Phages and bacterial strains:

Wild-type bacteriophage T4 is T4D. The amSF16 mutant contains an amber mutation in the dmd gene (Kai et al. 1996; Ueno and Yonesaki 2001). E. coli K-12 strain MH1 (sup⁰ araD139 hsdR ΔlacX74 rpsL) was described previously (Otsuka and Yonesaki 2005). TY0807 (MH1 araD⁺) was constructed by T4 GT7 phage transduction (Wilson et al. 1979) of araD⁺ from W3110 into MH1 and was screened for the ability to grow on M9 minimal medium containing arabinose as a sole carbon source. TY0802 (MH1 ΔrnlAB) and TY0809 (TY0807 ΔrnlAB) were constructed as described (Datsenko and Wanner 2000). Briefly, a fragment containing a chloramphenicol-resistance cassette flanked with the sequences upstream of rnlA and downstream of rnlB was amplified by PCR with pKD3 as a template and the primers 5′-attgtagagtttccccatatgtttctatgggatccaggaacatatgaatatcctccttag and 5′-gttaatatcatgccaaaagggcgaattctatactggttcgtgtgtaggctggagctgctt. The fragment was introduced into W3110 harboring pKD46, which encodes λ-phage Red. Chloramphenicol-resistant colonies were screened by PCR with primer 1 (5′-atgtttctatgggatccagg) and primer 2 (5′-gctatttgatcatattggac) to select ΔrnlAB∷cat cells. After ΔrnlAB∷cat was transferred into MH1 or TY0807 by T4 GT7 phage transduction (Wilson et al. 1979), the chloramphenicol-resistance cassette was removed by yeast Flp recombinase expressed from pCP20 (Cherepanov and Wackernagel 1995) to construct TY0802 and TY0809. BW25113, BW25113 ΔclpA∷kan, BW25113 ΔclpX∷kan, and BW25113 Δlon∷kan were kindly provided by the National BioResource Project (National Institute of Genetics, Kyoto, Japan).

Plasmids:

Plasmids constructed in this work are listed in Table 1. To clone the rnlAB ORFs, a DNA fragment was amplified by PCR with W3110 DNA as a template using primers 1 and 2 and ligated into the EcoRV site in pBluescript II SK+ to construct pMK02, in which rnlAB was oriented in the direction opposite of the lac promoter. The rnlA promoter region was amplified by PCR with W3110 DNA as a template using the primers 5′-cgatcgatgttgctgcttgg and 5′-cgaagcccagcccttgaccc. The resulting DNA fragment was digested with ClaI and BamHI and ligated into pBR322. The EcoRI site of this plasmid was disrupted by EcoRI digestion, blunted with T4 DNA polymerase (Nippon Gene), and self-ligated. The BamHI–SalI fragment in the constructed plasmid was replaced with the BamHI–SalI fragment from pMK02 to construct a plasmid containing the rnlAB ORFs with an rnlA promoter region to yield pMK05.

To remove rnlA, pMK05 was digested with BamHI and SphI, blunted with T4 DNA polymerase, and self-ligated to construct pMK08. In addition, a DNA fragment containing the rnlB ORF alone was amplified by PCR with pMK05 as a template using the primers 5′-cgggatccaaggacttatatattg and primer 2, digested with BamHI and EcoRI, and substituted for the rnlAB region in pMK05 to construct pMK09. A DNA fragment containing rnlB was excised with BamHI and HindIII from pMK09 and subcloned into pJK289 (Kato and Ikeda 1996) to construct pMK33.

A nonsense codon was introduced into rnlB in pMK05 by PCR using the primer 5′-ctgtcgttatgtgaaccggcttcgag (mutagenic sequence underlined) and the rnlA region in the resulting plasmid was removed in the same manner as pMK08 to construct pMK14.

To construct pMK19, a DNA fragment was amplified by PCR with pMK05 as a template using primers 3 (5′-cgggatccatgtttgaaatcaccgg) and 4 (5′-gctgcaggttcgtttagaaag), digested with BamHI and PstI, and ligated into pQE-80L (Qiagen). pMK20 was constructed in the following way: A duplex formed between 5′-aattaaagaggagaaag and 5′-gatcctttctcctcttt was ligated into pQE-80L previously digested with EcoRI and BamHI to remove the 6×His-tag region. The resulting plasmid was digested with BamHI and PstI and ligated with an rnlB-containing fragment obtained by PCR using pMK05 as a template with primers 3 and 4.

To construct pBAD24–rnlA, a DNA fragment was amplified by PCR with pMK05 as a template using the primers 5′-ggaggaattcaccatgacaatcaggagttacaaaaac and 5′-agtgctgcagtcataatttcaacctagctacgc, digested with EcoRI and PstI, and ligated into pBAD24 (Guzman et al. 1995). A DNA fragment containing N-terminal Flag-tag (8 amino acids: DYKDDDDK) fused to RnlA was amplified by PCR with pMK05 as a template using the primers 5′-ccatggtaccagattacaaggatgacgacgataagacaatcaggagttacaaaaac (Flag peptide sequence underlined) and 5′-aactgcagaactcaaacaatatataag, digested with KpnI and PstI, and ligated into pBAD24. The fragment containing flag–rnlA was excised from the resulting plasmid with BamHI and PstI and ligated into pBAD33 (Guzman et al. 1995) to construct pBAD33F–rnlA.

To construct pHU102, a DNA fragment was amplified by PCR with W3110 genomic DNA as a template using the primers 5′-gcgagctcatgacggctgaattg and 5′-ggggtaccttacatcattacgac, digested with SacI and KpnI, and ligated into pQE-80L.

mRNA functional decay rate:

Cells were grown at 30° in M9-glycerol medium, supplemented with 0.3% casamino acids, 1 μg/ml thiamine and 20 μg/ml tryptophan, and rifampicin was added to 500 μg/ml. Time 0 was set 2 min after the addition of rifampicin. [³⁵S]methionine/cysteine (American Radiolabeled Chemicals, St. Louis; >37 TBq/mmol) was added to a 100-μl culture to 3.7 MBq/ml at various times for 3 min. Labeled proteins from equal portions of cell culture were analyzed by electrophoresis through a 12.5% polyacrylamide gel containing 0.1% SDS and the half-life of bulk mRNA was calculated by plotting against time the sum of intensities for all bands present between the top and bottom of each lane. An autograph was taken with a Bio-Image analyzer (Fuji BAS-1800). The band intensities were quantified using the ImageJ program.

Northern blot analysis:

Cells were grown at 30° in LB medium. A 1.5-ml aliquot of cell culture was quickly chilled on ice and cells were harvested by centrifugation. Total RNA extraction and northern blotting were carried out according to Kai et al. (1996). Radioactive probes for ompA and lpp mRNAs were prepared according to Ueno and Yonesaki (2004). An autograph was taken with a Bio-Image analyzer (Fuji BAS-1800). The band intensities were quantified using the ImageJ program.

Pull-down of Flag–RnlA and His–RnlB:

S30 fractions of cell extract were prepared as described (Nirenberg 1963; Stanley and Wahba 1967) and an aliquot containing 20 mg protein was mixed with 20 μl of Ni-NTA agarose beads (Qiagen) or anti-Flag M2 agarose beads (Sigma-Aldrich) by end-over-end rotation for 8 hr at 4°. The agarose beads were sedimented at 15,000 rpm for 1 min at 4° and washed five times with 1 ml of TMCK-T buffer [10 mm Tris-HCl (pH 7.5), 10 mm magnesium acetate, 30 mm KCl, 0.02% Tween 20] containing 40 mm imidazole for Ni-NTA agarose or TMCK-T buffer containing 0.1 m NaCl for anti-Flag M2 agarose. Bound proteins were eluted (EL) from Ni-NTA agarose with 50 μl of TMCK-T buffer containing 400 mm imidazole or from anti-Flag M2 agarose with 50 μl of TMCK-T buffer containing 0.1 m NaCl and 0.2 mg/ml Flag peptide (Sigma-Aldrich).

Western blot analysis:

Proteins were separated by electrophoresis through a 15% polyacrylamide gel containing 0.1% SDS and transferred onto a PVDF membrane. Membranes were blocked for 60 min at room temperature with 5% skim milk in TBS-Tween [20 mm Tris-HCl (pH 8.0), 150 mm NaCl, 0.05% Tween-20] and incubated overnight at 4° with a mouse anti-His antibody (GE Healthcare, 1:3000) or mouse anti-Flag antibody (Sigma, 1:5000). Next, membranes were incubated for 60 min at room temperature with HRP-coupled sheep anti-mouse antibody (GE Healthcare, 1:10,000). Protein bands were detected with Immobilon Western Chemiluminescent Substrate (Millipore) and an LAS image analyzer (Fujifilm).

Mutations in rnlB:

To construct an rnlB mutant, we attempted to replace the chromosomal rnlB with rnlB∷kan in a plasmid with a temperature-sensitive DNA replication system (Link et al. 1997). Although this technique was easy and effective for disrupting chromosomal rnlA (Otsuka and Yonesaki 2005), it was difficult to obtain an rnlB∷kan mutant with this technique and the only candidate was found to have a frame-shifting deletion at nucleotide position 267 in the rnlA ORF in addition to rnlB∷kan. This strain completely lacked RNase LS activity, like rnlA∷kan cells. In addition, we found that JW5418 (ΔrnlB∷kan) from the National BioResource Project also lacked RNase LS activity because it had a deletion of five nucleotides at the C-terminal end of the rnlA ORF. In contrast, when cells were previously transformed with pMK33 carrying rnlB, the chromosomal rnlB could be efficiently replaced with rnlB∷kan. These observations suggested that rnlA in the absence of rnlB was deleterious to cells.

Effects of rnlA expression on the growth of ΔrnlAB cells:

The above result suggested that RnlA is toxic to cells and RnlB neutralizes the toxicity of RnlA. To investigate the toxicity of RnlA, TY0807 (wild type) and TY0809 (ΔrnlAB) cells were transformed with pBAD24–rnlA and transformants were cultivated on LB agar plates containing various concentrations of l-arabinose (Figure 1A). TY0809 cells grew as normally as TY0807 cells did in the absence of l-arabinose. However, the colony size of TY0809 was much smaller than that of TY0807 in the presence of 0.01% l-arabinose and the growth of TY0809 was almost completely inhibited in the presence of 0.1% l-arabinose. On the other hand, the induction of RnlA hardly affected the growth of TY0807, suggesting that endogenous RnlB counteracted the RnlA toxicity. Similarly, when TY0809 cells were cotransformed with pBAD24–rnlA and pMK33, the latter of which expressed RnlB constitutively, impairment of cell growth by pBAD24–rnlA in the presence of l-arabinose was not observed (Figure 1B).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.—

Effects of rnlA and rnlB on growth of ΔrnlAB cells. (A) An overnight culture of TY0807 (wild type) or TY0809 (ΔrnlAB) harboring pBAD24–rnlA was serially 10-fold diluted, and 1 μl of each dilution was spotted onto LB plates supplemented with 50 μg/ml ampicillin and various concentrations of l-arabinose as indicated below the figure. (B) An overnight culture of TY0809 harboring pBAD24–rnlA and either pJK289 (vector) or pMK33 was serially 10-fold diluted and spotted onto LB plates supplemented with 50 μg/ml ampicillin, 30 μg/ml kanamycin, and various concentrations of l-arabinose.

Effects of rnlA on mRNA stability in the absence of rnlB:

Because RnlA is an essential component of RNase LS and has an endoribonuclease activity in vitro (Otsuka et al. 2007), RnlA expression was anticipated to stimulate the degradation of E. coli mRNAs. To investigate this possibility, we measured functional mRNA half-lives with or without induction of RnlA (Figure 2A) by estimating the ability of mRNA to direct protein synthesis after transcription has been blocked. TY0809 cells harboring pBAD24–rnlA were grown to midlog phase and supplemented or not with 0.2% l-arabinose. After incubation for 4 min, rifampicin, a transcriptional inhibitor, was added to measure the functional half-lives of mRNAs. The half-life in the absence of l-arabinose was 8.1 min and was reduced to 3.8 min in the presence of l-arabinose. To examine the decay of individual transcripts, we measured the stability of lpp and ompA transcripts. These two transcripts are known to be stable species (Ueno and Yonesaki 2004) and were therefore chosen to facilitate measurement of mRNA decay accelerated by RnlA. After induction of rnlA, ompA mRNA was significantly reduced at 8 min and was almost undetectable at 32 min. Reduction of lpp mRNA was also significant after 16 min (Figure 2B). To confirm that mRNA degradation activity was increased after induction of rnlA, we examined the decay rate of these mRNAs. At 4 min after induction of rnlA, rifampicin was added to cell cultures and total RNAs were extracted at various times for Northern blotting. The half-lives were 28.4 min for lpp mRNA and 20.1 min for ompA mRNA, respectively, without induction of rnlA. The half-lives became shorter after induction of rnlA, 8.5 min for lpp mRNA and 3.7 min for ompA mRNA, respectively (Figure 2C).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.—

Effects of rnlA on functional and chemical decay of mRNA in the absence of rnlB. (A) TY0809 cells harboring pBAD24–rnlA were grown until the OD₆₀₀ reached 0.5 and were then divided into two subcultures, to one of which 0.2% l-arabinose (+) was added. After further incubation for 4 min, functional decay was measured as described in materials and methods. (B) TY0809 cells harboring pBAD24–rnlA were treated with l-arabinose when the OD₆₀₀ reached 0.5. Total RNAs were extracted at various times after induction and subjected to Northern blotting as described in materials and methods. (C) TY0809 cells harboring pBAD24–rnlA were treated for 4 min with or without l-arabinose when the OD₆₀₀ reached 0.5, and rifampicin was added to 500 μg/ml. Time 0 was set 2 min after addition of rifampicin. Total RNAs were extracted at various times and were subjected to Northern blotting as described in materials and methods. Three micrograms of total RNAs were applied to a 4% polyacrylamide gel containing 7 m urea.

Effect of rnlB on RNase LS activity:

The above results implied that RnlB suppresses RnlA or RNase LS activity. To clarify this possibility, we adopted a T4 dmd mutant to monitor the RNase LS activity. dmd encodes an inhibitor of RNase LS (Otsuka et al. 2007) and, when dmd is defective, RNase LS rapidly degrades T4 mRNA at late stages of infection, terminating T4 development (Kai et al. 1996). Cells were spread on an agar plate, a T4 suspension was spotted on the cells and T4 growth was visualized after overnight incubation (Figure 3). TY0802 (ΔrnlAB) cells showed plaques or a zone of clearing when either wild-type T4 or a dmd mutant was spotted, indicating vigorous growth of T4. MH1 (wild type) cells harboring no plasmid or a vector alone allowed vigorous growth of wild-type T4 but restricted the growth of T4 dmd; the efficiency of plating was reduced to 10⁻² and the plaque size was sharply reduced, indicating that T4 growth was inhibited by RNase LS (Kai et al. 1996). MH1 cells with pMK08 carrying rnlB, like TY0802 cells, supported the growth of a dmd mutant as well as of wild-type T4, suggesting that RNase LS activity was impaired by RnlB probably overexpressed from the plasmid. Consistently, this effect of pMK08 on the growth of T4 dmd was completely eliminated by introducing a stop codon into the rnlB ORF (pMK14).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.—

Effect of rnlB on growth of a T4 dmd mutant. From a solution containing T4 wild-type (+) or dmd mutant (−), 1 μl containing the number of phages indicated in the left margin was spotted onto a plate seeded with TY0802 cells or MH1 cells alone or with pBR322, pMK08, or pMK14. Photographs were taken after overnight incubation at 30°.

We further confirmed the suppressive role of RnlB for RNase LS activity using pMK20 carrying lacI and rnlB; the former was constitutively expressed and the latter was placed downstream of the lac promoter. MH1 harboring pMK20 was grown to midlog phase and rnlB was induced for 30 min with various concentrations of IPTG. The cells were infected with T4 phage and the number of progeny phage per infected cell (burst size) was measured after incubating for an additional 70 min (Table 2). The burst size of wild-type T4 was 100–110 in the range of IPTG concentrations adopted in this experiment. On the other hand, a dmd mutant could not propagate in the absence of IPTG, the burst size being <1. When IPTG was added to 5–50 μm, the burst size increased as IPTG concentration increased; at 50 μm, the burst size reached ∼100, comparable to the burst size of wild-type phage, suggesting that suppression of RNase LS activity was nearly complete. To assess the inhibitory effect of RnlB on RNase LS activity in term of mRNA cleavage, total RNA was extracted from MH1 cells harboring pMK20 at 20 min after infection by a dmd mutant and analyzed for cleavage in soc mRNA by primer extension analysis (Figure 4). In agreement with previous experiments (Kai and Yonesaki 2002; Otsuka and Yonesaki 2005), we identified RNase LS-specific cleavages in soc mRNA at nucleotide positions 135, 144, 145, 153, 172, 185, and 207 from the 5′ terminus. These RNase LS-specific cleavages decreased with increasing IPTG concentration and the amount of full-length soc mRNA increased in parallel. On the other hand, the cleavage at nucleotide position 59 generated by RNase E (Otsuka et al. 2003) was not affected by induction of RnlB. These results suggested that overexpression of RnlB specifically represses RNase LS activity.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.—

Suppression of RNase LS activity by rnlB. MH1 cells harboring pMK20 were grown in M9C medium at 30° until the OD₆₀₀ reached 0.6, whereupon IPTG was added and the culture was further incubated for 30 min. The cells were infected with T4 dmd at a multiplicity of infection of 10, total RNAs were extracted at 21 min after infection, and 5 μg each of total RNAs were used for primer extension as described in materials and methods. A set of sequence ladders for soc was obtained by dideoxy sequencing with same primer and run in parallel. The band indicated by F corresponds to full-length soc RNA. The bands at nucleotide positions 135, 144, 145, 153, 172, 185, and 207 correspond to RNase LS-specific cleavage sites. The band at position 59 corresponds to an RNase E-cleavage site.

Association of RnlA and RnlB:

To investigate an interaction between RnlA and RnlB, we performed a pull-down assay. For this purpose, we constructed pBAD33F–rnlA encoding Flag-tagged RnlA and pMK19 encoding His-tagged RnlB to facilitate their detection. Flag–RnlA showed toxicity similar to RnlA in TY0809 and His–RnlB could neutralize the toxicity, indicating that both tagged proteins are functional (Figure 5A). S30 extract was prepared from TY0809 cells harboring pBAD33F–rnlA and pMK19 or pHU102 encoding His–Rng as a control and mixed with Ni-NTA beads or anti-Flag M2 beads. After washing with buffer, bound proteins were eluted with imidazole or Flag peptide. In a pull-down experiment with Ni-NTA beads, Flag–RnlA was efficiently recovered in the bound protein fraction (EL) together with His–RnlB, but not with His–Rng (Figure 5B). In a reciprocal experiment with anti-Flag M2 beads, His–RnlB, but not His–Rng, could be recovered in the bound protein fraction together with Flag–RnlA (Figure 5C). These results suggested that RnlA and RnlB were specifically associated with each other.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.—

Association of RnlA and RnlB. (A) An overnight culture of TY0807 or TY0809 harboring pBAD33F–rnlA was serially 10-fold diluted, and 1 μl of each dilution was spotted onto an LB plate containing 30 μg/ml chloramphenicol with or without 0.1% l-arabinose (left two panels). An overnight culture of TY0809 harboring pBAD33F–rnlA and either pQE-80L (vector) or pMK19 was serially 10-fold diluted, and 1 μl of each dilution was spotted onto an LB plate containing 50 μg/ml ampicillin, 30 μg/ml chloramphenicol, and 10 μm IPTG with or without 0.1% l-arabinose (right two panels). (B and C) S30 fractions of TY0809 cells harboring pBAD33F–rnlA and either pMK19 or pHU102 were subjected to pull down with (B) Ni-NTA or (C) anti-Flag M2 agarose beads. Loaded (S30), unbound (UB) and eluted (EL) fractions were analyzed by Western blotting with anti-Flag (first row) or anti-His antibody (second and third rows) as described in the materials and methods. Induced proteins were shown with + above the figure.

Stabilities of RnlA and RnlB:

The relation between RnlA and RnlB reminded us of a type II TA system. Because type II TA systems are composed of a stable toxin and an unstable antitoxin, we investigated stabilities of RnlA and RnlB. TY0809 harboring pBAD33F–rnlA and pMK19 were grown to midlog phase and the two proteins were co-induced with l-arabinose and IPTG. After tetracycline was added to stop further translation, the amounts of RnlA and RnlB were monitored by Western blotting and their half-lives were measured. The half-lives were 27.6 min for Flag–RnlA and 2.1 min for His–RnlB, respectively (Figure 6A, left half). T4 infection rapidly shuts off host gene expression (Svenson and Karlstrom 1976; Kashlev et al. 1993; Ueno and Yonesaki 2004) and we accordingly anticipated that rapid degradation of RnlB also occurred upon T4 infection. In fact, His–RnlB was degraded rapidly after infection of a T4 dmd phage, while Flag–RnlA remained stable (Figure 6A, right half). To identify protease(s) responsible for rapid degradation of RnlB, we examined the effects of Lon and Clp, which are major proteases in E. coli (Figure 6B). Because the decay of RnlB was not significant in the first 5 min with the BW25113 strain adopted for this experiment, the half-life of RnlB was calculated on the basis of its amounts after 5 min and was 6.7 min in wild type and 4.5 min in ΔclpA. On the other hand, it was prolonged to 14 min in ΔclpX and 23 min in Δlon.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.—

Decay of RnlA and RnlB. (A) TY0809 cells harboring pBAD33F–rnlA and pMK19 were grown in LB at 30° and, when the OD₆₀₀ reached 0.5, Flag–RnlA and His–RnlB were co-induced with 0.1% l-arabinose and 0.05 mm IPTG for 30 min. Then tetracycline at the final concentration of 20 μg/ml or a T4 dmd mutant at a multiplicity of infection of 5 was added to the culture. S30 fractions were prepared from cells at various times after addition of tetracycline or infection of T4 and 15 μg of proteins were analyzed by Western blotting with anti-Flag (first row) or anti-His antibody (second row) as described in materials and methods. (B) BW25113, BW25113 ΔclpA∷kan, BW25113 ΔclpX∷kan, and BW25113 Δlon∷kan harboring pMK19 were grown in LB at 30° and, when the OD₆₀₀ reached 0.5, His–RnlB was induced with 0.05 mm IPTG for 30 min. S30 fractions prepared from cells at various times after addition of tetracycline were analyzed as in A.