|Home | About | Journals | Submit | Contact Us | Français|
Antibiotic-producing streptomycetes are rich sources of resistance mechanisms against endogenous and exogenous antibiotics. An ECF sigma factor σR (SigR) is known to govern the thiol-oxidative stress response in Streptomyces coelicolor. Amplification of this response is achieved by producing an unstable isoform of σR called σR′. In this work, we present evidence that antibiotics induce the SigR regulon via a redox-independent pathway, leading to antibiotic resistance. The translation-inhibiting antibiotics enhanced the synthesis of stable σR, eliciting a prolonged response. WblC/WhiB7, a WhiB-like DNA-binding protein, is responsible for inducing sigRp1 transcripts encoding the stable σR. The amount of WblC protein and its binding to the sigRp1 promoter in vivo increased upon antibiotic treatment. A similar phenomenon appears to exist in Mycobacterium tuberculosis as well. These findings reveal a novel antibiotic-induced resistance mechanism conserved among actinomycetes, and also give an explicit example of overlap in cellular damage and defense mechanisms between thiol-oxidative and anti- translational stresses.
Many actinomycetes, especially those of Streptomyces genus, are well recognized for undergoing complex developmental programs and producing diverse secondary metabolites. In soil environment where streptomycetes inhabit, thousands of bacterial species are estimated to reside in one gram of soil producing more than 104 bioactive small molecules1,2. In natural environment, streptomycetes have to deal with numerous growth-inhibitory antibiotics which are made by themselves (endogenous) or other organisms (exogenous). Therefore, while being major producers of antibiotics, actinomycetes are the major sources of antibiotic resistance mechanisms3. The mechanisms of antibiotic resistance found in clinical pathogens derive their origin from environmental bacteria as first identified for aminoglycoside resistance in Streptomyces4. Since then, various parallel examples, such as vanHAX gene cluster for vancomycin resistance, were reported in soil actinomycetes as well as in clinical strains5.
Whether living inside the human body or in natural environment, bacteria are exposed to wide concentration ranges of antibiotics. In most cases, they are exposed to non-lethal or sub-minimal inhibitory concentration (MIC) of antibiotics. Antibiotics at sub-MIC act as signals and stressors to elicit physiological and genetic changes to cope with antibiotic stress6,7,8. Antibiotic resistance phenotype is induced by sub-inhibitory antibiotics through modulating gene expression and physiology (intrinsic resistance) or through changing genetic information via mutation or horizontal transfer of resistance genes (acquired resistance). Modulation of bacterial gene expression to enhance intrinsic resistance is mediated via hosts of regulators. Some known transcriptional regulators include RNA polymerase sigma factors such as RpoS9, a redox-sensitive regulator such as SoxR10,11, or a WhiB-like factor (WblC/WhiB712,13). Unraveling the vast array of regulatory pathways and their networks are needed to understand and control resistance mechanisms.
Among regulators that respond to environmental changes, a group of alternative sigma factors called extra-cytoplasmic function (ECF) sigma factors are abundantly encoded in bacterial genomes14,15. They are also called group 4 sigma factors consisting of only σ2 and σ4 domains that recognize −35 and −10 regions, respectively, of cognate promoters16,17. In Streptomyces coelicolor, 50 such factors are encoded in the genome18. Among them, the role of only several factors has been elucidated, such as SigR (SCO521619), BldN (SCO332320), SigU (SCO296421), SigE (SCO335622), SigT (SCO389223) and SigQ (SCO490824).
The SigR system in S. coelicolor is activated by thiol-reactive chemicals that oxidize or alkylate cysteine thiols19,25. The induction mechanism involves the inactivation of its anti-sigma factor RsrA via forming disulfide bonds, and liberating active SigR26,27,28, which then positively regulates the expression of its own gene from the SigR-dependent upstream promoter (sigRp2) (Fig. 1A). The positively amplified sigRp2-derived SigR protein contains N-terminally extended 55 more amino acid residues, and is called σR′ to distinguish it from the apparently constitutive form σR expressed from the downstream promoter (sigRp1)29. A prominent difference between σR and σR′ is in their stability. Whereas σR is stable for hours, σR′ is short-lived with a half-life of ~10min29. Both σR and σR′ in their free state bind the core RNA polymerase and transcribe over 100 target genes to cope with the thiol-oxidative stress30. The SigR regulon includes the thiol-reducing systems which contribute to reactivating RsrA via disulfide reduction. It also includes proteases which degrade σR′, thereby turning off the response within an hour26,29. Therefore, the response of SigR-RsrA system to thiol-reactive chemical stresses is transient and is mediated by sensor RsrA and amplified σR′.
In this study, we demonstrate that multiple antibiotics induce the SigR system via yet another pathway of signal transduction, different from what conveys the thiol-perturbing signals. We show that the antibiotic induction of the SigR system proceeds via increasing the production of stable σR, and this induction is mediated by WblC/WhiB7. WblC is a WhiB-like protein conserved in actinomycetes31,32,33 and reported to confer resistance to antibiotics in Mycobacterium and Streptomyces12,34. WblC/WhiB7 proteins contain three functional domains such as an Fe-S cluster binding domain with four conserved cysteines, a G(V/I)WGG turn, and an AT-hook DNA binding domain35. The whiB7 gene is known to be induced by a variety of antibiotics via autoregulation, and WhiB7 may contribute to intrinsic resistance to antibiotics by activating antibiotic export, antibiotic inactivation and changes in thiol redox balance in mycobacteria13,36. Our work verifies sigRp1 promoter region as a novel binding site of WblC/WhiB7 in S. coelicolor, and suggests that the expression of SigR-homologous ECF sigma factor genes (sigE and sigH) in M. tuberculosis may also respond to antibiotics via WhiB7.
While performing hygromycin-chase experiment to measure the half-life of σR and σR′ proteins, we previously observed an increase in transcripts from the sigRp1 promoter29. This observation was unexpected since we used to regard the sigRp1 promoter as constitutive. We examined the effect of other antibiotics and compared it with that of thiol oxidant diamide. Figure 1B shows the induction profile of sigRp1 and sigRp2 transcripts after treatment with tetracycline (2μg/ml) or diamide (0.5mM) for up to 2h. Similarly to hygromycin, sigRp1 transcripts increased significantly by about 10-fold in response to tetracycline, in a prolonged fashion. This contrasts with the transient induction of sigRp2 transcripts by diamide as previously observed19,26. The antibiotic induction of sigRp1 transcription does not seem to be mediated by SigR itself, unlike sigRp2 transcription, since the induction occurred even in the ΔsigR mutant (MK1), where all transcriptions from the sigRp2 promoter disappeared (Fig. S1).
To investigate the induction of sigR mRNAs by antibiotics in further detail, we explored diverse antibiotics with different chemical structures and targets. Following 30min treatments at varying concentrations, the sigR transcripts were monitored by S1 mapping. The results demonstrated that translation-inhibiting antibiotics such as chloramphenicol, erythromycin, and lincomycin all induced sigRp1 expression significantly (Fig. 2A). Fusidic acid and streptomycin also induced sigRp1 transcripts (data not shown). On the other hand, ampicillin, norfloxacin, and rifampicin that affects cell wall, DNA replication, and transcription, respectively, failed to increase transcripts from sigRp1 (Fig. 2B). Rifampicin induced sigRp2 expression at 2μg/ml, as observed previously in a different S. coelicolor strain M60037. Thus, the sigRp1 expression is induced specifically by translation-inhibiting antibiotics. Determination of growth inhibitory concentrations for treated antibiotics (Fig. S1) indicated that the sigRp1 induction occurred at sub-inhibitory concentrations.
Whether the increase in sigRp1 transcripts leads to increased σR protein level in the presence of translation-inhibiting antibiotics was then examined. Analytical Western blot analysis with anti-SigR antibody revealed that erythromycin (0.25μg/ml) increased the level of σR, but not σR′ protein, continuously for up to 2h (Fig. 3A). This contrasts with the effect of thiol oxidant diamide which increased the amount of σR′ transiently by about 12-fold, without affecting the level of σR (Fig. 3A). Parallel detection of known amounts of σR protein enabled the estimation that σR increased steadily by erythromycin to about 3-fold level at 2h after treatment compared with the untreated level. The basal amounts of σR and σR′ proteins under non-treated condition were estimated to be about 23 (1.82μM) and 7 (0.56μM) fmole/μg proteins in cell extracts, respectively, assuming equal immune-specificity of σR and σR′ proteins to the antibody used. This corresponds to about 1.8 and 0.6μM in the cell for σR and σR′, respectively, assuming that about 43% of dry cell weight is from the protein, and that the wet cell weight is about 5.6 fold of the dry weight, and that cell density is 138. Following erythromycin treatment, there appeared a non-specific band which is absent in other antibiotic-treated samples (NS in Fig. 3A). The source of this protein band is not certain, except that it is not the product of the sigR gene, since it is observed in the ΔsigR mutant after erythromycin treatment. Treatments with chloramphenicol, lincomycin, and tetracycline caused similar increase in σR without changing the amount of σR′ (Fig. 3B). No increase in σR′ by antibiotics in spite of some increase in sigRp2 transcripts could be due to the unstable nature of σR′29.
We then examined the expression of a SigR-target gene trxB (SCO3890), which encodes thioredoxin reductase. Figure 3C shows that the SigR-dependent trxBp1 transcripts increased significantly by chloramphenicol and tetracycline treatments up to 80min, consistent with the steady increase in σR protein. Therefore, we conclude that the translation-inhibiting antibiotics induce the production of stable σR protein, which subsequently induces its target gene expression in a prolonged fashion.
To find clues to reveal mechanisms behind antibiotic induction of sigRp1, we scrutinized its flanking sequences. One prominent feature was a stretch of AT-rich sequence, which is not common in GC-rich actinomycetous genomes, located immediately upstream of the −35 region of the sigRp1 promoter (Fig. 4A). This sequence feature is present upstream of the whiB7 promoter in Mycobacterium species, and has been proposed as the binding site of a WhiB-like (Wbl) protein WhiB736,39. In S. coelicolor, WblC (SCO5190) is the orthologue of WhiB7 of M. tuberculosis, and the wblC gene also has a putative auto-regulatory WblC-binding signature similarly to the whiB7 gene of M. tuberculosis (Fig. 4A). The wblC and whiB7 mutants were reported to be hypersensitive to diverse antibiotics in S. lividans, S. coelicolor, and M. tuberculosis12,34. Inspection of the promoter region of sigR-homologous genes (sigE and sigH) in M. tuberculosis H37Rv also revealed the presence of putative WhiB7-binding sites immediately upstream of the promoters40 (Fig. 4A).
We investigated whether WblC is involved in inducing transcription from the sigRp1 promoter upon antibiotic treatment. The wild type and the ΔwblC mutant cells34 were treated with tetracycline for up to 3h, and examined for sigR-specific transcripts and their protein products by S1 mapping and Western blot analyses, respectively. Results in Fig. 4B demonstrated that WblC is critically required for the antibiotic induction of sigRp1 transcription. The sigRp2 transcription, however, was induced by tetracycline regardless of the wblC mutation. Immunoblot analysis revealed that the σR protein produced from the sigRp1 transcripts did not increase in ΔwblC mutant, in contrast to the wild type, where σR protein increased about 2.5-fold during the 2 to 3h treatments with tetracycline (Fig. 4C). These results clearly show that the increase in stable σR after antibiotic treatment depends almost entirely on WblC/WhiB7.
We then investigated how WblC is involved in antibiotic induction of sigRp1 or σR. For this purpose, polyclonal antibodies against WblC were raised in rabbits, and used to monitor WblC in cells treated with antibiotics. Figure 5A shows that the amount of WblC dramatically increased within an hour of erythromycin or tetracycline treatments. The WblC level decreased within 2h of antibiotic treatment. The decrease at 2h is more pronounced in erythromycin than tetracycline treated samples. With some slight differences in induction and shut-off kinetics, WblC was induced by other antibiotics such as hygromycin, chloramphenicol, and lincomycin to a maximal level within an hour, and then returned to the basal level within 2 or 3h (Fig. 5B).
Whether WblC binds directly to the sigRp1 promoter region in vivo was determined by chromatin immunoprecipitation (ChIP) analysis. The wild type and the ΔwblC mutant cells were treated with tetracycline (2μg/ml) for 1h, followed by fixation, cell lysis, DNA shearing, and immunoprecipitation with anti-WblC antibody as described in Materials and Methods. The amount of sigRp1 promoter DNA in the precipitate was estimated by quantitative real-time PCR (qRT-PCR), along with probe sets for the upstream sigRp2 or downstream rsrA regions. Figure 5C demonstrates that tetracycline increased WblC binding to the sigRp1 promoter region (from −84 to +7 nucleotide position, relative to the transcription start site of sigRp1) by more than 10-fold in the wild type cell, whereas no increased binding was observed in the ΔwblC mutant. In comparison, no significant binding of WblC to the sigRp2 or rsrA regions was observed following tetracycline treatments. Therefore, we can conclude that the antibiotic treatments increase the amount of WblC, which specifically binds to the sigRp1 promoter region and mediates increased expression of σR.
On the basis of induction by antibiotics, we hypothesized that the sigR gene functions in conferring resistance to antibiotics in S. coelicolor. So far, the revealed phenotypes of ΔsigR mutant are the sensitivity to thiol oxidant diamide19, sensitivity to electrophiles (Park JH, unpublished), and increased protein aggregation in cell extracts that reflects decreased protein quality control41. To assess antibiotic sensitivity, we spotted an equal number of spores from the wild type, ΔsigR, ΔwblC, and ΔsigR complemented with the chromosomally integrated sigR gene, on plates containing various antibiotics. Figure 6 shows that the ΔsigR and ΔwblC mutations do not cause sensitivity toward non-inducing antibiotics such as ampicillin, norfloxacin, or rifampicin. However, as predicted, the ΔsigR mutant was more susceptible to inducing antibiotics such as chloramphenicol, erythromycin, lincomycin, and tetracycline. The sensitivity was restored to the wild type level by complementation with the wild type sigR gene. The ΔwblC was more susceptible than the ΔsigR mutant to the inducing antibiotics except chloramphenicol. These results demonstrate that the sigR gene does play a critical role in ensuring cell viability in the presence of translation-inhibiting antibiotics.
M. tuberculosis (Mtb) has two close homologs of SigR from S. coelicolor (ScoSigR); SigE (Rv1221; MtbSigE) and SigH (Rv3223c; MtbSigH) with 37% and 72% identity, respectively. SigH is known to regulate the thioredoxin system and heat shock proteins upon oxidative and heat stresses42,43. SigE plays a role in response to oxidative and cell envelop stresses44. The presence of predicted WblC binding sites in the promoter regions of sigE and sigH (Fig. 4A) led us to examine the expression of these genes in Mtb upon antibiotic treatments. We treated Mtb H37Rv cells with 1 μg/ml each of erythromycin, streptomycin, or tetracycline for up to 3 days. Figure 7 demonstrates the results of S1 nuclease mapping of transcripts from the sigE (panel A) and sigH (panel B) genes. For Mtb_sigE gene, we detected transcripts from the two promoters (transcription start sites) as have been reported40. The sigEp1 promoter contains the WhiB7-binding motif and produces leaderless mRNA (Fig. 4A). The upstream promoter sigEp2 does not have WhiB7-binding motif but contains the promoter sequence feature recognizable by MtbSigE or MtbSigH45. We found that the sigEp1 transcripts increased significantly by all three antibiotics (Fig. 7A). The sigEp2 transcripts increased also by antibiotic treatments, but by less pronounced fold of induction. For Mtb_sigH gene, we detected transcripts from two promoters; one from the downstream sigHp1 as reported previously40 and the other from the upstream sigHp2 recognizable by MtbSigH42,46. A WhiB7-binding motif is present in the sigHp2 promoter (Fig. 4A). Results in Fig. 7B show that both sigHp1 and sigHp2 transcripts increased by antibiotics, even though not as much as the sigE transcripts. Based on these observations, we can predict that similar pathways of upregulating SigR-like sigma factors by antibiotics are present in M. tuberculosis, and MtbSigE may play a more significant role in orchestrating response against translational blocking antibiotics.
In this work, we demonstrated that the sigR gene expression is induced by translation-inhibiting antibiotics to produce a stable isoform of SigR, σR, which elevates its target gene expression for a prolonged period, in contrast to a transient induction of σR′ by thiol-oxidative stresses. We also found that the sigR gene confers resistance to these inducing antibiotics. Previously, we identified 108 direct target genes of SigR by using ChIP-chip analysis30. Since the ChIP experiment was done after diamide treatment for 30min, when the majority of the sigR gene product was σR′ (more than 80% of the total SigR; Fig. 3A), the SigR regulon we determined reflects primarily the promoters preferentially bound by σR′. Since σR′ differs from σR′ only by the N-terminal 55 amino acids, which may not affect promoter recognition, we consider the σR′-bound genes may not differ from σR-binding genes. Quite a number of SigR-target genes encode functions for thiol redox homeostasis, proteolysis, and ribosome modulation30,41.
Treatment with translation-inhibiting antibiotics will not only slow down the synthesis of new proteins, but also result in misfolded protein products due to mistranslation or protein truncation47,48. Stalled ribosomes uncoupled with transcription can cause mRNA cleavage, resulting in ribosome stuck at non-stop mRNA, which produces non-functional truncated protein upon ribosome rescue49,50. Therefore, the cellular damages caused by thiol-disturbing oxidative stress can overlap with those by translation-inhibiting antibiotics to quite an extent. In light of this, the functions of predicted ribosome-associated proteins of SigR regulon such as tmRNA (ssrA), RelA, HflX, peptide-releasing factor PrfA, EngA, and ObgE need be further investigated30.
Then, why is prolonged induction of SigR required to cope with antibiotics, whereas transient induction is sufficient to cope with oxidative stress? Our results implicate that S. coelicolor takes longer time to overcome antibiotic stress than thiol-oxidative stress. Thiol oxidants and electrophiles that elicit thiol-oxidative stress are efficiently removed in Streptomyces by mycothiol a functional equivalent of glutathione in actinomycetes25. Increased production and recycle of mycothiol, along with increased thiol-reducing systems, after thiol-oxidative stress will efficiently remove chemical stressors and return the thiol redox environment back to normal in a relatively short period of time. On the contrary, the antibiotics that bind to the ribosome is harder to be cleared from the cell, affecting cell physiology for longer period of time51. This may necessitate the utilization of stable regulator, such as stable σR, that can carry out the response for prolonged period of time.
We observed that the antibiotics that induced sigRp1 transcription also induced sigRp2 transcription, even though to a lesser extent (Figs 1B,B,2A2A and and4B).4B). The antibiotic induction of sigRp2 almost entirely depends on SigR, since no sigRp2 transcripts were observed in ΔsigR mutant (Fig. S1). Part of the reason that sigRp2 is induced by antibiotics is due to the secondary effect of increased σR that recognizes sigRp2. The results in Fig. 4B, which show that the sigRp2 is still induced by tetracycline in the ΔwblC mutant is not easy to explain. In the absence of WblC, no increase in σR is observed, and therefore, the sigRp2 induction is likely to occur via the pathway of inactivating RsrA (Fig. 1A). It can be speculated that somehow the intracellular environment of ΔwblC is more oxidized than the wild type following antibiotic treatment.
The more interesting question is how the production WblC protein is drastically elevated in the presence of translation-inhibiting antibiotics. The wblC/whiB7 mRNA contains unusually long 5′ UTR with possible ORF for a small protein. This feature appears conserved across actinomycetes52, and may play some role in elevating WblC expression upon slowing down translation. There is also a possibility that the wblC gene expression partly depends on SigR, as predicted from the presence of SigR-dependent promoter sequence upstream of the wblC gene. The finding that the extent of antibiotic induction of the sigRp1 transcription in the ΔsigR mutant reduced to about half of the wild type level supports this idea (Fig. S1). Further studies are in need to unravel the underlying mechanism.
Spores of S. coelicolor A3(2) strain M145, ΔsigRrsrA disruptant (MK1)29 and ΔwblC disruptant34 were inoculated in YEME liquid medium containing 5mM MgCl2•6H2O and 10% sucrose, and were grown at 30°C53. NA plates (0.8% nutrient broth, 2% agar powder) were used for spotting analysis. E. coli was grown in LB broth. The pSET152H plasmid and E. coli ET12567, a non-methylating strain containing pUZ8002 for donor functions, were used for complementation as recommended54. E. coli DE3/gold strain and pET15b plasmid was used for WblC over-expression. M. tuberculosis H37Rv cells were grown at 37°C in Middlebrooks 7H9 broth supplemented with 10% OADC and with or without antibiotics.
Antibiotics were obtained from Sigma-Aldrich and Duchefa biochemie. The solutions were prepared freshly before treatments.
S. coelicolor cells grown to OD600 of 0.3~0.4 in YEME were treated with various antibiotics or 0.5mM diamide for 30~120min. Harvested cells were disrupted by sonication in Kirby mix. RNA isolation and S1 nuclease protection assay were done as described previously29. For mycobacterial RNA preparation, harvested cells were re-suspended in TRIzol® Reagent (Ambion, Life Technologies, Carlsbad, CA, USA), mixed with acid-washed 425~600 glass beads (Sigma-Aldrich G8772), and lysed using a mini-bead beater (BioSpec, Bartlesville, OK, USA). Following chloroform extraction and isopropanol precipitation, the RNA pellet was resolved in RNA-free water (Ambion, Life Technologies, Carlsbad, CA, USA). Mycobacterial RNA was analyzed by S1 nuclease protection assay as described previously29.
Cell lysates were obtained by sonication, and its protein concentration was determined as described previously30. To detect SigR, cell extracts containing 25μg protein were diluted to the final concentration of 0.125μg/μl with lysis buffer (200μl) that contains 100μg BSA to serve as a protein buffer. Aliquots of 8μl containing 1μg crude protein extract and 4μg BSA were resolved on 15% SDS-PAGE. Immuno-detection was done by using polyclonal rabbit antibody against SigR and the anti-rabbit secondary antibody at 1:5000 dilution ratio, followed by ECL detection (Amersham Life Science). For detecting WblC, cell extracts containing 20μg protein were resolved on 15% SDS-PAGE. Immuno-detection was done by using polyclonal rabbit antibody against WblC and the anti-rabbit secondary antibody at 1:10000 dilution ratio. All experimental protocols that involve animals were approved by and done in accordance with the guidelines by Seoul National University Institutional Animal Care and Use Committees (SNUIACUC).
NA plates containing various antibiotics (20μg/ml ampicillin, 1μg/ml norfloxacin, 2μg/ml rifampicin, 10μg/ml chloramphenicol, 2μg/ml erythromycin, 10μg/ml lincomycin, or 2μg/ml tetracycline) were used to monitor sensitivity. An equal number of spores of wild type and mutant S. coelicolor strains were serially diluted by 10-fold and spotted on antibiotic-containing NA plates using a 48-pin replica plater (Sigma). The spotted plates were incubated at 30°C for up to 3 days before taking photos.
Exponentially grown cells (at OD600 of 0.3~0.4) were treated with 2μg/ml tetracycline for 1h, followed by fixation with 1% formaldehyde for 15min. 125mM glycine was subsequently added for 5min at room temperature. Harvested cells were washed twice with cold TBS wash buffer (20mM Tris-HCl, pH 7.5, 150mM NaCl). To break cells and shear DNA, cells were sonicated in RIPA buffer (50mM HEPES-KOH pH 7.5, 150mM NaCl, 1mM EDTA pH 8.0, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS, 1mM PMSF) with a sonicator (QSonica Q500) using a 3mm tip at 30% maximum power, with 5sec pulses for 15 times on ice. Following centrifugation at 13000rpm and 4°C for 10min to clear the cell debris, 50μl of each supernatant was set aside for input DNA control. To the cleared supernatant anti-WblC polyclonal rabbit antibody (5μl) was added, and incubated at 4°C for 1h, with gentle mixing by rotation. Subsequently, 20μl protein A/G beads (Santacruz) and 2μg BSA were added and rotated overnight at 4°C. The samples were centrifuged for 1min at 4°C and 3000rpm and the pellets were washed once with low salt wash buffer (50mM HEPES-KOH, pH 7.5, 150mM NaCl, 1mM EDTA, 1% Triton X-100, 0.1% sodium deoxycholate), once with high salt wash buffer (50mM HEPES-KOH, pH 7.5, 500mM NaCl, 1mM EDTA, 1% Triton X-100, 0.1% sodium deoxycholate), once with LiCl wash buffer (10mM Tris-HCl, pH 8.0, 250mM LiCl, 1mM EDTA, 1% NP-40, 1% sodium deoxycholate), and twice with TE buffer (10mM Tris-HCl, pH 8.0, 1mM EDTA). DNA was eluted by incubation in the elution buffer (10mM Tris-HCl, pH 8.0, 250mM NaCl, 1mM EDTA, 1% SDS) at 65°C for 30min, followed by treatment with 5μg proteinase K and 2μg RNaseA for 1h at 45°C. NaCl was added to final concentration of 350mM, and incubation continued at 65°C overnight for reverse-crosslinking. DNA was purified by phenol-chloroform extraction. The amount of sigRp1, sigRp2, and rsrA-specific DNA was quantified by qPCR (Agilent Stratagene Mx3000P), using primer sets which encompass the sigRp1 promoter region (from −125 to −34nt position, relative to the sigR start codon), sigRp2 promoter region (from −311 to −186nt position, relative to the sigR start codon), and rsrA (from +920 to +991nt position, relative to the sigR start codon).
How to cite this article: Yoo, J.-S. et al. Induction of a stable sigma factor SigR by translation-inhibiting antibiotics confers resistance to antibiotics. Sci. Rep. 6, 28628; doi: 10.1038/srep28628 (2016).
We thank Dr. Keith Chater (John Innes Institute) for providing the wblC mutant strain. This work was supported by a grant to J.-H. Roe (2014R1A2A1A01002846) from the Ministry of Science, ICT and Future Planning. J.-S. Yoo and G.-S. Oh were supported by B.K. Plus Fellowship for Biological Sciences at Seoul National University.
Author Contributions J.-S.Y., G.-S.O. and S.W.R. performed the experiments. J.-S.Y. and J.-H.R. conceived the work and wrote the manuscript.