#PAGE_PARAMS# #ADS_HEAD_SCRIPTS# #MICRODATA#

Global mistranslation increases cell survival under stress in Escherichia coli


Authors: Laasya Samhita aff001;  Parth K. Raval aff001;  Deepa Agashe aff001
Authors place of work: National Centre for Biological Sciences, Tata Institute of Fundamental research, Bangalore, India aff001
Published in the journal: Global mistranslation increases cell survival under stress in Escherichia coli. PLoS Genet 16(3): e32767. doi:10.1371/journal.pgen.1008654
Category: Research Article
doi: https://doi.org/10.1371/journal.pgen.1008654

Summary

Mistranslation is typically deleterious for cells, although specific mistranslated proteins can confer a short-term benefit in a particular environment. However, given its large overall cost, the prevalence of high global mistranslation rates remains puzzling. Altering basal mistranslation levels of Escherichia coli in several ways, we show that generalized mistranslation enhances early survival under DNA damage, by rapidly activating the SOS response. Mistranslating cells maintain larger populations after exposure to DNA damage, and thus have a higher probability of sampling critical beneficial mutations. Both basal and artificially increased mistranslation increase the number of cells that are phenotypically tolerant and genetically resistant under DNA damage; they also enhance survival at high temperature. In contrast, decreasing the normal basal mistranslation rate reduces cell survival. This wide-ranging stress resistance relies on Lon protease, which is revealed as a key effector that induces the SOS response in addition to alleviating proteotoxic stress. The new links between error-prone protein synthesis, DNA damage, and generalised stress resistance indicate surprising coordination between intracellular stress responses and suggest a novel hypothesis to explain high global mistranslation rates.

Keywords:

Point mutation – Proteases – Mutagenesis – Cellular stress responses – DNA damage – DNA repair – Transfer RNA

Introduction

The rate of protein mistranslation is amongst the highest known error rates in cellular biosynthetic processes, ranging from 1 in 10,000 to 1 in 100 mis-incorporated amino acids in E.coli [1,2]. As a result, 10 to 15% of all proteins in an actively growing E.coli cell are likely to carry at least one mis-incorporated amino acid [3,4], implying a high tolerance for mistakes. This is puzzling because mistranslation is thought to be deleterious, and cells have evolved several proofreading mechanisms to minimise error [reviewed in 5]. Multiple hypotheses may explain the occurrence of high mistranslation rates. First, high error rates may occur as an unavoidable tradeoff: fast growth requires rapid translation, and proofreading would increase translation times [68]. However, other evidence contradicts this hypothesis. For instance, across natural isolates of E.coli, mis-incorporation of leucine from a poly-U mRNA chain is not correlated with translation and growth rates [9]. Second, mechanisms that increase robustness (e.g. folding via chaperones and acquiring stabilizing amino acid changes) may minimize the phenotypic impact of mistranslated proteins, weakening selection against mistranslation and allowing high error rates to persist [10,11].

The above hypotheses refer to the generation and maintenance of baseline mistranslation rates. In addition, mistranslation can also increase transiently (over and above baseline levels) under stress. Indeed, a body of work shows that cells elevate mistranslation levels under specific stresses (stress-induced mistranslation) [reviewed in 12,13]. In some cases, this elevated mistranslation is advantageous, because it produces specific mistranslated proteins that are beneficial. For instance, in Mycobacterium smegmatis, artificially increasing specific amino acid substitutions at glutamate and aspartate tRNAs generates a mixed population of wild type and mistranslated RNA polymerase molecules [14], imparting greater resistance to rifampicin (an antibiotic that targets RNA polymerase). More generally, high global (rather than protein-specific) mistranslation rates could confer a selective advantage by generating a “statistical proteome”–a bet-hedging strategy whereby a few cells with some mistranslated proteins (rather than all cells with one specific mistranslated protein as in the previous example) can survive an environmental stress [15,16]. Such a mixed proteome can occur both as a consequence of baseline mistranslation (as shown by proteome analyses [1]) or through artificial or stress-induced mistranslation. One example of general, stress-induced proteome-wide beneficial mistranslation comes from work on mis-methionylation in E. coli. In anaerobic environments or upon exposure to low concentrations of the antibiotic chloramphenicol, the methionyl tRNA synthetase enzyme (responsible for adding methionine onto target tRNAs) loses its succinyl modifications, increasing survival [17].

Such examples of beneficial induced mistranslation have motivated the idea that a transient increase in mistranslation rates can evolve under positive selection. However, across natural E. coli isolates, even baseline mistranslation rates vary at least 10-fold [9], suggesting that error rates do not converge to a single optimum. Could this variability reflect divergent selection for a high level of specific mistranslated proteins, causing increased basal mistranslation rates as a by-product? Mistranslation is typically deleterious due to protein misfolding and subsequent loss of function [reviewed in 13,18]. Thus, the potential benefit of a few specific mistranslated proteins may only rarely outweigh the larger overall cost of global mistranslation. As such, higher baseline mistranslation rates may be selectively favoured only when specific severe stresses are encountered very frequently. It remains unclear whether high basal and non-specific mistranslation rates can evolve under positive selection.

Here, we propose a fresh hypothesis for the persistence of high baseline mistranslation rates, that bypasses the need for specific mistranslated proteins and invokes a fitness benefit of global mistranslation under stress. Our hypothesis relies on prior work showing that specific kinds of mistranslation activate cellular stress responses. Such stress responses are beneficial for immediate survival in a hostile environment, but are energetically costly; and hence are not constitutively active. In E. coli, mistranslation induced by a mutation in the ribosomal protein S4 triggers an RpoS-mediated oxidative stress response [19], and ageing colonies carrying an editing-defective amino-acyl tRNA synthetase activate the SOS response [20]. Similarly, in yeast, increased translation errors activate the unfolded protein response and environmental stress response [21]. Thus, we hypothesize that mistranslation arising from diverse mechanisms could provide an indirect fitness advantage by activating stress responses, without relying on a direct benefit due to specific mistranslated proteins. Prior studies show that in some cases, the beneficial effect of mistranslation in conjunction with stress responses relies on increased mutagenesis [22]. In others, there is a general increase in mutation frequency, with or without an identified benefit [23,24]. Such an increase in mutagenesis can increase the probability of sampling beneficial mutations, but should also impose the associated cost of simultaneously acquiring more deleterious mutations. However, the sampling of beneficial mutations can also be increased by a simple increase in population size. We propose that the activation of stress responses is advantageous because it increases immediate cell survival and prevents a dramatic decline in population size. As a result, the population has a greater probability of sampling beneficial mutations without altering mutation frequency. Such a non-genetic mechanism lends itself to broad generalization, because it does not rely on specific mistranslated proteins; and populations do not incur the costs of constitutive stress responses or an increased genetic load due to deleterious mutations. Depending on the magnitude of the stress and the rate at which it is encountered, this hypothesis may therefore help understand the impact of mistranslation on cellular fitness.

To test our hypothesis, we manipulated baseline mistranslation levels in E. coli, both genetically and environmentally. We initiated our study using a strain with genetically depleted initiator tRNA (tRNAi) content (henceforth “Mutant”, carrying only one of four wild type “WT” tRNAi genes [25]). As central players in translation, cellular tRNA levels have a major impact on mistranslation rates [26,27], and are rapidly altered in response to environmental change [2830]. tRNAi levels are especially interesting because translation initiation is a rate limiting step [31], and tRNAi levels change under various stresses. For instance, in E. coli, amino acid starvation is accompanied by transcriptional downregulation of tRNAi during the stringent response [32], while mammalian cells reduce tRNAi levels on exposure to stressors such as the toxin VapC [33] and high temperature [34]. Specifically, depleting tRNAi increases mistranslation rates by allowing promiscuous non-AUG initiation by elongator tRNAs [27,33]. We therefore tested whether mistranslation resulting from tRNAi depletion leads to a general survival advantage.

We first carried out a Biolog screen [35] comparing WT and Mutant growth across 24 antibiotics with various modes of action. The Mutant consistently showed higher growth in the presence of Novobiocin (S1 Fig), a fluoroquinolone antibiotic that inhibits DNA gyrase. Further work showed that increasing baseline mistranslation rates via multiple mechanisms conferred protection against several kinds of DNA damage, via induction of the well-studied bacterial SOS response. Increased mistranslation brings cells closer to the intracellular molecular threshold for SOS induction, such that mistranslating cells sense and repair DNA damage sooner than the wild type. The resulting increase in early survival facilitates the eventual emergence of genetic resistance under antibiotic stress. Importantly, decreasing baseline mistranslation levels has the opposite effect. The mistranslation-induced SOS response is also beneficial under other stresses such as elevated temperature. Thus, we have uncovered a novel link between mistranslation and DNA damage that integrates two major cellular pathways and suggests a new role for mistranslation in cellular survival.

Results

Mistranslation increases resistance to DNA damage by enhancing early cell survival

A preliminary screen using Biolog plates indicated that the mistranslating Mutant performed better than the WT when exposed to the DNA damaging antibiotic Novobiocin (S1A–S1C Fig). In most other antibiotics included in the Biolog array plate, the Mutant and WT showed similar performance. We therefore further tested WT and Mutant under different kinds of DNA damage. Compared to WT, the mistranslating Mutant with depleted tRNAi showed higher survival under three kinds of DNA damage, imposed by exposure to UV radiation (base dimerization), hydrogen peroxide (base oxidation) or the antibiotic ciprofloxacin (‘Cip’, a potent DNA gyrase inhibitor that causes double stranded DNA breaks) (Fig 1A–1C). Thus, mistranslation seemed to be linked to surviving DNA damage. We explored the reasons behind this intriguing link, focusing on Cip resistance (i.e. the number of Cip-resistant colonies per unit viable count). While the Mutant has a higher mistranslation rate, it also has a lower rate of steady state protein synthesis because of lower tRNAi content [36]. To test whether the Mutant’s enhanced survival arises simply from a reduction in protein synthesis rates, we treated WT with the antibiotic kasugamycin (which inhibits initiation without altering tRNA levels). However, this treatment did not improve WT Cip resistance significantly (S2 Fig). The Mutant also grows more slowly than the WT in rich media (LB) (mean Mutant growth rate = 0.93/h; WT = 1.15/h; S3A and S3B Fig); such slow growth can enhance survival under stress as a by-product. However, decreasing WT growth by culturing it in glycerol (which causes a 5-fold higher doubling time) had no impact on Cip resistance (S3C Fig). Finally, the Mutant did not produce more GyrA protein (subunit A of DNA gyrase, which is the target of Cip) (S4 Fig); hence, this could not explain enhanced Cip resistance.

Mistranslation confers resistance to DNA damage.
Fig. 1. Mistranslation confers resistance to DNA damage.
(a) Survival of WT and Mutant cultures (n = 9) exposed to 20 J/m2 of UV-C radiation for 5 s. The plot shows mean % survival relative to the number of colonies on a non-irradiated control plate. Mann-Whitney U test, Mutant>WT, U = 9.5, P = 0.0046 (b) Time course of survival (viable counts) of cultures of WT and Mutant (n = 4) treated with 5 mM hydrogen peroxide. At 3 h, Mutant>WT, Mann-Whitney U test, U = 0, P = 0.029 (c) Resistance of WT and Mutant cultures inoculated from single colonies (n = 6), pulsed with 20 ng/mL ciprofloxacin (Cip) for 1 h, and plated on LB agar plates with vs. without 50 ng/mL Cip (Cip-50). The plot shows the average proportion of resistant colonies relative to total viable counts. Mann-Whitney U test, U = 0, P = 0.002. All cells were sampled from mid log phase cultures (OD600nm ~0.6) of the relevant strain. In all figures, asterisks indicate a significant difference. (d) Schematic showing the distribution of CipR-conferring point mutations in the QRDR (Quinolone resistance determining region) of the gyrA gene, found in WT and Mutant upon whole genome sequencing. The number of independent colonies harbouring each mutation is noted in parentheses.

Whole genome sequencing showed that each Cip-resistant (CipR) colony of WT and Mutant (after 24 h on Cip plates) had a single mutation within the well-known QRDR (Quinolone Resistance Determining Region) of the gyrA gene (Fig 1D). However, as described above, Mutant cultures produced more resistant colonies per unit viable cell. Thus, while WT and Mutant cells both acquired beneficial mutations in gyrA, the Mutant was either sampling them more frequently or earlier than the WT, resulting in a higher final number of CipR colonies. A standard rifampicin (Rif) resistance assay [20,24] showed that WT and Mutant had similar basal mutation frequencies (Fig 2A), suggesting that mutation frequency could not explain differential Cip resistance. While assessing Cip resistance, we incubated plates for 24 h, as per the standard protocol for E. coli. However, prior work showed that an E. coli strain with compromised aminoacyl tRNA editing activity had increased mutation frequency, producing more Rif resistant colonies over time(from 60 – 180h of incubation [20]). Thus, we also carried out the MAC (Mutation accumulation in ageing colonies) assay. We did not see any difference in mutation frequency between WT and Mutant even after 144 h of incubation (S5 Fig), emphasizing that the observed survival differences do not rely on mutagenesis. Instead, we found that Mutant cells had higher early survival upon exposure to Cip (after 2 h; Fig 2B), and sequencing confirmed that cells sampled at this point did not have any gyrA QRDR mutations. Therefore, this early survival was not due to genetic resistance, but implies a form of tolerance. Together, these results suggested that mistranslation indirectly enhanced Cip resistance by increasing early survival.

Mistranslation increases early survival under DNA damage, without altering mutation frequency.
Fig. 2. Mistranslation increases early survival under DNA damage, without altering mutation frequency.
(a) Overnight cultures from single colonies of WT and Mutant were plated on LB agar with 50 μg/mL of rifampicin (Rif-50, n = 15 plates per strain, with paired control plates for viable counts). The plot shows the mean number of resistant colonies per unit viable count from LB plates. Mann-Whitney U test: WT vs Mutant, ns, U = 100.5, P = 0.63. (b) Early survival (tolerance) of WT, mistranslating and hyper-accurate strains treated with Cip-50 for 2 h, from cultures (n = 3) treated with Cip-50 for 2 h. We estimated viable counts before and after exposure. The plot shows mean % survival in each case. t tests: Mutant>WT, t = 6.13, P = 0.02; WT(canavanine)>WT, t = 6.1, P = 0.005; WT(norleucine)>WT, t = 9.6, P = 0.002; WT(hyper-accurate) vs. WT, ns, t = 2.3, P = 0.08; Mutant(hyper-accurate)WT: U = 1, P = 0.0043; WT(canavanine)>WT: U = 1, P = 0.0043; WT(norleucine)>WT: U = 0, P = 0.0022; WT(hyper-accurate)600nm ~0.6) of single colonies of the relevant strain, unless indicated otherwise.

To test the generality of this result, we manipulated mistranslation levels by (a) increasing WT mistranslation by adding the non-proteinogenic amino acids canavanine or norleucine to the growth medium [19,37] and (b) reducing global mistranslation via hyper-accurate ribosomes. Point mutations in the ribosomal protein S12 increase the fidelity of codon–anticodon recognition, reducing overall translation errors [38]. To generate hyper-accurate ribosomes, we introduced the mutation K42R, which has almost no impact on growth rate but increases translation fidelity by ~2 fold [39]. Other mutations confer higher fidelity (e.g. K42T) but impose significant fitness costs [39]; hence we did not use them. To estimate the frequency of translational errors due to hyper-accurate ribosomes, we used a modified version of the previously described dual luciferase assay [40] that measures the frequency with which specific codons are misread. We found that both WT and Mutant mistranslation levels decreased ~8 to 10-fold, confirming that the altered ribosomes indeed conferred higher fidelity (S6 Fig). Note that basal misincorporation rates for WT and Mutant are comparable, which is not surprising given that the Mutant is thus far only known to carry out non-AUG initiation [27,33], and no other forms of mistranslation.

With the above manipulations, we found that increasing mistranslation rates consistently increased early survival (Fig 2B) and Cip resistance in WT (Fig 2D); whereas suppressing basal mistranslation decreased early survival and Cip resistance in both WT and Mutant (Fig 2B and 2D). Mutant survival was 5 times higher than WT, and the number of final CipR colonies obtained per unit viable count in the Mutant were comparable: ~4.9 times that of WT. This indicates that mistranslation (including basal mistranslation in the WT) indeed increased Cip resistance via an increase in early survival (tolerance), allowing the maintenance of a larger population, and therefore enhancing the number of gyrA mutants sampled in liquid cultures that were eventually selected on the Cip plates (Fig 2C). Note that in our whole genome sequencing, we do not see any evidence of increased mutagenesis upon Cip exposure (every colony sampled only had a single point mutation in gyrA). Having established this, we next focused on understanding the molecular basis of our finding.

Mistranslation mediates ciprofloxacin resistance via the SOS response

In response to DNA damage, bacterial cells induce the SOS response, which controls the expression of several DNA repair pathways [41]. Briefly, DNA damage generates single stranded DNA that binds to the protein RecA. Activated RecA stimulates cleavage of LexA (a repressor), which in turn induces the SOS response, de-repressing several DNA repair genes (Fig 3A). When we blocked SOS induction by replacing the WT lexA allele with a non-degradable allele [lexA3; 42] and challenged cultures with Cip, both WT and Mutant had lower early survival (Fig 3B) and negligible Cip resistance (Fig 3D), as expected in the absence of an intact DNA repair response. Thus, the mistranslation-induced increase in tolerance leading to Cip resistance depends on the SOS response.

Mistranslation increases early survival and ciprofloxacin resistance via the SOS response.
Fig. 3. Mistranslation increases early survival and ciprofloxacin resistance via the SOS response.
(a) Schematic of the SOS response in E.coli (b-c) Early survival of indicated strains inoculated from single colonies (n = 3), treated with 50 ng/mL Cip for 2h, and then plated on plain LB. Plots show the mean % survival calculated from total viable counts before and after Cip treatment. (b) Effect of inactivating the SOS response using the lexA3 allele. Unpaired t test, WTlexA3 or WT(inactive SOS)lexA3 or Mutant(inactive SOS)recN gene. Unpaired t test, WΔrecN recNrecN vs MutantΔrecN, ns, t = 0.96, P = 0.43 (d) Resistance of cultures inoculated from single colonies (n = 6) of indicated strains, pulsed with 20 ng/mL ciprofloxacin (Cip-20) for 1 h, and plated on LB agar with vs. without 50 ng/mL Cip (Cip-50). The plot shows the mean proportion of resistant colonies relative to total viable counts. Mann-Whitney U test: Mutant>WT, U = 1, P = 0.0043; WT(inactive SOS)recNrecN

The SOS response has two opposing aspects: rapid DNA repair, and increased mutagenesis due to the activation of error-prone polymerases. The latter temporarily elevates mutation rate, increasing the supply of beneficial mutations [43]. However, as shown above, WT and Mutant had similar basal mutation frequencies (Fig 2A). Therefore, we reasoned that the increased early survival of mistranslating strains must be aided by faster or more efficient repair and recombination. To test this, we deleted RecN–a key member of the SOS-linked recombination mediated repair pathway [44]. The deletion decreased early survival upon Cip exposure (Fig 3C) and led to a complete loss of Cip resistance (Fig 3D), indicating that repair and recombination functions indeed underlie the increased Cip resistance observed in the mistranslating Mutant.

All antibiotics–especially fluoroquinolones such as ciprofloxacin–are thought to act at least in part through the generation of reactive oxygen species (ROS) [45] that damage proteins and increase mutagenesis [24,45,46]. Therefore, we tested whether differential mitigation of reactive oxygen species could explain survival differences in our strains. However, despite prior treatment with an anti-oxidant (glutathione), mistranslating cells continued to show a survival advantage under Cip (S7 Fig). In conjunction with the lack of evidence for increased mutagenesis in our mistranslating strains (Figs 2A, S5 and 1E), this suggests that mistranslation does not alter the role of ROS in Cip toxicity, and hence cannot explain our results.

Mistranslation enhances Cip resistance by allowing faster induction of the SOS response

Since both WT and Mutant rely on the SOS response for Cip resistance, the ~5-fold greater survival of the Mutant (Fig 1D) continued to be a puzzle. We hypothesized that the survival advantage arose from differential induction of SOS due to mistranslation, allowing rapid DNA repair. Consistent with this hypothesis, greater mistranslation was associated with slightly (though not significantly) higher basal RecA levels across multiple experimental blocks (Fig 4A). However, all mistranslating strains rapidly degraded LexA when exposed to Cip for just 1 hour; conversely, hyper-accurate strains were unable to degrade LexA in this period (Fig 4B). These results suggest that even in the absence of DNA damage, RecA was already elevated in mistranslating strains, positioning the cell closer to the SOS induction threshold and leading to rapid degradation of LexA upon exposure to Cip (Fig 4D). To test this, we induced the SOS response in WT and Mutant and monitored the time course of LexA degradation. The Mutant degraded LexA within 10–20 minutes of SOS induction, while the WT took twice as long (Fig 4C and S8 Fig). Similarly, LexA was degraded at lower concentrations of Cip in the Mutant (S9 Fig). Note that the Mutant is not already ‘stressed’; it has similar basal LexA levels to the WT (S9 Fig). Thus, the Mutant only starts degrading LexA upon encountering DNA damage.

Mistranslation brings cells closer to the threshold for SOS activation.
Fig. 4. Mistranslation brings cells closer to the threshold for SOS activation.
(a) A representative blot showing RecA protein levels from mid log phase cultures of WT and mistranslating strains (n = 3). Quantification of mean blot band normalized to total protein is represented in arbitrary units, relative to WT. Paired t tests: WT vs. Mutant, ns, t = 3.5, P = 0.07; WT(canavanine) >WT, t = 5.1, P = 0.03; WT vs WT(norleucine), ns, t = 2.2, P = 0.1 (b) A representative blot showing LexA protein levels in WT and mistranslating strains (n = 3). Quantification of mean blot band normalized to total protein is represented in arbitrary units, relative to WT. Paired t tests: MutantWT, t = 12.4, P = 0.006; Mutant(hyper-accurate)> Mutant, t = 10.4, P = 0.009 (c) Time course of LexA degradation from 0 to 50 min after exposure to Cip-20. One blot and bands normalised to total protein are shown here; for more experimental blocks, see S8 Fig. (d) The proposed model of the state of WT and mistranslating strains with respect to the SOS activation threshold.

Together, these observations suggest that in mistranslating strains, (i) LexA is degraded faster upon encountering the DNA damaging stress, and (ii) LexA is degraded at a lower magnitude of the stress (Fig 4D and S9 Fig). Thus, we demonstrate a direct causal relationship between mistranslation, induction of the SOS response, and enhanced survival under Cip.

Mistranslation activates the SOS response via Lon protease

In our experiments, we increased mistranslation in distinct ways, and consistently observed increased survival under DNA damage. Canavanine and norleucine respectively replace arginine and leucine in the proteome, whereas tRNAi depletion causes mis-initiation with elongator tRNAs. The parallel outcomes from these diverse modes of mistranslation suggested a general mechanistic link between mistranslation and SOS response. Based on prior studies, we suspected that Lon–a key protease across eubacteria–may represent such a link. In mistranslating E. coli cells, Lon alleviates the associated deleterious effects by degrading aggregated and non-functional proteins [3]. In Pseudomonas, Lon is essential for RecA accumulation and induction of the SOS response, and is suggested to degrade RecA repressors such as RecX and RdgC [47].

We therefore hypothesized that our mistranslating strains may have higher amounts of Lon, in turn accumulating RecA and bringing cells physiologically closer to the threshold for SOS induction. Because Lon is part of the E. coli heat shock regulon [48], we also suspected a general increase in the heat shock response. Indeed, our mistranslating cells had higher levels of Lon protease (Fig 5A), as well as the heat shock transcription factor sigma 32 (S10 Fig; also independently reported recently by [49]). Over-expressing Lon enhanced early survival (Fig 5B) and resistance to Cip (Fig 5C), and reduced LexA levels in both WT and Mutant upon SOS induction (Fig 5D). For technical reasons, we were unable to knock out Lon in our wild type strain KL16. Hence, we deleted Lon in E. coli MG1655. While MG1655 had comparable ciprofloxacin resistance to our WT (S11 Fig), deleting Lon decreased Cip resistance (Fig 5C) and increased LexA levels in SOS-induced cells (Fig 5D). Over-expressing Lon also elevated basal RecA levels (in the absence of any DNA damage), further supporting our hypothesis (S12 Fig). Finally, since Lon is part of the heat shock regulon, we predicted that prior exposure to high temperature should induce Lon and increase resistance to DNA damage. True to expectation, cells grown at 42°C for three hours had higher Cip resistance (Fig 6A). Together, these results demonstrate a key role for Lon in mediating Cip resistance by inducing the SOS response.

Mistranslation induces the SOS response via Lon protease.
Fig. 5. Mistranslation induces the SOS response via Lon protease.
(a) A representative blot showing Lon protein levels in WT and mistranslating strains (n = 3). Quantification of mean blot band normalized to total protein is represented in arbitrary units, relative to WT. Paired t tests: Mutant>WT, t = 4.2, P = 0.01; WT(canavanine)>WT, t = 6.5, P = 0.003; WT vs. WT(norleucine), ns, t = 2.5, P = 0.05 (b) Early survival of WT and WT+Lon cultures (n = 3) treated with 50 ng/mL Cip for 2 h and then plated on plain LB. The plot shows the mean % survival calculated from total viable counts before and after Cip treatment. Paired t test, WT+Lon>WT, t = 5.1, P = 0.03 (c) Resistance of MG1655, MGΔlon (MG-Lon), WT (KL16), WT+Lon, Mutant and Mutant+Lon on Cip-50 cultures (n = 4) pulsed with Cip-20 for 1 h and plated on LB agar with vs. without Cip-50. The plot shows the mean proportion of resistant colonies relative to total viable counts. Paired t tests: MGΔlonWT, t = 5.5, P = 0.01; Mutant vs. Mutant+Lon, ns, t = 1.9, P = 0.14 (d) A representative blot showing LexA protein levels in WT and mistranslating strains (n = 3). Quantification of mean blot band normalized to total protein is represented in arbitrary units relative to WT. SOS was inactivated using the lexA3 allele. Paired t tests: MutantWT, t = 5.5, P = 0.03; MG vs MGΔlon, ns, t = 3.2, P = 0.08. Asterisks indicate a significant difference between strains. All cells were sampled from mid log phase cultures (OD600nm ~0.6) derived from single colonies of the relevant strain.
Mistranslation-induced SOS response enhances survival in other stresses.
Fig. 6. Mistranslation-induced SOS response enhances survival in other stresses.
(a) Survival of WT and Mutant mid log phase cultures (OD600nm~0.6) grown overnight at 37°C from single colonies (n = 4), sub-cultured and grown at 37°C or 42°C (to induce the heat shock response while cells entered log phase) for 3.5 h, and plated on LB agar with vs. without 50 ng/mL Cip (Cip-50). The plot shows the mean proportion of resistant colonies relative to total viable counts. Mann-Whitney tests: WT(42°C)>WT(37°C), U = 0, P = 0.0286; Mutant(42°C)>Mutant(37°C), U = 0, P = 0.0286 (b) Total viable counts of the indicated strains at 0, 12 and 24 h after exposure to 42° C (n = 4, for Mutant (no mistranslation) at 24h, n = 3). SOS was inactivated using the lexA3 allele. At 24 h, Mann-Whitney tests: Mutant>WT, U = 0, P = 0.0286; Mutant(inactive SOS) vs. WT, ns, U = 4.5, P = 0.4; Mutant(hyper-accurate) vs. WT, ns, U = 8, P>0.99.

Mistranslation-induced SOS response enhances survival in other stresses

As mentioned above, Lon is part of the heat shock regulon; hence we tested whether mistranslation also increased survival under high temperature. As predicted, we found that the Mutant had greater survival at high temperature, especially in the stationary phase of growth (after 12 hours; Fig 6B). We also observed this growth advantage in WT cells treated with norleucine and canavanine, although the results with canavanine were variable (S13 Fig). Importantly, Lon alone could not explain the greater survival at high temperature, which required both mistranslation and a functional SOS response (Fig 6B). Lon levels at 42°C were comparable across WT and Mutant (S14 Fig), and LexA was degraded in the Mutant but not in the WT (S15 Fig). Interestingly, RecN is also important for high temperature survival of both WT and Mutant, suggesting that survival is influenced by functional DNA repair (S16 Fig). The clear dependence of the survival advantage on SOS induction suggests cross-talk between mistranslation, the heat shock response and SOS induction.

Our results thus show that mistranslation-induced early activation of the SOS response is responsible for multiple stress resistance phenotypes. We suggest a model whereby mistranslation of various kinds increases Lon protease levels, triggering an early induction of the SOS response that enhances cell survival not only under DNA damage, but also under other stresses (Fig 7).

Proposed model for stress resistance mediated by faster SOS activation in mistranslating strains.
Fig. 7. Proposed model for stress resistance mediated by faster SOS activation in mistranslating strains.

Discussion

High global translation error rates have remained an enduring puzzle, given the large overall costs of generalized mistranslation compared to the benefits of generating specific mistranslated proteins in particular environments. Notably, this problem has mired the hypothesis that mistranslation rates may evolve under positive selection. Here, we diminish this barrier by demonstrating that higher levels of basal mistranslation enhance early survival under DNA damage by allowing rapid induction of the SOS response, making a larger pool of cells available for subsequent genetic change. Note that basal WT levels of mistranslation are critical for survival: introducing hyper-accurate ribosomes reduces WT resistance to ciprofloxacin. We further show that this effect is mediated by increased levels of a key protease that deals with aberrant proteins as well as several important regulatory enzymes. Since survival does not rely on the chance generation of specific mistranslated proteins, cells effectively bypass the deleterious load of high mistranslation, triggering a stress response that alleviates damage to cellular components. Most stresses that are commonly encountered by bacteria (oxidative stress, high temperature, radiation, starvation) induce either mistranslation (through damage to proteins), or DNA damage, or both. Hence, it seems fitting that these two phenomena should be linked, and should operate through a common effector molecule. Note that in our mistranslating strains, rapid SOS activation is a by-product of mistranslation, and we do not propose that high mistranslation rates have evolved primarily to activate stress responses. Instead, we propose that the large benefit of mistranslation under diverse stresses (that can be mitigated via the SOS response) may weaken selection against high basal mistranslation. We predict that this should be especially important when stress is encountered frequently or is highly deleterious; this needs to be explicitly tested in further work.

The early tolerance and subsequent high resistance that we observed does not rely on mutagenesis. In contrast, prior work shows that increasing basal mistranslation generates faulty mismatch repair proteins, increasing mutation frequency in E. coli [24]. A body of work describing translation stress induced mutagenesis (TSIM) also suggested that altered DNA Polymerase III proteins could underlie the increased mutagenesis observed in strains with mutated glyV tRNA anticodons [23]. Lastly, earlier work linking faulty tRNA aminoacylation with activation of the SOS response suggested that the impact of mistranslation on fitness was mediated entirely through elevated global mutagenesis [20]. In contrast to this known mutagenic impact of SOS induction (generating “hopeful monsters”), we demonstrate that mistranslation is generally beneficial under DNA damage because it enhances the rapid DNA repair modules controlled by the SOS response. In our case, this effect is non-specific, acts in the short-term, is induced by different kinds of mistranslation, and leads to increased levels of a key protein quality control molecule. Our proposed model (Fig 7) thus suggests that high global mistranslation could be beneficial under multiple stresses; and could potentially evolve under positive selection. Of course, the generality of this hypothesis needs to be validated with further experiments. More broadly, our results imply that generalized non-genetic changes can facilitate subsequent genetic adaptation by increasing short-term survival. This has been an attractive hypothesis with limited and protein-specific prior support. In Saccharomyces cerevisiae, ribosomal frameshifting alters localization of a specific protein, a phenotype that is then fixed by genetic mutations in a few generations [50]. Similarly, changes in Hsp90 levels in Candida albicans [51] and at least one phenotype conferred by the prion PSI+ in wild yeasts can be stabilised over evolutionary time [52]. Here, we provide clear evidence that non-directed phenotypic changes can increase stress resistance by enhancing the subsequent sampling of beneficial mutations.

Our work helps to synthesize a diverse body of prior results into a cohesive framework. Mistranslation leads to misfolding and protein aggregation, which is typically deleterious [reviewed in 53]. However, a recent study showed that cells carrying intracellular protein aggregates were more stress-resistant [54]. Interestingly, these cells had upregulated chaperones such as DnaK and proteases such as ClpP, suggesting that protein aggregation can also precipitate stress resistance. Prior work also shows that Lon is required to alleviate the toxic effects of mistranslation [3]. However, thus far only two studies [49] [55] have linked the heat shock response with mistranslation. We now add to this work, clearly showing that the heat shock response is activated by general mistranslation and corroborating prior predictions that aberrant proteins activate heat shock chaperones [56]. Thus, our work broadly connects mistranslation with the heat shock response, and with increased stress resistance.

Our model could also explain the puzzling observation that Lon protease function determines sensitivity to high concentrations of quinolone antibiotics (which target DNA gyrase, with no direct connection to Lon) such as nalidixic acid [57] and levofloxacin [58]. We show that impairing Lon hinders the ability of cells to rapidly induce the SOS response and repair damaged DNA. The central role of Lon in bridging mistranslation and the SOS response is also supported by previously observed links between the heat shock response and the SOS response. For example, in Listeria monocytogenes, heat shock directly triggers the SOS response [59]. In E. coli, the heat shock chaperone GroE positively regulates the error-prone DNA Polymerase IV, hitherto thought to be regulated only via the SOS response [60]. Finally, when exposed to levofloxacin, E. coli cells express both the SOS response and heat shock genes [58]. These results corroborate our observation that mistranslation and SOS activation together increase heat resistance, though we do not know precisely how this phenotype is regulated. Altogether, we suggest that Lon acts as a key molecule that coordinates several aspects of stress responses, encompassing toxin-antitoxin regulation, survival under anaerobic conditions, SOS, heat shock, antibiotic resistance, and cell division [reviewed in 48].

Finally, our model suggests a general framework to understand the evolution of baseline as well as broad stress-induced mistranslation rates, given the potential generality of the underlying mechanism. We suggest that the “optimal” level of mistranslation rate is set by a fundamental tradeoff between the cost (vs. benefit) of global mistranslation, with the nature of the tradeoff being shaped by the magnitude and frequency of adverse conditions. Thus, mistranslation rates cannot increase indefinitely because they impose a substantial cellular cost under normal conditions (e.g. in our case, the Mutant has ~20% lower growth rate than the WT-S3A and S3B Fig). Similarly, WT cells do not constitutively activate the SOS response because this would lead to a disruption of cell division and increased mutation accumulation [61]. Instead, we suggest a delicate balance whereby cells are poised closer to the threshold for SOS activation; increasing the speed of response once the stress arrives, but not before. DNA damage is a ubiquitous stress for bacterial cells, faced both internally (through the accumulation of reactive oxygen species and replication errors [reviewed in 61,62]) and externally (through UV radiation and more recently, sub-inhibitory antibiotics in soil and water [63]). We speculate that mistranslation levels that extract a cost under normal growth but provide a clear advantage under DNA damage could well be favoured by natural selection. As discussed in the Introduction, generalised mistranslation is already known to activate stress responses. Therefore, the benefit of high mistranslation need not be restricted to DNA damage alone, but could extend to cells facing frequent oxidative stress and anaerobic conditions [19,64]. We therefore suggest that cells could tolerate high mistranslation levels because of the potential benefit under environmental stresses that are encountered frequently.

The SOS response is among the best studied pathways in E. coli, inducing DNA repair genes in response to double strand breaks and stalled replication forks generated by severe DNA damage. Yet, we continue to unravel new phenotypes controlled by this response. Instead of being directed solely at DNA repair, the SOS response is turning out to be central for several stressful situations, including high mechanical pressure and antibiotic tolerance [41,65,66]. Similarly, the causes and impacts of mistranslation also continue to be extensively explored, with new details surfacing each day. At the moment, we cannot determine whether mistranslation was co-opted by the DNA damage response as a trigger, or vice-versa. Irrespective of which response evolved first, it is clear that diverse cellular mechanisms are linked in unexpected ways, co-ordinating the cellular response to multiple stresses. While these phenomena are independently well studied, we can now connect them using a single effector molecule, Lon protease. Our study raises the question of whether such novel links could themselves be evolving in different directions, leading to cross talk between mutation-independent phenotypic variation and genetic change in response to stress.

Methods

Strains and growth media

See S1 Text

Initial screening of WT and mutant strains using phenotype micro arrays

To screen for phenotypic differences between the WT and the mutant across a range of environments, we employed phenotype microarrays. See S1 Text for details.

Assaying survival under DNA damage

We assessed strain survival under three kinds of DNA damage: exposure to UV light (base dimerization), hydrogen peroxide (base oxidation) and ciprofloxacin (double stranded DNA breaks). See S1 Text for details.

Dual luciferase assay to assess miscoding

To assess the frequency of translational errors, we used a modified version of a previously described dual luciferase assay. See S1 Text for details.

Measuring mutation frequency using rifampicin resistance

We estimated mutation frequency in WT and mutant cells using spontaneous mutations conferring rifampicin resistance. See S1 Text for details.

Measuring protein production

To measure levels of candidate proteins responsible for the observed cellular response to stress, we used western blots with a chemiluminescent detection system. See S1 Text for details.

Whole genome sequencing

To identify mutations responsible for ciprofloxacin resistance we sequenced whole genomes of the resistant colonies. See S1 Text for details.

Supporting information

S1 Text [docx]
Supplementary methods.

S1 Fig [a]
Phenotype microarray for WT and Mutant.

S2 Fig [ksg]
Reducing WT protein synthesis does not impact Cip resistance.

S3 Fig [a]
Reducing WT growth rate does not impact Cip resistance.

S4 Fig [tif]
The increased Cip resistance in the Mutant is not mediated by GyrA upregulation.

S5 Fig [tif]
WT and Mutant show similar mutation frequencies in ageing colonies (MAC).

S6 Fig [a]
Hyper-accurate strains show lower mistranslation.

S7 Fig [tif]
Treatment with an anti-oxidant does not eliminate increased Cip resistance due to mistranslation.

S8 Fig [tif]
Time course of LexA degradation by WT and mutant.

S9 Fig [a]
The mutant activates SOS at lower levels of stress.

S10 Fig [tif]
Mistranslation elevates the heat shock transcription factor sigma 32.

S11 Fig [kl16]
WT (KL16) and MG1655 show comparable ciprofloxacin resistance.

S12 Fig [tif]
Over-expression of Lon protease elevates RecA levels.

S13 Fig [a]
Mistranslation provides an early but variable survival advantage under high temperature.

S14 Fig [tif]
WT and mutant have comparable Lon levels under heat stress.

S15 Fig [tif]
Mutant degrades LexA under heat stress.

S16 Fig [tif]
RecN provides an early survival advantage under high temperature.


Zdroje

1. Mordret E, Yehonadav A, Barnabas GD, Cox J, Dahan O, Geiger T, et al. Systematic detection of amino acid substitutions in proteome reveals a mechanistic basis of ribosome errors. bioRxiv. 2018.

2. Ellis N, Gallant J. An estimate of the global error frequency in translation. Mol Gen Genet. 1982;188(2):169–72. doi: 10.1007/bf00332670 6759868

3. Ruan B, Palioura S, Sabina J, Marvin-Guy L, Kochhar S, Larossa RA, et al. Quality control despite mistranslation caused by an ambiguous genetic code. Proceedings of the National Academy of Sciences of the United States of America. 2008;105(43):16502–7. doi: 10.1073/pnas.0809179105 18946032

4. Jakubowski H, Goldman E. Editing of errors in selection of amino acids for protein synthesis. Microbiol Rev. 1992;56(3):412–29. 1406490

5. Zaher HS, Green R. Fidelity at the molecular level: lessons from protein synthesis. Cell. 2009;136(4):746–62. doi: 10.1016/j.cell.2009.01.036 19239893

6. Wohlgemuth I, Pohl C, Rodnina MV. Optimization of speed and accuracy of decoding in translation. EMBO J. 2010;29(21):3701–9. doi: 10.1038/emboj.2010.229 20842102

7. Banerjee K, Kolomeisky AB, Igoshin OA. Elucidating interplay of speed and accuracy in biological error correction. Proceedings of the National Academy of Sciences of the United States of America. 2017;114(20):5183–8. doi: 10.1073/pnas.1614838114 28465435

8. Zhang J, Ieong KW, Johansson M, Ehrenberg M. Accuracy of initial codon selection by aminoacyl-tRNAs on the mRNA-programmed bacterial ribosome. Proceedings of the National Academy of Sciences of the United States of America. 2015;112(31):9602–7. doi: 10.1073/pnas.1506823112 26195797

9. Mikkola R, Kurland CG. Selection of laboratory wild-type phenotype from natural isolates of Escherichia coli in chemostats. Mol Biol Evol. 1992;9(3):394–402. doi: 10.1093/oxfordjournals.molbev.a040731 1584010

10. Rajon E, Masel J. Evolution of molecular error rates and the consequences for evolvability. Proceedings of the National Academy of Sciences of the United States of America. 2011;108(3):1082–7. doi: 10.1073/pnas.1012918108 21199946

11. Bratulic S, Gerber F, Wagner A. Mistranslation drives the evolution of robustness in TEM-1 beta-lactamase. Proceedings of the National Academy of Sciences of the United States of America. 2015;112(41):12758–63. doi: 10.1073/pnas.1510071112 26392536

12. Netzer N, Goodenbour JM, David A, Dittmar KA, Jones RB, Schneider JR, et al. Innate immune and chemically triggered oxidative stress modifies translational fidelity. Nature. 2009;462(7272):522–6. doi: 10.1038/nature08576 19940929

13. Mohler K, Ibba M. Translational fidelity and mistranslation in the cellular response to stress. Nat Microbiol. 2017;2:17117. doi: 10.1038/nmicrobiol.2017.117 28836574

14. Javid B, Sorrentino F, Toosky M, Zheng W, Pinkham JT, Jain N, et al. Mycobacterial mistranslation is necessary and sufficient for rifampicin phenotypic resistance. Proceedings of the National Academy of Sciences of the United States of America. 2014;111(3):1132–7. doi: 10.1073/pnas.1317580111 24395793

15. Woese CR. On the evolution of the genetic code. Proceedings of the National Academy of Sciences of the United States of America. 1965;54(6):1546–52. doi: 10.1073/pnas.54.6.1546 5218910

16. Moghal A, Mohler K, Ibba M. Mistranslation of the genetic code. FEBS Lett. 2014;588(23):4305–10. doi: 10.1016/j.febslet.2014.08.035 25220850

17. Jones TE, Alexander RW, Pan T. Misacylation of specific nonmethionyl tRNAs by a bacterial methionyl-tRNA synthetase. Proceedings of the National Academy of Sciences of the United States of America. 2011;108(17):6933–8. doi: 10.1073/pnas.1019033108 21482813

18. Ribas de Pouplana L, Santos MA, Zhu JH, Farabaugh PJ, Javid B. Protein mistranslation: friend or foe? Trends Biochem Sci. 2014;39(8):355–62. doi: 10.1016/j.tibs.2014.06.002 25023410

19. Fan Y, Wu J, Ung MH, De Lay N, Cheng C, Ling J. Protein mistranslation protects bacteria against oxidative stress. Nucleic acids research. 2015;43(3):1740–8. doi: 10.1093/nar/gku1404 25578967

20. Bacher JM, Schimmel P. An editing-defective aminoacyl-tRNA synthetase is mutagenic in aging bacteria via the SOS response. Proceedings of the National Academy of Sciences of the United States of America. 2007;104(6):1907–12. doi: 10.1073/pnas.0610835104 17264207

21. Paredes JA, Carreto L, Simoes J, Bezerra AR, Gomes AC, Santamaria R, et al. Low level genome mistranslations deregulate the transcriptome and translatome and generate proteotoxic stress in yeast. BMC Biol. 2012;10:55. doi: 10.1186/1741-7007-10-55 22715922

22. <SOS_editing defective synthetase_paul schimmel.pdf>.

23. Al Mamun AAM, Gautam S, Humayun MZ. Hypermutagenesis in mutA cells is mediated by mistranslational corruption of polymerase, and is accompanied by replication fork collapse. Molecular Microbiology. 2006;62(6):1752–63. doi: 10.1111/j.1365-2958.2006.05490.x 17427291

24. Krisko A, Radman M. Phenotypic and genetic consequences of protein damage. PLoS genetics. 2013;9(9):e1003810. doi: 10.1371/journal.pgen.1003810 24068972

25. Samhita L, Shetty S, Varshney U. Unconventional initiator tRNAs sustain Escherichia coli. Proceedings of the National Academy of Sciences of the United States of America. 2012;109(32):13058–63. doi: 10.1073/pnas.1207868109 22829667

26. Kramer EB, Farabaugh PJ. The frequency of translational misreading errors in E. coli is largely determined by tRNA competition. RNA. 2007;13(1):87–96. doi: 10.1261/rna.294907 17095544

27. Samhita L, Virumae K, Remme J, Varshney U. Initiation with elongator tRNAs. Journal of bacteriology. 2013;195(18):4202–9. doi: 10.1128/JB.00637-13 23852868

28. Dittmar KA, Sorensen MA, Elf J, Ehrenberg M, Pan T. Selective charging of tRNA isoacceptors induced by amino-acid starvation. EMBO Rep. 2005;6(2):151–7. doi: 10.1038/sj.embor.7400341 15678157

29. Svenningsen SL, Kongstad M, Stenum TS, Munoz-Gomez AJ, Sorensen MA. Transfer RNA is highly unstable during early amino acid starvation in Escherichia coli. Nucleic acids research. 2017;45(2):793–804. doi: 10.1093/nar/gkw1169 27903898

30. Zhong J, Xiao C, Gu W, Du G, Sun X, He QY, et al. Transfer RNAs Mediate the Rapid Adaptation of Escherichia coli to Oxidative Stress. PLoS genetics. 2015;11(6):e1005302. doi: 10.1371/journal.pgen.1005302 26090660

31. Gualerzi CO, Pon CL. Initiation of mRNA translation in bacteria: structural and dynamic aspects. Cell Mol Life Sci. 2015;72(22):4341–67. doi: 10.1007/s00018-015-2010-3 26259514

32. Nagase T, Ishii S, Imamoto F. Differential transcriptional control of the two tRNA(fMet) genes of Escherichia coli K-12. Gene. 1988;67(1):49–57. doi: 10.1016/0378-1119(88)90007-8 2843439

33. Winther KS, Gerdes K. Enteric virulence associated protein VapC inhibits translation by cleavage of initiator tRNA. Proceedings of the National Academy of Sciences of the United States of America. 2011;108(18):7403–7. doi: 10.1073/pnas.1019587108 21502523

34. Watanabe K, Miyagawa R, Tomikawa C, Mizuno R, Takahashi A, Hori H, et al. Degradation of initiator tRNAMet by Xrn1/2 via its accumulation in the nucleus of heat-treated HeLa cells. Nucleic acids research. 2013;41(8):4671–85. doi: 10.1093/nar/gkt153 23471000

35. Bochner BR, Gadzinski P, Panomitros E. Phenotype microarrays for high-throughput phenotypic testing and assay of gene function. Genome Res. 2001;11(7):1246–55. doi: 10.1101/gr.186501 11435407

36. Samhita L, Nanjundiah V, Varshney U. How many initiator tRNA genes does Escherichia coli need? Journal of bacteriology. 2014;196(14):2607–15. doi: 10.1128/JB.01620-14 24816600

37. Karkhanis VA, Mascarenhas AP, Martinis SA. Amino acid toxicities of Escherichia coli that are prevented by leucyl-tRNA synthetase amino acid editing. Journal of bacteriology. 2007;189(23):8765–8. doi: 10.1128/JB.01215-07 17890314

38. Agarwal D, Gregory ST, O'Connor M. Error-prone and error-restrictive mutations affecting ribosomal protein S12. J Mol Biol. 2011;410(1):1–9. doi: 10.1016/j.jmb.2011.04.068 21575643

39. Chumpolkulwong N, Hori-Takemoto C, Hosaka T, Inaoka T, Kigawa T, Shirouzu M, et al. Effects of Escherichia coli ribosomal protein S12 mutations on cell-free protein synthesis. Eur J Biochem. 2004;271(6):1127–34. doi: 10.1111/j.1432-1033.2004.04016.x 15009191

40. O'Connor M, Thomas CL, Zimmermann RA, Dahlberg AE. Decoding fidelity at the ribosomal A and P sites: influence of mutations in three different regions of the decoding domain in 16S rRNA. Nucleic acids research. 1997;25(6):1185–93. doi: 10.1093/nar/25.6.1185 9092628

41. Michel B. After 30 years of study, the bacterial SOS response still surprises us. PLoS biology. 2005;3(7):e255. doi: 10.1371/journal.pbio.0030255 16000023

42. Little JW, Harper JE. Identification of the lexA gene product of Escherichia coli K-12. Proceedings of the National Academy of Sciences of the United States of America. 1979;76(12):6147–51. doi: 10.1073/pnas.76.12.6147 160562

43. Radman M. SOS repair hypothesis: phenomenology of an inducible DNA repair which is accompanied by mutagenesis. Basic Life Sci. 1975;5A:355–67. doi: 10.1007/978-1-4684-2895-7_48 1103845

44. Uranga LA, Reyes ED, Patidar PL, Redman LN, Lusetti SL. The cohesin-like RecN protein stimulates RecA-mediated recombinational repair of DNA double-strand breaks. Nat Commun. 2017;8:15282. doi: 10.1038/ncomms15282 28513583

45. Dwyer DJ, Belenky PA, Yang JH, MacDonald IC, Martell JD, Takahashi N, et al. Antibiotics induce redox-related physiological alterations as part of their lethality. Proceedings of the National Academy of Sciences of the United States of America. 2014;111(20):E2100–9. doi: 10.1073/pnas.1401876111 24803433

46. Rodríguez-Rosado AI, Valencia EY, Rodríguez-Rojas A, Costas C, Galhardo RS, Blázquez J, et al. Reactive oxygen species are major contributors to SOS-mediated mutagenesis induced by fluoroquinolones. BioRxiv. 2018.

47. Breidenstein EB, Bains M, Hancock RE. Involvement of the lon protease in the SOS response triggered by ciprofloxacin in Pseudomonas aeruginosa PAO1. Antimicrobial agents and chemotherapy. 2012;56(6):2879–87. doi: 10.1128/AAC.06014-11 22450976

48. Tsilibaris V, Maenhaut-Michel G, Van Melderen L. Biological roles of the Lon ATP-dependent protease. Res Microbiol. 2006;157(8):701–13. doi: 10.1016/j.resmic.2006.05.004 16854568

49. Evans CR, Fan Y, Ling J. Increased mistranslation protects E. coli from protein misfolding stress due to activation of a RpoS-dependent heat shock response. FEBS Lett. 2019.

50. Yanagida H, Gispan A, Kadouri N, Rozen S, Sharon M, Barkai N, et al. The Evolutionary Potential of Phenotypic Mutations. PLoS genetics. 2015;11(8):e1005445. doi: 10.1371/journal.pgen.1005445 26244544

51. Cowen LE, Lindquist S. Hsp90 potentiates the rapid evolution of new traits: drug resistance in diverse fungi. Science. 2005;309(5744):2185–9. doi: 10.1126/science.1118370 16195452

52. Halfmann R, Jarosz DF, Jones SK, Chang A, Lancaster AK, Lindquist S. Prions are a common mechanism for phenotypic inheritance in wild yeasts. Nature. 2012;482(7385):363–8. doi: 10.1038/nature10875 22337056

53. Drummond DA, Wilke CO. The evolutionary consequences of erroneous protein synthesis. Nat Rev Genet. 2009;10(10):715–24. doi: 10.1038/nrg2662 19763154

54. Govers SK, Mortier J, Adam A, Aertsen A. Protein aggregates encode epigenetic memory of stressful encounters in individual Escherichia coli cells. PLoS biology. 2018;16(8):e2003853. doi: 10.1371/journal.pbio.2003853 30153247

55. Hoffman KS, Berg MD, Shilton BH, Brandl CJ, O'Donoghue P. Genetic selection for mistranslation rescues a defective co-chaperone in yeast. Nucleic acids research. 2017;45(6):3407–21. doi: 10.1093/nar/gkw1021 27899648

56. Neidhardt FC, VanBogelen RA. Positive regulatory gene for temperature-controlled proteins in Escherichia coli. Biochem Biophys Res Commun. 1981;100(2):894–900. doi: 10.1016/s0006-291x(81)80257-4 7023474

57. Malik M, Capecci J, Drlica K. Lon protease is essential for paradoxical survival of Escherichia coli exposed to high concentrations of quinolone. Antimicrobial agents and chemotherapy. 2009;53(7):3103–5. doi: 10.1128/AAC.00019-09 19414573

58. Yamaguchi Y, Tomoyasu T, Takaya A, Morioka M, Yamamoto T. Effects of disruption of heat shock genes on susceptibility of Escherichia coli to fluoroquinolones. BMC Microbiol. 2003;3:16. doi: 10.1186/1471-2180-3-16 12911840

59. van der Veen S, Hain T, Wouters JA, Hossain H, de Vos WM, Abee T, et al. The heat-shock response of Listeria monocytogenes comprises genes involved in heat shock, cell division, cell wall synthesis, and the SOS response. Microbiology. 2007;153(Pt 10):3593–607. doi: 10.1099/mic.0.2007/006361-0 17906156

60. Layton JC, Foster PL. Error-prone DNA polymerase IV is regulated by the heat shock chaperone GroE in Escherichia coli. Journal of bacteriology. 2005;187(2):449–57. doi: 10.1128/JB.187.2.449-457.2005 15629916

61. Baharoglu Z, Mazel D. SOS, the formidable strategy of bacteria against aggressions. FEMS Microbiol Rev. 2014;38(6):1126–45. doi: 10.1111/1574-6976.12077 24923554

62. Ganai RA, Johansson E. DNA Replication-A Matter of Fidelity. Mol Cell. 2016;62(5):745–55. doi: 10.1016/j.molcel.2016.05.003 27259205

63. Cycon M, Mrozik A, Piotrowska-Seget Z. Antibiotics in the Soil Environment-Degradation and Their Impact on Microbial Activity and Diversity. Front Microbiol. 2019;10:338. doi: 10.3389/fmicb.2019.00338 30906284

64. Schwartz MH, Waldbauer JR, Zhang L, Pan T. Global tRNA misacylation induced by anaerobiosis and antibiotic exposure broadly increases stress resistance in Escherichia coli. Nucleic acids research. 2016.

65. Dorr T, Lewis K, Vulic M. SOS response induces persistence to fluoroquinolones in Escherichia coli. PLoS genetics. 2009;5(12):e1000760. doi: 10.1371/journal.pgen.1000760 20011100

66. Aertsen A, Van Houdt R, Vanoirbeek K, Michiels CW. An SOS response induced by high pressure in Escherichia coli. Journal of bacteriology. 2004;186(18):6133–41. doi: 10.1128/JB.186.18.6133-6141.2004 15342583


Článek vyšel v časopise

PLOS Genetics


2020 Číslo 3
Nejčtenější tento týden
Nejčtenější v tomto čísle
Kurzy Podcasty Doporučená témata Časopisy
Přihlášení
Zapomenuté heslo

Zadejte e-mailovou adresu, se kterou jste vytvářel(a) účet, budou Vám na ni zaslány informace k nastavení nového hesla.

Přihlášení

Nemáte účet?  Registrujte se

#ADS_BOTTOM_SCRIPTS#