Helicobacter pylori 5′ureB-sRNA, a cis-Encoded Antisense Small RNA, Negatively Regulates ureAB Expression by Transcription Termination

Abstract

Urease is an essential component of gastric acid acclimation by Helicobacter pylori. The increased level of urease in gastric acidity is due, in part, to acid activation of the two-component system consisting of the membrane sensor HP0165 (ArsS) and its response regulator HP0166 (ArsR), which regulates transcription of the seven genes in two separate operons (ureAB and ureIEFGH) of the urease gene cluster. Recently, we identified a novel cis-encoded antisense small RNA, 5′ureB-sRNA, targeted at the 5′ end of ureB, which downregulates ureAB expression by truncation of the ureAB transcript at neutral pH. It is not known whether the truncated transcript is due to transcription termination or processing of the full-length mRNA by codegradation of a ureAB mRNA-sRNA hybrid complex. S1 nuclease mapping assays show that the truncated transcript is due to transcription termination. Further studies using an in vitro transcription assay found that 5′ureB-sRNA promotes premature termination of transcription of ureAB mRNA. These results suggest that the antisense small RNA 5′ureB-sRNA downregulates ureAB expression by enhancing transcription termination 5′ of ureB. With this mechanism, a limited amount of 5′ureB-sRNA is sufficient to regulate the relatively high level of ureAB transcript.

INTRODUCTION

Helicobacter pylori is a Gram-negative bacterial pathogen that colonizes the acidic environment of the human stomach and causes chronic superficial gastritis and peptic ulcer disease (1, 2). Persistent infection with H. pylori, if untreated, lasts for the lifetime of the infected individual and predisposes the individual to gastric malignancies such as adenocarcinoma and mucosa-associated lymphoid tissue (MALT) lymphoma (3, 4). The ureAB genes of H. pylori, located in the urease gene cluster, which consists of two operons (ureAB and ureIEFGH) (5), encode the urease structural subunits UreA and UreB of the apoenzyme and are essential for the survival of H. pylori at low pH, together with UreI, which is a regulator of urea entry to the bacterial cytoplasmic urease (6–9). UreA and UreB are expressed at very high levels in H. pylori, accounting for as much as 8% of the total bacterial protein (10, 11). Urease activity generates the buffers NH₃ and HCO₃⁻ from the metabolism of ambient urea, thereby maintaining both cytoplasmic and periplasmic pH to enable the organism to survive and grow in the stomach (12–14).

As a key virulence and acid acclimation factor, urease gene expression in H. pylori is highly regulated in response to environmental pH changes. Transcription of the ureAB genes was found to be positively regulated by the NikR protein in response to increasing concentrations of Ni²⁺ in the surrounding medium (15, 16), through which the environmental acidity might be indirectly sensed (17). Recent studies have shown that the transcriptional induction of urease genes (ureAB and ureI) in response to low pH is mediated mainly by the HP0165-0166 two-component system (ArsSR), since the pH-induced upregulation was largely abolished in an ArsS-deficient mutant (18–20). Accordingly, the phosphorylated response regulator ArsR (ArsR∼P) was found to bind to extended regions overlapping both the P_ureA and P_ureI promoters (20). However, although the upregulatory response is suitable in gastric acidity, the homeostasis of urease expression needs to be tightly regulated, as uncontrolled expression at neutral pH leads to decreased fitness of the bacterium due to overalkalization (12, 21). In recent studies (22), we identified a novel urease-regulatory mechanism that downregulates ureB expression, which is controlled primarily by the ArsSR two-component system. An antisense small RNA (5′ureB-sRNA) encoded in the 5′ region of the H. pylori ureB cistron was found to downregulate ureAB expression by increasing the level of a 1.4-kb truncated ureAB transcript lacking 3′ureB under relatively neutral conditions, since the increased levels of the 5′ureB-sRNA (transcribed by an intracistronic antisense promoter that may be induced by unphosphorylated ArsR) associate with increased levels of a truncated form (1.4 kb) of the ureAB transcript and this truncated form is more abundant in cells growing at neutral pH. In contrast, at low pH the level of 5′ureB-sRNA decreases, and then the intact form (2.7 kb) of the ureAB transcript predominates, allowing full expression of the urease apoenzyme.

Small, noncoding RNAs (sRNAs) are increasingly recognized as being crucial for the regulatory networks of all organisms, including bacteria (23). In the last decade, there has been an explosion in the identification of regulatory sRNAs encoded on bacterial chromosomes. The majority of characterized sRNAs act by base pairing with target mRNAs. Bacterial base-pairing sRNAs fall into two categories (23): cis-encoded sRNAs are located in the same DNA region and are therefore fully complementary to their targets over a long sequence stretch, whereas trans-encoded sRNAs are located in another chromosomal location and are only partially complementary to their target mRNAs. The regulatory mechanism of antisense sRNAs may be advantageous over other types of regulation, since the antisense sRNAs could provide a specific mechanism whereby the levels of a particular protein need to be repressed or expressed under very select circumstances. In addition, many of the characterized antisense sRNA targets are subject to extensive regulation, and these antisense sRNAs provide yet one more level of control (24).

The currently known regulatory mechanisms employed by cis-encoded antisense sRNAs include transcription attenuation/termination, translation inhibition, and promotion or inhibition of mRNA degradation (25). Our previous studies (22) suggest that the 5′ureB-sRNA initiates the downregulation of ureB by base pairing with a coding region of the ureB transcript which does not include the Shine-Dalgarno sequence. Therefore, the possibility that this sRNA interferes with the initiation of translation through competition with 16S rRNA for the Shine-Dalgarno sequence is eliminated (26, 27). However, it is still not clear whether the 1.4-kb truncated ureB transcript resulting from 5′ureB-sRNA is due to transcription termination or processing of the full-length mRNA by codegradation of a ureAB-sRNA hybrid complex.

In the current study, we show that the antisense sRNA 5′ureB-sRNA downregulates ureAB expression by interacting with the ureAB mRNA, resulting in premature termination of transcription 5′ of ureB through a transcriptional attenuation mechanism. We also provide in vitro evidence suggesting that with this mechanism, a limited amount of 5′ureB-sRNA is sufficient to regulate the relatively high level of ureAB transcript.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

H. pylori strain 43504 was obtained from the American Type Culture Collection (ATCC). A 5′ureB-sRNA overexpression strain [HP43504/pTM-P_cagA-5′ureB-sRNA(+)] (22) was constructed by introducing plasmid pTM-P_cagA-5′ureB-sRNA(+) into H. pylori strain 43504 via natural transformation as described below. Primary plate cultures of H. pylori were grown from glycerol stocks on Trypticase soy agar (TSA) plates with 5% sheep blood (Fisher Scientific) for 2 to 3 days in a microaerobic environment (5% O₂, 10% CO₂, and 85% N₂) at 37°C. In preparation for an experiment, bacteria were scraped from the plates, suspended in 1 mM phosphate HP buffer (138 mM NaCl, 5 mM KCl, 1 mM CaCl₂, 0.5 mM MgCl₂, 10 mM glucose, 1 mM glutamine) (pH 7.0), and transferred to fresh plates for 24 h. For exposure to experimental low-pH conditions, the overnight culture of H. pylori strain 43504 on TSA plates supplemented with 5% sheep blood was suspended in brain heart infusion (BHI) medium (Difco) to an optical density at 600 nm (OD₆₀₀) of 0.20 to 0.25. The pH of the BHI medium was adjusted to 7.4 or pH 4.5 using concentrated HCl, followed by filtration to remove any precipitate. H. pylori was then incubated in the presence of 5 mM urea with shaking (120 rpm) under microaerobic conditions at 37°C for 30 min. The medium pH was measured after 30 min of incubation to ensure that there was no pH change at the end of the experiment. Escherichia coli strains were grown in Luria-Bertani (LB) broth. When necessary, antibiotics were added to the following final concentrations: ampicillin, 100 μg/ml; kanamycin 50 μg/ml (for E. coli) or 20 μg/ml (for H. pylori).

Construction of overexpression strain for 5′ureB-sRNA.

pTM117, which was reported as a transcriptional reporter and complementation vector in H. pylori (28), was used for the overexpression of 5′ureB-sRNA by insertion of a 593-bp DNA fragment containing the promoter region of cagA at KpnI/NcoI sites (pTM-P_cagA). The cagA promoter fragment was prepared by PCR from the intergenic region between cagB and cagA (29) of the H. pylori strain 26695 genome with the primer pair cagBA-IGR-5′P-KpnI/cagBA-IGR-3′P-NcoI (Table 1). For the plasmid used for overexpression of 5′ureB-sRNA [pTM-P_cagA-5′ureB-sRNA(+)], the fusion fragment of 5′ureB-sRNA and a transcriptional terminator identified from H. pylori (HP0092-T1) (31) was prepared in a two-step PCR process by amplification of the 5′ureB-sRNA fragment with primer 5′-ureB sRNA(+)-5′P-NcoI and fusion primer HP0092-T1-5′P-rev–5′-ureB sRNA(+)-3′P (Table 1) and of the HP0092-T1 fragment with fusion primer 5′-ureB sRNA(+)-3′P–HP0092-T1-5′P and HP0092-T1-3′P-PstI (Table 1). The two fragments were subsequently used together as templates in a sewing PCR with primer pair 5′-ureB sRNA(+)-5′P-NcoI/HP0092-T1-3′P-PstI to enable the fusion, followed by cloning of the fusion fragment into pTM-P_cagA at NcoI/PstI sites. The plasmid pTM-P_cagA-5′ureB-sRNA(+) was introduced into H. pylori strain 43504 via natural transformation (32), and the transformants were selected on BHI plates containing kanamycin.

Table 1.

Oligonucleotide primers and probes used in this study

Name	Sequence (5′ to 3′)a	Siteb	Strand	Positionc
cagBA-IGR-5′P	tttgggtaccTTTTTAATCGTCTCAGGTTCA	KpnI	+	579301–579321
cagBA-IGR-3′P	ctagccatggGACTATCGGTATCTTATTGGTATCA	NcoI	−	579869–579892
5′-ureB sRNA(+)-5′P	CTTGCcAtgGCTGTAGGGATTTGTTGGGGTG	NcoI	+	76785–76815
5′-ureB sRNA(+)-3′P	CTGAGAGAAGGCATGAGCCA		−	77067–77086
HP0092-T1-5′P-rev	AGTTTAATTACCAGTGGATA		+	98324–98343
HP0092-T1-5′P	TATCCACTGGTAATTAAACT		−	98324–98343
HP0092-T1-3′P	ATTTcTGCaGATCCGCCTTATTTCCTCTCT	PstI	+	98206–98235
HP0547-5′P(548-571)	AGGCTCATTTCTTATTTCTTGTTC		+	579635–579658
HP0547-3′P(1785-1806)	TGTTTCTCCTTACTATACCTAG		−	579899–579920
P-PcagA	CTAGGTATAGTAAGGAGAAACA		+	579899–579920
HP0072-5′P(1111-1133)	CTACAGGCGATAAAGTGAGATTG		−	77168–77190
HP0072-3′P(1902-1923)	TAATATCAGGAGCGTGTCCGCC		+	76378–76399
HP0072-5′P(1266-1290)	CTAATCATCACTAACGCTTTAATCG		−	77011–77035
HP0072-3′P(2653-1673)	CAATGTGAGCGGTAGTGTCGT		+	75628–75648
HP0072-5′P(1215-1234)	CTGAGAGAAGGCATGAGCCA		−	77067–77086
HP0072-3′P(1486-1456)	CTTGCAAAAGCTGTAGGGATTTGTTGGGGTG		−	76785–76815
HP0072-5′P(2159-2182)	TATGGGTCGTGTGGGTGAAGTTAT		−	76119–76142
HP0072-3′P(2653-2673)	CAATGTGAGCGGTAGTGTCGT		+	75628–75648
Mut-YUAN-up	CAAAGACATGCAAGATGGCGgaAcAAACAATCTTAGCGTAGGTC		−	76891–76934
Mut-YUAN-dn	GACCTACGCTAAGATTGTTTgTtcCGCCATCTTGCATGTCTTTG		+	76891–76934
5′-ureB 2-1S	CTAATCATCACTAACGCTTTAATCGTGGATTACACCGGTATTTATAAAGCGGATA		−	76981–77035
5′-ureB 2-1AS	CTAATCATCACTAACGCTTTAATCGTGGATTACACCGGTATTTATAAAGCGGATA		+	76981–77035

^aSequences in uppercase letters are derived from the genome sequences of H. pylori 26695 (30). Sequences introduced for cloning or mutation purposes are given in lowercase letters, and restriction recognition sites are underlined.

^bRestriction recognition sites.

^cNucleotide positions refer to the genome sequence of H. pylori 26695 (30).

RNA preparation.

Total RNA was isolated from H. pylori strains using TRIzol reagent (Invitrogen, CA) combined with RNeasy columns (Qiagen, CA). The bacterial pellet was resuspended in 500 μl of TRIzol reagent (Invitrogen) and lysed at room temperature for 5 min before 100 μl of chloroform was added. After centrifugation at 12,000 × g for 10 min at 4°C, the supernatant was mixed with 250 μl ethanol and applied to an RNeasy spin column (Qiagen), and RNA purification was done following the manufacturer's instructions (beginning with the application to the column). The RNA concentration was quantified by absorbance at 260 nm using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies), and the quality was evaluated by capillary electrophoresis using an Agilent 2100 Bioanalyzer with the RNA 6000 Nano assay kit (Agilent Technologies).

Northern blot analysis.

Total RNA (5 μg) was fractionated in 6% polyacrylamide-urea gels (Invitrogen) and electrically transferred to Zeta Probe GT membranes (Bio-Rad Laboratories). For verification of equal loading, the RNA on the gel was visualized by ethidium bromide staining and photographed to compare the rRNA band intensities before the RNA was transferred. For preparation of the oligonucleotide probes, the strand-specific sense oligonucleotide 5′ureB 2-1S (55 nucleotides [nt]) and antisense oligonucleotide 5′ureB 2-1AS (55 nt) were synthesized (Eurofins MWG Operon) and 5′-end labeled radioactively with T4 polynucleotide kinase (Promega) and [γ-³²P]ATP. The blots were hybridized with ³²P-labeled strand-specific oligonucleotide probe 5′-ureB 2-1AS (for detection of ureAB mRNA) or 2-1S (for detection of 5′ureB-sRNA) in ULTRAhyb-Oligo hybridization buffer (Ambion) at 42°C overnight and then washed with 2× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate)–0.5% SDS. The hybridized blots were autoradiographed using a 445 SI PhosphorImager (Molecular Dynamics, Sunnyvale, CA). The bands representing each transcript in the hybridized blots and 23S and 16S rRNAs in ethidium bromide-stained gels were quantified using ImageJ analysis system (available from rsb.info.nih.gov/ij/). The relative expression level for each transcript was normalized to its corresponding total intensity of 23S and 16S rRNAs.

S1 nuclease mapping.

S1 mapping of the mRNA in the ureB coding region spanning the potential termination site was performed according to standard protocols (33). For the 5′-end-labeled probe, the primer HP0072-3′P(1902-1923) was labeled using [γ-³²P]ATP and T4 polynucleotide kinase and then was used to amplify a DNA fragment with primer HP0072-5′P(1266-1290) and 26695 genomic DNA as the template, followed by digestion with restriction enzyme AgeI (to generate the 625-bp probe). For the 3′-end-labeled probe, a DNA fragment generated by PCR with primers HP0072-3′P(2653-2673) and HP0072-5′P(1266-1290) was digested with AgeI and the overhang was filled in using [α-³²P]dCTP and the Klenow fragment of DNA polymerase I, followed by digestion with restriction enzyme BamHI (to generate the 718-bp probe). Single-stranded DNA probes from both 5′-end-labeled AgeI fragment (625 nt) and 3′-end-labeled BamHI fragment (718 nt) were separated with a 6% polyacrylamide minigel containing 8 M urea and recovered from the gel by the crash-and-soak procedure. Fifty micrograms of total RNA from wild-type H. pylori and 25,000 cpm of 5′-end-labeled or 3′-end-labeled probe were hybridized at 37°C overnight, followed by digestion with S1 nuclease at 45°C for 1 h. The S1 nuclease digestion products were examined by electrophoresis on a 5% polyacrylamide sequencing gel.

Production of 5′ureB-sRNA.

Plasmid pCR4-5′ureB-sRNA(302), used for making in vitro-transcribed 5′ureB-sRNA, was constructed by cloning a PCR-amplified fragment coding for the entire 5′ureB-sRNA with primers HP0072-5′P(1215-1234)/HP0072-3′P(1486-1516) into the TA cloning site of pCR4-TOPO (Invitrogen). The orientation of the insert was determined by sequencing. The NotI-linearized plasmid was used as the template in an in vitro transcription reaction with T3 RNA polymerase (MAXIscript in vitro transcription kit; Ambion) to synthesize 5′ureB-sRNA. The in vitro-transcribed sRNA was purified with the RNA cleanup protocol from the RNeasy Minikit (Qiagen). The RNA concentration was quantified by absorbance at 260 nm using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies).

In vitro transcription assays.

An 813-bp 5′ureB fragment under control of the cagA promoter (P_cagA) was used as the template for in vitro transcription assays. The fusion fragment of P_cagA and an 813-bp ureB fragment (nucleotides 47 to 859 of the ureB coding sequence) was constructed in a two-step PCR process by amplification of a 286-bp P_cagA fragment with primer pair HP0547-5′P(548-571)/HP0547-3′P(1785-1806) (Table 1) and of a 5′ureB fragment with fusion primer P_cagA-HP0072-5′P(1111-1133) and HP0072-3′P(1902-1923) (Table 1). The two fragments were subsequently used together as templates in a sewing PCR with primer pair HP0547-5′(548-571)/HP0072-3′P(1902-1923) to enable the fusion. For a control template, a 515-bp 3′ureB fragment (nucleotides 1123 to 1608 of the ureB coding sequence) was amplified with fusion primer P_cagA-HP0072-5′P(2159-2182) and HP0072-3′P(2653-2673) and fused to the downstream region of the P_cagA fragment in a sewing PCR with primer pair HP0547-5′(548-571)/HP0072-3′P(2653-2673). For in vitro transcription assays, a gel-purified transcriptional fusion construct containing P_cagA (0.1 μg) was transcribed in a volume of 5 μl containing 40 mM Tris-HCl (pH 7.9), 10 mM MgCl₂, 150 mM KCl, 0.01% Triton X-100, 10 mM dithiothreitol, 500 μM ATP, 500 μM CTP, 500 μM GTP, 50 μM UTP, 500 nM [α³²P]UTP (800 Ci/mmol), 2 U of RNasin RNase inhibitor, and 0.5 U of E. coli RNA polymerase (USB). Different amounts of purified 5′ureB-sRNA (100 to 550 fmol) were added to the reaction mixtures before RNA polymerase. Rat liver mRNA (1 μg) was used to control for specificity of 5′ureB-sRNA. After incubation for 30 min at 37°C, the reaction was stopped by adding an equal volume of stop solution containing 95% formamide and 18 mM EDTA and heated at 90°C for 3 min. The samples were then separated by 6% urea-PAGE and visualized with a 445 SI PhosphorImager (Molecular Dynamics, Sunnyvale, CA).

For mutation of the YUNR U-turn motif in stem-loop structure II, the fusion fragment P_cagA-5′ureB(813) was cloned into the TA cloning site of pCR4-TOPO (Invitrogen) to generate pCR4-P_cagA-5′ureB(813). Substitution of three bases on the 5′ureB sequence (positions 322, 323, and 325 of the ureB coding sequence), to produce the Mut-YUNR mutant [pCR4-P_cagA-5′ureB(813)M], was carried out with the QuikChange site-directed mutagenesis kit (Stratagene) using pCR4-P_cagA-5′ureB(813) as the template and the mutagenic oligonucleotide primers Mut-YUAN-up and Mut-YUAN-dn (Table 1). The EcoRI fragment (containing the mutated P_cagA-5′ureB template) from the resulting plasmid pCR4-P_cagA-5′ureB(813)M was used in in vitro transcription assays. The mutation was confirmed by DNA sequencing.

RESULTS

5′ureB-sRNA regulates ureAB expression by controlling the levels of intact 2.7-kb and truncated 1.4-kb ureAB transcripts in response to environmental pH change.

In order to examine the regulatory role of the cis-encoded 5′ureB-sRNA in ureAB expression, overnight cultures of wild-type H. pylori (strain 43504) and a 5′ureB-sRNA overexpression strain [HP43504/pTM-P_cagA-5′ureB-sRNA(+)] (22) were treated under different pH conditions (pH 7.4 and 4.5) for 30 min, and Northern analysis was performed on equal amounts of RNA fractionated in 6% polyacrylamide-urea gels with a strand-specific deoxyoligonucleotide sense probe (5′ureB 2-1S) detecting 5′ureB-sRNA and an antisense probe (5′ureB 2-1AS) detecting ureAB transcripts. As shown in Fig. 1, in the H. pylori wild-type strain, the level of the 2.7-kb intact ureAB transcript increases at pH 4.5, corresponding to the decrease of the truncated 1.4-kb transcript, compared to the levels of transcripts at pH 7.4 (Fig. 1C and D). Meanwhile, at pH 4.5 the expression of the 5′ureB-sRNA is significantly decreased compared to that at pH 7.4 (Fig. 1A and B). In the 5′ureB-sRNA overexpression strain, while the level of 5′ureB-sRNA is highly increased under both pH conditions compared to that in the wild-type strain (Fig. 1A and B), the level of the truncated 1.4-kb ureAB transcript significantly increases, corresponding to the decrease of the intact 2.7-kb intact transcript, under both pH conditions (Fig. 1C and D). However, the pH 4.5-induced increase of the intact 2.7-kb ureAB transcript is still detectable in the 5′ureB-sRNA overexpression strain, even though it is significantly weaker than that in the wild-type strain (Fig. 1C and D). These results suggest that the acid-responsive expression of ureAB is controlled by the cis-encoded antisense RNA 5′ureB-sRNA, which provides fine-tuning of ureAB expression by regulating the level of the 1.4-kb truncated ureAB transcript at the expense of the 2.7-kb full-length transcript.

Fig 1.

pH-responsive expression profile of an antisense sRNA, 5′ureB-sRNA, in H. pylori by Northern blotting. (A and C) Total RNAs from wild-type (WT) H. pylori 43504 and a transformant with pTM-P_cagA-5′ureB-sRNA(+) that expressed a high level of 5′ureB-sRNA (sRNA+) were harvested after treatment at pH 7.4 (lanes 1) and pH 4.5 (lanes 2) for 30 min. RNA samples (5 μg) were separated in 6% polyacrylamide-urea gels and then transferred to Zeta-Probe GT membranes. The sRNA was detected with an oligonucleotide sense probe (5′-ureB 2-1S) corresponding to the 5′ ureB with overnight exposure (A), and the ureAB transcripts were detected with a strand-specific oligonucleotide antisense probe (5′-ureB 2-1AS) with 30 min of exposure (C). The gels (stained with ethidium bromide) are shown as loading controls (right panels). (B and D) Relative transcript levels normalized to the corresponding intensity of 23S and 16S rRNAs for ∼290-nt 5′ureB-sRNA (B) and intact 2.7-kb and truncated 1.4-kb ureAB transcripts (D) from wild-type H. pylori 43504 and sRNA overexpression strains under different pH conditions. Arbitrary transcript units were determined by separately quantifying 16S and 23S rRNAs in ethidium bromide-stained gels and ureAB sRNAs from Northern blots at distinct settings for contrast and background subtraction of each sample. Since the bands representing each transcript in the hybridized blots and the bands representing 23S and 16S rRNAs in ethidium bromide-stained gels were quantified in two separate measurements with different settings in contrast and background subtraction, the arbitrary transcript units shown in the bar graphs do not reflect the actual relative levels between the transcripts tested and rRNAs.

The level of 5′ureB-sRNA even at neutral pH is very low compared to that of ureAB (the Northern results were detectable for 5′ureB-sRNA only after overnight exposure to be comparable to those for ureAB, which needed only 30 min of exposure, as shown in Fig. 1).

The truncated 1.4-kb ureAB transcript results from transcription termination.

The Northern blot analysis presented in Fig. 1 showed a higher level of truncated 1.4-kb ureAB transcript at neutral pH compared to under acidic conditions. The results from the 5′ureB-sRNA overexpression strain suggested that 5′ureB-sRNA is needed for these increased levels of truncated ureAB transcript and, concomitantly, the decreased levels of the full-length ureAB mRNA. However, these studies cannot determine whether the truncated ureAB transcript is due to termination of transcription or processing of the full-length mRNA followed by rapid degradation of the ureAB mRNA. In order to discriminate between these two possibilities, both 3′-end and 5′-end S1 nuclease mapping were performed. As shown in Fig. 2, S1 nuclease mapping using 5′-end-labeled probe detected only the band (625 nt) representing the full-length ureB transcripts, without detection of smaller degraded bands. The smaller band that appeared in the 5′-end S1 mapping is not a specific S1 nuclease-digested product, since a similar band was also found in the probe control experiment in which no S1 nuclease was included in the reaction mixture (Fig. 2C). Using the 3′-end-labeled probe, while a band (718 nt) representing full-length transcript corresponding to the unterminated ureB mRNA was detected, two major smaller bands (∼220 nt and ∼195 nt) were also detected after S1 nuclease treatment. These two specific S1 nuclease-digested products may represent the 3′ end of the ureAB gene up to the termination site. Therefore, the results are consistent with a termination mechanism but not with degradation of the processed RNA, in which low-molecular-weight bands corresponding to protection of the labeled probe by transcript representing both the 3′ and 5′ ends of the processed mRNAs should be detected along with a band corresponding to the full-length ureAB transcript resulting from protection by the unprocessed mRNAs (Fig. 2B).

Fig 2.

3′-end and 5′-end S1 nuclease mapping of the ureB 5′ region. (A) Physical map of the ureAB gene cluster with an antisense sRNA complementary to the 5′ one-third of ureB. The corresponding transcripts for the ureAB genes (intact 2.7 kb and truncated 1.4 kb) are shown at the top. The 3′-end-labeled and 5′-end-labeled probes that were used for S1 nuclease mapping are shown at the bottom at the corresponding position. The stars indicate the positions of the radioactive label. Note that the probes are single-stranded cDNA fragments complementary to the ureB coding sequence and therefore oriented in the opposite direction from the ureB gene. (B) Scheme of the expected banding patterns in the S1 nuclease mapping experiments resulting from degradation of the intact mRNA of ureAB or transcription termination at a site downstream of the sRNA coding region. In the case of codegradation, both 3′- and 5′-end-labeled probes should detect low-molecular-weight bands corresponding to protection of the labeled probe by transcripts representing both the 3′ and 5′ ends of the degrading RNA, in addition to a band corresponding to the full-length transcript resulting from protection by the nondegraded mRNAs. In the case of termination, the 5′-end-labeled probe should detect only the full-length mRNA species that have not been terminated, while the 3′-end-labeled probe should detect a band corresponding to unterminated full-length mRNA and smaller distinct bands representing the 3′ end of the transcript up to the termination site. (C) Experimental results of the 5′- and 3′-end S1 nuclease mapping using a ³²P-labeled probe and total RNA from wild-type H. pylori (WT). A control experiment (probe) using the same probes only but without S1 nuclease treatment is also shown.

The transcription termination of ureAB is induced by 5′ureB-sRNA.

Although the S1 nuclease mapping results suggested a transcription termination mechanism, the possibility of the codegradation still could not be completely eliminated, especially if the codegradation happened in an extremely rapid manner. To provide further evidence for transcription termination and to confirm the possible effect of 5′ureB-sRNA on ureAB transcription, we cloned the DNA sequence coding for this antisense sRNA into pCR4-TOPO. The recombinant plasmid was used to synthesize 5′ureB-sRNA by in vitro transcription with T3 RNA polymerase. After purification, 5′ureB-sRNA was added to an in vitro transcription assay mixture in which an 813-bp 5′ureB DNA fragment that spans the potential termination site was fused downstream of a cagA promoter (P_cagA) and used as the template for E. coli RNA polymerase. As shown in Fig. 3A, a smaller band (∼400 nt) representing the terminated transcript is formed in the presence of even a small amount of 5′ureB-sRNA (150 fmol), and its level progressively increases with increasing amounts of 5′ureB-sRNA, becoming the only transcript detectable at higher 5′ureB-sRNA amounts (350 fmol), while the band (∼800 nt) representing the full-length runoff transcript of ureB gradually disappears with increasing amounts of 5′ureB-sRNA. When no 5′ureB-sRNA is added (lane 3) or when rat liver mRNA (lane 2) is added to the reaction mixture, only the full-length transcript is detected. Under the same experimental conditions, the transcription product of a control template (3′ureB) is not affected by 5′ureB-sRNA, with only the band representing full-length transcript (∼500 nt) being detectable (Fig. 3B). These results suggest that the antisense small RNA (5′ureB-sRNA) regulates the transcription of the target gene (ureAB) by a specific transcription termination mechanism.

Fig 3.

In vitro transcription of 5′ ureB as a function of increasing amounts of purified 5′ureB-sRNA. (A) An 813-bp DNA fragment corresponding to 5′ ureB (nucleotides 46 to 859 of the ureB coding sequence) fused downstream of the cagA promoter (P_cagA) was used as the template with E. coli RNA polymerase, [α-³²P]UTP, and NTPs, without addition (lane 3) or with addition of increasing amounts of 5′ureB-sRNA from 100 to 400 fmol (lanes 4 to 10). The reaction mixtures were incubated with and without in vitro-synthesized 5′ureB-sRNA for 30 min at 37°C. (B) A similar in vitro transcription assay with a 515-bp DNA fragment corresponding to 3′ ureB (nucleotides 1123 to 1608 of the ureB coding sequence) fused downstream of P_cagA was used as a control. Lanes 1, marker (ϕX174 DNA/HinfIII). Lanes 2, control with 1 μg rat liver mRNA but without 5′ureB-sRNA.

The amount of the ∼800-nt transcript representing the full-length ureB from an in vitro transcription reaction under the same conditions except for no addition of 5′ureB-sRNA was estimated on an ethidium bromide-stained gel by comparison with known amounts of purified ∼800-nt transcript from a separate reaction without radioactive labeling. About 475 fmol of ∼800-nt transcript was produced in a 5-μl transcription reaction mixture (30-min reaction) without 5′ureB-sRNA. This suggests that a relatively small amount of 5′ureB-sRNA (150 to 300 fmol) is enough to start terminating the higher-level transcription of its target mRNA.

In our in vitro transcription system, the E. coli RNA polymerase was used due to the unavailability of H. pylori RNA polymerase. A previous study has reported that the P_cagA from H. pylori can be activated in vitro by purified E. coli RNA polymerase, which shows activity similar to that of H. pylori RNA polymerase (29). In interpreting terminator signals, we assume that the RNA polymerases from both E. coli and H. pylori also share a similar property.

Potential secondary structure of the 5′ureB-sRNA.

To better understand how this cis-encoded antisense 5′ureB-sRNA regulates the expression of ureAB, analysis of the secondary structure of full-length 5′ureB-sRNA and prediction of the sense mRNA (ureAB) in the complementary region were performed with the Vienna RNA Package program RNAfold web server (http://www.tbi.univie.ac.at/). The sequences of both sRNA and mRNA predict several stem-loop structures, and three of these structures have sequence complementarities in the loops between the sRNA and mRNA (Fig. 4). These structures are designated I, II, and III in the sense mRNA and I*, II*, and III* in 5′ureB-sRNA. A YUNR U-turn motif (UUAA) is found in the stem-loops II and II*. The YUNR U-turn motif (a motif containing a pyrimidine [Y] followed by a uracil [U], any nucleotide [N], and a purine [R]) is important for RNA-RNA interactions by serving as an initial point of contact between the antisense sRNA and the mRNA target and leading to complete duplex formation (34).

Fig 4.

(A) Nucleotide sequence of the region encompassing the 3′ end of ureA and the 5′ end of ureB. The complete antisense sRNA, 5′ureB-sRNA (green line beneath the sequence), including the 5′ end (+1) and the putative −10 promoter sequence (red arrow) for 5′ureB-sRNA, is indicated. The nucleotides of loops of the stem-loop structures are highlighted (yellow for ureB mRNA and blue for 5′ureB-sRNA). (B) The putative stem-loop structures were generated using the RNAfold web server (http://www.tbi.univie.ac.at/). The complementary stem-loop structures (I-I*, II-II*, and III-III*) are shown. Position 1 in the sense RNA structure corresponds to nucleotide 151 of the ureB coding sequence (numbered on the right side of the sequence in panel A). A YUNR U-turn motif is indicated in stem-loop structures II and II*.

To investigate the role of the YUNR U-turn motif in 5′ureB-sRNA-mediated transcription termination, we mutagenized pCR4-P_cagA-5′ureB(813), creating pCR4-P_cagA-5′ureB(813)M, which carries a three-base substitution in stem-loop structure II of 5′ureB mRNA (at positions 322, 323, and 325 of the ureB coding sequence). This mutation was designed to eliminate the YUNR U-turn motif that may facilitate the interaction between 5′ureB-sRNA and its target mRNA. As shown in Fig. 5, 5′ureB-sRNA is not able to cause transcription termination when the mutated 5′ureB mRNA (Mut-YUNR) is synthesized, except when larger amount of 5′ureB-sRNA (450 fmol or more) is added. These results suggested that the YUNR U-turn motif in stem-loop structure II may play a role in the initial pairing between 5′ureB-sRNA and ureB mRNA.

Fig 5.

The mutation in the YUNR U-turn motif in stem-loop structure II reduces the efficiency of transcriptional termination mediated by 5′ureB-sRNA. (A) Stem-loop structure II with a YUNR U-turn motif in 5′ureB mRNA and base exchanges to create the mutated 5′ureB mRNA (Mut-YUNR). (B) An 813-bp 5′ ureB DNA fragment carrying the mutation (Mut-YUNR) fused downstream of P_cagA was used as the template for in vitro transcription with E. coli RNA polymerase, [α-³²P]UTP, and NTPs, without addition (lane 2) or with addition of increasing amounts of 5′ureB-sRNA from 200 to 550 fmol (lanes 3 to 10). The reaction mixtures were incubated with or without in vitro-synthesized 5′ureB-sRNA for 30 min at 37°C. Lane 1, marker (ϕX174 DNA/HinfIII).

DISCUSSION

Emerging evidence suggests that the regulation of gene expression through cis-encoded antisense RNAs (asRNAs) constitutes a distinct level of control in bacteria (35). Different levels of regulation are especially important for genes that must be tightly controlled or are critical in multiple cellular responses. In H. pylori, UreA and UreB are the enzymatically active subunits of the urease enzyme necessary for cytoplasmic and periplasmic pH homeostasis and are present as inactive apoenzyme until nickel insertion by UreE, -F, -G, and -H. Transcription of the ureAB operon is positively regulated by the ArsRS two-component system (18, 20). The periplasmic histidine kinase sensor ArsS perceives lowering of medium pH via protonation of its histidine residue 94 (H94) in the periplasmic input domain (36) and triggers autophosphorylation and subsequent phosphorylation of its cognate response regulator ArsR. The phosphorylated response regulator ArsR (ArsR∼P) then binds to extended regions overlapping the P_ureA promoters (20) to activate the transcription of ureAB into a 2.7-kb mRNA coding for functional UreA and UreB. However, the activity of the urease could be lethal at a relatively neutral pH due to overalkalization of the medium (12, 21). Thus, an additional level of regulation of urease activity in response to relatively neutral pH is needed. We have identified a cis-encoded asRNA (5′ureB-sRNA) (22) that negatively regulates ureAB transcription by leading to the accumulation of a truncated 1.4-kb ureAB transcript at neutral pH. A direct interaction between the promoter for 5′ureB-sRNA and unphosphorylated ArsR was also found (22), suggesting that unphosphorylated ArsR may activate the expression of the 5′ureB-sRNA. In the current study, we show that overexpression of 5′ureB-sRNA significantly increases the level of truncated 1.4-kb ureAB transcript at both pH 7.4 and pH 4.5 (Fig. 1), confirming that this cis-encoded asRNA (5′ureB-sRNA) provides an additional point of control for urease activity by regulating the level of 1.4-kb truncated ureAB transcript.

The chromosomally encoded cis-asRNAs can regulate transcription of the genes carried on the opposite strand in different ways, including transcription termination, codegradation, control of translation, transcriptional interference, and enhanced stability of their respective target transcripts (35). The fact that the truncated 1.4-kb ureAB transcript is longer than the distance (1,247 bp) between the two convergent promoters P_ureA and P_{5′ureB-sRNA} eliminates the possibility that the negative regulation of ureAB is caused by transcriptional interference, which refers to a direct negative influence of one transcriptional process on a second transcriptional process occurring in cis (37). It was speculated that the transient formation of 5′ureB-sRNA/ureAB-mRNA duplexes makes them a target for selective codegradation, which would result in a 1.4-kb truncated ureAB transcript that lacks most of the 3′ ureB transcript (22). A previous study (5) had shown that several species of mRNA, including the 2.7-kb ureAB transcript from the urease gene cluster of H. pylori are more stable at lower pH than at neutral pH and therefore suggested a pH-dependent posttranscriptional regulatory mechanism for urease gene expression by mRNA decay. However, no further studies have been done to show how ureAB mRNA decay is facilitated, and we could not rule out transcription termination that may be responsible for the ureAB transcript truncation.

In this study, we show that this truncation is indeed caused by transcription termination at the 5′ one-third of ureB, resulting in truncated 1.4-kb ureAB transcript that lacks the 3′ two-thirds of ureB. The transcription was investigated in vitro as function of increasing amounts of purified 5′ureB-sRNA using a cagA promoter-driven 813-bp 5′ureB DNA fragment that spans the potential termination site as the template. A terminated transcript (∼400 nt) appeared only in the presence of the 5′ureB-sRNA in place of the full-length transcript (∼800 nt) (Fig. 3A), and the transcription of the control gene (3′ureB) is not affected in vitro even at high concentrations of 5′ureB-sRNA (Fig. 3B), suggesting that the transcription termination of 5′ureB is highly specific to 5′ureB-sRNA.

Antisense-RNA-mediated transcriptional attenuation is a replication control mechanism discovered first in Gram-positive bacteria for the staphylococcal plasmid pT18 (38) and later for the streptococcal plasmids pIP501 (39) and pAMβ1 (40), in which the antisense sRNA can regulate transcription by binding and folding the target RNA so that a Rho-independent terminator structure forms (41). Recently, some evidence has shown that in Gram-negative bacteria an antisense sRNA can cause premature termination of transcription of the target gene by a transcriptional attenuation mechanism (42, 43). The 427-nt antisense sRNA RNAβ, encoded opposite the fatDCBA-angRT iron transport biosynthesis operon in the fish pathogen Vibrio anguillarum, is complementary to the 3′ region of fatA and the 5′ end of angR. The interaction of the asRNA with the growing fatDCBA transcript leads to transcription termination at a potential hairpin close to the fatA stop codon, which results in increased levels of the fatA portion of the mRNA compared to the downstream angRT portion, providing a mechanism for discoordinate expression within an operon (43). Another example is the Shigella flexneri virulence gene icsA, which is downregulated by the antisense sRNA RnaG with both transcriptional interference and transcription attenuation (42). The 5′ part of the nascent icsA RNA forms two long hairpin motifs that seemingly are similar to an antiterminator structure. The binding of RnaG to the actively transcribed icsA mRNA prevents the formation of the antiterminator and promotes the formation of a terminator hairpin that leads to transcription termination (42).

The physical interaction of the asRNA with its target is an essential prerequisite for any asRNA regulatory mechanism. Antisense RNAs can initiate base pairing with their target mRNAs via stem-loops containing a YUNR (where Y is pyrimidine, U is uracil, N is any nucleoside, and R is purine) U-turn motif, which mediates fast RNA pairing in the majority of RNA-controlled systems (34). Using computational analysis of the secondary structure of full-length 5′ureB-sRNA and sense mRNA in its complementary region, we found three stem-loop structures (Fig. 4) that have sequence complementarities in the loops between the antisense sRNA and the sense mRNA, of which stem-loop II contains a YUNR U-turn motif (UUAA). The in vitro transcription assay (Fig. 5) with a mutated 5′ureB mRNA (Mut-YUNR) suggests that this structure may play an important role in facilitating rapid bimolecular interaction between 5′ureB-sRNA and its target mRNA.

Although S1 nuclease mapping and in vitro transcription assays have suggested that the 5′ureB-sRNA downregulates ureAB expression by enhancing transcription termination 5′ of ureB, the computational search around the potential termination site on ureB did not reveal any typical transcription terminator structures with a Rho-independent stem-loop or Rho-binding site (26). This appears to be similar to a report by Stork et al. (43), in which a novel mechanism for transcription termination by an antisense sRNA without Rho-independent terminator structures or Rho binding sites has been suggested. It has been speculated that, as indicated by the in vitro study with the complementary oligoribonucleotides (44), the antisense sRNA that hybridizes to the nascent RNA strand may form part of the terminator and lead to destabilization of the RNA polymerase-template-transcript elongation complex and subsequently termination.

Given that the level of the target mRNA ureAB is significantly higher than that of the regulatory asRNA 5′ureB-sRNA, under any circumstances, and the active termination of ureB transcription by 5′ureB-sRNA in an in vitro transcription system, we believe that transcription termination is a more efficient mechanism in vivo than codegradation in tight control of UreAB expression. With the transcription termination mechanism, a limited amount of 5′ureB-sRNA is sufficient to regulate the relatively high level of ureAB transcript.