlica, K., and X. Zhao. 2007. Mutant selection window hypothesis updated. Clin. Infect. Dis. 44:681 688. 104 Duffy, S, P. E. Turner, and C. L . Burch. 2006. Pleiotropic costs of nic he expansion in the RNA 757. Ellis, C. N., and V. S. Cooper. 2010. Experimental adaptation of Burkholderia cenocepacia to onion medium reduces host range. Appl. Environ. Microbiol. 76:2387 2396. Enne, V. I., A. A. Dels ol, G. R. Davis, S. L. Hayward, J. M. Roe, and P. M. Bennett. 2005. Assessment of the fitness impacts on Escherichia coli of acquisition of antibiotic resistance genes encoded by different types of genetic element. J. Antimicrob. Chemother. 56:544 551. EU CAST. Breakpoint tables for interpretations of MICs and zone diameters, version 10.0, 2020. https://www.eucast.org/fileadmin/src/media/PDFs/EUCAST_files/Breakpoint_tables/v_10.0_Bre akpoint_Tables.pdf. Ferenci, T. 200 5. Maintaining a healthy SPANC balance through regulatory and mutational adaptation. Mol. Microbiol. 57:1 8. Fisher, R. A. 1934. Statistical Methods for Research Workers. 5th ed. Oliver and Boyd, Edinburgh. Giangrossi, M., S. Zattoni, A. Tramonti, D. De Biase, and M. Falconi. 2005. Antagonist ic role of H - NS and GadX in the regulation of the glutamate decarboxylase - dependent acid resistance system in Escherichia coli . J. Biol. Chem. 280:21498 21505. Gillet, R., and B. Felden. 2001. Transfer RNA Ala recogni zes transfer - messenger RNA with specifi city; a functional complex prior to entering the ribosome? EMBO J. 20:2966 2976. Gopalakrishnan, V., N. P. Krishnan, E. McClure, J. Pelesko, D. Crozier, D. F. K. Williamson, N. Webster, D. Ecker, D. Nichol, S. Mandal , R. A. Bonomo, and J. G. Scott. 2019. A low - cost, open source, self - contained bacterial EVolutionary biorEactor (EVE). bioRxiv 729434. Govan, J. R. W., and V. Deretic . 1996. Microbial pathogenesis in cystic fibrosis: Mucoid Pseudomonas aeruginosa and Burkholderia cepacia . Microbiol. Rev. 60:5 39 574. Hall, A. R., and R. C. MacLean. 2011. Epistasis buffers the fitness effects of rifampicin - resistance mutations in Pseudo monas aeruginosa . Evolution 65:2370 2379. Han, F, S. Pu, F. Wang, J. Meng, and B. Ge. 2009. Fitness cost of macrolide resistan ce in Campylobacter jejuni . Int. J. Antimicrob. Agents. 34:462 466. Hansen, E., R. J. Woods, and A. F. Read. 2017. How to use a chemotherapeutic agent when resistance to it threatens the patient. PLoS Biol. 15:e2001110. Hansen, E., J. Karslake, R. J. Woo ds, A. F. Read, and K. B. Wood. 2020. Antibiotics can be used to contain drug - resistant bacteria by maintaining sufficiently larg e sensitive populations. PLoS Biol. 18:e3000713. 105 Hughes, D., and D. I. Andersson. 2017. Evolutionary trajectories to antibioti c resistance. Annu. Rev. Microbiol. 71:579 596. Huseby, D. L., F. Pietsch, G. Brandis, L. Garoff, A. Tegehall, and D. Hughes. 20 17. Mutation supply and relative fitness shape the genotypes of ciprofloxacin - resistant Escherichia coli . Mol. Biol. Evol. 34:1 029 1039. Imamovic, L., and M. O. A. Sommer. 2013. Use of collateral sensitivity networks to design drug cycling protocols that avoid resistance development. Sci. Transl. Med. 5:204ra132. Imamovic, L., M. M. H. Ellabaan, A. M. Dantas Machado, L. Citterio , T. Wulff, S. Molin, H. Krogh Johansen, and M. O. A. Sommer. 2018. Drug - driven phenotypic convergence supports rational treatmen t strategies of chronic infections. Cell 172:1 14. Iram, S., E. Dolson, J. Chiel, J. Pelesko, N. Krishnan, Ö. Güngör, B. Kuzne ts - speck, S. Deffner, E. Ilker, J. G. Scott, and M. Hinczewski. 2020. Controlling the speed and trajectory of evolution with coun terdiabatic driving. Nat. Phys., in press. Jayaraman, R. 2011. Hypermutation and stress adaptation in bacteria. J. Genet. 90:3 83 391. Jeong, H., V. Barbe, C. H. Lee, D. Vallenet, D. S. Yu, S. - H. Choi, A. Couloux, S. - W. Lee, S. H. Yoon, L. Cattolico, C. - G . Hur, H. - S. Park, B. Ségurens, S. C. Kim, T. K. Oh, R. E. Lenski, F. W. Studier, P. Daegelen, and J. F. Kim. 2009. Genome sequ ences of Escherichia coli B strains REL606 and BL21(DE3). J. Mol. Biol. 394:644 652. Jiang, X., A. B. Hall, T. D. Arthur, D. R. Plichta, C. T. Covington, M. Poyet, J. Crothers, P. L. Moses, A. C. Tolonen, H. Vlamakis, E. J. Alm, and R. J. Xavier. 2019. In vertible promoters mediate bacterial phase variation, antibiotic resistance, and host adaptation in the gut. Science 363:181 187. Jochumsen, N., R. L. Marvig, S. Damkiær, R. L. Jensen, W. Paulander, S. Molin, L. Jelsbak, and A. Folkesson. 2016. The evolut ion of antimicrobial peptide resistance in Pseudomonas aeruginosa is shaped by strong epistatic interactions. Nat. Commun. 7:13002. Kassen, R., and T. Bataillon. 2006. Distribution of fitness effects among beneficial mutations before selection in experime ntal populations of bacteria. Nat. Genet. 38:484 488. Kaznatcheev, A., J. Peacock, D. Basanta, A. Marusyk, and J. G. Scott. 2019. Fibroblasts and alectinib switch the evolutionary games played by non - small cell lung cancer. Nat. Ecol. Evol. 3:450 456. Ke seler, I. M., A. Mackie, A. Santos - Zavaleta, R. Billington, C. Bonavides - Martínez, R. Caspi, C. Fulcher, S. Gama - Castro, A. Kothari, M. Krummenacker, M. Latendresse, L. Muñiz - Rascado, Q. Ong, S. Paley, M. Peralta - Gil, P. Subhraveti, D. A. Velázquez - Ramírez , D. Weaver, J. Collado - Vides, I. Paulsen, and P. D. Karp. 2017. The EcoCyc database: reflecting new knowledge about Escherichia coli K - 12. Nucleic Acids Res. 45:D543 D550. 106 Khan, A. I., D. M. Dinh, D. Schneider, R. E. Lenski, and T. F. Cooper. 2011. Negat ive epistasis between beneficial mutations in an evolving bacterial population. Scienc e 332:1193 1196. Koschwanez, J. H, K. R. Foster, and A. W. Murray. 2013. Improved use of a public good selects for the evolution of undifferentiated multicellularity. eLi fe 2:e00367. Kryazhimskiy, S., D. P. Rice, E. R. Jerison, and M. M. Desai. 2014. Glob al epistasis makes adaptation predictable despite sequence - level stochasticity. Science 344:1519 1522. Kugelberg, E., S. Löfmark, B. Wretlind, and D. I. Andersson. 2005. Reduction of the fitness burden of quinolone resistance in Pseudomonas aeruginosa . J. Antimicrob. Chemother. 55:22 30. Lahti, D. C., N. A. Johnson, B. C. Ajie, S. P. Otto, A. P. Hendry, D. T. Blumstein, R. G. Coss, K. Donohue, and S. A. Foster. 2009. Rel axed selection in the wild. Trends. Ecol. Evol. 24:487 496. Lamrabet , O., M. Martin, R. E. Lenski, and D. Schneider. 2019. Changes in intrinsic antibiotic susceptibility during a long - term evolution experiment with Escherichia coli . mBio 10:e00189 - 19. La ng, G. I. 2018. Measuring mutation rates using the Luria - Delbrück flu ctuation assay. Methods Mol. Biol. 1672:22 31. Lázár, V., G. P. Singh, R. Spohn, I. Nagy, B. Horváth, M. Hrtyan, R. Busa - Fekete, B. Bogos, O. appanos, B. Kégl, B. Papp, and C. Pál. 2013. Bacterial evolution of a ntibiotic hypersensitivity. Mol. Syst. Biol. 9:700. Leiby, N, and C. J. Marx. 2014. Metabolic erosion primarily through mutation accumulation, and not tradeoffs, drives limited evolution of substrate specificity in Escherichia coli . PLoS. Biol. 12:e100178 9. Lenski, R. E., and J. E. Bouma. 1987. Effects of segregation and selection on instability of pACYC184 in Escherichia coli B. J. Bacteriol. 169:5314 5316. Lenski, R. E. 1988. Experime ntal studies of pleiotropy and epistasis in Escherichia coli . I. Vari ation in competitive fitness among mutants resistant to virus T4. Evolution 42:425 432. Lenski, R. E., M. R. Rose, S. C. Simpson, and S. C. Tadler. 1991. Long - term experimental evolution in Escherichia coli . I. Adaptation and divergence during 2,000 gener ations. Am. Nat. 138:1315 1341. Lenski, R. E., S. C. Simpson, and T. T. Nguyen. 1994. Genetic analysis of a plasmid - encoded, host genotype - specific enhancement of bacterial fitness. J. B acteriol. 176:3140 3147. Lenski, R. E. 1997. The cost of antibiotic resistance from the perspective of a bacterium. Pp. 131 151 in D. J. Chadwick and G. Cardew, eds. Antibiotic Resistance: Origins, Evolution, Selection and Spread. Wiley, Chichester, UK. 107 Lenski, R. E. 2004. Phenotypic and genomic evolution during a 20,000 - generation experiment with the bacterium Escherichia coli . Plant Breed. Rev. 24:225 265. Lenski, R. E., M. J. Wiser, N. Ribeck, Z. D. Blount, J. R. Nahum, J. J. Morris, L. Zaman, C. B. Tu rner, B. D. Wade, R. Maddamsetti, A. R. Burmeister, E. J. Baird, J. Bundy, N. A. Grant, K. J. Card, M. Rowles, K. Weatherspoon, S. E. Papoulis, R. Sullivan, C. Clark, J. S. Mulka, and N. Hajela. 2015. Sustained fitness gains and variability in fitness traj ectories in the long - term evolution experiment with Escherichia coli . Proc. R. Soc. B 282:20152292. Lenski, R. E. 2017. Convergence and divergence in a long - term experiment with bacteria. Am. Nat. 190: S57 S68. E. Barrick. 2018. Innovation in an E. coli evolution experiment is contingent on maintaining adaptive potential until competition subsides. PLoS Genet. 14:e1007348. Levin, B. R., M. Lipsitch, V. Perrot, S. Schrag, R. Antia, L. Simonsen, N. Moore Walker, a nd F. M. Stewart. 1997. The population genetics of antibiotic resistance. Clin. Infect. Dis. 24:S9 S16. Levin, B. R., V. Perrot, and N. Walker. 2000. Compensatory mutations, antibiotic resistance and the population genetics of adaptive evolution in bacter ia. Genetics 154:985 997. Lindgren, P. K., Å. Karlsson, and D. Hughes. 2003. Mutation rate and evolution of fluoroquinolone resistance in Escherichia coli isolates from patients with urinary tract infections. Antimicrob. Agents Chemother. 47:3222 3232. L inkevicius, M., L. Sandegren, and D. I. Andersson. 2013. Mechanisms and fitness costs of tigecycline resistance in Escherichia coli . J. Antimicrob. Chemother. 68:2809 2819. Linkevicius, M., J. M. Anderssen, L. Sandegren, and D. I. Andersson. 2016. Fitness of Escherichia coli mutants with reduced susceptibility to tigecycline. J. Antimicrob. Chemother. 71:1307 1313. Lucht, J. M., P. Dersch, B. Kempf, and E. Bremer. 1994. Interactions of the nucleoid - associated DNA - binding protein H - NS with the regulatory r egion of the osmotically controlled proU operon of Escherichia coli . J. Biol. Chem. 269:6578 6586. Luo, N. S. Pereira, O. Sahin, J. Lin, S. Huang., L. Michel, and Q. Zhang. 2005. Enhanced in vivo fitness of fluoroquinolone - resistant Campylobacter jejuni i n the absence of antibiotic selection pressure. Proc. Natl. Acad. Sci. 102:541 546. Luria, S. E., and M. Delbrück. 1943. Mutations of bacteria from virus sensitivity to virus resistance. Genetics 28:491 511. Macía, M. D., D. Blanquer, B. Togores , J. Sauleda, J. L. Pérez, and A. Oliver. 2005. Hypermutation is a key factor in development of multiple - antimicrobial resistance in Pseudomonas aeruginosa strains causing chronic lung infections. Antimicrob. Agents Chemother. 49:3382 3386. 108 MacLean, R. C. , G. G. Perron, and A. Gardner. 2010. Diminishing returns from beneficial mutations and pervasive epistasis shape the fitness landscape for rifampicin resistance in Pseudomonas aeruginosa . Genetics 186:1345 1354. Maltas, J., and K. B. Wood. 2019. Pervasiv e and diverse collateral sensitivity profiles inform optimal strategies to limit antibiotic resistance. PLoS Biol. 17:e3000515. Marcusson, L. L., N. Frimodt - Møller, and D. Hughes. 2009. Interplay in the selection of fluoroquinolone resistanc e and bacteria l fitness. PLoS Pathog. 5:e1000541. Martin, L. W., C. L. Robson, A. M. Watts, A. R. Gray, C. E. Wainwright, S. C. Bell, K. A. Ramsay, T. J. Kidd, D. W. Reid, B. Brockway, and I. L. Lamont. 2018. Expression of Pseudomonas aeruginosa antibioti c resistance g enes varies greatly during infections in cystic fibrosis patients. Antimicrob. Agents Chemother. 62:e01789 - 18. Martínez, J. L., and F. Baquero. 2000. Mutation frequencies and antibiotic resistance. Antimicrob. Agents Chemother. 44:1771 1777. Mehrotra, R. , R. De Gaudio, and M. Palazzo. 2004. Antibiotic pharmacokinetic and pharmacodynamic considerations in critical illness. Intensive Care Med. 30:2145 2156. Melnyk, A. H., A. Wong, and R. Kassen. 2015. The fitness costs of antibiotic resistanc e mutations. E vol. Appl. 8:273 283. Meyer, S. E., S. Granett, J. U. Jung, and M. R. Villarejo. 1990. Osmotic regulation of PhoE porin synthesis in Escherichia coli . J. Bacteriol. 172:5501 5502. Meyer, J. R, D. T. Dobias, S. J. Medina, L. Servilio, A. Gup ta, and R. E. Lenski. 2016. Ecological speciation of bacteriophage lambda in allopatry and sympatry. Science 354:1301 1304. Moore, F. B. - G., D. E. Rozen, and R. E. Lenski. 2000. Pervasive compensatory adaptation in Escherichia coli . Proc. R. Soc. B 267:51 5 522. Moxon , E. R., P. B. Rainey, M. A. Nowak, and R. E. Lenski. 1994. Adaptive evolution of highly mutable loci in pathogenic bacteria. Curr. Biol. 4:24 33. Moxon, R., and E. Kussell. 2017. The impact of bottlenecks on microbial survival, adaptation, and phenotypic switching in host pathogen interactions. Evolution 71:2803 2816. Nguyen, T. N. M., Q. G. Phan, L. P. Duong, K. P. Bertrand, and R. E. Lenski. 1989. Effects of carriage and expression of the Tn 10 tetracycline - resistance operon on the fitness of Escherichi a coli K12. Mol. Biol. Evol. 6:213 225. Nichol, D., P. Jeavons, A. G. Fletcher, R. A. Bonomo, P. K. Maini, J. L. Paul, R. A. Gatenby, A. R. A. Anderson, and J. G. Scott. 2015. Steering evolution with sequential therapy to prevent the emergence of bacteria l antibiotic resistance. PLoS Comput. Biol. 11:e1004493. 109 Nichol, D., J. Rutter, C. Bryant, A. M. Hujer, S. Lek, M. D. Adams, P. Jeavons, A. R. A. Anderson, R. A. Bonomo, and J. G. Scott. 2019. Antibiotic collateral sensitivity is contingent on the repeatab i lity of evolution. Nat. Commun. 10:334. Nilsson, A. I., A. Zorzet, A. Kanth, S. Dahlström, O. G. Berg, and D. I. Andersson. 2006. Reducing the fitness cost of antibiotic resistance by amplification of initiator tRNA genes. Proc. Natl. Acad. Sci. 103:6976 6 981. Nishino, K., and A. Yamaguchi. 2004. Role of histone - like protein H - NS in multidrug resistance of Escherichia coli . J. Bacteriol. 186:1423 1429. Oliver, A., R. Cantón, P. Campo, F. Baquero, and J. Blázquez . 2000. High frequency of hypermutable Pseud omonas aeruginosa in cystic fibrosis lung infection. Science 288:1251 1253. - resistant infections globally: Final report and recommendations. The Review on Antimicrobial Resistance . London: Wellcome Trust and UK Government. Orr, H. A. 2005. The genetic theory of adaptation: A brief history. Nat. Rev. Genet. 6:119 127. Oz, T., A. Guvenek, S. Yildiz, E. Karaboga, Y. T. Tamer, N. Mumcuyan, V. B. Ozan, G. H. Senturk, M. Cokol, P. Yeh, and E. Toprak. 2014. Strength of selection pressure is an important parameter contributing to the complexity of antibiotic resistance evolution. Mol. Biol. Evol. 31:2387 2401. Palaci, M., R. Dietze, D. J. Hadad, F. K. C. Ribeiro, R. L. Peres, S. A. Vinha s, E. L. N. Maciel, V. D. V. Dettoni, L. Hor ter, W. H. Boom, J. L. Johnson, and K. D. Eisenach. 2007. Cavitary disease and quantitative sputum bacillary load in cases of pulmonary tuberculosis. J. Clin. Microbiol. 45:4064 4066. Palmer, A. C., R. Chait, an d R. Kishony. 2018. Nonoptimal gene expressi on creates latent potential for antibiotic resistance. Mol. Bol. Evol. 35:2669 2684. Passagem - Santos, D., S. Zacarias, and L. Perfeito. 2018. Power law fitness landscapes and their ability to predict fitness. He redity 121:482 498. Phan, K., and T. Ferenc i. 2017. The fitness costs and trade - off shapes associated with the exclusion of nine antibiotics by OmpF porin channels. ISME J. 11:1472 1482. 008. A practical guide to measuring mutation rates in antibiotic resistance. Antimicrob. Agents Chemother. 52:1209 1214. Pugsley, A. P., and J. P. Rosenbusch. 1983. OmpF porin synthesis in Escherichia coli strains B and K - 12 carrying heterologous ompB and /or ompF loci. FEMS Microbiol. Lett. 16:143 147. Quandt, E. M., J. Gollihar, Z. D. Blount, A. D. Ellington, G. Georgiou, and J. E. Barrick. 2015. Fine - tuning citrate synthase flux potentiates and refines metabolic innovation in the Lenski evolution experi ment. eLife 4:e09696. 110 Raeside, C., J. Gaffé, D. E. Deatherage, O. Tenaillon, A. M. Briska, R. N. Ptashkin, S. Cruveiller, C. Médigue , R. E. Lenski, J. E. Barrick, and D. Schneider. 2014. Large chromosomal rearrangements during a long - term evolution experiment with Escherichia coli . mBio 5:e01377 - 14. Ratcliff, W. C, R. F. Denison, M. Borrello, and M. Travisano. 2012. Experimental evolu tion of multicellularity. Proc. Natl. Acad. Sci. 109:1595 1600. Reboud, X, and G. Bell. 1997. Experimental evolution in Chlamydomonas . III. Evolution of specialist and generalist types in environments that vary in space and time. Heredity 78:507 514. Rey nolds, M. G. 2000. Compensatory evolution in rifampicin - resistant Escherichia coli . Genetics 156:1471 1481. Rozen, D. E., L. McGee, B. R. Levin, and K. P. Klugman. 200 7. Fitness costs of fluoroquinolone resistance in Streptococcus pneumoniae . Antimicrob. Agents Chemother. 51:412 416. Sandegren, L., and D. I. Andersson. 2009. Bacterial gene amplification: implications for the evolution of antibiotic resistance. Nat. Rev . Microbiol. 7:578 588. Santos - Lopez, A., C. W. Marshall, M. R. Scribner, D. J. Snyder, and V. S. Cooper. 2019. Evolutionary pathways to antibiotic resistance are dependent upon environmental structure and bacterial lifestyle. eLife 8:1 23. Schneider, D. , E. Duperchy, J. Depeyrot, E. Coursange, R. E. Lenski, and M. Blot. 2002. Genomic compa risons among Escherichia coli strains B, K - 12, and O157:H7 using IS elements as molecular markers. BMC Microbiol. 2:1 8. Schrag, S. J., V. Perrot, and B. R. Levin. 199 7. Adaptation to the fitness costs of antibiotic resistance in Escherichia coli . Proc. R . Soc. B. 264:1287 1291. Sjölund, M., K. Wreiber, D. I. Andersson, M. J. Blaser, and L. Engstrand. 2003. Long - term persistence of resistant Enterococcus species after antibiotics to eradicate Helicobacter pylori . Ann. Intern. Med. 139:483 487. Smith, J. J., S. M. Travis, E. P. Greenberg, and M. J. Welsh. 1996. Cystic fibrosis airway epithelia fail to kill bacteria because of abnormal airway surface fluid. Cell 85:229 2 36. Sniegowski, P. D., P. J. Gerrish, and R. E. Lenski. 1997. Evolution of hi gh mutation rates in experimental populations of E . coli . Nature 387:703 705. Sokal, R. R., and F. J. Rohlf. 1994. Biometry: The Principles and Practices of Statistics in Biolo gical Research. 3rd ed. W. H. Freeman and Company, New York. Sommer, M. O. A. , G. M. Church, and G. Dantas. 2010. The human microbiome harbors a diverse reservoir of antibiotic resistance genes. Virulence 1:299 303. 111 Stranieri, I., K. A. Kanunfre, J. C. Ro drigues, L. Yamamoto, M. I. V. Nadaf, P. Palmeira, and T. S. Okay. 2018. Asses sment and comparison of bacterial load levels determined by quantitative amplifications in blood culture - positive and negative neonatal sepsis. Rev. Inst. Med. Trop. Sao Paulo 60 :e61. Studier, F. W., P. Daegelen, R. E. Lenski, S. Maslov, and J. F. Kim. 20 09. Understanding the differences between genome sequences of Escherichia coli B strains REL606 and BL21(DE3) and comparison of the E. coli B and K - 12 Genomes. J. Mol. Biol. 394: 653 680. Taddei, F., M. Radman, J. Maynard - Smith, B. Toupance, P. H. Gouyon, and B. Godelle. 1997. Role of mutator alleles in adaptive evolution. Nature 387:700 702. Tenaillon, O., F. Taddei, M. Radman, and I. Matic. 2001. Second - order selection in bacte rial evolution: selection acting on mutation and recombination rates in the co urse of adaptation. Res. Microbiol. 152:11 16. Tenaillon, O., A. Rodriguez - Verdugo, R. L. Gaut, P. McDonald, A. F. Bennett, A. D. Long, and B. S. Gaut. 2012. The molecular diver sity of adaptive convergence. Science 335:457 461. Tenaillon, O., J. E. Barri ck, N. Ribeck, D. E. Deatherage, J. L. Blanchard, A. Dasgupta, G. C. Wu, S. Wielgoss, S. Cruveiller, C. Médigue, D. Schneider, and R. E. Lenski. 2016. Tempo and mode of genome ev olution in a 50,000 - generation experiment. Nature 536:165 170. Thanassi, D. G ., L. W. Cheng, and H. Nikaido. 1997. Active efflux of bile salts by Escherichia coli . J. Bacteriol. 179:2512 2518. Toprak, E., A. Veres, J. - B. Michel, R. Chait, D. L. Hartl, an d R. Kishony. 2012. Evolutionary paths to antibiotic resistance under dynamically sustained drug selection. Nat. Genet. 44:101 105. Travisano, M., J. A. Mongold, A. F. Bennett, and R. E. Lenski. 1995. Experimental tests of the roles of adaptation, chance, and history in evolution. Science 267:87 90. Trindade, S., A. Sousa, K. B. Xavier, F. Dionisio, M. G. Ferreira, and I. Gordo. 2009. Positive epistasis drives the acquisition of multidrug resistance. PLoS Genet. 5:e1000578. Tubulekas, I., and D. Hughes. 1993. Suppression of rpsL phenotypes by tuf mutations reveals a unique relationship between translation elongation and growth rate. Mol. Microbiol. 7:275 284. Turner, P. E, and S. F. Elena. 2000. Cost of host radiation in an RNA virus. Genetics 156: 1465 1470. Children 2009: Maternal and Newborn Health. New York: UNICEF. van der Woude, M. W., and A. J. Bäumler. 2004. Phase and antigenic variation in bacteria. Clin. Microbiol. Rev. 17:581 611. 112 Vogwill, T., and R. C. MacLean. 2015. The genetic basis of the fitne ss costs of antimicrobial resistance: a meta - analysis approach. Evol. Appl. 8:284 295. Vogwill, T., M. Kojadinovic, and R. C. MacLean. 2016. Epistasis between antibiotic resistance muta tions and genetic background shape the fitness effect of resistance ac ross species of Pseudomonas . Proc. R. Soc. B 283:20160151. Wasik, B. R, A. Bhushan, C. B. Ogbunugafor, and P. E. Turner. 2015. Delayed transmission selects for increased survival of ves icular stomatitis virus. Evolution 69:117 125. Wenger, J. W., J. Piot rowski, S. Nagarajan, K. Chiotti, G. Sherlock, and F. Rosenzweig. 2011. Hunger artists: Yeast adapted to carbon limitation show trade - offs under carbon sufficiency. PLoS Genet. 7:e100220 2. Wielgoss, S., J. E. Barrick, O. Tenaillon , M. J. Wiser, W. J. Dittmar, S. Cruveiller, B. Chane - Woon - Ming, C. M digue, R. E. Lenski, and D. Schneidera. 2013. Mutation rate dynamics in a bacterial population reflect tension between adaptation and geneti c load. Proc. Natl. Acad. Sci. 110:222 227. Wiser, M. J., N. Ribeck, and R. E. Lenski. 2013. Long - term dynamics of adaptation in asexual populations. Science 342:1364 1367. Wohlleben, W., Y. Mast, E. Stegmann, and N. Ziemert. 2016. Antibiotic drug discovery. Microb. Biotechnol. 9:541 548. Wong, A. 2017. Epistasis and the evolution of antimicrobial resistance. Front. Microbiol. 8:246. Woods, R., D. Schneider, C. L. Winkworth, M. A. Riley, and R. E. Lenski. 2006. Tests of parallel molecular evolution in a long - term experiment with Escherichia coli . Proc. Natl. Acad. Sci. 103:9107 9112. Yen, P., and J. A. Papin. 2017. History of antibiotic adaptation influences microbial evolutionary dynamics during subsequent treatment. PLoS Biol. 15:e2001586.