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). T4 GT7 phage was used for transduction (Wilson et al. 1979). The E. coli strains MH1 (sup⁰ hsdR ΔlacX74 rpsL) and MG1655 were used as wild types. Strain MH1 ΔiscR∷kan was constructed by GT7-mediated transduction of a kanamycin-resistance marker from strain JW2515 (Baba et al. 2006) provided by National BioResource Project (NIG, Japan). E. coli strain MH1 ΔrnlAB has a deletion of rnlA and rnlB (M. Koga and T. Yonesaki, unpublished results). MG1655 Δisc has a deletion of the isc operon from iscR to iscX (formerly ORF3) (Tokumoto and Takahashi 2001).

Plasmids:

To clone iscR, a DNA fragment was amplified by PCR using the primers 5′-GAGATCCACGCCCTGCTTCTTACATTC and 5′-CGGATCCTGATCAGTACCTTCAAGTTC, and DNA from E. coli W3110 strain as a template. The amplified fragment was digested with NruI and EcoRI and blunted. The resulting fragment (1.3 kb) containing the only ORF of iscR, its promoter, and the Shine-Dalgarno sequence was ligated into pBluescript II KS+ previously digested with EcoRV to construct pBS477NE, where iscR was oriented in the opposite direction as the lac promoter. The 531-bp DNA fragment containing the Shine-Dalgarno sequence and the iscR ORF was obtained by digestion of pBS477NE with EcoRI and SmaI and cloned into pRKNSE (Nakamura et al. 1999) to construct pRKSmE, where iscR was oriented in the same direction as the lac promoter. pRKC92AR, pRKC98AR, and pRKC104AR encoded alanine-substituted IscR at Cys92, Cys98, and Cys104, respectively. These mutants were generated by PCR using mutagenic oligonucleotides and pRKSmE as a template.

To construct pQ-ORF2-95 for expression of His-tagged IscR, a DNA fragment containing iscR was PCR amplified with MG1655 DNA as a template and the primers 5′-GGATCCAGACTGACATCTAAAGGGCG and 5′-GTCGACTATTAAGCGCGTAACTTAACGTC, digested with BamHI and SalI, and ligated into pQE30 (Qiagen).

A DNA fragment containing the N-terminal 39 amino acids of RnlA and 67 bp upstream of rnlA was amplified by PCR with W3110 DNA as a template. The antisense downstream primer was 5′-GGAATTCCTGGCATAGGACCAATTG and the sense upstream primer was 5′-GGAATTCCATACACTTCAG. The amplified fragment was digested with EcoRI and ligated into pRS552 (Simons et al. 1987) to construct an rnlA–lacZ fusion gene, generating pRSrnlA67. pPN–lacZ containing 200 bp upstream of the initiation codon and the N-terminal 50 amino acids of RnlA fused to β-galactosidase was constructed in a similar manner. pRSrnlB-long was constructed as follows. A DNA fragment was amplified by PCR with W3110 DNA as a template and the primers 5′-GGAATTCCATATCTCACTGGAGCCA and 5′-CGGATCCCCTGAGACGTTAATTCCGG. The resulting fragment containing 854 bp upstream of rnlB and the N-terminal 11 amino acids of RnlB was digested with EcoRI and BamHI and ligated into pRS552.

Burst size assay:

MH1 cells with or without different plasmids were grown to 5 × 10⁸ cells/ml in LB medium and infected with wild-type T4 or a dmd mutant at a multiplicity of 0.1 at 30°. At 8 min, the cells were diluted 10⁻⁴-fold with fresh LB medium and further incubated for 70 min. After cells were lysed with chloroform, the total number of progeny phages was counted by plating. The burst size is the ratio of the number of progeny to the number of input phage and infection of wild-type cells with wild-type T4 gives a burst size of 80–110. The relative burst size is expressed as a percentage of wild-type T4 for wild-type cells harboring no plasmid or a vector plasmid. Each value is the mean of at least three independent experiments with a standard deviation.

RNA purification, Northern blot, primer extension, and RT–PCR analysis:

Isolation of total RNA and northern blot analysis were carried out as described previously (Kai et al. 1996). For primer extension analysis to identify transcription start sites of rnlA and rnlB, we used primers 5′-GGAATTCCTGGCATAGGACCAATTG for rnlA and 5′-CGGATCCCCTGAGACGTTAATTCCGG for rnlB, respectively. For RT–PCR analysis, reverse transcription was performed by incubation at 42° for 45 min with 2 μg of total RNA, 100 units of ReverTra Ace reverse transcriptase (TOYOBO), 1 mm of dNTPs, and 5 pmol of gene-specific primers in 20 μl of reverse transcription buffer. Primers, rnlA-dw N134 (5′-CTGGCATAGGACCAATT), ompA-dw (5′-TTGGATTTAGTGTCTGCACG), lpp dw (5′-TTATCTTGCGGTATTTAGTAG), or b2531-up (5′-TTTATCGCCGCCCTGGCAGCC) were used for reverse transcription of rnlA, ompA, lpp, or iscR, respectively. PCR amplification was performed with 0.2 mm dNTPs, 2 μl of reverse transcription mixture for rnlA, 1 μl for ompA and lpp, or 1.5 μl for iscR, 10 pmol of sense and antisense primers, and 0.3 units of KOD Dash (TOYOBO) in 25 μl of PCR buffer. A thermal cycle of 95° for 30 s, 51° for 15 s, and 72° for 30 s was repeated 20 times for rnlA, 16 times for lpp, or 12 times for ompA. A thermal cycle of 95° for 15 s, 60° for 15 s, and 72° for 30 s was repeated 20 times for iscR. Primers for PCR were 5′-ATGTTTCTATGGGATCCAGG and rnlA-dw N134 for rnlA, 5′-ATGAAAAAGACAGCTATCGC and ompA-dw for ompA, 5′-TTATCTTGCGGTATTTAGTAG and lpp-dw for lpp, 5′-GTCTGCGTAAAAATGGTCTG and b2531-up for iscR, respectively. The products from each reaction were separated through a 5% polyacrylamide gel.

Western blot analysis:

Proteins were separated on a 15% polyacrylamide gel containing SDS and electroblotted onto Immuno-Blot PVDF membrane (Bio-Rad). The membranes were probed with a rabbit antibody against IscR and successively with horseradish peroxidase-conjugated donkey anti-rabbit IgG (GE Biosciences). The blot was developed with Immobilon Western Chemiluminescent HRP Substrate (Millipore).

β-Galactosidase assay:

Twenty microliters of bacterial culture was mixed with 80 μl of permeabilization buffer (100 mm Na₂HPO₄, 20 mm KCl, 2 mm MgSO₄, 0.8 mg/ml CTAB, 0.4 mg/ml sodium deoxycholate, and 5.4 μl/ml β-mercaptoethanol) to lyse cells, and 600 μl of substrate solution (60 mm Na₂HPO₄, 40 mm NaH₂PO₄, 1 mg/ml ONPG, 2.7 μl/ml β-mercaptoethanol) was added and incubated at room temperature for 5–15 min. After stopping the reaction by adding 700 μl of 1m Na₂CO₃, absorbance was measured at OD_420nm. The Miller unit is calculated using the following formula: 1 Miller unit = 1,000A₄₂₀/(bacterial density at A₆₀₀)(ml of bacterial culture)(min of reaction time).

Purification of IscR:

MH1 cells harboring pQ-ORF2-95 were grown at 37° to a density of 3 × 10⁸ cells/ml in 100 ml of LB containing 50 μg/ml of ampicillin. After addition of 0.2 mm IPTG, cells were incubated for 2 hr and harvested by centrifugation. All the following procedures were performed below 4°. Cells were suspended with 5 ml of lysis buffer (50 mm NaH₂PO₄, 300 mm NaCl, 10 mm imidazole, 10% (v/v) glycerol, and 1 mm PMSF, pH 8.0) and disrupted with an ultrasonic generator, UD-201 (Tomy Seiko). After centrifugation at 30,000 × g for 30 min, the cleared lysate was loaded onto a Ni-beads column (Ni-NTA Superflow, Qiagen) according to the manufacturer's instruction. The fraction containing His-tagged IscR was eluted with 250 mm imidazole and dialyzed against TEGX buffer [20 mm Tris·Cl (pH 8.0), 0.5 mm EDTA (pH 8.0), 10% (v/v) glycerol, 0.2% Triton-X 100] containing 0.1 m NaCl. After the sample was loaded onto an Affi-Gel Blue column (Bio-Rad), the column was washed with TEGX buffer containing 0.3 m NaCl and His-tagged IscR was eluted with TEGX buffer containing 1 m NaCl. The purified protein was precipitated with ammonium sulfate, dialyzed against TEGX buffer containing 0.1 m NaCl and 50% (v/v) glycerol, and stored at −20°.

Gel mobility shift assay:

A DNA probe (222 bp) containing the rnlA promoters was prepared by PCR with pBRMpNO carrying from yfjM to rnlB as a template and the primers 5′-[³²P]CTGGCATAGGACCAATT and 5′-GGAATTCCGGAGTGCCCCCTATTC. A DNA probe (448 bp) containing the rnlB promoter was prepared by PCR with pBRMpNO as a template and the primers 5′-[³²P] CGGATCCCCTGAGACGTTAATTCCGG and 5′-CCGCATCTCCACGTGACTGC. For each binding assay, 4 × 10³ cpm of gel-purified DNA probe and the indicated amount of His-tagged IscR were added to the binding reaction mixture (15 μl) consisting of 20 mm Tris-HCl (pH 8.0), 50 mm KCl, 4 mm MgCl₂, 0.5 mm EDTA, 0.02 mg/ml BSA, 5 mm DTT, 10% (v/v) glycerol, and 200 ng of poly-[dI-dC] (Roche). After incubating at room temperature for 20 min, half of the reaction mixture was loaded on an 8% polyacrylamide gel in 0.25 × TBE and electrophoresed at 4° for 2–2.5 hr at 150 V. The gel was analyzed by Bio-Image Analyzer (Fuji BAS-1800).

Overexpression of IscR partially restores the growth of a T4 dmd mutant:

To investigate the effect of iscR on the growth of a T4 dmd mutant, wild-type T4 or a dmd mutant was plated to measure the efficiency of plating (Figure 1A). A T4 dmd mutant grew normally on MH1 ΔrnlAB cells lacking RNase LS, but its efficiency of plating decreased to <10⁻⁴ on parental MH1 cells or MH1 ΔiscR cells harboring a vector plasmid, pBluescript II SK− (pBSSK−). Deletion of iscR has no effect on the growth of wild-type T4. Introduction of iscR cloned in a plasmid, pBS477NE, allowed T4 dmd mutants to form plaques on MH1 or MH1 ΔiscR cells with an efficiency of plating of nearly 1, as on MH1 ΔrnlAB cells. Because the high-copy plasmid-encoded iscR but not the genome-encoded iscR restored the growth of a T4 dmd mutant, this effect could be caused by overproduction of IscR. Indeed, overproduction of IscR was clearly demonstrated by Western blot; more than 10-fold IscR was accumulated in cells harboring pBS477NE than in cells harboring pBSSK− (Figure 1C).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.—

Effect of IscR on RNase LS activity. (A) A solution containing wild-type T4 or dmd mutant phages was serially diluted in 10-fold steps, and 2-μl samples containing the number of phages indicated on the left were spotted onto a plate seeded with MH1 or its derivative cells carrying a plasmid as shown below the figure. Photographs were taken after overnight incubation at 30°. (B) MH1 cells with pBSSK− or pBS477NE were infected with wild-type T4 or a dmd mutant and newly synthesized proteins were pulse labeled for 3 min at the times indicated above the figure, as described previously (Ueno and Yonesaki 2001). Labeled proteins from an equal portion of infected cells were separated by 12.5% SDS–PAGE and analyzed by Bio-Image Analyzer (Fuji BAS-1800). The numbers in the right margin show representative late genes encoding the proteins. The 23* indicates a processed product encoded by gene 23. (C) MH1 cells with pBSSK− or pBS477NE were grown to 5 × 10⁸ cells/ml in LB containing 50 μg/ml of ampicillin and total proteins from an equal portion of cells were analyzed by Western blot with an antibody against IscR. (D) MH1 cells with pBSSK− or pBS477NE grown in LB at 30° to a mid-log phase were infected with wild-type T4 or a dmd mutant and total RNAs were extracted at 25 and 35 min after infection. Two micrograms of each RNA was analyzed by Northern blot with a ³²P-labeled DNA probe for soc as described previously (Kai and Yonesaki 2002).

We also examined the effect of overexpressed IscR on progeny production by measuring a number of progeny per infected cell (burst size). A T4 dmd mutant grew very poorly on MH1 cells, with a burst size only 2.3% of that of wild-type phages (Table 1). The burst size of a T4 dmd mutant on MH1 cells harboring pBS477NE was reproducibly higher (21%) than on MH1 cells. Thus, suppression by overexpression of IscR was remarkable, though partial, as with the std-1, -3, and std-4 mutations (Otsuka et al. 2003). The burst size of the wild-type phage in MH1 cells harboring pBS477NE was reduced to ∼86%, compared to parental MH1 cells.

Because deprivation of intracellular iron is known to induce the expression of IscR, we measured the burst size of a T4 dmd mutant on MH1 cells in the presence of 0.25 mm 2, 2′-dipyridyl, a membrane-permeant ferrous iron chelator that causes deprivation of intracellular iron (supporting information, Table S1). Treatment with this reagent partially restored the growth of a T4 dmd mutant, although it reduced the growth ability of wild-type T4. This result suggested that an increase of intracellular IscR counteracts RNase LS activity.

Next, we investigated the effect of IscR on gene expression in a T4 dmd mutant. Gene expression at each stage after T4 infection was examined by pulse labeling newly synthesized proteins (Figure 1B). Consistent with previous reports, protein synthesis by a T4 dmd mutant at late times (23 and 40 min) was impaired in MH1 cells carrying pBSSK−. In contrast, protein synthesis in MH1 cells harboring pBS477NE infected with a dmd mutant was almost as high as in MH1 cells infected with wild-type T4.

Suppression of the growth defect of a T4 dmd mutant should accompany the stabilization of late-gene mRNAs. To confirm this expectation, the steady-state level of T4 late-gene soc mRNA was examined by Northern blot, because its degradation has been extensively characterized (Kai and Yonesaki 2002). As shown in Figure 1D, the amount of soc mRNA at late stages (25 and 35 min) after T4 dmd infection to MH1 cells with a vector alone was hardly detected because of rapid degradation. However, when a T4 dmd mutant infected MH1 cells with pBS477NE, soc mRNA was present as abundantly as in MH1 cells infected with wild-type T4. Our primer extension analysis also confirmed the loss of RNase LS-mediated specific cleavages in MH1 cells with pBS477NE (Figure S1). All of the above results clearly indicated that overexpression of IscR decreased the RNase LS activity to allow the growth of a T4 dmd mutant.

Ligation of the Fe-S cluster with IscR is not required for inhibition of RNase LS activity:

We examined whether ligation of the IscR with Fe-S cluster is necessary for the suppression of the growth defect of a T4 dmd mutant. Cys92, Cys98, and Cys104 are predicted Fe-S cluster ligands in IscR and indeed, IscR-C92A with an alanine substitution did not contain an Fe-S cluster, when it was purified under anaerobic condition (Nesbit et al. 2009). Cells harboring a plasmid expressing IscR with individual alanine substitutions of these amino acids were examined for supporting the growth of a T4 dmd mutant (Table 2). pRKSmE expressing wild-type IscR raised the burst size of a T4 dmd mutant from 1.1 without plasmid to 28.3 with plasmid. Each of these substitutions (pRKC92AR, pRKC98AR, or pRKC104AR) resulted in a burst size similar to that for wild-type IscR. In addition, soc mRNA in infection of MH1 cells harboring these mutant plasmids with a T4 dmd mutant was accumulated as abundantly as that in cells harboring pRKSmE (Figure 2). These results demonstrated that ligation of the IscR with Fe-S cluster had almost no effect on IscR-mediated repression of RNase LS activity.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.—

Effect of an Fe-S cluster in IscR on RNase LS activity. MH1 cells with the pRKNSE vector, pRKSmE, pRKC92AR, pRKC98AR, or pRKC104AR grown in LB medium at 30° to a mid-log phase were infected with a T4 dmd mutant, total RNAs were extracted at the indicated times after infection, and soc mRNA was analyzed by Northern blot as described previously (Kai and Yonesaki 2002).

Genes downstream of iscR in isc operon are not involved in inhibition of RNase LS activity:

iscR is part of the isc operon and negatively regulates its own expression (Schwartz et al. 2001). Repression of downstream genes in this operon by IscR might be responsible for inhibition of RNase LS activity. To investigate this possibility, we measured progeny production of a T4 dmd mutant on MG1655 and MG1655 Δisc. The deletion of isc operon did not alleviate the growth defect of a T4 dmd mutant (Table S2). This result suggested that components of iscS to iscX were not involved in the inhibition of RNase LS activity.

Transcription start sites for rnlA and rnlB were identified by primer extension analysis:

We tested whether IscR can directly inhibit RNase LS activity in vitro using S30 extract from MH1 cells, but IscR could not inhibit RNA cleavages by RNase LS (Figure S2). Next, we investigated another possibility that IscR regulates RNase LS activity via transcriptional control of rnlA and/or rnlB. To identify promoters experimentally, we mapped the 5′ ends of rnlA and rnlB transcripts. Unfortunately, we could not detect endogenous rnlA and rnlB transcripts by primer extension or Northern blot analysis because their expression was very weak. To circumvent this problem, we constructed two plasmid derivatives of pBR322: pPN-lacZ containing 200 bp upstream of the translation start site of rnlA and 150 bp encoding the N-terminal 50 amino acids of RnlA, and pRSrnlB-long containing 854 bp upstream of the translation start site of rnlB and 33 bp encoding the N-terminal 11 amino acids of RnlB. MH1 cells transformed with either of these plasmids were incubated until log phase and then total RNA was extracted. Primer extension analysis identified four 5′ termini specific to rnlA transcripts 65, 46, 32, and 24 bp upstream of the initiation codon (Figure 3A). In addition to a strong signal for the last site, this site solely matches a transcription site from the promoter predicted by the GENETYX program (Figure 7A). Therefore, rnlA mRNA would be mainly transcribed from 24 bp upstream of the rnlA initiation codon. This idea was confirmed by β-galactosidase reporter assays in the next section. On the other hand, the other three sites might correspond to unexpected reverse transcriptase stop sites or RNA processing sites in transcripts originating from a weak plasmid promoter. Two 5′ termini specific to rnlB transcripts were also detected 123 and 124 bp upstream of the initiation codon (Figure 3B), which are within the rnlA ORF. Two starts depending on a single promoter are often observed in E. coli (for example, refer to Otsuka et al. 2003). Although a promoter corresponding to these sites was not predicted by GENETYX, another program, BPROM (http://linux1.softberry.com/berry.phtml?topic=bprom&group=programs&subgroup=gfindb), identified a promoter matching these sites as transcription start sites (Figure 7B). Indeed, gene expression driven by a sequence containing this region was confirmed in the next section.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.—

Identification of rnlA and rnlB transcription start sites. (A) Total RNAs were extracted from MH1 cells with pPN–lacZ grown in LB medium at 30° when the OD₆₀₀ reached 0.7. Ten micrograms of RNA was used for primer extension analysis as described in materials and methods. (B) Total RNA was extracted from MH1 cells harboring pRSrnlB-long (rightmost lane) when the OD₆₀₀ reached 0.65 and was analyzed using primer extension analysis. A set of sequence ladders for rnlA or rnlB was obtained by dideoxy sequencing method with the same primer and run in parallel in each figure. Arrows indicate bands detected by primer extension.

IscR represses transcription from rnlA promoter but not from rnlB promoter:

To investigate whether IscR regulates transcription from the rnlA and rnlB promoters, we examined the effect of an iscR deletion on the expression of β-galactosidase from pRSrnlA67, from which the expression of rnlA–lacZ translational fusion gene depended on 67 bp of sequence upstream of rnlA, or pRSrnlB-long (Figure 4). First, a plasmid containing no promoter for lacZ resulted in a low level of β-galactosidase activity (background value of 75). On the other hand, β-galactosidase activity expressed from the rnlA and rnlB promoters was significantly higher than background, indicating that both inserted DNAs provided promoter activity. Second, deletion of iscR led to three- to fourfold increase of β-galactosidase activity driven by the rnlA promoter at any cell growth stage. In contrast, deletion of iscR had no effect on the activity driven by the rnlB promoter. Third, β-galactosidase activity driven by the rnlA promoter increased with cell growth in both wild-type and iscR-deleted cells, but the activity driven by the rnlB promoter was nearly constant. Interestingly, β-galactosidase activity driven by the rnlB promoter was considerably higher than that driven by the rnlA promoter at any cell growth stage; the former activity was more than eightfold the latter even in the absence of iscR. This result indicates that transcription activity from the rnlA promoter is much weaker than from the rnlB promoter, so that RnlB would be synthesized more vigorously than RnlA in wild-type cells. Putting it all together, we conclude that the rnlA and rnlB promoters are regulated independently and that IscR represses transcription from the rnlA promoter (refer to Figure 7A).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.—

β-Galactosidase reporter assay with rnlA or rnlB promoter. MH1 or MH1 ΔiscR cells with pRSrnlA67 (A) or pRSrnlB-long (B) were grown in LB medium at 30° and β-galactosidase activities were measured as described in materials and methods, when the OD₆₀₀ reached the value indicated in the figure. The β-galactosidase activity is shown as Miller units. Each value is given after subtracting the background value of 75, which was obtained for cells harboring pRS552, and shows the mean ± SD from two or three independent experiments.

IscR can bind to rnlA promoter:

Since the above results suggested that IscR specifically binds to the rnlA promoter, we performed gel mobility shift assays with different concentrations of purified His-tagged IscR and ³²P-labeled DNA probes. The DNA probe for rnlA contained a region from −99 to +117, and the DNA probe for rnlB from −410 to +33. As shown in Figure 5, a shifted rnlA DNA probe was apparent in the presence of 50 nm IscR, although 200 nm IscR did not significantly shift the rnlB DNA probe.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.—

Binding of His-tagged IscR to rnlA promoter region. His-tagged IscR at increasing concentrations from 0 to 200 nm was incubated at room temperature for 20 min with a ³²P-labeled DNA probe (4 × 10³ cpm) containing the rnlA promoters or the rnlB promoter. The mixtures were separated through an 8% polyacrylamide gel in 0.25 × TBE at 4°. F and C indicate a free probe and a DNA–protein complex, respectively.

IscR represses transcription from chromosomal rnlA:

Finally, we confirmed that IscR represses transcription of rnlA in the chromosome. The amount of endogenous rnlA mRNA in MH1 cells harboring pBSSK− or pBS477NE was measured by RT–PCR (Figure 6A). Simultaneously, lpp and ompA mRNAs were measured as loading controls and iscR mRNA was also measured to confirm its overexpression. As expected, MH1 cells harboring pBS477NE overexpressed iscR mRNA by approximately threefold (bottom panel) and instead underexpressed rnlA mRNA by approximately twofold (lane 4), compared to MH1 cells with pBSSK− (lane 3). On the other hand, the amount of endogenous rnlA mRNA in MH1 ΔiscR cells was approximately twofold more than that in MH1 cells (Figure 6B). Finally, we observed that MH1 cells under oxidative stress overproduced iscR mRNA by approximately threefold and simultaneously underexpressed rnlA mRNA by ∼1.5-fold (Figure 6C).

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.—

Negative correlation between the levels of rnlA mRNA and iscR mRNA in vivo. (A) MH1 cells (lane 1), MH1 ΔrnlAB cells (lane 2), and MH1 cells with pBSSK− (lane 3) or pBS477NE (lane 4) were grown to mid-log growth phase in LB medium and total RNAs were extracted. RT–PCR analyses for rnlA, ompA, lpp, and iscR mRNAs were performed as described in materials and methods. Various amounts of pBRMpNO carrying sequence from yfjM to rnlB were used as a template to demonstrate a semiquantitative profile of PCR conditions; lane 5, 2 pg; lane 6, 4 pg; lane 7, 8 pg; and lane 8, 16 pg. An asterisk indicates a band of nonspecific product. Lane 9 shows size markers. (B) MH1 or MH1 ΔiscR cells were grown to mid-log growth phase in LB medium and total RNAs were analyzed for rnlA and iscR mRNAs by RT–PCR as described above. (C) MH1 cells were grown to mid-log growth phase in M9C medium and treated with or without 1 mm of H₂O₂ for 10 min. Total RNAs were analyzed by RT–PCR for rnlA and iscR mRNAs as described above.