SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

6H57, a 69-nucleotide-long small RNA, was isolated in shotgun cloning using an RNA sample derived from early stationary-phase cells. The 6H57 gene is located in a 798-bp intergenic region between two acid resistance-related genes, hdeD and gadE, and is encoded on the strand opposite these flanking genes. In this study, we carried out stringent Northern blotting to determine target mRNAs of 6H57. A band approximately 1300 nucleotides in length was detected using a probe containing a partial sequence of 6H57 and was confirmed to be the gadE mRNA T3, which has a 566-nucleotide-long 5′ untranslated region. These results show that 6H57 is an antisense RNA of gadE mRNA T3 and can base pair with a −380 to −312 region of the translation initiation site of gadE. We analyzed the transcription of 6H57 and showed that 6H57 transcription is dependent on GadE in the early stationary phase. Furthermore, 6H57 is induced in the exponential growth phase by an acid stimulus of pH 5.5. A 189-bp DNA fragment containing the upstream region of the 6H57 gene showed clear promoter activities in these culture conditions. These results suggest that 6H57 plays several roles in acid resistance, and we renamed it acid resistance-related small RNA.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Escherichia coli are highly resistant to acid and can even survive at pH 2.0 for several hours (Small et al. 1994 and Lin et al. 1995, 1996). This extraordinary acid resistance (AR) ability depends on the growth medium (Foster 2004). E. coli are known to possess four AR systems. AR1 is an RpoS-dependent glucose-repressed system and involves F0/F1 proton-translocating ATPase, although the precise mechanisms underlying the function of this system remain unclear. AR2, AR3 and AR4 are based on similar mechanisms and require the presence of glutamate, arginine, and lysine, respectively. AR2, the glutamate-dependent AR (GDAR) system, is the most efficient of the four systems. In the GDAR system, glutamate decarboxylase isozymes, GadA and GadB, replace the α-carboxyl group of glutamate with a proton recruited from the cytoplasm. γ-amino butyric acid is the end product of this reaction and is expelled by the membrane-bound antiporter GadC in exchange for new glutamates. GadE, formerly known as YhiE, is a LuxR family member and is an essential transcriptional activator in the GDAR system. GadE binds to the gadE box located 63 bp upstream of the gadA and gadBC transcriptional initiation sites.

We previously carried out shotgun cloning to isolate small RNAs (sRNAs) that appeared in the early stationary phase and identified a novel sRNA, 6H57 (Horie et al. 2007). 6H57 was found to be encoded on the reverse strand of two flanking genes, hdeD and gadE (Fig. 1A). The highest level of cellular 6H57 was observed in the early stationary phase (Fig. 1B); however, in the exponential growth phase, 6H57 was not observed. A cloned 6H57 fragment was 69 nucleotides in length and predicted to have three stem–loop structures; one located in the 3′ region appeared to be a ρ-independent transcriptional terminator (Fig. 1C). In addition, a promoter (−35: TTGTGA, −10: TAAAGT) was predicted to be located 7–35 bp upstream of 6H57 (Fig. 1D).

image

Figure 1.  Genomic location and expression of arrS. (A) Genetic map of arrS and its flanking genes. The gray box, open boxes, thick lines and thin arrows represent the arrS gene, flanking genes, oligonucleotide probes or a riboprobe, and primers for RT-PCR, respectively. (B) Total RNA (8 μg) isolated from W3110 cells cultivated for 2, 6 or 24 h was analyzed using the [γ-32P] ATP-labeled oligonucleotide probe c-6H57. (C) RNA secondary structure prediction for ArrS by the mfold. (D) A promoter predicted for arrS. The open boxes and the shaded letters represent the −35 and −10 regions of the promoter and the nucleotide sequence of arrS, respectively.

Download figure to PowerPoint

In this study, we clarified that 6H57 is an antisense RNA of a gadE transcript, T3, which has an extraordinarily long 5′ untranslated region (UTR). In addition, we investigated the regulation of 6H57 transcription and showed that 6H57 is transcribed, not only in the stationary but also in the exponential growth phase via an acid stimulus. The transcription of 6H57 was dependent on GadE in the early stationary phase and not under the acidic condition in the exponential growth phase. The cellular level of 6H57 was much higher in the early stationary phase than that observed under the acidic condition in the exponential growth phase. This result is in agreement with promoter activities in these two conditions. Based on these results, we renamed 6H57 ‘acid resistance-related sRNA (ArrS)’.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Detection of target mRNAs of ArrS

We first attempted to detect mRNAs containing a sequence that was completely complementary to ArrS using the oligonucleotide probe s-6H57 in stringent Northern blotting. The 1300-nucleotide bands indicated with an arrow were obtained when total RNAs derived from the exponential growth-phase cells or early stationary-phase cells were used (Fig. 2A, lanes 1 and 2). Furthermore, these bands were hybridized with an oligonucleotide probe c-gadE, which is specific to the open reading frame (ORF) of one flanking gene, gadE (Fig. 2B, lanes 1 and 2). However, an oligonucleotide probe c-hdeD, which marked another flanking gene, hdeD, did not anneal (Fig. 2C). These results show that the 1300-nucleotide band contains the antisense sequence of ArrS and the gadE coding sequence. Sayed & Foster (2009) showed that RNA derived from early stationary-phase cells grown in LB medium at pH 5.5 contains three gadE transcripts of 900, 1100 or 1380 nucleotides. In Fig. 2B lane 2, three bands appear to represent the 900-, 1100-, and 1380-nucleotide gadE transcripts. The band indicated by an arrow at approximately 1300-nucleotide appears to be the 1380-nucleotide gadE transcript. The 1380-nucleotide transcript, named T3 by Sayed and Foster, has a 566-nucleotide-long 5′ UTR (Fig. 1A).

image

Figure 2.  Detection of mRNAs complementary to ArrS by Northern blotting. Total RNA (8 μg) isolated from W3110 cells cultivated for 2, 6 or 24 h was analyzed using the oligonucleotide probes (A) s-6H57, (B) c-gadE or (C) c-hdeD. Arrow head indicates the 1300-nucleotide band.

Download figure to PowerPoint

We confirmed that this band is the gadE transcript T3 through circularization RT-PCR. For this experiment, we used total RNA extracted from the exponential growth-phase cells because the RNA sample did not contain any gadE transcripts other than the 1300-nucleotide transcript (Fig. 2B, lane 1). In this method, the 5′ and 3′ ends of mRNAs are ligated to each other by RNA ligase. The junction regions are amplified by RT-PCR, and then sequences of cDNAs are determined to map both the 5′ and 3′ ends of mRNAs. We determined the sequences of six cDNA clones and found that the 5′ end-points were located at −566 and −564 for five and one of the clones, respectively. These results show that the 1300-nucleotide gadE mRNA hybridized with s-6H57 is the gadE transcript T3. This indicates that ArrS could pair with the gadE mRNA T3 at 312–380 nucleotides upstream from the start codon.

Induction of ArrS by an acid stimulus in the exponential growth phase

GadE is a transcriptional activator that functions in the GDAR system (Foster 2004). In experiments of gadE–lacZ operon fusions, the gadE P3 promoter for transcript T3 had a 2.5-fold higher activity in the stationary phase than that in the exponential growth phase, and a 4-fold higher activity at pH 5.5 than at pH 7.8 in the exponential growth phase (Sayed & Foster 2009).

We investigated whether transcription of antisense RNA ArrS was induced by an acid stimulus during the exponential growth phase (Fig. 3). The pH of the exponentially growing cell culture in E media containing 0.5% glucose and 2% polypeptone was approximately 6.9. Under neutral pH, ArrS was not detected during the exponential growth phase (panel B, lanes 1–3). We added acid to the culture to adjust the pH to 5.5 as shown in the experimental procedure. Levels of gadE transcript T3 increased after the pH shift (panel A, lanes 4–6). Moreover, transcription of arrS was shown to be induced at 5 min, and a strong ArrS signal pointed by an arrow was obtained 30 min after the pH shift (panel B, lanes 5–6). In lane 6, smaller bands other than full-length ArrS indicated by an arrow were also detected. Similar bands, which were also observed in Figs 1B, 4A,B, 5B and 6A,C, possibly originated from an intact ArrS via an endoribonucleolytic cleavage at the single-strand position(s) of the structured ArrS (Fig. 1C).

image

Figure 3.  Expression of gadE mRNA T3 and ArrS under the acidic condition in the exponential growth phase. W3110 cells were cultivated for 2 h, and then acid was added to the culture to obtain a pH of 5.5. RNA samples were removed 1, 5 and 30 min after the addition of acid. (A) gadE mRNA T3 was analyzed using primers gadE RT2 and gadE (−558) in comparative RT-PCR. (B) Sixteen microgram of total RNAs were fractionated, and ArrS was analyzed using a non-RI-labeled oligonucleotide probe, c-6H57-49, in Northern blotting. An arrow indicates the full size of ArrS.

Download figure to PowerPoint

image

Figure 4.  Effects of gadE deletion on cellular levels of ArrS. (A) BW25113 (parent strain), ΔgadE or ΔgadE/pCA24N-gadE was grown for 5.5 h (early stationary phase), and IPTG was added. Samples were removed 30 min after addition of IPTG. (B) BW25113 (parent strain) and ΔgadE were grown for 1.5 h (exponential growth phase), and acid was added to the culture to obtain a pH of 5.5. Samples were removed 30 min after the addition of acid. ArrS was detected using ribo-5H57 as a probe. An arrow indicates the full size of ArrS.

Download figure to PowerPoint

image

Figure 5.  Promoter activity of arrS and cellular levels of ArrS in exponential growth-phase cells under the acidic condition or in early stationary-phase cells. (A) Luciferase assay was carried out using the luciferase fusion reporter plasmid p6H57-luc#1. (−) vector, (+) recombinant. Cell lysates derived from cells of JM109/p6H57-luc#1 cultivated with an acid stimulus in the exponential growth phase or in the early stationary phase were used for the assay. (B) Sixteen microgram and 2 or 16 μg of total RNAs derived from the exponentially growing cells and early stationary-phase cells, respectively, were analyzed using ribo-5H57 as a probe. *An appropriate amount of NaOH was added to the culture to adjust the pH to 7.0. An arrow indicates the full size of ArrS.

Download figure to PowerPoint

image

Figure 6.  Requirement of sigma factors for arrS transcription. ArrS levels of BW25113 or its isogenic deletion mutants of ΔfecI,ΔrpoN, ΔrpoS,ΔrpoS/pCA24N-gadE and ΔrpoN/pCA24N-gadE were analyzed. (A) Early stationary phase. (B) and (C) The exponential growth phase with the acid stimulus.

Download figure to PowerPoint

Participation of GadE in arrS transcription

Next, we investigated whether GadE and/or the process of gadE transcription modulate arrS transcription. First, we confirmed the participation of GadE in arrS transcription using a gadE deletion mutant of the KO collection (Fig. 4). In this mutant, the gadE ORF is replaced with a kanamycin gene, but promoter sequences of gadE remain unchanged as shown in Fig. 1A (Baba et al. 2006). The level of ArrS was clearly lower in the ΔgadE mutant than that in the parent strain in the early stationary phase (panel A, lanes 1 and 3). This elimination was restored by GadE provided in trans by pCA24-gadE (lane 2). The restoration was, however, not to the level found in the parent strain. The discrepancy may be related to excess GadE in the ΔgadE/pCA24-gadE strain relative to the parent strain. In 2009, Sayed and Foster reported that GadE could repress the gadE promoter P2 and P3. In contrast to arrS expression in early stationary phase, the expression induced by an acid stimulus in the exponential growth phase scarcely changed in the ΔgadE mutant (panel B, lanes 2 and 4).

Is the transcription of antisense RNA ArrS observed in the early stationary phase and under the acidic condition in the exponential growth phase believed to occur via a cis effect of gadE T3 transcription? To contradict this possibility, we cloned a 189-bp fragment into a site upstream of luc ribosome-binding site (RBS) in pSPluc + NF and carried out an arrS promoter assay (Fig. 5A). The 189-bp fragment contains a 163-bp upstream sequence of arrS and a 26-bp 5′ sequence that codes for ArrS, but does not contain any gadE promoter sequences on the opposite strand. The 189-bp fragment represents high promoter activity in early stationary-phase cells (column 6). In contrast, the activity observed under an acid stimulus was low (column 4). There is no termination signal upstream from the promoter cloning sites in the pSPluc + NF. This may be a reason for the high fluorescence level for the vector control strain in column 5. The results of Fig. 5A strongly suggest that antisense RNA ArrS could be transcribed depending on its own promoter in the early stationary phase and under acidic conditions in the exponential growth phase.

Next, we compared the cellular levels of ArrS between cells in the early stationary phase and those under the acidic condition in the exponential growth phase by Northern blotting (Fig. 5B). Sixteen micrograms of total RNA derived from exponentially growing cells 30 min after the shift to the acidic condition and 2 μg of that derived from early stationary-phase cells were loaded in lanes 2 and 3, respectively. This showed that the level of ArrS was extremely high in the early stationary phase than that observed under the acidic condition in the exponential growth phase. These results of Northern blotting are in agreement with those of the promoter assay (Fig. 5A). The pH value was ∼5.9 in the early stationary phase culture without the addition of acid. A similar result was obtained when the pH of the culture was adjusted to 5.5 (lane 4). An intriguing question raised is whether ArrS observed in early stationary phase a consequence of just a response to acid. We examined the level of ArrS under neutral pH in the early stationary phase (lane 5), and it was observed that the level was lower under neutral pH than under a pH of 5.5. These results showed that a low pH is a stimulus that induces the arrS transcription in the early stationary phase. Moreover, on comparing lane 6 with lane 1, it was clearly observed that some factors other than low pH are indispensable for the high expression of arrS in the early stationary phase.

Dependence of rpo genes in arrS transcription

Next, we investigated the dependence of arrS transcription on sigma factors (Fig. 6). Transcription of arrS decreased in an rpoS deletion mutant in the early stationary phase and under the acidic condition in the exponential growth phase (Fig. 6A, lane 2 and Fig. 6B, lane 4, respectively). Transcription of arrS has been shown to be dependent on GadE in the early stationary phase as in Fig. 4A. In addition, it has been reported that gadE transcription depends on RpoS indirectly through a positive regulator, GadX/GadW (Sayed et al. 2007). Therefore, the decrease in arrS transcription in the rpoS mutant may be because of GadE elimination. Next, we introduced the ASKA gadE clone in the rpoS mutant (Fig. 6A, lane 4 and Fig. 6C, lane 4). Transcription of arrS was not restored by gadE over-expression.

An rpoN mutation also affected arrS transcription under the acidic condition in the exponential growth phase (Fig. 6B, lane 3), and gadE over-expression suppressed the elimination of arrS transcription found in the rpoN mutant (Fig. 6C, lane 5). In contrast, the rpoN mutation did not affect arrS transcription in the early stationary phase (Fig. 6A, lane 3).

Effects of ArrS over-expression on the level of gadE mRNA T3

We carried out arrS over-expression by a plac-arrS transcriptional fusion to speculate on the functions of ArrS in gadE expression. In this fusion, a fragment containing a sequence of −142 to −1 relative to the transcription start site (+1) of lacZ was ligated to a fragment carrying the code for ArrS and its 3′ flanking sequence. This fusion did not contain the lac operator sequence; accordingly, expression of arrS would be constitutive.

As shown in lanes 3 and 4 of Fig. 7A, large amount of ArrS was synthesized in the cells containing the fusion gene. The bands in this figure were detected after a short exposure (10 min) in autoradiography, so the signal of ArrS in early stationary-phase cells without the fusion gene was extremely low (lane 2). We investigated the level of gadE mRNAs in these cells via comparative RT-PCR (panel B). The primer set RT2/gadE (−558) and RT1/gadE (−19) were used to detect gadE mRNAs containing the 566-nucleotide-long 5′ UTR and the gadE ORF, respectively (Fig. 1A). The 760-bp band amplified with the primer set RT1/gadE (−558) is only obtained if the gadE mRNA T3 exists. As shown before in lanes 1 and 2 of Fig. 2B, the gadE mRNA T3 is expressed without acid stimuli in the exponential growth phase or in the early stationary phase. The 760-bp bands were amplified with the primer set RT1/gadE (−558) in cells without the fusion gene (lanes 1 and 3). Similarly, the 412-bp bands and 221-bp bands were obtained with the primer set RT2/gadE (−558) and RT1/gadE (−19), respectively (lanes 5, 7, 9 and 11). The bands in lanes 1, 3, 5 and 7 disappeared because of the over-expression of arrS (lanes 2, 4, 6 and 8). On the other hand, the 221-bp bands amplified with the primer set RT1/gadE (−19) had not disappeared by the ArrS over-expression (lanes 10 and 12). These results strongly suggest that the excess ArrS could cleave or degrade gadE mRNA T3; however, the gadE mRNAs that contain the gadE coding region but not the 5′ UTR still remain.

image

Figure 7.  Over-expression of arrS in the exponential growth and early stationary phases. Cells of JM109/pBR-plac-arrS were cultivated in LB medium and total RNAs were extracted to analyze levels of ArrS (A) and gadE mRNAs (B). (A) Northern blotting of ArrS. The c-6H57-49 was used as a probe. The signal was obtained via an exposure for 10 min. (B) Comparative RT-PCR for gadE mRNA. The reaction was run for 18 cycles. Lanes 1-8, primer sets RT1/gadE (−558) and RT2/gadE (−558) were used to amplify gadE mRNAs, which contain the 5′ UTR of gadE mRNA T3; lanes 9-12, a primer set RT1/(−19) was used to amplify gadE mRNAs, which contain the code for GadE. L, exponential growth phase; S, early stationary phase.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

In this study, we confirmed that ArrS is an antisense RNA of the gadE mRNA T3. The gadE mRNA T3 is observed whenever ArrS is expressed. Does ArrS increase depend only on the cis effect of T3 transcription? The possibility of this appears to be small, based on the following results. First, the 189-bp fragment containing a 163-bp-long upstream sequence of arrS and a 26-bp-long 5′ coding sequence of ArrS showed clear promoter activity in the early stationary phase and under the acidic condition in the exponential growth phase in the promoter assay (Fig. 5A). The promoter activity was approximately 10-fold higher in the early stationary phase than that under the acidic condition in the exponential growth phase, and these results correspond with the cellular levels of ArrS in each condition (Fig. 5B). The DNA fragment used in the promoter assay did not contain any gadE promoters on the opposite strand, indicating that the reporter gene could be transcribed by the arrS promoter independent of the cis effect via gadE transcription. Second, arrS transcription was not observed in the gadE deletion mutant of the KO collection in the early stationary phase. This deficiency in arrS transcription was, however, restored in cells provided with GadE in trans (Fig. 4A). We examined the transcription from the gadE P3 promoter in this mutant by comparative RT-PCR. Transcription from the P3 promoter occurred in the mutant, and the level was higher in the mutant than in the parent strain (data not shown). It was reported that transcription from the gadE P3 promoter is negatively regulated by GadE (Ma et al. 2004; Sayed & Foster 2009). Therefore, increasing transcriptions from the gadE P3 promoter can be interpreted as a result of GadE deficiency in the gadE deletion mutant. These results demonstrate that transcription of arrS does not occur without GadE, even if transcription from the gadE P3 promoter increases in the gadE deletion mutant during the early stationary phase. Similarly, ArrS does not appear without an acid stimulus in the exponential growth phase, even though transcription of T3 occurs in this growth phase (Figs 2B and 3A).

Model for control of arrS transcription

In Fig. 8, we propose a model for the control of arrS expression. In this model, arrS expression occurs in two ways: first, as shown in panel A, an unknown stimulus originates in early stationary phase induces arrS transcription (Fig. 5B, lanes 1 and 6). The transcription begins with its own promoter depending on RpoS, a sigma factor required for expression of many genes in the stationary phase or in stress responses. The consensus sequence for sigma D is found upstream of arrS (Fig. 1D). Sigma S and sigma D sometimes share promoter sequences. GadE is indispensable as a transcriptional activator in the system in panel A, but the stimulus still remains unknown. We found a complementary sequence of ArrS in RBS of fdoG mRNA in in silico analysis. fdoG is a gene of the catalytic subunit of formate dehydrogenase-O, the physiological role of which is believed to provide cells the ability to rapidly adapt to anaerobic conditions. Anaerobic stress may induce arrS transcription because oxygen concentration in the cultures is transiently lowered in the early stationary phase. Second, the arrS transcription occurs by acid stresses (panel B). GadE is dispensable in this case. The arrS transcription in this system depends on rpoS and rpoN. The mechanisms for the regulation by rpoS and rpoN still remain unclear. The arrS transcription probably depends directly on rpoS and indirectly on rpoN. It is not clear why the rpoN deletion was suppressed by GadE under low pH (Fig. 6C). It might not be the suppression of the rpoN deletion by GadE, but rather the induction of arrS via the system in panel A by GadE over-expression. Riordan et al. (2010) reported that gadE expression increased in the exponential growth phase in an rpoN deletion mutant of E. coli O157:H7. In their study, the effect of RpoN deficiency on gadE expression was not observed in the stationary phase. They also showed that AR increased in the rpoN deletion mutant using a model stomach system. It is interesting how rpoN participates in AR.

image

Figure 8.  A model for regulation of arrS transcription. Broken lines with arrows indicate transcription of gadE mRNA T3 or ArrS. The size of the arrows represents levels of transcription. (A) arrS transcription induced by an unknown stimulus occurs in early stationary phase. The transcription depends on σS and is positively regulated by GadE. (B) arrS transcription induced by acid stimuli depending on σS and σN.

Download figure to PowerPoint

Position of arrS gene in the 750-bp sensory integration region of gadE

gadE transcription is modulated by at least nine regulatory factors: EvgA, YdeO, GadE, TorR, Hns, PhoP, TrmE, GadX and GadW (Bordi et al. 2003; Masuda & Church 2003; Tucker et al. 2003; Gong et al. 2004; Ma et al. 2004; Zwir et al. 2005; Sayed et al. 2007). The factor used depends on the culture media and the growth phase of the cells. The binding sites of these transcriptional regulators appear to converge in the intergenic region located between hdeD and gadE. Three gadE promoters termed P1, P2 and P3 are also included in this region (Sayed & Foster 2009). Transcriptional lacZ fusion to gadE promoter regions showed that P1 and P3 are active in the stationary phase and under acidic conditions in the exponential growth phase, whereas P2 is not. In contrast, the fragment P3P2 (−754 to −312) that contains P3 and P2 promoters is the most effective in lacZ fusions (Sayed & Foster 2009). In their fusion system, the native gadE region is retained because the fusion genes are inserted in the trp operon on the chromosome. This means that native arrS can be transcribed with its promoter. Of note, the fragment P3P2 contains a sequence at −380 to −312, which is complementary to the sequence of ArrS. It is possible that arrS, which was mapped in the sensory integration region, may also play several roles in gadE expression.

Possible functions of antisense RNA ArrS

Cis-encoded antisense RNAs are known to base pair with the 5′ UTRs, ORFs or 3′ UTRs of their target mRNA (Fozo et al. 2008). However, antisense RNA that base pair with sequences located 380–312 bp upstream of the translation initiation site have not been identified to date. With the exception of antitoxin RNA found in E. coli, GadY represents one of the few examples of a cis-encoded antisense RNA (Opdyke et al. 2004). This sRNA is encoded on the opposite strand of gadX and base pairs with the 3′ UTR of gadX mRNA to increase its stability. sRNAs are known to regulate translation of target mRNAs either via a controlling ribosome that binds to RBS or by modulating the stability of mRNAs.

How does ArrS function in expression of gadE or some other genes? We constructed a plac-arrS recombinant plasmid by which the excess ArrS of an exact size was synthesized constitutively (Fig. 7A). The arrS over-expression resulted in a lower level of gadE mRNA T3 (Fig. 7B, lanes 2 and 4); however, the level of smaller gadE mRNAs, which contain the coding region for GadE but not the long 5′ UTR, did not change (lanes 10 and 12). Transcription with the P1 or P2 promoter was not observed at least in the exponentially growing cells cultivated under a neutral pH (Fig. 2B, lane 1). This finding indicates that the smaller gadE mRNAs detected with the primer set RT1/gadE (−19) were generated from gadE mRNA T3 via processing (Fig. 7B, lane 10). These results imply that ArrS participates in the processing of gadE mRNA T3. This processing may regulate the translation of gadE mRNAs. Remarkable differences in phenotype have not been observed after the arrS over-expression so far. GadE proteolysis is constitutively observed at a low pH or during the early stationary phase (Heuveling et al. 2008). It is possible that GadE degradation is necessary to avoid the over-expression of AR genes. Through a search of bacterial genomes, we found homologues of ArrS only in Shigella. Furthermore, the GadE homologue is found only in Shigella strains (Hommais et al. 2004). The regulatory mechanisms for ArrS in AR remain to be clarified.

In conclusion, ArrS is an antisense RNA of gadE mRNA T3 and can base pair with a 69-nucleotide sequence in the 566-nucleotide-long 5′ UTR of T3. The arrS is transcribed from its own promoter depending on GadE and RpoS in the early stationary phase and on RpoN and RpoS by acid stimuli in the exponential growth phase.

Experimental procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Bacterial strains, plasmids and culture conditions

Bacterial strains and plasmids used in this study are listed in Table 1. Cells were grown overnight at 37 °C in LB or E media containing 0.5% glucose and 2% polypeptone (Vogel & Bonner 1956), inoculated into fresh media and then cultivated for 2 (exponential growth phase) or 6 h (early stationary phase). The doubling time of W3110 and BW25113 in this E medium was approximately 30 min. An appropriate amount of H2SO4 and NaOH was added to the cultures to adjust their pH values to 5.5 and 7.0, respectively; then, the pH values were confirmed. The cells were removed 30 min after the pH shift.

Table 1.   Bacterial strains and plasmids used in this study
Strains or plasmidRelevant genotype or characteristicsSource or references
Strains
 W3110K-12 Fλ in (rrnD-rrnE)Niki, NBRP (NIG, Japan): E. coli
 BW25113K-12 lacIqrrnBT14 DlacZWJ16hsdR514ΔaraBA-DAH33ΔrhaBADLD78Niki, NBRP (NIG, Japan): E. coli, Datsenko & Wanner 2000
Keio strains
 JW3480BW25113 ΔgadENiki, NBRP (NIG, Japan): E. coli, Baba et al. 2006
 JW5437BW25113 ΔrpoS
 JW3169BW25113 ΔrpoN
 JW4253BW25113 ΔfecI
 JM109recA1 endA1 gyrA96 thi-1 hsdR117 (rk- mk-) e14-(mcrA-)supE44 relA1Δ(lac-proAB)/F’[traD36, proAB+lacIqlacZΔM15]Takara Bio Inc.
Plasmids
 pCR2.1-TOPO Invitrogen
 JW3480pCA24N-gadENiki, NBRP (NIG, Japan): E. coli,Kitagawa et al. 2005
 pSP-luc + NF Promega
 p6H57-luc#1A 189 bp arrS promoter fragment was cloned into the NdeI –KpnI sites of pSP-luc + NF plasmidThis study
 pMW118 Nippon Gene
 pMW118-6H57A 245-bp arrS fragment was cloned into the EcoRI-KpnI site of pMW118This study
 pBR-plac-arrSA 142-bp lac promoter fragment was ligated to a 173-bp arrS fragment then cloned into the HindIII-BamHI site of pBR322This study

Oligonucleotides

Oligonucleotides used as probes or primers are listed in Table S1 in Supporting Information.

Northern blotting

Total RNA was extracted using hot phenol (Aiba et al. 1981). Northern blotting for gadE or hdeD was carried out as follows: 8 μg of total RNA was fractionated by electrophoresis at 80 V for 2–3 h on a 1.5% agarose gel containing 6% formaldehyde. RNA was then transferred to a nylon membrane (Hybond N+; Amersham Pharmacia Biotech) in 20 × SSC. Oligonucleotide probes s-6H57, c-gadE and c-hdeD were labeled with [γ-32P] ATP by T4 polynucleotide kinase (Takara Bio Inc.), and hybridization was carried out in ULTRAhyb-Oligo Hybridization Buffer (Ambion) at 42 °C. Autoradiograms were obtained using a FLA3000G Imaging Analyzer (Fuji). Northern analysis for ArrS was carried out as follows: total RNA (2–16 μg) was fractionated using a 6% polyacrylamide gel containing 7 m urea and electroblotted on a Clear Blot Membrane N Plus (Atto). The BrightStar psoralen-biotin nonisotopic labeling kit (Ambion) and the BrightStar biodetect nonisotopic detection kit (Ambion) were used for oligonucleotide probe c-6H57-49. Alternatively, riboprobe ribo-6H57 and ribo-5S were synthesized by the DIG RNA Labeling Kit (SP6/T7) (Roche). 6H57 (+1)-22 and pT7-6H57L, and 5S (−29) and pT7-5SL were used as primers for ribo-6H57 and ribo-5S, respectively. The DIG Luminescent Detection Kit (Roche) was used for signal detection.

Comparative RT-PCR

Total RNA was treated with RQ1 RNase-free DNase (Promega) in the presence of RNasin (Promega) to remove genomic DNA. RT was carried out at 43 °C using DNA-free total RNA (10 μg), 2 pmol of oligonucleotide primer gadE-RT2, and 200 units of SuperScript II reverse transcriptase (Invitrogen) in 20 μL of reaction mixture. RNA was then degraded with two units of RNase H (Invitrogen) at 37 °C for 20 min. PCR was carried out in 1 × Takara Taq buffer containing 2 μL of cDNA, 0.2 mm of dNTP, 1 μm of PCR primer, 1 μm of RT primer and 0.5 units of Takara Taq (Takara Bio Inc.) for 18 cycles. PCR products were electrophoresed on 2% agarose gels.

Circularization RT-PCR

We carried out circularization RT-PCR to determine the 5′ and 3′ ends of gadE mRNA (Couttet et al. 1997). We used a modified procedure for analyzing bacterial mRNA (Yonesaki 2002). DNA-free total RNA (6.3 μg) was dephosphorylated using E. coli A19 alkaline phosphatase (Takara) and phosphorylated using T4 polynucleotide kinase (Takara) to convert the 5′ termini of all mRNA to a monophosphorylated state. The RNA was ligated at 12 °C overnight using T4 RNA ligase to achieve self-circularization or intermolecular ligation. cDNA synthesis and PCR were carried out as described earlier using gadE (−408) RT and gadE569 as primers. The PCR products were then cloned in a pCR2.1-TOPO vector (Invitrogen), and the sequences surrounding the 5′ 3′ junctions of the mRNA molecules were determined.

Construction of a reporter plasmid for the promoter assay

The luciferase fusion reporter plasmid p6H57-luc#1 was constructed as follows. The 189-bp fragment containing a 163-bp upstream sequence and a 26-bp 5′ coding sequence of arrS was amplified using 6H57 (−163) NdeI and 6H57-26 KpnI as primers. The PCR fragment was cloned in the NdeI–KpnI sites of pSP-luc + NF plasmid (Promega), and then the sequence was confirmed.

Luciferase assay

Cells of JM109/p6H57-luc#1 were cultivated under the acidic condition or in the early stationary phase and used for the luciferase assay with a Luciferase assay system (Promega). The luciferase assay was carried out according to the manufacturer’s protocol. In brief, 50 μL of JM109/p6H57-luc#1 culture was mixed with 40 μL of JM109 culture. The cell suspension was added to 10 μL of 1 m K2HPO4 (pH 7.8) and 20 mm EDTA and then stored for 10 min at −80 °C. Freshly prepared lysis mix (300 μL) was added and mixed well. After incubation for 10 min at room temperature, 10 μL of the cell lysate was mixed with 50 μL of luciferase assay reagent and the light produced by Lumat LB9507 (EG & G Berthold) was measured. The promoter activities were represented as RLU/cell.

Over-expression of arrS

We constructed a pBR-plac-arrS recombinant plasmid to over-express arrS in the exponential growth and early stationary phases. First, a 245-bp fragment containing arrS was amplified using 6H57-UKpn and 6H57-LEco2 as primers. The PCR fragment was digested with KpnI and EcoRI and then inserted into an EcoRI/KpnI site of a pMW118 to obtain a pMW118-6H57. A plac-arrS transcriptional fusion was constructed by the SOEing method (Horton 1993). A lac promoter fragment containing a sequence of −1 to −142 bp relative to the transcription start site (+1) of lacZ was amplified using lacZ-UHin and lacZP-6H57AN22 as primers. A 173-bp arrS fragment containing an arrS coding region (+1 to +69) and a 3′ flanking region was amplified using 6H57 (+1) 22-nt and 6H57-LEco2 as primers. The lac promoter fragment and the arrS fragment were partially annealed and then a plac-arrS region was amplified using lacZ-UHin and 6H57-LBam as primers. The plac-arrS fragment was cloned into a HindIII/BamHI site of pBR322 to form a pBR-plac-arrS. The sequence was confirmed and then introduced into cells of JM109. Cells of JM109/pBR-plac-arrS were cultivated until the exponential growth phase or an early stationary phase in LB medium was achieved, and the total RNAs were extracted and used in Northern blotting or RT-PCR.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

We thank Reiko Ohki for helpful discussions, Hironori Niki for the ASKA clones and KO strains. This work was supported by a Grant-in-Aid from the Research Promotion Award of Faculty of Health Sciences, Kyorin University.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Table S1 Oligonucleotides used in this study

FilenameFormatSizeDescription
GTC_1516_sm_TableS1.doc39KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.