Antibiotic exposure rapidly selects for more resistant bacterial strains, and both a drug's chemical structure and a bacterium's cellular network affect the types of mutations acquired.
To better characterize the genetic determinants of antibiotic susceptibility, we exposed a transposon-mutagenized library of Escherichia coli to each of 17 antibiotics that encompass a wide range of drug classes and mechanisms of action. Propagating the library for multiple generations with drug concentrations that moderately inhibited the growth of the isogenic parental strain caused the abundance of strains with even minor fitness advantages or disadvantages to change measurably and reproducibly. Using a microarray-based genetic footprinting strategy, we then determined the quantitative contribution of each gene to E. coli's intrinsic antibiotic susceptibility. We found both loci whose removal increased general antibiotic tolerance as well as pathways whose down-regulation increased tolerance to specific drugs and drug classes. The beneficial mutations identified span multiple pathways, and we identified pairs of mutations that individually provide only minor decreases in antibiotic susceptibility but that combine to provide higher tolerance.
Our results illustrate that a wide-range of mutations can modulate the activity of many cellular resistance processes and demonstrate that E. coli has a large mutational target size for increasing antibiotic tolerance. Furthermore, the work suggests that clinical levels of antibiotic resistance might develop through the sequential accumulation of chromosomal mutations of small individual effect.
Citation: Girgis HS, Hottes AK, Tavazoie S (2009) Genetic Architecture of Intrinsic Antibiotic Susceptibility. PLoS ONE 4(5): e5629. https://doi.org/10.1371/journal.pone.0005629
Editor: Christophe Herman, Baylor College of Medicine, United States of America
Received: February 15, 2009; Accepted: April 23, 2009; Published: May 20, 2009
Copyright: © 2009 Girgis et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: AKH was assisted by fellowship #08-1090-CCR-EO from the New Jersey State Commission on Cancer Research. S.T. was supported by grants from the NSF Early Career Development (CAREER) Program, Defense Advanced Research Projects Agency and National Institute of General Medical Sciences (P50 GM071508). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Antibiotic tolerance, the decreased efficacy of antimicrobial agents in eliminating infections, is a serious and worsening global problem in human health , . During the long history of chemical warfare between microbes, the genomes of many bacteria have evolved to encode multiple counter-measures . Moreover, a level of antibiotic tolerance that allows some bacteria to survive an initial exposure gives the population the opportunity to accumulate mutations, leading to higher levels of tolerance and potentially to full clinical resistance . Much of the literature on antibiotic resistance focuses on tolerance to the high antibiotic levels typically used in a clinical setting (see  for an exception).
Even in clinical practice, however, bacteria commonly experience sub-inhibitory drug concentrations, which may be capable of reducing the growth rate but are lower than the minimum inhibitory concentration (MIC). The cyclical dosing regimen for most antibiotics, for example, may cause the drug's plasma concentration to approach the MIC for short intervals during treatment. Furthermore, micro-niches within the host, such as epidermis, lungs, and joints, may attain significantly lower drug concentrations than the plasma . Finally, patient non-compliance with the prescribed frequency and duration of antibiotic use can allow plasma levels to fall below the MIC. In such circumstances, selection for more tolerant variants is strong.
Outside clinical settings, environments containing antibiotics, especially at sub-inhibitory concentrations, abound. Soil contains numerous antibiotic-producing species , which generate compounds with roles in killing competitors as well as in inter- and intra-species signaling . Antibiotics also enter the soil through the use of manure from livestock whose feed was supplemented with antibiotics , and wastewater can contain multiple drugs at concentrations in the range of ng/L, even after treatment . The rise in environmental antibiotic levels resulting from the widespread use of antibiotics has selected for resistant strains in both soil  and aqueous  environments.
Diverse mechanisms including drug target modification, enzymatic drug inactivation, and intracellular drug concentration reduction can lead to antibiotic resistance , , . A variety of sources, such as lateral gene transfer and chromosomal mutations, can provide the underlying genetic changes, and clinically resistant strains often contain multiple alterations. The plasmids, transposons, and mobile chromosome cassettes that contribute to antibiotic resistance, including methicillin resistance, in Staphylococcus aureus are examples of well-studied extrinsic elements that confer antibiotic resistance . Similarly, E. coli strains can receive contributions to quinolone resistance from gyrA and parC mutations as well as a reduction in outer membrane porins and an increase in drug efflux pump activity . Cases where multiple mutations work collectively to give resistance, while of clear importance, are not well understood, in part because the range of mutations that can increase a particular bacterium's antibiotic tolerance is rarely known.
The effect of mutation in altering genetic programs and modulating susceptibility, however, is starting to receive attention. Multiple groups have assayed the drug susceptibility of bacteria containing families of simple genetic perturbations such as single gene deletions or transposon insertions , , , , . These studies found that large numbers of genes influence drug susceptibility, including many whose dominant function is not resistance. Moreover, while some genes, such as those that code for drug efflux pumps, contribute to resistance to a wide range of compounds, many others are drug or drug class specific.
Most previous work focused on identifying genes whose removal changes the MIC by some predetermined threshold, usually 2-fold. The increasing number of examples of multiple genetic perturbations combining to confer high levels of resistance , , however, suggests that mutations of even mild individual benefit are of potential clinical relevance. As such, we decided to take a complementary approach to identifying mutations that alter antibiotic susceptibility.
Rather than testing MIC directly, we searched a library of transposon-mutagenized cells  for those with a competitive advantage at drug concentrations that moderately reduced the parental strain's growth rate. Competitive growth of the mutants for multiple generations caused the abundance of strains with even minor fitness advantages or disadvantages to change measurably and reproducibly. Applying microarray-based genetic footprinting  to the resulting population allowed us to identify and quantify the genetic determinants contributing to both antibiotic sensitivity and tolerance.
The experimental conditions proved to be particularly conducive to finding mutants fitter than the wildtype in the presence of antibiotics, enabling us to find many genes whose inactivation increases antibiotic tolerance. Analysis of individual mutants revealed a non-perfect correlation between a strain's fitness advantage in the conditions of the selection and its MIC in similar conditions. Explanations for and implications of the variations are discussed.
Results and Discussion
Selection of Transposon Insertion Mutants under Sub-inhibitory Antibiotic Exposure
To characterize the contribution of all genetic loci to antibiotic tolerance, we exposed a collection of mutants, each with a single transposon insertion, to sub-inhibitory drug concentrations (Figure 1). We chose 17 antibiotics that possess a range of mechanisms of action (Table 1). These drugs inhibit various cellular functions including the synthesis of proteins, nucleic acids, folic acid, and cell wall , . As intermediate antibiotic concentrations are typically the most selective for resistance , during the selections, we used antibiotic concentrations that impaired but did not completely inhibit the growth of the wild-type strain (Table 1). Using a sensitive microarray-based strategy , we then quantified the change in prevalence of different transposon insertion mutants in parallel.
(A) An aliquot of a library containing ∼5×105 mutants each with a single transposon insertion  was taken from frozen stock, grown overnight in LB, pelleted, washed, and resuspended at 2% inoculum in fresh M9-media containing an antibiotic at the chosen concentration (Table 1). Each day, an aliquot was frozen, and 2% of the culture was transferred to fresh media to continue the selection. Genetic footprinting was performed on frozen samples to amplify the region of genomic DNA adjacent to the transposon in each of the mutants . DNA was subsequently labeled and hybridized with a reference of labeled genomic DNA to spotted microarrays . (B) Dose response curves were used to select drug concentrations. For each antibiotic, fresh media containing various drug concentrations was inoculated with overnight culture of the wild-type strain. Growth was monitored using OD600 readings. Shown are the curves for amikacin; curves for all other antibiotics are in Figure S1. Typically, we selected moderately inhibitory drug concentrations that reduced the growth after 14 hours by 30–50%. (C) Separation of DNA on an agarose gel provided a qualitative depiction of the population diversity after each day of selection. Shown are the amplified Tn-adjacent DNA from all seven days of one of the ampicillin selections. Selections performed without antibiotic showed no discernable banding pattern (Figure S2). Gel images for all selections with antibiotics are in Figure S3.
The final distribution was compared to both the original, unselected library and to the library selected in the same media without antibiotics (See Materials and Methods). Comparison to populations grown in M9 media without antibiotics allowed media and antibiotic effects to be distinguished. As some external stresses such as low-level antibiotic exposure can ameliorate the deleterious nature of some mutations , the abundance of slow growing mutants may increase in cultures with antibiotics compared to cultures without antibiotics. Comparison to the unselected library allowed the identification of insertion locations whose prevalence had increased or decreased during the experiment and reduced the chances of classifying slow-growing mutants as less susceptible. Dataset S1 lists the loci whose disruption caused a significant fitness change in each of the drugs tested. Heatmaps displaying the data for loci whose disruptions were significant in individual drugs and classes of drugs are in Figures S4–pone.0005629.s006pone.0005629.s007pone.0005629.s008pone.0005629.s009pone.0005629.s010pone.0005629.s011S11.
Validation of Global Fitness Profiling
Comparison to the literature indicates that we found many genes whose removal is known to increase susceptibility (See Tables S1–pone.0005629.s014pone.0005629.s015S4). However, attempts to set significance thresholds that captured all such genes resulted in an unacceptable number of false positives. As shown later in this work, some mutations confer such a high degree of fitness in particular antibiotics that strains possessing these mutations rapidly take over the population. The existence of such highly fit mutants makes it difficult to distinguish neutral transposon insertion locations from some with a slightly deleterious effect because mutants with either are quickly lost from the population. Consequently, when selecting genes for additional study, an emphasis was placed on those whose removal increases antibiotic tolerance.
To quantify the strength of the identified loci, we subjected the wild-type parental strain to direct competition with mutant strains in conditions that mimicked those of the selection (See Materials and Methods and Text S1). The mutants tested behaved as expected from the en mass library selections (Figure 2).
Selection rates (generations/day) were calculated as (log2(A(t1)/A(t0))−log2(B(t1)/B(t0)))/(t1−t0) . A(t0) and B(t0) are, respectively, the mutant and the wild-type population sizes at t0, the beginning of the competition, and A(t1) and B(t1) are the mutant and the wild-type population sizes at t1, the end of the competition. Shown are the average and standard deviation of three repetitions. The selection rate for the trpA::kan mutant in amikacin was calculated after two days of enrichment to correspond with the samples hybridized. The trpA::kan strain's reliance on tryptophan from lysed wild-type cells prevents the mutant from taking over the culture, and during additional transfers, the wild-type strain showed a competitive advantage. Selection rates for other strains were insensitive to the competition duration.
We also determined the MICs for a representative set of individual mutants lacking genes judged to have a significant role in antibiotic susceptibility. The relative efficiency of our MIC measurement method (see Text S1) allowed us to analyze a larger set of strains than would have been possible using only the more labor-intensive direct competitions. As the initial selection process did not require a change in MIC, the MICs provide complementary information about each mutation's effect, and we did not expect all beneficial (or deleterious) disruptions to cause an increase (or decrease) in MIC. The most common result was a change in MIC of 1.5 fold (the smallest change detectable with our technique) in the expected direction—an increase (or decrease) for a locus whose disruption was beneficial (or deleterious). We observed increases up to ∼5-fold and decreases down to ∼7-fold (Tables S5–pone.0005629.s018S7).
Genes Exert Class-Specific Effects on Antibiotic Susceptibility
Excluding the 30 genes identified as having general roles in antibiotic susceptibility (see below), only antibiotics of the same class have a substantial number of beneficial or deleterious transposon insertion locations in common (Figure 3). Some functional classes of antibiotics, particularly the β-lactams and the aminoglycosides, have strong class-specific signatures, while other classes, such as the quinolones, have few loci in common. The number of loci impacting growth at the sub-inhibitory drug concentrations tested varied greatly from the low hundreds for the aminoglycosides down to zero for fusidic acid and erythromycin, inhibitors of the 50S subunit of the ribosome to which E. coli is intrinsically resistant . The mutational target size observed likely depends on both the antibiotic concentration assayed and the structure of E. coli's cellular network. For example, redundant pathways may mask the contribution of some genes to antibiotic tolerance.
Squares on the main diagonals indicate the number of genes whose disruption caused a significant fitness effect (See Materials and Methods). Genes causing general changes in antibiotic susceptibility (Figure 7) were excluded. The lower left (upper right) triangle reports on genes whose disruption was beneficial (deleterious) to E. coli in the presence of the indicated antibiotic. Off-diagonal squares indicate how many genes caused significant fitness changes in both antibiotics when disrupted. The shading shows the likelihood of an overlap of the indicated size or larger occurring by chance and was calculated using the hypergeometric distribution. P-values were corrected for multiple testing. Erythromycin and fusidic acid are not shown as the only genes whose disruption affected fitness caused general changes in susceptibility.
The observation that disruptions caused increases in fitness during drug exposure is characteristic of antagonistic pleiotropy , which reflects an evolutionary tradeoff for increased drug tolerance at the consequence of other traits. A large number of disruptions fell into this category (336), suggesting that the bacterium contains a large mutational target size for increasing drug tolerance. To elucidate the pathways and mechanisms contributing to susceptibility and tolerance, we analyzed the identified loci for three drug classes – the aminoglycosides (Figure 4), the β-lactams (Figure 5), and the folic acid biosynthesis inhibitors (Figure 6) – in more detail. In each case, the data indicate the major pathways involved and the many ways that mutations can modulate those pathways.
(A) The heatmaps, in which hierarchical clustering was used to order both the genes and the drugs , show loci whose disruption changed susceptibility to all four aminoglycosides tested (See Materials and Methods). Table S1 lists the genes with annotations. (B) Of the 73 transposon insertions regions identified as beneficial in all four aminoglycosides, the 48 shown are expected to reduce Fenton reaction-based oxidative damage. Following exposure to lethal concentrations of bactericidal antibiotics, the oxidative electron transport chain depletes the NADH pool, generating high levels of superoxide, which removes iron from iron-sulfur clusters . The free iron subsequently generates hydroxyl radicals through the Fenton reaction . Removal of key catabolic enzymes should shrink the NADH pool and reduce the flux through the electron transport chain. The media used lacks cysteine, the sulfur donor for iron-sulfur center synthesis , so disruption of cysteine biosynthesis should reduce the availability of sulfur for iron-sulfur centers. The iron-sulfur center synthesis genes shown are not specific for NADH dehydrogenase I, and their disruption should reduce the number of iron-sulfur clusters throughout the cell. Q: ubiquinone; FMN: flavin mononucleotide; FAD: flavin adenine dinucleotide
(A) The heatmaps shows loci whose disruption changed susceptibility to all three β-lactams tested (See Materials and Methods). Hierarchical clustering was used to order both genes and drugs . Table S2 lists the genes with annotations. (B) Both transposon insertions that disrupt genes that encode flagella components as well as insertions that indirectly reduce flagella synthesis by activating the Rcs system are beneficial. The core components of the Rcs system are RcsC, a hybrid sensor kinase, RcsD, a histidine phosphotransferase, and RcsB, a DNA-binding response regulator . Other components are RcsF, a lipoprotein that activates RcsC , , and RcsA, a transcription factor that forms a heterodimer with RcsB . Together, RcsA and RcsB repress transcription of flhDC, the master regulator of flagella synthesis . RcsA is a target of the Lon protease , and insertions in lon, which stabilize RcsA, are beneficial. RcsC and RcsD both transfer phosphate to as well as remove phosphate from RcsB, resulting in higher activation of the Rcs system in rcsC or rcsD mutants than in wildtype , . Insertions in mdoG and mdoH, which encode proteins that synthesize osmoregulated periplasmic glucans (OPGs), reduce motility by activating the Rcs system . The beneficial effects of mdoG, modH, and rcsC disruptions are not limited to β-lactams (Figure 7).
Sulfamonomethoxine inhibits FolP, a dihydropteroate synthase ; trimethoprim inhibits FolA, the cell's main dihydrofolate reductase (DHFR) . FolM, which also acts as a DHFR, is not inhibited by trimethoprim . Mutants lacking folM or folX are less sensitive to both to trimethoprim and sulfamonomethoxine.
Mutations Beneficial in the Presence of Aminoglycosides
Excluding loci with a general effect on antibiotic tolerance, disruptions of 73 loci were beneficial in all four aminoglycosides tested while disruptions of 14 loci were deleterious (Figure 4A). Using an aerobic environment – a condition where many antibiotics, especially aminoglycosides, are particularly effective  – likely increased the number and types of beneficial mutations identified. The set of beneficial disruption loci is enriched for genes with products involved in quinone binding (p = 1.7×10−15), ubiquinone biosynthesis (p = 1.1×10−5), glycolysis (p = 1.5×10−7), L-serine biosynthesis (p = 1.2×10−3), and cytochrome-c oxidase activity (p = 1.2×10−3). Additionally, the set is enriched for genes down-regulated in response to oxygen depravation (p = 3.5×10−16). Forty-eight of the beneficial mutations likely reduce the number of harmful hydroxyl radicals created by the Fenton reaction, part of a death-inducing chain reaction triggered by bactericidal antibiotics that includes NADH depletion, superoxide generation, and iron-sulfur center destabilization  (Figure 4B). Kohanski et al.  found that a set of E. coli strains with similar mutations had increased growth in the presence of gentamycin. Similarly, Schurek et al.  found that mutants of Pseudomonas aeruginosa with disruptions in homologous genes had increased tobramycin resistance
To understand the magnitude of the tolerance changes, we competed a nuoG::kan strain with the wild-type strain. While the nuoG::kan strain was at a slight disadvantage in the media without drugs, in the presence of amikacin, the mutant rapidly took over the culture (Figure 2). In contrast, the nuoG::kan strain and other mutants that perturbed the pathway leading to death from Fenton reaction-derived hydroxyl radicals had at most modest (less than 3-fold) increases in MICs and frequently failed to exhibit any increase in MIC (Table S5). Since oxidative respiration processes are more active in the shaken cultures used for the competitions than in static MIC-measurement plates, the fitness effect of removing genes that contribute to the Fenton pathway likely depends strongly on the oxygen level.
Disruptions expected to interfere with the death pathway mediated by the Fenton reaction were widely beneficial only in selections with aminoglycosides, even though all bactericidal antibiotics are thought to trigger the pathway . Furthermore, since the selections were performed at antibiotic concentrations that reduced, but did not completely inhibit, the wild-type strain's growth, observing genetic interactions with a putative death process was initially puzzling.
An explanation for the apparent paradox came from the observation that during the enrichments in amikacin and streptomycin, disruptions in tryptophan biosynthesis genes that should have been lethal in the growth media, which lacked tryptophan, were strongly beneficial. We confirmed that a trpA::kan mutant does not grow in the media, with or without antibiotics, and that when trpA::kan and wild-type strains compete without antibiotics, the wild-type strain rapidly takes over (Figure 2). When trpA::kan and wild-type strains compete in the presence of amikacin, however, the trpA::kan strain remains an appreciable part of the population (Figure 2). Presumably, a portion of the wild-type cells are dying, lysing, and releasing enough tryptophan to support the trpA::kan mutant. Thus, in the aminoglycoside enrichments, although the overall population was growing, individual cells were dying. The existence of similar concentration regimes for other bactericidal drugs remains an interesting area for future inquiry.
Several beneficial disruptions that likely reduce NADH accumulation and lower the metabolic flux through the Fenton reaction may also provide secondary benefits. First, disruption of electron transport reduces the uptake of aminoglycosides . Second, mutations that keep cAMP-CRP (cyclic AMP bound to the CRP transcription factor) levels low reduce the transcription of the cAMP-CRP regulon, which has been suggested to include transporters with affinity for aminoglycosides . Disruptions of both cyaA, which encodes adenylate cyclase, and crp were beneficial during the aminoglycoside selections, and a cyaA::kan mutant has a higher MIC than the wildtype in three of the four aminoglycosides tested (Table S5). (crp does not appear in Figure 4 due to lack of data in gentamycin and tobramycin; cyaA is not in Figure 4 because its deletion was generally advantageous.) Transposon insertions in ptsH, ptsI, and crr, which are expected to lower cAMP levels , were also advantageous. Additionally, synthesis of the large cAMP-CRP regulon, which is expected to occur when the glucose in the media is exhausted , may not be the optimal allocation of cellular resources during antibiotic challenge. Salmonella strains lacking either cyaA or crp are more resistant to a wide range of antibiotics , suggesting that similar phenomena may occur with other organisms and antibiotics.
Other beneficial disruptions seem to alter the timing and magnitude of the stringent response, a program E. coli uses to redirect energy from rRNA and tRNA transcription to the creation of amino acid biosynthesis enzymes in response to amino acid starvation . To accomplish the transition, Lon protease bound to polyphosphate degrades ribosomal proteins, freeing amino acids that can be incorporated into the needed enzymes , and guanosine tetraphosphate (ppGpp) and guanosine pentaphosphate (pppGpp) bind RNA polymerase, altering promoter selectivity . Transposon insertions near gpp or spoT, genes whose products affect the levels and ratio of ppGpp and pppGpp , are beneficial in aminoglycosides (Figure 4 and Table S5). Similarly, insertions near ppk, which encodes polyphosphate kinase , or lon are beneficial in tetracyclines (Table S6).
The advantageous character of cpxA and cpxP disruptions points to a role for the Cpx system, which helps E. coli respond to extracytoplasmic stress , in aminoglycoside susceptibility. At the core of the Cpx system are the CpxA histidine kinase, its cognate response regulator, CpxR, and CpxP, a periplasmic repressor of CpxA . Recent work indicates that the presence of the wild-type Cpx system increases the number of hydroxyl radicals, the final output of the Fenton reaction death pathway, possibly through crosstalk with the Arc system . Somewhat surprisingly, however, in the absence of CpxA, the cellular pool of CpxR is partially phosphorylated, and the Cpx pathway is active, not off . Furthermore, some mutations that constitutively activate the Cpx system cause pleiotrophic effects including amikacin  and kanamycin resistance , possibly through the increased expression of drug efflux pumps . Different levels of Cpx-Arc crosstalk between the various mutant and wild-type strains may account for these observations.
Mutations Affecting Susceptibility to β-lactams
Excluding loci with a general effect on antibiotic tolerance, disruption of 33 loci was beneficial in all three β-lactams tested, while disruption of 12 loci was deleterious (Figure 5A). Many of the beneficial disruptions turn on the Rcs signaling pathway (Figure 5B), which Laubacher and Ades  previously found to contribute to β-lactam tolerance. Laubacher and Ades also showed that the tolerance is not dependent on the Rcs system's role in increasing capsule synthesis ; they did not, however, evaluate the importance of the system's function in suppressing flagella-based motility . Since many other disruptions beneficial in β-lactams are in genes whose products participate in flagella assembly, we hypothesize that suppression of flagella synthesis is responsible for the beneficial effect of Rcs system activation.
To confirm that the lack of flagella confers an advantage in β-lactams, we focused on a fliN::kan mutant in the presence of ampicillin. In direct competition with the parental strain, the mutant has a small advantage in media without drug, likely due to the high energetic cost of motility  (Figure 2). With ampicillin, however, the advantage is much larger (Figure 2), possibly because energy is a more valuable commodity in antibiotic-stressed cells. Alternatively, since assembling a flagellum requires peptidoglycan hydrolysis  and β-lactams inhibit peptidoglycan transpeptidation, the combination of stresses may interact synergistically. Interestingly, the MIC of the filN::kan mutant as well as that of several other strains with increased fitness in ampicillin is within measurement error of that of the wild-type strain (data not shown), highlighting the ability of the selection method employed to identify mutations of small effect.
Disruptions expected to work synergistically with β-lactams to disrupt peptidoglycan integrity are particularly harmful (Figure 5A). For example, loss of either mltB or slt, which encode membrane-bound lytic murein transglycosylases , is deleterious. Disruptions of ampG, which encodes a transporter involved in recycling murein , or of ampC, which encodes a β-lactamase resistance protein , are also detrimental.
Disruption of folX or folM is Beneficial in Drugs that Inhibit Folate Biosynthesis
The only locus whose disruption was beneficial in both sulfamonomethoxine and trimethoprim, two drugs that inhibit key steps in folate metabolism (Figure 6), was folM, which encodes one of E. coli's two dihydrofolate reductases (DHFRs). E. coli's other DHFR, FolA, is inhibited by trimethoprim , and folA is essential in minimal media unless thyA is also knocked out and the media is supplemented with thymidine . In contrast, folM deletion strains have no major growth defects, and trimethoprim does not inhibit FolM . A folM::kan mutant has a 5-fold higher MIC in sulfamonomethoxine than the wild-type strain, but no detectable change in MIC in trimethoprim (Table S2). To assess the strain's fitness more sensitively, we subjected the folM::kan strain to direct competition with the wild-type strain and found that the folM::kan mutation is neutral without drug and beneficial in the presence of trimethoprim (Figure 2). That the effects in sulfamonomethoxine are stronger than those in trimethoprim is consistent with the original transposon enrichment experiments, which needed two and four days to find loci affecting susceptibility to sulfamonomethoxine and trimethoprim, respectively. The beneficial nature of folM deletions in sulfamonomethoxine and trimethoprim is surprising, as deleting folM would naively be expected to reduce the available amount of DHFR, making a bad situation worse. We hypothesize that E. coli may respond to a lack of FolM by increasing FolA levels, ameliorating the effects of sulfamonomethoxine and trimethoprim.
Another gene connected to folate biosynthesis whose disruption was beneficial during the trimethoprim enrichment is folX. Although the folX data slightly missed the significance thresholds for the sulfamonomethoxine enrichment, deleting folX gave a ∼2-fold increase in MIC in sulfamonomethoxine (Table S2). Like folM, a folX deletion did not change the MIC in trimethoprim, but the mutant strain did have a competitive advantage over the wild-type strain in trimethoprim (Figure 2). FolX catalyzes the conversion of 7,8-dihydroneopterin triphosphate to dihydromonapterin-triphosphate , which redirects 7,8-dihydroneopterin triphosphate away from the synthesis of tetrahydrofolate (Figure 6). Thus, folX mutations likely allow metabolic compensation  by increasing the flux of metabolites through the folate biosynthesis pathway. In fact, increased flux from enhanced p-aminobenzoate production is a common mechanism of sulfamonomethoxine resistance .
Loci Conferring a General Increase or Decrease in Antibiotic Susceptibility
Transposon insertions in or near 30 genes provided a significant change in fitness in at least three antibiotics with distinct targets (Figure 7). While some drugs of the same class, such as the aminoglycosides have similar fitness profiles, a mutant's behavior in the presence of a drug cannot generally be determined based solely on knowledge of the drug's mechanism of action. This is especially true for drugs of the same class, such as tetracycline and doxycycline, which have distinct chemical properties that restrict them to different routes of entry into the cell. In particular, the comparatively hydrophilic tetracycline passes through OmpF porins while the more hydrophobic doxycycline diffuses through the outer membrane .
Yellow (blue) indicates that transposon insertions in or near a gene were beneficial (deleterious). Black indicates no significant effect; gray indicates missing data. Antibiotics with the same target are written in the same color. Sulfamonomethoxine and trimethoprim inhibit different enzymes in the folic acid biosynthesis pathway; placing them in separate classes did not alter the results. Z-scores were calculated as described in Materials and Methods.
E. coli generates multiple barriers to protect itself from different classes of harmful, foreign compounds, and we found, as expected, that many of the loci responsible for general alterations in susceptibility encode enzymes that synthesize cell envelope components. In particular, perturbations to the negatively-charged lipopolysaccharides (LPS) and enterobacteria common antigen (ECA) that protect the outer membrane ,  were widely deleterious. Defective LPS is known to increase sensitivity to hydrophobic antibiotics and polycationic compounds such as aminoglycosides . ECA, on the other hand, is thought to provide protection against organic acids , but its role in antibiotic tolerance had not been well explored; Tamae et al.  did, however, report that loss of rffA increases susceptibility to gentamycin. Strains with defective ECA (i.e., wzxE::kan) and LPS (i.e., rfaG::kan) generally had lower MICs than wild-type in the same antibiotics in which they were depleted during the selections (Table S7). Direct competitions with a wild-type strain confirmed that a ΔwzxE mutation is neutral in media without drug and deleterious in both nalidixic acid and amikacin (Figure 2).
Rarely was the disruption of a locus beneficial in the presence of some drugs and deleterious in others. Notable exceptions are yrbB and yrbE, whose products belong to a system that prevents mislocalized phospholipids from accumulating in the outer membrane's outer leaflet . We confirmed that a yrbE::kan strain has a higher MIC than the wildtype in bleomycin, has similar tolerance to the wildtype in tetracycline, and is more susceptible than the wildtype to nalidixic acid, lomefloxacin, and doxycycline (Table S7). Loss of the Yrb system likely has little effect in tetracycline because tetracycline enters the cell using porins rather than passing directly through the outer membrane . Defects in the Yrb system likely reduce the negative charge on the outer membrane, which would decrease the permeability to positively charged bleomycin and increase the permeability towards the more neutral and negatively charged nalidixic acid, lomefloxacin, and doxycycline. The mechanism may be similar to how expression of the PmrA regulon, which makes the LPS less negative, increases E. coli's tolerance to the positively charged polymyxin B but also increases susceptibility to anionic detergents .
To increase antibiotic tolerance, bacteria often reduce the intracellular drug concentration by increasing the expression of efflux pumps that use either ATP or membrane potential to expel toxic agents . Not surprisingly, disruption of genes that control the levels of the AcrB/AcrA/TolC system, E. coli's main drug efflux pump , alters tolerance to multiple antibiotics. Disruption of the acrB gene is widely deleterious (Figure 7). Disruptions of phoP or rob, which encode transcriptional activators of the acrAB operon , , are also deleterious, while disruption of acrR, which encodes a transcriptional repressor of the acrAB operon , is beneficial. Disruption of no other drug pump was generally deleterious.
Disruption of E. coli's Trk potassium transport system was beneficial in a wide range of antibiotics. The low-affinity transporter contains TrkA, TrkE (SapD), and either TrkH or TrkG proteins , , and the Tn-enrichment experiment indicated that disruptions in sapD, trkA, and trkH are beneficial in piperacillin, doxycycline, tetracycline, and nalidixic acid (Figure 7). Work with a sapD::kan mutant indicated that removal of the system increases the MIC by about 1.5-fold in doxycycline, tetracycline, and nalidixic acid (Table S7). In direct competitions between a ΔsapD mutant and the parental strain, the ΔsapD mutation was slightly deleterious in the absence of drug, beneficial in the presence of nalidixic acid, and neutral with tetracycline (Figure 2). The connection between potassium transport and antibiotic tolerance merits further study.
Notably, disruptions of two genes of unknown function, yecR and yfgC, are deleterious (Figure 7). Tamae et al.  found that removal of yfgC decreases the MIC in vancomycin, rifampicin, and ampicillin. We confirmed the result for ampicillin, and we found that the MIC is also lower than wild-type in fusidic acid, doxycycline, and trimethoprim (Table S7). yfgC has homology to peptidases, and PSORTb  predicts that the protein is in the inner membrane. yecR is regulated by FlhDC  and has homology to lipoproteins. Strong homologs to yecR are found only in other Enterobacter species; homologs for yfgC are slightly more widespread, appearing in multiple gamma-proteobacteria.
Resistance through Accumulation of Mutations of Small Effect
To explore the potential for E. coli to acquire higher levels of antibiotic tolerance through the sequential accumulation of the identified chromosomal mutations, we constructed several double mutants. For each of two drugs, from among the genes that gave measurable MIC increases when removed singly, we chose pairs that were putatively in different pathways. As expected, the double mutants exhibited higher MICs than the parental strain and both of the two single mutant derivatives. In particular, ybjC::kan and ompR::kan mutants have MICs 2.25-fold greater than the wild-type strain in nitrofurantoin, while a ΔompR ybjC::kan double mutant has a MIC 5-fold greater than the wild-type strain. The ybjC::kan allele's beneficial effects likely come from reduced expression of the downstream nfsA gene, which encodes an oxygen-insensitive nitroreductase that converts nitrofurantoin into toxic intermediates . Similarly, MICs of sapD::kan and lon::kan mutants in tetracycline are 1.5-fold higher than the wild-type strain, and the MIC of a ΔsapD lon::kan mutant is 2.25-fold greater than the wild-type parent.
A wide variety of mutations, including single base pair changes, can cause the null phenotypes attained in this work through transposon insertions and gene replacements. Furthermore, for most antibiotics, the set of beneficial disruptions spans multiple pathways. Thus, E. coli's current genome is likely mutationally close to one conferring significantly higher antibiotic tolerance. A clinical S. aureus strain was observed acquiring 35 chromosomal mutations on the way to vancomycin resistance , and in the laboratory, multiple weak chromosomal mutations have been combined to give higher resistance in both P. aeruginosa  and Helicobacter pylori , suggesting that the phenomenon is quite general.
In this work, we competitively grew transposon insertion mutants of E. coli in batch cultures with drug concentrations that had a moderate impact on the parental strain's growth rate. Propagating the mutant collection for a sufficiently long duration allowed us to identify both beneficial and deleterious mutations of a wide range of strengths. Our analyses reveal that E. coli has a large mutational target size for altering its antibiotic tolerance.
As the disruption of the genes identified in this study pushes cells from the growth regime of moderate inhibition towards one of the extremes of no inhibition or full inhibition, the products of the loci and the pathways in which they reside are promising starting points for the development of adjuvant therapies. For example, if a gene's deletion causes hypersensitivity to an antibiotic, that antibiotic and a drug targeting the corresponding gene product may act synergistically. Similarly, when disrupting a pathway increases bacterial fitness in the presence of a particular antibiotic, stimulating the pathway might enhance the antibiotic's efficacy. The development of such adjuvant therapies has the potential to expand the usefulness of the limited set of antibiotics currently available.
With whole-genome sequencing becoming increasingly affordable, this work should provide a wealth of data for interpreting mutations present in drug resistant, pathogenic strains. As approximately half of the genes identified as altering fitness in the presence of antibiotics increase tolerance when disrupted, it will be important to learn how frequently and in what combinations the adaptive building blocks revealed here appear in clinical and environmental settings. The bulk of the loci identified occur in multiple species, and future work will be needed to discover how specific the beneficial and deleterious nature of each perturbation is to the wiring of E. coli's cellular network. Our observations should provide a scaffold for understanding the contribution of chromosomal mutations to antibiotic resistance as well as an aid in the development of novel therapeutics.
Materials and Methods
Bacterial Strains and Growth Conditions
All experiments were performed using E. coli MG1655 . Transposon insertion mutants were generated in a MG1655 ΔlacZ strain as described in a previous study . All experiments were conducted in M9 salts  supplemented with 0.4% glucose, 0.1% casamino acids, 1 mM MgSO4, 0.1 mM CaCl2, and 1.5 μM thiamine. LB media contained 0.1% Bacto Tryptone, 0.05% yeast extract, and 0.05% NaCl. All antibiotics were purchased from Sigma. Unless otherwise noted, cultures were shaken at 37°C.
Transposon Library Enrichments, DNA manipulation, and Hybridization
Genetic footprinting and subsequent hybridization to DNA spotted arrays were performed as described in Girgis et al. . As a starting point, we hybridized DNA from the day in which minor banding patterns began to emerge (Figure 1C) and then adjusted the chosen day as necessary. On early days during a selection, a mutant's fitness did not have a measurable effect on its prevalence, while on very late days, only a few types of mutants remained, and the relative fitness of the mutants that completely dropped out could not be discerned. In a few cases (Table 1), data from adjacent days of roughly equal suitability were included in the analysis. Other hybridizations from days not ultimately chosen exhibited either extreme selection or little to no selection and were excluded. The distinct behavior of the library in each antibiotic necessitated the choice of different days for different antibiotics (Table 1). To reduce the chance of spontaneous mutations overtaking the cultures, to remove the need for additional sets of controls for comparison, and to focus on transposon insertions causing larger effects, no samples from days 5–7 were chosen. Samples from at least two independent replicate selections were hybridized for each antibiotic. As controls, six samples from independent selections in the absence of any drug were hybridized.
Determining Significant Changes
Ratios (transposon signal/genomic DNA signal) from the antibiotic enrichments were compared to both the ratios from the original unselected library and to ratios from enrichments of the transposon library performed in identical media without antibiotics. Two z-scores were calculated for each ratio, r, where z = (x−μ)/σ, x = log2(r), and μ and σ are the mean and standard deviation, respectively, of the log2 ratios for the gene from reference hybridizations. One z-score used five reference hybridizations of the unselected library (from Girgis et al. ) and the other used six reference hybridizations of the library selected in the same media without antibiotics. All six no-antibiotic samples came from independent selections; three selections lasted two days, and three lasted four days.
To identify the most reproducible fitness effects, we considered all of the z-scores for each gene for a given antibiotic. (Antibiotics with two and three hybridizations had four and six z-scores, respectively.) When all of the z-scores had the same sign, we assigned the gene the z-score in the set that was closest to zero (representing the smallest effect). When a gene had z-scores of different signs, the gene was assigned a score of 0, indicating no consistent fitness effect. Supplementary information contains normalized ratios (Dataset S2), z-scores relative to the unselected library (Dataset S3), z-scores relative to the enrichments performed in the media without antibiotics (Dataset S4), combined z-scores (Dataset S5), and the combined z-scores considered significant (Dataset S1).
The significance threshold was set so that two false positives are expected per antibiotic. False positives were estimated by treating randomly chosen reference samples as data and repeating the analysis procedure. (See Text S1.)
Determining Genes Common to Aminoglycosides and β-lactams
In identifying loci important to fitness in specific antibiotic classes, care was taken to prevent the exclusion of genes that barely missed the significance cutoff for a subset of the drugs. As such, a locus was considered to be beneficial or deleterious in aminoglycosides if i) the prevalence of mutants where the locus was disrupted changed significantly during the enrichments for at least two of the four drugs and ii) the z-scores for the locus for all four drugs had the same sign. For example, disruption of a locus was classified as beneficial in aminoglycosides if the gene had positive z-scores in all four drugs, and the z-scores reached the significance level for at least two drugs. β-lactams were treated similarly except that the disruption of a locus was required to cause a significant fitness change during the enrichments for at least one of three drugs. Loci with a general effect on antibiotic tolerance were excluded from the sets.
Rather than choosing one of the many mutants in the library with a transposon inserted in a particular gene, we corroborated behavior observed during the selections using strains where the gene of interest had either been replaced with a kanamycin (kan) resistance cassette or removed to create an in-frame deletion. To construct the strains, P1vir transduction  was used to move the necessary alleles from the Keio collection  to MG1655 . To create unmarked, in-frame deletions, the kan cassette was removed using FLP recombinase . In rare cases, both the original transposon insertions as well as the kanamycin resistance cassette can produce polar effects, resulting in mutants with phenotypes distinct from the null phenotype of the disrupted or replaced gene.
Data was clustered with Cluster  and visualized using Treeview . Data manipulations were performed using Perl and Matlab. iPAGE (Hani Goodarzi, unpublished data) was used to examine sets of genes for enrichments in GO category, transcription factor regulon, and stress response membership. Annotations came from EcoCyc  and genome-tools .
Additional Materials and Methods
(0.07 MB PDF)
Dose Response Curves Used to Select Drug Concentrations. For each antibiotic, fresh media containing various drug concentrations was inoculated with overnight culture of the wild-type strain. Cultures were shaken at 37°C, and growth was monitored using OD600 readings. Blue indicates the concentrations chosen for the enrichments.
(0.18 MB PDF)
Gel images from enrichments done in the study media in the absence of antibiotics. Shown are the amplified Tn-adjacent DNA from all seven days for each of the seven repetitions. DNA was amplified as described in Girgis et al.  and separated on a 2% agarose gel. Yellow rectangles indicate samples hybridized. From the bottom, marker sizes are 100, 200, 300, 400, 500, 650, 850, and 1000 bases.
(1.48 MB PDF)
Gel images from Tn-insertion library enrichments done in the presence of antibiotics. Shown are the amplified Tn-adjacent DNA from all seven days for each of the three repetitions done for each antibiotic. DNA was amplified as described in Girgis et al.  and separated on a 2% agarose gel. Yellow rectangles indicate samples hybridized. From the bottom, marker sizes are 100, 200, 300, 400, 500, 650, 850, and 1000 bases.
(2.10 MB PDF)
Loci whose disruption was significant in at least one quinolone. Yellow (blue) indicates that transposon insertions in or near a gene were beneficial (deleterious). Black indicates no significant effect. Z-scores were calculated as described in Materials and Methods.
(0.21 MB PDF)
Loci whose disruption was significant in at least one tetracycline. Yellow (blue) indicates that transposon insertions in or near a gene were beneficial (deleterious). Black indicates no significant effect; gray indicates missing data. Z-scores were calculated as described in Materials and Methods.
(0.24 MB PDF)
Loci whose disruption was significant in at least one folic acid biosynthesis inhibitor. Yellow (blue) indicates that transposon insertions in or near a gene were beneficial (deleterious). Black indicates no significant effect; gray indicates missing data. Z-scores were calculated as described in Materials and Methods.
(0.13 MB PDF)
Loci whose disruption was significant in at least one inhibitor of the 50S subunit of the ribosome. Yellow (blue) indicates that transposon insertions in or near a gene were beneficial (deleterious). Black indicates no significant effect. Z-scores were calculated as described in Materials and Methods.
(0.08 MB PDF)
Loci whose disruption was significant in bleomycin. Yellow (blue) indicates that transposon insertions in or near a gene were beneficial (deleterious). Z-scores were calculated as described in Materials and Methods.
(0.13 MB PDF)
Loci whose disruption was significant in at least one β-lactam. Yellow (blue) indicates that transposon insertions in or near a gene were beneficial (deleterious). Black indicates no significant effect. Z-scores were calculated as described in Materials and Methods. Note that this set of loci is distinct from the set of loci whose disruption caused significant changes in all the beta-lactams tested (Table S2).
(0.10 MB PDF)
Loci whose disruption was significant in nitrofurantoin. Yellow (blue) indicates that transposon insertions in or near a gene were beneficial (deleterious). Z-scores were calculated as described in Methods.
(0.09 MB PDF)
Loci whose disruption was significant in at least one aminoglycoside. Due to the large size of the set, genes whose disruption was only significant in tobramycin are not shown. Data for tobramycin is available in Dataset S1. Yellow (blue) indicates that transposon insertions in or near a gene were beneficial (deleterious). Black indicates no significant effect; gray indicates missing data.
(0.24 MB PDF)
Loci that changed susceptibility to all aminoglycosides tested.
(0.08 MB PDF)
Loci that changed susceptibility to all beta-lactams tested.
(0.07 MB PDF)
Genes identified in this work as having a general role in antibiotic susceptibility.
(0.07 MB PDF)
Additional genes identified in both this study and previous work.
(0.07 MB PDF)
MIC changes in aminoglycosides.
(0.06 MB PDF)
Additional class-specific MIC changes (non-aminoglycosides).
(0.07 MB PDF)
MIC changes for mutants with altered susceptibility to multiple drug classes.
(0.07 MB PDF)
Z-scores for loci with a significant effect on antibiotic susceptibility.
(0.90 MB XLS)
Normalized ratios (transposon signal/genomic DNA signal)
(4.23 MB XLS)
Z-scores for individual hybridization computed relative to five hybridizations of the original, unselected library.
(3.58 MB XLS)
Z-scores for individual hybridizations computed relative to six hybridization of the library cultured in the same media (M9 with glucose and casamino acids) without antibiotics.
(3.58 MB XLS)
We thank members of the Tavazoie lab for critical reading of the manuscript and Hani Goodarzi for use of unpublished software. We thank Juliana C. Malinverni and Thomas J. Silhavy for helpful discussions and for sharing unpublished data.
Conceived and designed the experiments: HG ST. Performed the experiments: HG AKH. Analyzed the data: AKH. Wrote the paper: HG AKH ST.
- 1. Levy SB, Marshall B (2004) Antibacterial resistance worldwide: causes, challenges and responses. Nat Med 10: S122–129.SB LevyB. Marshall2004Antibacterial resistance worldwide: causes, challenges and responses.Nat Med10S122129
- 2. Walsh C (2003) Where will new antibiotics come from? Nat Rev Microbiol 1: 65–70.C. Walsh2003Where will new antibiotics come from?Nat Rev Microbiol16570
- 3. Lenski RE, Riley MA (2002) Chemical warfare from an ecological perspective. Proc Natl Acad Sci U S A 99: 556–558.RE LenskiMA Riley2002Chemical warfare from an ecological perspective.Proc Natl Acad Sci U S A99556558
- 4. Drlica K, Zhao X (2007) Mutant selection window hypothesis updated. Clin Infect Dis 44: 681–688.K. DrlicaX. Zhao2007Mutant selection window hypothesis updated.Clin Infect Dis44681688
- 5. Davies J, Spiegelman GB, Yim G (2006) The world of subinhibitory antibiotic concentrations. Curr Opin Microbiol 9: 445–453.J. DaviesGB SpiegelmanG. Yim2006The world of subinhibitory antibiotic concentrations.Curr Opin Microbiol9445453
- 6. Rybak MJ (2006) Pharmacodynamics: relation to antimicrobial resistance. Am J Med 119: S37–44; discussion S62–70.MJ Rybak2006Pharmacodynamics: relation to antimicrobial resistance.Am J Med119S3744; discussion S62–70
- 7. D'Costa VM, McGrann KM, Hughes DW, Wright GD (2006) Sampling the antibiotic resistome. Science 311: 374–377.VM D'CostaKM McGrannDW HughesGD Wright2006Sampling the antibiotic resistome.Science311374377
- 8. Fajardo A, Martinez JL (2008) Antibiotics as signals that trigger specific bacterial responses. Curr Opin Microbiol 11: 161–167.A. FajardoJL Martinez2008Antibiotics as signals that trigger specific bacterial responses.Curr Opin Microbiol11161167
- 9. Thiele-Bruhn S (2003) Pharmaceutical antibiotic compounds in soils - a review. J Plant Nutr Soil Sci 166: 145–167.S. Thiele-Bruhn2003Pharmaceutical antibiotic compounds in soils - a review.J Plant Nutr Soil Sci166145167
- 10. Watkinson AJ, Murby EJ, Costanzo SD (2007) Removal of antibiotics in conventional and advanced wastewater treatment: Implications for environmental discharge and wastewater recycling. Water Res 41: 4164–4176.AJ WatkinsonEJ MurbySD Costanzo2007Removal of antibiotics in conventional and advanced wastewater treatment: Implications for environmental discharge and wastewater recycling.Water Res4141644176
- 11. Onan LJ, LaPara TM (2003) Tylosin-resistant bacteria cultivated from agricultural soil. FEMS Microbiol Lett 220: 15–20.LJ OnanTM LaPara2003Tylosin-resistant bacteria cultivated from agricultural soil.FEMS Microbiol Lett2201520
- 12. Schwartz T, Kohnen W, Jansen B, Obst U (2003) Detection of antibiotic-resistant bacteria and their resistance genes in wastewater, surface water, and drinking water biofilms. FEMS Microbiol Ecol 43: 325–335.T. SchwartzW. KohnenB. JansenU. Obst2003Detection of antibiotic-resistant bacteria and their resistance genes in wastewater, surface water, and drinking water biofilms.FEMS Microbiol Ecol43325335
- 13. McKeegan KS, Borges-Walmsley MI, Walmsley AR (2002) Microbial and viral drug resistance mechanisms. Trends Microbiol 10: S8–14.KS McKeeganMI Borges-WalmsleyAR Walmsley2002Microbial and viral drug resistance mechanisms.Trends Microbiol10S814
- 14. Walsh C (2000) Molecular mechanisms that confer antibacterial drug resistance. Nature 406: 775–781.C. Walsh2000Molecular mechanisms that confer antibacterial drug resistance.Nature406775781
- 15. Wright GD (2003) Mechanisms of resistance to antibiotics. Curr Opin Chem Biol 7: 563–569.GD Wright2003Mechanisms of resistance to antibiotics.Curr Opin Chem Biol7563569
- 16. Deurenberg RH, Vink C, Kalenic S, Friedrich AW, Bruggeman CA, et al. (2007) The molecular evolution of methicillin-resistant Staphylococcus aureus. Clin Microbiol Infect 13: 222–235.RH DeurenbergC. VinkS. KalenicAW FriedrichCA Bruggeman2007The molecular evolution of methicillin-resistant Staphylococcus aureus.Clin Microbiol Infect13222235
- 17. Chenia HY, Pillay B, Pillay D (2006) Analysis of the mechanisms of fluoroquinolone resistance in urinary tract pathogens. J Antimicrob Chemother 58: 1274–1278.HY CheniaB. PillayD. Pillay2006Analysis of the mechanisms of fluoroquinolone resistance in urinary tract pathogens.J Antimicrob Chemother5812741278
- 18. Gomez MJ, Neyfakh AA (2006) Genes involved in intrinsic antibiotic resistance of Acinetobacter baylyi. Antimicrob Agents Chemother 50: 3562–3567.MJ GomezAA Neyfakh2006Genes involved in intrinsic antibiotic resistance of Acinetobacter baylyi.Antimicrob Agents Chemother5035623567
- 19. Fajardo A, Martinez-Martin N, Mercadillo M, Galan JC, Ghysels B, et al. (2008) The neglected intrinsic resistome of bacterial pathogens. PLoS ONE 3: e1619.A. FajardoN. Martinez-MartinM. MercadilloJC GalanB. Ghysels2008The neglected intrinsic resistome of bacterial pathogens.PLoS ONE3e1619
- 20. Tamae C, Liu A, Kim K, Sitz D, Hong J, et al. (2008) Determination of antibiotic hypersensitivity among 4,000 single-gene-knockout mutants of Escherichia coli. J Bacteriol 190: 5981–5988.C. TamaeA. LiuK. KimD. SitzJ. Hong2008Determination of antibiotic hypersensitivity among 4,000 single-gene-knockout mutants of Escherichia coli.J Bacteriol19059815988
- 21. Breidenstein EB, Khaira BK, Wiegand I, Overhage J, Hancock RE (2008) Complex ciprofloxacin resistome revealed by screening a Pseudomonas aeruginosa mutant library for altered susceptibility. Antimicrob Agents Chemother 52: 4486–4491.EB BreidensteinBK KhairaI. WiegandJ. OverhageRE Hancock2008Complex ciprofloxacin resistome revealed by screening a Pseudomonas aeruginosa mutant library for altered susceptibility.Antimicrob Agents Chemother5244864491
- 22. Schurek KN, Marr AK, Taylor PK, Wiegand I, Semenec L, et al. (2008) Novel genetic determinants of low-level aminoglycoside resistance in Pseudomonas aeruginosa. Antimicrob Agents Chemother 52: 4213–4219.KN SchurekAK MarrPK TaylorI. WiegandL. Semenec2008Novel genetic determinants of low-level aminoglycoside resistance in Pseudomonas aeruginosa.Antimicrob Agents Chemother5242134219
- 23. Albert TJ, Dailidiene D, Dailide G, Norton JE, Kalia A, et al. (2005) Mutation discovery in bacterial genomes: metronidazole resistance in Helicobacter pylori. Nat Methods 2: 951–953.TJ AlbertD. DailidieneG. DailideJE NortonA. Kalia2005Mutation discovery in bacterial genomes: metronidazole resistance in Helicobacter pylori.Nat Methods2951953
- 24. El'Garch F, Jeannot K, Hocquet D, Llanes-Barakat C, Plesiat P (2007) Cumulative effects of several nonenzymatic mechanisms on the resistance of Pseudomonas aeruginosa to aminoglycosides. Antimicrob Agents Chemother 51: 1016–1021.F. El'GarchK. JeannotD. HocquetC. Llanes-BarakatP. Plesiat2007Cumulative effects of several nonenzymatic mechanisms on the resistance of Pseudomonas aeruginosa to aminoglycosides.Antimicrob Agents Chemother5110161021
- 25. Girgis HS, Liu Y, Ryu WS, Tavazoie S (2007) A comprehensive genetic characterization of bacterial motility. PLoS Genet 3: 1644–1660.HS GirgisY. LiuWS RyuS. Tavazoie2007A comprehensive genetic characterization of bacterial motility.PLoS Genet316441660
- 26. Scott GM, Kyi MS (2001) Handbook of Essential Antibiotics. Amsterdam: Harwood Academic. GM ScottMS Kyi2001Handbook of Essential AntibioticsAmsterdamHarwood Academic
- 27. Walsh C (2003) Antibiotics: Actions, Origins, Resistance. Washington, D.C.: American Society for Microbiology. C. Walsh2003Antibiotics: Actions, Origins, ResistanceWashington, D.C.American Society for Microbiology
- 28. Baquero F (2001) Low-level antibacterial resistance: a gateway to clinical resistance. Drug Resist Updat 4: 93–105.F. Baquero2001Low-level antibacterial resistance: a gateway to clinical resistance.Drug Resist Updat493105
- 29. Kishony R, Leibler S (2003) Environmental stresses can alleviate the average deleterious effect of mutations. J Biol 2: 14.R. KishonyS. Leibler2003Environmental stresses can alleviate the average deleterious effect of mutations.J Biol214
- 30. Leclercq R, Cantón R, Giske C, Heisig P, Livermore D, et al. (2008) R. LeclercqR. CantónC. GiskeP. HeisigD. Livermore2008Expert rules in antimicrobial susceptibility testing. European Committee on Antimicrobial Susceptibility Testing. Expert rules in antimicrobial susceptibility testing. European Committee on Antimicrobial Susceptibility Testing.
- 31. Williams GC (1957) Pleiotropy, Natural Selection, and the Evolution of Senescence. Evolution 11: 398–411.GC Williams1957Pleiotropy, Natural Selection, and the Evolution of Senescence.Evolution11398411
- 32. Verklin RM Jr, Mandell GL (1977) Alteration of effectiveness of antibiotics by anaerobiosis. J Lab Clin Med 89: 65–71.RM Verklin JrGL Mandell1977Alteration of effectiveness of antibiotics by anaerobiosis.J Lab Clin Med896571
- 33. Kohanski MA, Dwyer DJ, Hayete B, Lawrence CA, Collins JJ (2007) A common mechanism of cellular death induced by bactericidal antibiotics. Cell 130: 797–810.MA KohanskiDJ DwyerB. HayeteCA LawrenceJJ Collins2007A common mechanism of cellular death induced by bactericidal antibiotics.Cell130797810
- 34. Kohanski MA, Dwyer DJ, Wierzbowski J, Cottarel G, Collins JJ (2008) Mistranslation of membrane proteins and two-component system activation trigger antibiotic-mediated cell death. Cell 135: 679–690.MA KohanskiDJ DwyerJ. WierzbowskiG. CottarelJJ Collins2008Mistranslation of membrane proteins and two-component system activation trigger antibiotic-mediated cell death.Cell135679690
- 35. Taber HW, Mueller JP, Miller PF, Arrow AS (1987) Bacterial uptake of aminoglycoside antibiotics. Microbiol Rev 51: 439–457.HW TaberJP MuellerPF MillerAS Arrow1987Bacterial uptake of aminoglycoside antibiotics.Microbiol Rev51439457
- 36. Holtje JV (1978) Streptomycin uptake via an inducible polyamine transport system in Escherichia coli. Eur J Biochem 86: 345–351.JV Holtje1978Streptomycin uptake via an inducible polyamine transport system in Escherichia coli.Eur J Biochem86345351
- 37. Plumbridge J (2002) Regulation of gene expression in the PTS in Escherichia coli: the role and interactions of Mlc. Curr Opin Microbiol 5: 187–193.J. Plumbridge2002Regulation of gene expression in the PTS in Escherichia coli: the role and interactions of Mlc.Curr Opin Microbiol5187193
- 38. Deutscher J (2008) The mechanisms of carbon catabolite repression in bacteria. Curr Opin Microbiol 11: 87–93.J. Deutscher2008The mechanisms of carbon catabolite repression in bacteria.Curr Opin Microbiol118793
- 39. Alper MD, Ames BN (1978) Transport of antibiotics and metabolite analogs by systems under cyclic AMP control: positive selection of Salmonella typhimurium cya and crp mutants. J Bacteriol 133: 149–157.MD AlperBN Ames1978Transport of antibiotics and metabolite analogs by systems under cyclic AMP control: positive selection of Salmonella typhimurium cya and crp mutants.J Bacteriol133149157
- 40. Srivatsan A, Wang JD (2008) Control of bacterial transcription, translation and replication by (p)ppGpp. Curr Opin Microbiol 11: 100–105.A. SrivatsanJD Wang2008Control of bacterial transcription, translation and replication by (p)ppGpp.Curr Opin Microbiol11100105
- 41. Kuroda A, Nomura K, Ohtomo R, Kato J, Ikeda T, et al. (2001) Role of inorganic polyphosphate in promoting ribosomal protein degradation by the Lon protease in E. coli. Science 293: 705–708.A. KurodaK. NomuraR. OhtomoJ. KatoT. Ikeda2001Role of inorganic polyphosphate in promoting ribosomal protein degradation by the Lon protease in E. coli.Science293705708
- 42. Jishage M, Kvint K, Shingler V, Nystrom T (2002) Regulation of sigma factor competition by the alarmone ppGpp. Genes Dev 16: 1260–1270.M. JishageK. KvintV. ShinglerT. Nystrom2002Regulation of sigma factor competition by the alarmone ppGpp.Genes Dev1612601270
- 43. Cashel M, Gentry DR, Hernandez VJ, Vinella D (1996) The stringent response. In: Neidhardt FC, III RC, Ingraham JL, Lin ECC, Low KB, et al., editors. Escherichia coli and Salmonella: Cellular and Molecular Biology. 2nd ed. Washington, D.C.: ASM Press. pp. 1458–1496.M. CashelDR GentryVJ HernandezD. Vinella1996The stringent response.FC NeidhardtRC IIIJL IngrahamECC LinKB LowEscherichia coli and Salmonella: Cellular and Molecular Biology. 2nd edWashington, D.C.ASM Press14581496
- 44. Akiyama M, Crooke E, Kornberg A (1992) The polyphosphate kinase gene of Escherichia coli. Isolation and sequence of the ppk gene and membrane location of the protein. J Biol Chem 267: 22556–22561.M. AkiyamaE. CrookeA. Kornberg1992The polyphosphate kinase gene of Escherichia coli. Isolation and sequence of the ppk gene and membrane location of the protein.J Biol Chem2672255622561
- 45. Ruiz N, Silhavy TJ (2005) Sensing external stress: watchdogs of the Escherichia coli cell envelope. Curr Opin Microbiol 8: 122–126.N. RuizTJ Silhavy2005Sensing external stress: watchdogs of the Escherichia coli cell envelope.Curr Opin Microbiol8122126
- 46. Wolfe AJ, Parikh N, Lima BP, Zemaitaitis B (2008) Signal integration by the two-component signal transduction response regulator CpxR. J Bacteriol 190: 2314–2322.AJ WolfeN. ParikhBP LimaB. Zemaitaitis2008Signal integration by the two-component signal transduction response regulator CpxR.J Bacteriol19023142322
- 47. Rainwater S, Silverman PM (1990) The Cpx proteins of Escherichia coli K-12: evidence that cpxA, ecfB, ssd, and eup mutations all identify the same gene. J Bacteriol 172: 2456–2461.S. RainwaterPM Silverman1990The Cpx proteins of Escherichia coli K-12: evidence that cpxA, ecfB, ssd, and eup mutations all identify the same gene.J Bacteriol17224562461
- 48. Newman EB, Malik N, Walker C (1982) L-serine degradation in Escherichia coli K-12: directly isolated ssd mutants and their intragenic revertants. J Bacteriol 150: 710–715.EB NewmanN. MalikC. Walker1982L-serine degradation in Escherichia coli K-12: directly isolated ssd mutants and their intragenic revertants.J Bacteriol150710715
- 49. Hirakawa H, Nishino K, Hirata T, Yamaguchi A (2003) Comprehensive studies of drug resistance mediated by overexpression of response regulators of two-component signal transduction systems in Escherichia coli. J Bacteriol 185: 1851–1856.H. HirakawaK. NishinoT. HirataA. Yamaguchi2003Comprehensive studies of drug resistance mediated by overexpression of response regulators of two-component signal transduction systems in Escherichia coli.J Bacteriol18518511856
- 50. Laubacher ME, Ades SE (2008) The Rcs phosphorelay is a cell envelope stress response activated by peptidoglycan stress and contributes to intrinsic antibiotic resistance. J Bacteriol 190: 2065–2074.ME LaubacherSE Ades2008The Rcs phosphorelay is a cell envelope stress response activated by peptidoglycan stress and contributes to intrinsic antibiotic resistance.J Bacteriol19020652074
- 51. Francez-Charlot A, Laugel B, Van Gemert A, Dubarry N, Wiorowski F, et al. (2003) RcsCDB His-Asp phosphorelay system negatively regulates the flhDC operon in Escherichia coli. Mol Microbiol 49: 823–832.A. Francez-CharlotB. LaugelA. Van GemertN. DubarryF. Wiorowski2003RcsCDB His-Asp phosphorelay system negatively regulates the flhDC operon in Escherichia coli.Mol Microbiol49823832
- 52. Macnab RM (1996) Flagella and motility. In: Neidhardt FC, Curtiss R, Ingraham JL, Lin ECC, Low KB, et al., editors. Escherichia coli and Salmonella: Cellular and molecular biology. 2nd ed. Washington, D.C.: American Society for Microbiology Press. pp. 123–145.RM Macnab1996Flagella and motility.FC NeidhardtR. CurtissJL IngrahamECC LinKB LowEscherichia coli and Salmonella: Cellular and molecular biology. 2nd edWashington, D.C.American Society for Microbiology Press123145
- 53. Nambu T, Minamino T, Macnab RM, Kutsukake K (1999) Peptidoglycan-hydrolyzing activity of the FlgJ protein, essential for flagellar rod formation in Salmonella typhimurium. J Bacteriol 181: 1555–1561.T. NambuT. MinaminoRM MacnabK. Kutsukake1999Peptidoglycan-hydrolyzing activity of the FlgJ protein, essential for flagellar rod formation in Salmonella typhimurium.J Bacteriol18115551561
- 54. von Rechenberg M, Ursinus A, Holtje JV (1996) Affinity chromatography as a means to study multienzyme complexes involved in murein synthesis. Microb Drug Resist 2: 155–157.M. von RechenbergA. UrsinusJV Holtje1996Affinity chromatography as a means to study multienzyme complexes involved in murein synthesis.Microb Drug Resist2155157
- 55. Jacobs C, Huang LJ, Bartowsky E, Normark S, Park JT (1994) Bacterial cell wall recycling provides cytosolic muropeptides as effectors for beta-lactamase induction. Embo J 13: 4684–4694.C. JacobsLJ HuangE. BartowskyS. NormarkJT Park1994Bacterial cell wall recycling provides cytosolic muropeptides as effectors for beta-lactamase induction.Embo J1346844694
- 56. Cayley PJ, Dunn SM, King RW (1981) Kinetics of substrate, coenzyme, and inhibitor binding to Escherichia coli dihydrofolate reductase. Biochemistry 20: 874–879.PJ CayleySM DunnRW King1981Kinetics of substrate, coenzyme, and inhibitor binding to Escherichia coli dihydrofolate reductase.Biochemistry20874879
- 57. Herrington MB, Chirwa NT (1999) Growth properties of a folA null mutant of Escherichia coli K12. Can J Microbiol 45: 191–200.MB HerringtonNT Chirwa1999Growth properties of a folA null mutant of Escherichia coli K12.Can J Microbiol45191200
- 58. Giladi M, Altman-Price N, Levin I, Levy L, Mevarech M (2003) FolM, a new chromosomally encoded dihydrofolate reductase in Escherichia coli. J Bacteriol 185: 7015–7018.M. GiladiN. Altman-PriceI. LevinL. LevyM. Mevarech2003FolM, a new chromosomally encoded dihydrofolate reductase in Escherichia coli.J Bacteriol18570157018
- 59. Haussmann C, Rohdich F, Lottspeich F, Eberhardt S, Scheuring J, et al. (1997) Dihydroneopterin triphosphate epimerase of Escherichia coli: purification, genetic cloning, and expression. J Bacteriol 179: 949–951.C. HaussmannF. RohdichF. LottspeichS. EberhardtJ. Scheuring1997Dihydroneopterin triphosphate epimerase of Escherichia coli: purification, genetic cloning, and expression.J Bacteriol179949951
- 60. Hartl DL, Dykhuizen DE (1984) The population genetics of Escherichia coli. Annu Rev Genet 18: 31–68.DL HartlDE Dykhuizen1984The population genetics of Escherichia coli.Annu Rev Genet183168
- 61. Then RL (1982) Mechanisms of resistance to trimethoprim, the sulfonamides, and trimethoprim-sulfamethoxazole. Rev Infect Dis 4: 261–269.RL Then1982Mechanisms of resistance to trimethoprim, the sulfonamides, and trimethoprim-sulfamethoxazole.Rev Infect Dis4261269
- 62. Thanassi DG, Suh GS, Nikaido H (1995) Role of outer membrane barrier in efflux-mediated tetracycline resistance of Escherichia coli. J Bacteriol 177: 998–1007.DG ThanassiGS SuhH. Nikaido1995Role of outer membrane barrier in efflux-mediated tetracycline resistance of Escherichia coli.J Bacteriol1779981007
- 63. Hancock RE (1984) Alterations in outer membrane permeability. Annu Rev Microbiol 38: 237–264.RE Hancock1984Alterations in outer membrane permeability.Annu Rev Microbiol38237264
- 64. Mannel D, Mayer H (1978) Isolation and chemical characterization of the enterobacterial common antigen. Eur J Biochem 86: 361–370.D. MannelH. Mayer1978Isolation and chemical characterization of the enterobacterial common antigen.Eur J Biochem86361370
- 65. Barua S, Yamashino T, Hasegawa T, Yokoyama K, Torii K, et al. (2002) Involvement of surface polysaccharides in the organic acid resistance of Shiga Toxin-producing Escherichia coli O157:H7. Mol Microbiol 43: 629–640.S. BaruaT. YamashinoT. HasegawaK. YokoyamaK. Torii2002Involvement of surface polysaccharides in the organic acid resistance of Shiga Toxin-producing Escherichia coli O157:H7.Mol Microbiol43629640
- 66. Malinverni JC, Silhavy TJ (2009) An ABC transport system that maintains lipid asymmetry in the Gram-negative outer membrane. Proc Natl Acad Sci U S A. JC MalinverniTJ Silhavy2009An ABC transport system that maintains lipid asymmetry in the Gram-negative outer membrane.Proc Natl Acad Sci U S AIn press. In press.
- 67. Froelich JM, Tran K, Wall D (2006) A pmrA constitutive mutant sensitizes Escherichia coli to deoxycholic acid. J Bacteriol 188: 1180–1183.JM FroelichK. TranD. Wall2006A pmrA constitutive mutant sensitizes Escherichia coli to deoxycholic acid.J Bacteriol18811801183
- 68. Lewis K (1994) Multidrug resistance pumps in bacteria: variations on a theme. Trends Biochem Sci 19: 119–123.K. Lewis1994Multidrug resistance pumps in bacteria: variations on a theme.Trends Biochem Sci19119123
- 69. Okusu H, Ma D, Nikaido H (1996) AcrAB efflux pump plays a major role in the antibiotic resistance phenotype of Escherichia coli multiple-antibiotic-resistance (Mar) mutants. J Bacteriol 178: 306–308.H. OkusuD. MaH. Nikaido1996AcrAB efflux pump plays a major role in the antibiotic resistance phenotype of Escherichia coli multiple-antibiotic-resistance (Mar) mutants.J Bacteriol178306308
- 70. Minagawa S, Ogasawara H, Kato A, Yamamoto K, Eguchi Y, et al. (2003) Identification and molecular characterization of the Mg2+ stimulon of Escherichia coli. J Bacteriol 185: 3696–3702.S. MinagawaH. OgasawaraA. KatoK. YamamotoY. Eguchi2003Identification and molecular characterization of the Mg2+ stimulon of Escherichia coli.J Bacteriol18536963702
- 71. Rosenberg EY, Bertenthal D, Nilles ML, Bertrand KP, Nikaido H (2003) Bile salts and fatty acids induce the expression of Escherichia coli AcrAB multidrug efflux pump through their interaction with Rob regulatory protein. Mol Microbiol 48: 1609–1619.EY RosenbergD. BertenthalML NillesKP BertrandH. Nikaido2003Bile salts and fatty acids induce the expression of Escherichia coli AcrAB multidrug efflux pump through their interaction with Rob regulatory protein.Mol Microbiol4816091619
- 72. Ma D, Alberti M, Lynch C, Nikaido H, Hearst JE (1996) The local repressor AcrR plays a modulating role in the regulation of acrAB genes of Escherichia coli by global stress signals. Mol Microbiol 19: 101–112.D. MaM. AlbertiC. LynchH. NikaidoJE Hearst1996The local repressor AcrR plays a modulating role in the regulation of acrAB genes of Escherichia coli by global stress signals.Mol Microbiol19101112
- 73. Harms C, Domoto Y, Celik C, Rahe E, Stumpe S, et al. (2001) Identification of the ABC protein SapD as the subunit that confers ATP dependence to the K+-uptake systems TrkH and TrkG from Escherichia coli K-12. Microbiology 147: 2991–3003.C. HarmsY. DomotoC. CelikE. RaheS. Stumpe2001Identification of the ABC protein SapD as the subunit that confers ATP dependence to the K+-uptake systems TrkH and TrkG from Escherichia coli K-12.Microbiology14729913003
- 74. Schlösser A, Meldorf M, Stumpe S, Bakker EP, Epstein W (1995) TrkH and its homolog, TrkG, determine the specificity and kinetics of cation transport by the Trk system of Escherichia coli. J Bacteriol 177: 1908–1910.A. SchlösserM. MeldorfS. StumpeEP BakkerW. Epstein1995TrkH and its homolog, TrkG, determine the specificity and kinetics of cation transport by the Trk system of Escherichia coli.J Bacteriol17719081910
- 75. Gardy JL, Laird MR, Chen F, Rey S, Walsh CJ, et al. (2005) PSORTb v.2.0: expanded prediction of bacterial protein subcellular localization and insights gained from comparative proteome analysis. Bioinformatics 21: 617–623.JL GardyMR LairdF. ChenS. ReyCJ Walsh2005PSORTb v.2.0: expanded prediction of bacterial protein subcellular localization and insights gained from comparative proteome analysis.Bioinformatics21617623
- 76. Stafford GP, Ogi T, Hughes C (2005) Binding and transcriptional activation of non-flagellar genes by the Escherichia coli flagellar master regulator FlhD2C2. Microbiology 151: 1779–1788.GP StaffordT. OgiC. Hughes2005Binding and transcriptional activation of non-flagellar genes by the Escherichia coli flagellar master regulator FlhD2C2.Microbiology15117791788
- 77. Whiteway J, Koziarz P, Veall J, Sandhu N, Kumar P, et al. (1998) Oxygen-insensitive nitroreductases: analysis of the roles of nfsA and nfsB in development of resistance to 5-nitrofuran derivatives in Escherichia coli. J Bacteriol 180: 5529–5539.J. WhitewayP. KoziarzJ. VeallN. SandhuP. Kumar1998Oxygen-insensitive nitroreductases: analysis of the roles of nfsA and nfsB in development of resistance to 5-nitrofuran derivatives in Escherichia coli.J Bacteriol18055295539
- 78. Mwangi MM, Wu SW, Zhou Y, Sieradzki K, de Lencastre H, et al. (2007) Tracking the in vivo evolution of multidrug resistance in Staphylococcus aureus by whole-genome sequencing. Proc Natl Acad Sci U S A 104: 9451–9456.MM MwangiSW WuY. ZhouK. SieradzkiH. de Lencastre2007Tracking the in vivo evolution of multidrug resistance in Staphylococcus aureus by whole-genome sequencing.Proc Natl Acad Sci U S A10494519456
- 79. Blattner FR, Plunkett G 3rd, Bloch CA, Perna NT, Burland V, et al. (1997) The complete genome sequence of Escherichia coli K-12. Science 277: 1453–1474.FR BlattnerG. Plunkett 3rdCA BlochNT PernaV. Burland1997The complete genome sequence of Escherichia coli K-12.Science27714531474
- 80. Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, et al. (1994) Current protocols in molecular biology. New York, NY: Wiley Interscience. FM AusubelR. BrentRE KingstonDD MooreJG Seidman1994Current protocols in molecular biologyNew York, NYWiley Interscience
- 81. Silhavy TJ, Berman ML, Enquist LW (1984) Experiments with gene fusions. Plainview, NY: Cold Spring Harbor Press. TJ SilhavyML BermanLW Enquist1984Experiments with gene fusionsPlainview, NYCold Spring Harbor Press
- 82. Baba T, Ara T, Hasegawa M, Takai Y, Okumura Y, et al. (2006) Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol 2: 2006 0008.T. BabaT. AraM. HasegawaY. TakaiY. Okumura2006Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection.Mol Syst Biol22006 0008
- 83. Datsenko KA, Wanner BL (2000) One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci U S A 97: 6640–6645.KA DatsenkoBL Wanner2000One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products.Proc Natl Acad Sci U S A9766406645
- 84. Eisen MB, Spellman PT, Brown PO, Botstein D (1998) Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci U S A 95: 14863–14868.MB EisenPT SpellmanPO BrownD. Botstein1998Cluster analysis and display of genome-wide expression patterns.Proc Natl Acad Sci U S A951486314868
- 85. Karp PD, Keseler IM, Shearer A, Latendresse M, Krummenacker M, et al. (2007) Multidimensional annotation of the Escherichia coli K-12 genome. Nucleic Acids Res 35: 7577–7590.PD KarpIM KeselerA. ShearerM. LatendresseM. Krummenacker2007Multidimensional annotation of the Escherichia coli K-12 genome.Nucleic Acids Res3575777590
- 86. Lee W, Chen SL (2002) Genome-tools: a flexible package for genome sequence analysis. Biotechniques 33: 1334–1341.W. LeeSL Chen2002Genome-tools: a flexible package for genome sequence analysis.Biotechniques3313341341
- 87. Travisano M, Lenski RE (1996) Long-term experimental evolution in Escherichia coli. IV. Targets of selection and the specificity of adaptation. Genetics 143: 15–26.M. TravisanoRE Lenski1996Long-term experimental evolution in Escherichia coli. IV. Targets of selection and the specificity of adaptation.Genetics1431526
- 88. Johnson DC, Dean DR, Smith AD, Johnson MK (2005) Structure, function, and formation of biological iron-sulfur clusters. Annu Rev Biochem 74: 247–281.DC JohnsonDR DeanAD SmithMK Johnson2005Structure, function, and formation of biological iron-sulfur clusters.Annu Rev Biochem74247281
- 89. Majdalani N, Gottesman S (2005) The Rcs phosphorelay: a complex signal transduction system. Annu Rev Microbiol 59: 379–405.N. MajdalaniS. Gottesman2005The Rcs phosphorelay: a complex signal transduction system.Annu Rev Microbiol59379405
- 90. Majdalani N, Heck M, Stout V, Gottesman S (2005) Role of RcsF in signaling to the Rcs phosphorelay pathway in Escherichia coli. J Bacteriol 187: 6770–6778.N. MajdalaniM. HeckV. StoutS. Gottesman2005Role of RcsF in signaling to the Rcs phosphorelay pathway in Escherichia coli.J Bacteriol18767706778
- 91. Castanié-Cornet MP, Cam K, Jacq A (2006) RcsF is an outer membrane lipoprotein involved in the RcsCDB phosphorelay signaling pathway in Escherichia coli. J Bacteriol 188: 4264–4270.MP Castanié-CornetK. CamA. Jacq2006RcsF is an outer membrane lipoprotein involved in the RcsCDB phosphorelay signaling pathway in Escherichia coli.J Bacteriol18842644270
- 92. Gottesman S, Trisler P, Torres-Cabassa A (1985) Regulation of capsular polysaccharide synthesis in Escherichia coli K-12: characterization of three regulatory genes. J Bacteriol 162: 1111–1119.S. GottesmanP. TrislerA. Torres-Cabassa1985Regulation of capsular polysaccharide synthesis in Escherichia coli K-12: characterization of three regulatory genes.J Bacteriol16211111119
- 93. Torres-Cabassa AS, Gottesman S (1987) Capsule synthesis in Escherichia coli K-12 is regulated by proteolysis. J Bacteriol 169: 981–989.AS Torres-CabassaS. Gottesman1987Capsule synthesis in Escherichia coli K-12 is regulated by proteolysis.J Bacteriol169981989
- 94. Majdalani N, Hernandez D, Gottesman S (2002) Regulation and mode of action of the second small RNA activator of RpoS translation, RprA. Mol Microbiol 46: 813–826.N. MajdalaniD. HernandezS. Gottesman2002Regulation and mode of action of the second small RNA activator of RpoS translation, RprA.Mol Microbiol46813826
- 95. Patel OG, Mberu EK, Nzila AM, Macreadie IG (2004) Sulfa drugs strike more than once. Trends Parasitol 20: 1–3.OG PatelEK MberuAM NzilaIG Macreadie2004Sulfa drugs strike more than once.Trends Parasitol2013