Introduction

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 ~10 min29. 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′.

Figure 1: Induction of sigR transcription by antibiotics via downstream p1 promoter.
figure 1

(A) The regulatory loops in activating SigR regulon. Two isoforms of SigR, σR and σR′, are produced from the two promoters of sigR gene, sigRp1 and sigRp2, respectively. Under cytoplasmic reducing environment, the reduced RsrA binds SigR (primarily the abundant σR), inhibiting SigR-directed transcription. Upon oxidative stress by thiol-oxidants such as diamide, di-sulfide bonds are formed in RsrA, and SigR is released from sequestration. The released SigR directs transcription of more than 100 genes (SigR regulon) that includes its own gene (from the upstream sigRp2 promoter). In contrast to σR that is very stable, σR′ with 55 more N-terminal amino acids is very unstable. (B) Difference between induction by thiol oxidant diamide and antibiotic tetracycline. The wild-type cells were sampled at 0, 30, 60 and 120 min after treatment with tetracycline (2 μg/ml, or 4.16 μM) or diamide (0.5 mM) for S1 nuclease mapping of sigR-specific RNAs. The rRNA in each RNA sample were resolved in parallel. Results from three independent experiments were quantified to present values for average fold induction with standard error of the mean (s.e.m.).

Full size image

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.

Results

Induction of the sigR gene expression by translation-inhibiting antibiotics

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 (2 μg/ml) or diamide (0.5 mM) for up to 2 h. 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 30 min 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 2 μg/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.

Figure 2: Translation-inhibiting antibiotics induce sigRp1 transcription.
figure 2

(A) Effect of antibiotics that target translation. S. coelicolor cells were sampled at 30 min after treatments with chloramphenicol (0 to 40 μg/ml), erythromycin (0 to 2 μg/ml), or lincomycin (0 to 40 μg/ml). S1 nuclease protection assay for sigR-specific transcripts were done. The rRNA in each sample were presented as a control. The asterisk (*) denotes inhibitory concentration of the antibiotics (see Fig. S1). (B) Effect of antibiotics that target other cellular processes; cell wall synthesis (ampicillin from 0 to 200 μg/ml), DNA replication (norfloxacin from 0 to 80 μg/ml), and transcription (rifampicin from 0 to 8 μg/ml). The sigR-specific RNA analysis was done as in panel (A).

Full size image

Antibiotic treatment increases σR protein and steadily induces target gene expression

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 (0.25 μg/ml) increased the level of σR, but not σR′ protein, continuously for up to 2 h (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 2 h 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 (1.82 μM) and 7 (0.56 μM) 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 0.6 μM 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.

Figure 3: Steady increase in σR protein by antibiotic treatments and prolonged induction of its target promoter (trxBp1).
figure 3

(A) Steady vs. transient increase in SigR proteins by antibiotic or thiol oxidant. S. coelicolor cells were treated with either erythromycin (0.25 μg/ml) or diamide (0.5 mM) for up to 2 h, followed by western blot analysis with antibody against SigR. The positions of σR and σR′ were marked by arrows. A non-specific band (NS) produced in erythromycin-treated samples was also indicated. Analytical western blotting of indicated amounts of purified SigR (His-σR; from 0.25 to 5 ng) with anti-SigR antibody was done in parallel to quantify the amount of SigR-specific protein bands. (B) Western blot analyses of SigR proteins following treatments with chloramphenicol, lincomycin, and tetracycline for 2 h. (C) S1 mapping analysis of trxB transcripts. The wild-type (M145) and ΔsigRrsrA (MK1) cells were treated with chloramphenicol (17 μg/ml) or tetracycline (5 μg/ml) for up to 80 min, and analyzed for trxB transcripts. The trxBp1 promoter is under the control of SigR.

Full size image

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 80 min, 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.

Antibiotic induction of stable σR depends on WblC/WhiB7

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).

Figure 4: Antibiotic induction of sigRp1 transcription and σR production depends on WblC/WhiB7.
figure 4

(A) The presence of AT-rich sequence adjacent to the −35 element of promoters for sigRp1 in S. coelicolor and its homologous genes (sigE and sigH) in M. tuberculosis. The AT-rich sequences upstream of the wblC in S. coelicolor, and the whiB7 promoters known to bind WhiB7 protein in mycobacteria, were also presented. (B) Antibiotic induction of sigRp1 transcripts depends on WblC/WhiB7. S1 nuclease mapping was done, following treatment of the wild type and ΔwblC cells with 2 μg/ml tetracycline for up to 2 h. Results from three independent experiments were quantified to present average values for relative fold change and s.e.m. (C) Western blot analysis of SigR proteins from cells similarly treated as in panel (B). Results from three independent experiments were quantified.

Full size image

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 3 h, 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 3 h treatments with tetracycline (Fig. 4C). These results clearly show that the increase in stable σR after antibiotic treatment depends almost entirely on WblC/WhiB7.

Antibiotics increase the amount and the binding of WblC to sigRp1 promoter in vivo

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 2 h of antibiotic treatment. The decrease at 2 h 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 3 h (Fig. 5B).

Figure 5: Increase in the amount of WblC protein and its binding to the sigRp1 promoter in vivo upon antibiotic treatments.
figure 5

(A) Western blot analysis of WblC. The wild-type and ΔwblC cells were sampled after antibiotic treatments; Ery for 0.25 μg/ml erythromycin, and Tet for 2 μg/ml tetracycline. The WblC-specific band was detected slightly below the 15 kDa marker, coinciding with its predicted size (13.2 kDa). NS denotes non-specific band. (B) Western blot analysis of WblC in wild type cells treated with other translation-inhibiting antibiotics; Chl for 10 μg/ml chloramphenicol, Hyg for 5 μg/ml hygromycin, and Lin for 5 μg/ml lincomycin. (C) Chromatin immunoprecipitation with anti-WblC polyclonal antibody followed by q-PCR with gene-specific primer sets for sigRp1, sigRp2, and rsrA genes. The wild-type and ΔwblC cells were processed for immunoprecipitation after 1 h treatment with or without 2 μg/ml tetracycline. The enrichment of each region was estimated by quantitative real-time PCR. The relative average fold with s.e.m. were presented (y-axis), by taking the value for untreated ΔwblC sample as 1. The asterisk (*) denotes p ≤ 0.05 by Student’s t-test (n = 3).

Full size image

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 (2 μg/ml) for 1 h, 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.

SigR confers resistance to translation-inhibiting antibiotics

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.

Figure 6: SigR confers resistance to translation-inhibiting antibiotics.
figure 6

An equal number of spores of the wild-type, ΔsigR, ΔwblC, and ΔsigR+sigR complemented strains were serially diluted by 10-fold and spotted on NA plates with or without antibiotics. Concentrations of antibiotics in the plates were 20 μg/ml ampicillin, 1 μg/ml norfloxacin, 2 μg/ml rifampicin, 10 μg/ml chloramphenicol, 2 μg/ml erythromycin, 10 μg/ml lincomycin, or 2 μg/ml tetracycline. Plates were incubated for 40 to 72 h.

Full size image

Induction of sigR-homologous genes (sigE and sigH) by antibiotics in M. tuberculosis

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.

Figure 7: Induction of sigR-homologous gene (sigE and sigH) expression by antibiotics in M. tuberculosis (Mtb).
figure 7

Transcripts from the sigE panel (A) and sigH panel (B) genes of Mtb H37Rv strain were analyzed by S1 mapping. Mtb RNAs were obtained from cells grown in Middlebrooks 7H9 broth, and either non-treated or treated with 1 μg/ml of antibiotics for 24, 48, and 72 h: Ery, erythromycin; Str, streptomycin; Tet, tetracycline. Results from more than four independent experiments (n = 4 for sigE and n = 6 for sigH) were quantified to estimate changes in the level of transcripts, taking the level of untreated sample as 1. The average fold changes with s.e.m. were presented for transcripts from the p1 (downstream) and p2 (upstream) promoters of sigE and sigH genes. The sigEp1 and sigHp2 promoters (marked with *) contain putative WhiB7 binding motifs.

Full size image

Discussion

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 30 min, 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,2A and 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.

Materials and Methods

Strains, plasmids, and growth conditions

Spores of S. coelicolor A3(2) strain M145, ΔsigRrsrA disruptant (MK1)29 and ΔwblC disruptant34 were inoculated in YEME liquid medium containing 5 mM 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 and reagents

Antibiotics were obtained from Sigma-Aldrich and Duchefa biochemie. The solutions were prepared freshly before treatments.

RNA preparation and S1 nuclease protection assay

S. coelicolor cells grown to OD600 of 0.3~0.4 in YEME were treated with various antibiotics or 0.5 mM diamide for 30~120 min. 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.

Immuno-blot analysis

Cell lysates were obtained by sonication, and its protein concentration was determined as described previously30. To detect SigR, cell extracts containing 25 μg protein were diluted to the final concentration of 0.125 μg/μl with lysis buffer (200 μl) that contains 100 μg BSA to serve as a protein buffer. Aliquots of 8 μl containing 1 μg crude protein extract and 4 μg 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 20 μg 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).

Spotting assay to monitor antibiotic sensitivity

NA plates containing various antibiotics (20 μg/ml ampicillin, 1 μg/ml norfloxacin, 2 μg/ml rifampicin, 10 μg/ml chloramphenicol, 2 μg/ml erythromycin, 10 μg/ml lincomycin, or 2 μg/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.

Chromatin immuno-precipitation

Exponentially grown cells (at OD600 of 0.3~0.4) were treated with 2 μg/ml tetracycline for 1 h, followed by fixation with 1% formaldehyde for 15 min. 125 mM glycine was subsequently added for 5 min at room temperature. Harvested cells were washed twice with cold TBS wash buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl). To break cells and shear DNA, cells were sonicated in RIPA buffer (50 mM HEPES-KOH pH 7.5, 150 mM NaCl, 1 mM EDTA pH 8.0, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS, 1 mM PMSF) with a sonicator (QSonica Q500) using a 3 mm tip at 30% maximum power, with 5 sec pulses for 15 times on ice. Following centrifugation at 13000 rpm and 4 °C for 10 min to clear the cell debris, 50 μl of each supernatant was set aside for input DNA control. To the cleared supernatant anti-WblC polyclonal rabbit antibody (5 μl) was added, and incubated at 4 °C for 1 h, with gentle mixing by rotation. Subsequently, 20 μl protein A/G beads (Santacruz) and 2 μg BSA were added and rotated overnight at 4 °C. The samples were centrifuged for 1 min at 4 °C and 3000 rpm and the pellets were washed once with low salt wash buffer (50 mM HEPES-KOH, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% sodium deoxycholate), once with high salt wash buffer (50 mM HEPES-KOH, pH 7.5, 500 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% sodium deoxycholate), once with LiCl wash buffer (10 mM Tris-HCl, pH 8.0, 250 mM LiCl, 1 mM EDTA, 1% NP-40, 1% sodium deoxycholate), and twice with TE buffer (10 mM Tris-HCl, pH 8.0, 1 mM EDTA). DNA was eluted by incubation in the elution buffer (10 mM Tris-HCl, pH 8.0, 250 mM NaCl, 1 mM EDTA, 1% SDS) at 65 °C for 30 min, followed by treatment with 5 μg proteinase K and 2 μg RNaseA for 1 h at 45 °C. NaCl was added to final concentration of 350 mM, 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 −34 nt position, relative to the sigR start codon), sigRp2 promoter region (from −311 to −186 nt position, relative to the sigR start codon), and rsrA (from +920 to +991 nt position, relative to the sigR start codon).

Additional Information

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).