A Novel Fur- and Iron-Regulated Small RNA, NrrF, Is Required for Indirect Fur-Mediated Regulation of the sdhA and sdhC Genes in Neisseria meningitidis

Bacterial strains, plasmids, and culture conditions. N. meningitidis strains used in the present study are listed in Table 1. N. meningitidis MC58 was the wild-type (wt) strain used in the present study. The MC-Fko strain has the fur gene replaced with a kanamycin resistance cassette. A fur-complemented strain MC-Fko-C has a copy of the fur gene inserted heterologously into the chromosome between the NMB1074 and NMB1075 genes. The construction and characterization of both of these strains was described previously (24). The JM01 strain was constructed by replacing the nrrF locus with an erythromycin resistance cassette via homologous recombination with the pGemNKo:Erm plasmid derived from pSLFur-C1 (6). Positive transformants were confirmed by PCR and reverse transcription-PCR (RT-PCR) (data not shown). Both the pGemFkoB:Km and the pSLFur-C1 plasmids were generous gifts from Vincenzo Scarlato's laboratory at Novartis Vaccines (Sienna, Italy).

Growth conditions were identical for all experiments. Bacteria were initially grown on GCB agar plates overnight in a 5% CO₂ environment at 37°C and then used to inoculate cultures in a chemically defined minimal medium (16) supplemented with 0.042% NaHCO₃. Cultures were incubated at 37°C with shaking and grown to an optical density at 600 nm of 0.7 to 1.2, split, and inoculated with 100 μM ferric nitrate (FeNO₃) or a 100 μM concentration of an iron chelator, desferal (deferoxamine mesylate USP; Sigma), to create iron-replete or iron-deplete conditions, respectively. Cultures were grown for an additional 30 min and then pelleted by centrifugation at 3,000 rpm for 10 min. Cultures were supplemented with erythromycin (5 μg/ml) or kanamycin (150 μg/ml) when appropriate.

RNA purification.Total RNA was isolated using an RNeasy kit (QIAGEN, Inc.) and treated with DNase I to remove contaminating DNA. RNA was quantified by using a ND-1000 Nanodrop spectrophotometer (Nanodrop Technologies).

Bioinformatic prediction of Fur-regulated sRNA coding regions.The N. meningitidis MC58 genome (27) was divided into open reading frames (ORFs) and intergenic (IG) regions. ORFs were excluded from further analysis. IG regions were parsed for a 24-bp consensus neisserial Fur binding site (TAAAATAAWAATAATTATCATTAT) (16, 24) using the pattern search algorithm Fuzznuc, available as part of the EMBOSS software suite. Nine mismatches from the consensus were allowed. Separate analysis using the Transterm program (10) from The Institute for Genomic Research parsed the IG regions for predicted rho-independent terminator sequences. The two databases were cross-referenced to identify regions where a Fur-binding site occurred within 0 to 250 bp of the beginning of a predicted rho-independent terminator sequence. These regions were considered putative Fur-regulated sRNA coding regions.

Northern blot analysis and RT-PCR.For Northern blot analysis, 3 to 10 μg of total RNA was mixed with glyoxal loading dye (Ambion), loaded onto 1.5% agarose gels, and separated by gel electrophoresis. RNA was transferred to Hybond Nytran+ membranes (Amersham Biosciences) by downward passive transfer with 20× SSC buffer (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate) and then probed with a 50-bp DNA oligonucleotide complementary to the NrrF RNA (Table 1). Blots were prehybridized in ULTRAhyb (Ambion) for 30 min, and then γ-³²P-end-labeled probes were added to a final concentration of 200 ng/ml. Blots were hybridized overnight at 42°C, subsequently washed according to the NorthernMax-Gly protocol (Ambion), and used to expose 100NIF X-ray film (Fuji Medical). All Northern blotting experiments were repeated four times.

For RT-PCR experiments, Superscript III One-Step reagents were used (Invitrogen). The primers used are listed in Table S1 in the supplemental material. A total of 125 ng of total RNA was used per reaction, and primers were added to a final concentration of 2 ng/μl. Reactions were run as follows: 1 cycle of 2 min at 95°C, followed by 25 to 28 cycles of 30 s at 95°C, 30 s at 50°C, and 30 s at 72°C, followed by a final cycle of 72°C for 5 min. Samples were electrophoresed on a 1% agarose gel for visualization. RT-PCR of the rmp gene transcript was used as a loading control. All RT-PCR experiments were repeated a minimum of three times.

Quantitative RT-PCR.One-Step Quantitect SYBR green RT-PCR reagents were used for all experiments (QIAGEN, Inc.). We used 50 ng of total RNA per reaction, and primers were added to a final concentration of 2 ng/μl. The total reaction volumes were 25 μl. The primers used are listed in Table S1 in the supplemental material. Experiments were carried out with an Applied Biosystems 7000 cycler. Statistical analysis was carried out by using GraphPad InStat software. Ratios were calculated as the level of expression of a gene under iron-deplete conditions divided by the level of expression under iron-replete conditions. The data presented are averages of three independent experiments, and error bars represent standard deviations.

Electrophoretic gel mobility shift assays (EMSAs).Probes were made by PCR from genomic DNA using the primers listed in Table S1 in the supplemental material. All probes contained BamHI restriction sites at the 5′ end. The PCR products were purified, restriction digested, backfilled using Klenow large fragment with [α-³²P]dATP, and run through Sephadex G-25 columns (Amersham Biosciences) to remove unincorporated nucleotides. Portions (4 pmol) of radiolabeled probe were incubated with or without various amounts of recombinant meningococcal Fur for 30 min at room temperature and then electrophoresed on a nondenaturing 6% polyacrylamide gel. Gels were dried and used to expose 100NIF X-ray film (Fuji Medical).

For competition experiments, Fur was incubated for 15 min with a 25-fold excess unlabeled probe or with a 150-fold excess of inhibitor oligonucleotide (Oligo-I) or mutated inhibitor oligonucleotide (Oligo-ΔI) (Table 1). Oligo-I consisted of a double-stranded,100-bp oligonucleotide that has the same sequence as a region of the nrrF promoter encompassing the predicted Fur binding site. Oligo-ΔI has the identical sequence except that the 24-bp region of the predicted Fur-binding site was changed as follows: T→C, A→G, G→A, and C→T.

Computational prediction of NrrF regulatory targets.The program TargetRNA was used to identify candidate message targets of NrrF regulation (28). The following default parameter settings for TargetRNA were used when we searched for potential targets of NrrF action: the individual base-pair model of hybridization scoring was used, the putative terminator stem-loop of NrrF was removed, the search for base-pair binding in messages was restricted to a region from 30 nucleotides upstream of translation initiation to 20 nucleotides downstream of translation initiation, G-U base pairs were not considered when the hybridization seeds of a minimum number of consecutive base pairs were determined, and potential base pair binding interactions were considered significant only if their P value fell below 0.01. Further, messages were considered as potential targets only if their predicted interaction with NrrF contained a hybridization seed of at least eight consecutive base pairs.

Identification of putative Fur-regulated sRNA coding regions in the N. meningitidis MC58 genome.In an effort to identify Fur-regulated sRNAs in the N. meningitidis MC58 genome, we utilized a bioinformatics approach tailored for N. meningitidis as described in Materials and Methods. Briefly, the genome was partitioned into ORFs and IG regions. Only IG regions were examined since the majority of bacterial sRNA molecules discovered to date have been demonstrated to be coded for within IG regions. It was reasoned that a Fur-regulated sRNA would likely have at least two characteristic features; a Fur-binding site in its promoter region and a rho-independent terminator tail at the end of its coding sequence. Rho-independent terminators are a hallmark of bacterial sRNAs and have been used reliably in screens for bacterial sRNA molecules in other organisms (2, 3, 18, 30). Regions where our analysis indicated a predicted Fur-binding site within 0 to 250 bp of the beginning of a predicted rho-independent terminator sequence were considered candidate Fur-regulated sRNA coding regions. Nineteen regions meeting these criteria were identified (Table 2). Three additional regions meeting the above criteria were excluded from the data set since they code for the tRNA-Pro-2, tRNA-Asn-2, and tRNA-ser-4 molecules (27). Increasing the stringency of the Fur-binding site parameter to a requirement of 18/24-bp identity with the consensus reduced the output by fifteen to only four candidate sRNAs, and a requirement of 19/24-bp identity reduced the number of candidate sRNA coding regions from four to only one. The candidate sRNA coding regions were ranked from mc01 to mc19 (Table 2) according to the degree to which the predicted Fur-binding site of the coding region matched the consensus Fur-binding site. It was reasoned that this ordering would reflect the likelihood that these regions corresponded to true Fur-regulated coding regions.

Transcription of N. meningitidis candidate sRNAs.The transcription levels of all 19 candidate sRNAs were evaluated first by S1 protection analysis. N. meningitidis MC58 wt bacteria were grown under iron-replete and iron-depleted conditions, as well as the fur mutant (MC-Fko) strain grown under iron-replete conditions, and total RNA was isolated. It was expected that a Fur-regulated sRNA would be repressed under iron-replete conditions and transcribed under iron-depleted conditions in the wt strain. Furthermore, it was expected that in the fur mutant strain, a Fur-repressed sRNA would be transcribed even under iron-replete conditions. We performed a multiprobe screening of our 19 sRNA candidates using up to five probes in the same S1 protection reaction. Surprisingly, a number of the probe sets examined in this manner detected what appeared to be novel transcripts, although at this stage, it was impossible to tell which corresponding transcript of the probe set had been detected (data not shown). Thus, the expression of individual transcripts was further evaluated by RT-PCR. This analysis reconfirmed the transcription of the mc01, mc03, mc05, mc08, and mc15 candidate small RNAs (Fig. 1A). The transcription of the, mc05, mc08, and mc15 transcripts appeared to be constant under both iron-replete and iron-depleted growth conditions, and thus these candidate sRNAs were not pursued further in these studies. Interestingly, transcription of the mc01 candidate sRNA was shown to depend on growth in the presence of differing iron conditions and exhibited the pattern of expression expected of a Fur-regulated transcript (Fig. 1A). Initial RT-PCR analysis, however, did not allow us to determine whether or not transcription of the mc03 transcript varied subtly with regard to differing iron conditions, thus, transcription of these transcripts was more closely evaluated by quantitative RT-PCR. This analysis demonstrated that the mc01 transcript was transcribed under iron-deplete conditions at over a 100-fold times the level under iron-replete conditions (Fig. 1B), indicating it was iron-regulated. The mc03 transcript, however did not appear to vary significantly in response to iron as the ratio of expression under iron-deplete conditions compared to the expression level under iron-replete conditions was 0.66 with a standard deviation of 0.21. Since this value was close to 1, which would indicate no difference in expression levels between conditions, we concluded that the variation observed in the RT-PCR experiment was likely due to experimental variation, and this transcript was not pursued further in the present study. The genomic positions and adjoining open reading frames of each newly identified transcript are indicated (Fig. 1C).

• Open in new tab
• Download powerpoint

FIG. 1.

Expression of sRNA candidates. (A) RT-PCR analysis of transcript levels of select sRNA candidates from N. meningitidis MC58 wt bacteria grown under iron-replete (+) and iron-depleted (−) conditions. The rmp gene transcript was used as a loading control. RNA was isolated 30 min after the cultures were inoculated with ferric nitrate or desferal. (B) Real-time quantitative PCR analysis of mc01 and mc03 expression levels in a wt strain grown under either iron replete (+) or iron-depleted (−) conditions. The expression of each transcript was normalized to the rmp gene. Ratios were calculated as the expression level of each transcript under iron-depleted conditions divided by the expression level under iron-replete conditions. The data presented represent the average of three independent experiments, and error bars represent the standard deviations. (C) Schematic showing the genomic localization of each candidate sRNA with the adjacent ORFs and their orientation indicated by arrows. The position of the ORF-IG border is denoted. The predicted sRNA coding region is indicated by the black boxes, with the in silico predicted length in white numbering.

Transcription of mc01 was more closely examined by Northern blot analysis and RT-PCR with RNA isolated from N. meningitidis MC58 wt, fur mutant (MC-Fko), and complement fur mutant (MC-Fko-C) strains at 30 min, 60 min (data not shown), and 180 min after cultures were inoculated with either ferric nitrate (iron replete) or the iron chelator, desferal (iron depleted). As expected, mc01 expression was repressed under iron-replete conditions but expressed under iron-depleted conditions in the wt strain, confirming that transcription is iron regulated (Fig. 2, lanes 1 and 2, respectively). In the fur mutant strain (MC-Fko), however, mc01 was transcribed under both iron-replete and iron-depleted conditions (Fig. 2, lanes 3 and 4), demonstrating that mc01 is an iron-responsive, Fur-regulated transcript. Lastly, in the complemented fur mutant strain (MC-Fko-C), the iron-replete repression of the mc01 transcript resembled that seen in the wt strain (Fig. 2A, lane 5 versus lane 1), further confirming the regulation of mc01 by the Fur protein. Based on the regulation of mc01 by both iron and Fur it was subsequently renamed NrrF (for neisserial regulatory sRNA responsive to iron [Fe]).

• Open in new tab
• Download powerpoint

FIG. 2.

Expression of nrrF in response to iron and Fur. (A) Northern blot analysis of RNA from N. meningitidis MC58 (wt), fur mutant (MC-Fko), and complemented fur mutant (MC-Fko-C) strains grown under iron-replete (+) and iron-depleted (−) conditions. RNA was isolated 30 min after cultures were inoculated with either ferric nitrate or desferal. Blots were probed with a 50 nucleotide γ-³²P-end-labeled DNA oligonucleotide complementary to the nrrF transcript. Ethidium bromide staining of rRNA was used as a loading control and is shown below the Northern blot in panel A. Size determinations were estimated from ethidium bromide staining of a 0.1- to 1-kb RNA ladder run in an adjacent lane (not shown) and are indicated by arrows. (B) RT-PCR analysis of RNA from cultures as described above at 30 min postinoculation. The rmp gene transcript was examined as a loading control. (C) RT-PCR analysis of RNA from cultures as described above at 180 min postinoculation.

Characterization of Fur binding to the nrrF promoter.The predicted Fur-binding site of the nrrF coding region identified in the initial bioinformatics screen had the highest degree of identity to the consensus Fur-binding site (19 of 24 bp) of any of the 19 candidate regions. Thus, the finding that nrrF was Fur regulated served to validate the initial bioinformatics screen. To better define the specificity of Fur binding to the predicted Fur box sequence of nrrF, we performed EMSAs using different concentrations of recombinant meningococcal Fur protein (Fig. 3). As expected, Fur was shown not to bind a negative control probe, rmp (lanes 1 and 2), and to bind a positive control probe, fur, which contains a Fur binding site (lanes 3 and 4), demonstrating its functionality. Using a PCR-amplified probe containing a 206-bp region of the nrrF promoter encompassing the Fur box sequence, preincubation with 30 nM Fur resulted in a band shift (single arrow, lane 7). The concentration of Fur required to produce a shift of the nrrF probe was similar to that required to produce shifts with other promoter-probes to which Fur binds (5, 24). At increasing concentrations of Fur, a supershifted band appeared (double arrows, lane 8), suggesting both a high-affinity and a lower-affinity interaction between Fur and the nrrF probe.

• Open in new tab
• Download powerpoint

FIG. 3.

EMSA analysis of meningococcal Fur binding to rmp, fur, and nrrF promoter probes. The probes used are indicated below the lanes. rmp is a negative control probe that does not contain a Fur box. fur is a positive control probe containing a Fur box. nrrF is a probe containing a 206-bp region of the nrrF operator. The probes were incubated with N. meningitidis Fur as follows (lanes 1 to 11, respectively): no Fur (−), 150 nM (+), no Fur (−), 150 nM (+), no Fur (−), 5 nM, 30 nM, 75 nM, 150 nM, 600 nM, and 1.5 μM. In lane 12, 150 nM Fur was preincubated with a 25-fold excess (+25X) of unlabeled nrrF probe for 15 min. In lane 13, 150 nM Fur was preincubated for 15 min with a 150-fold excess of a double-stranded DNA oligonucleotide (+I) with the same sequence as a 100-bp region of the nrrF operator encompassing the predicted Fur-binding site. In lane 14, 150 nM Fur was preincubated for 15 min with a 150-fold excess of an oligonucleotide (+ΔI) identical to the previous oligonucleotide except where the 24-bp predicted Fur-binding site has been changed (T→C, A→G, G→A, and C→T). A band corresponding to a high-affinity complex is indicated by a single arrow. A band corresponding to a lower-affinity complex is indicated by double arrows.

The interaction of the Fur protein with the nrrF probe was inhibited by preincubation with a 25 M excess of unlabeled probe (lane 12), demonstrating the specificity of Fur for the nrrF probe. The interaction of Fur with the predicted 24-bp Fur-binding site was also investigated. A 100-bp, double-stranded oligonucleotide (Oligo-I) was designed and synthesized such that it had the same sequence as a 100-bp region of the nrrF promoter encompassing the Fur-binding site. When a 150-fold excess of this oligonucleotide was preincubated for 15 min with the Fur protein before addition of the nrrF probe it prevented the formation of a supershifted band at a concentration of 150 nM Fur (Fig. 3, lane 13). An otherwise identical oligonucleotide (Oligo-ΔI) in which the 24-bp region comprising the predicted Fur-binding site was changed (T→C, A→G, G→A, and C→T) was no longer able to prevent the formation of a supershifted band at 150 nM Fur. Thus, we conclude that the interaction of the Fur protein with the nrrF promoter probe involves the 24-bp region predicted to be the Fur-binding site of the nrrF promoter.

Bioinformatic analysis of NrrF regulatory targets.Since many small RNAs in other bacteria act as posttranscriptional regulators via base pair binding to message targets (14, 19, 29), we investigated the possibility of such a regulatory action for NrrF. Using a recently available computational tool for predicting message targets of small RNA regulation (28), a number of candidate regulatory targets of NrrF were identified. The region around the ribosome-binding site of each message in the N. meningitidis genome was evaluated for its likelihood of interacting with NrrF. Eleven messages were identified as having significant (P < 0.01) base pair binding potential to NrrF. We chose to evaluate the regulation of 2 of the 11 predicted targets, sdhA and sdhC, because in E. coli the putative orthologues of the sdhA and sdhC genes are posttranscriptionally regulated by the Fur-regulated small RNA, RyhB (20). The other nine putative targets are not known to be sRNA regulated in other organisms and were not pursued. The predicted base-pairing interaction of NrrF with the sdhA and sdhC mRNAs is illustrated in Fig. 4A.

• Open in new tab
• Download powerpoint

FIG. 4.

NrrF-dependent and NrrF-independent transcriptional regulation. (A) Proposed alignment from TargetRNA analysis of potential interactions between the NrrF sRNA and the sdhA and sdhC mRNAs. The translational start codon is underlined and shown in boldface. (B and C) RT-PCR analysis of total RNA isolated from a wt and a nrrF mutant (JM01) grown under iron-replete (+) and iron-depleted (−) conditions. Transcript levels of the sdhA, sdhC, and nrrF transcripts were determined as indicated. The rmp gene transcript was used as a loading control. RNA was isolated 30 min (B) or 180 min (C) after cultures were inoculated with either ferric nitrate or desferal. (D and E) RT-PCR analysis of total RNA isolated from wt or nrrF mutant (JM01) strains grown under iron-replete (+) or iron-depleted (−) conditions. Transcript levels of the bfrA, sodB, aniA, and nrrF genes were determined at 30 min (D) or 180 min (E) after cultures were inoculated with either ferric nitrate or desferal. The rmp gene transcript was a loading control.

NrrF-dependent regulation of genes in N. meningitidis.Having identified the sdhA and sdhC genes as potential regulatory targets of NrrF from our in silico analysis, we used RT-PCR to examine the role of NrrF on sdhA and sdhC regulation. For this analysis an nrrF deletion mutant was constructed by homologous recombination with pGemNeko:Erm derived from pGemFkoB:Km (Table 1). The wt strain and the nrrF mutant (JM01) were grown in parallel under iron-replete and iron-depleted conditions. The nrrF mutant grew as well as the wt strain under both conditions as measured by the optical density (data not shown).

Primers were designed to amplify a 288-bp region of the sdhA gene and a 151-bp region of the sdhC gene (see Table S1 in the supplemental material), and RT-PCRs were carried out using total RNA isolated 30 or 180 min after the inoculation of cultures with either ferric nitrate or desferal. In the wt strain both sdhA and sdhC showed iron-depleted repression at the 30- and 180-min time points (Fig. 4B and C, lanes 2) corresponding with an increase in nrrF transcription. In the nrrF mutant strain (JM01), however, sdhA and sdhC were not repressed under iron-depleted conditions at either the 30- or 180-min time points (Fig. 4B and C, lanes 4 versus lanes 3) and, as expected, the NrrF transcript was not expressed under either condition in the nrrF mutant. The rmp gene was used as a loading control. In addition, preliminary quantitative real-time PCR experiments confirmed an approximately five- to sixfold increase in transcript level of the sdhA and sdh genes in the wt strain grown under iron-replete conditions but no change in transcript levels in the nrrF mutant (data not shown). These results confirmed that the NrrF sRNA is required for the iron-depleted repression of the sdhA and sdhC transcripts at as early as 30 min after the inoculation of cultures with desferal and that this pattern of repression carries out to at least the 180-min time point. Transcription of the sdhA and sdhC transcripts was also evaluated at the 60-min time point and showed an identical pattern of expression (data not shown). Judging from the genomic organization of the sdh genes, it is likely that they are transcribed as a continuous, polycistronic sdhDCAB mRNA; however, this has not been experimentally determined. The observation that both the sdhC and sdhA genes were repressed under iron-depleted conditions in the wt strain suggests that if the sdh genes are transcribed as polycistronic mRNA in N. meningitidis, then regulation by the NrrF sRNA does not lead to the accumulation of a truncated transcript containing the sdhC gene.

We also examined the transcript levels of the sodB, bfrA, and aniA genes, which have previously been demonstrated to be iron activated and Fur regulated in N. meningitidis (5). Meningococcal Fur has been demonstrated by EMSA and footprinting to bind to the promoter regions of the sodB and aniA genes (7, 15), suggesting that they are directly activated by Fur. The sodB and bfrA genes, however, have both been shown to be indirectly Fur activated via an sRNA mechanism in other organisms (4, 20, 30), suggesting that they could also be regulated in a similar manner in N. meningitidis. To determine whether NrrF played a role in the regulation of the N. meningitidis sodB, bfrA, and aniA genes, we thus examined the presence of these transcripts in the nrrF mutant strain (JM01) at 30, 60, and 180 min after the inoculation of cultures with either ferric nitrate or desferal. Our analysis showed that in the case of all three genes, iron-responsive regulation did not require NrrF, since these transcripts continued to exhibit iron-depleted repression in the nrrF mutant at the 30-min, 60-min (data not shown), and 180-min time points (Fig. 4D and E). We also carried out a preliminary analysis of the transcript levels of these genes by quantitative real-time PCR experiments, and these experiments further confirmed that the transcript levels of bfrA, sodB, and aniA are not influenced by the mutation in nrrF (data not shown). We therefore conclude that N. meningitidis must possess at least one other mechanism that is distinct from NrrF-mediated indirect regulation for activating gene transcription under iron-replete growth conditions in a Fur-dependent manner.

In silico analysis of NrrF.Similar to N. meningitidis, N. gonorrhoeae has also been demonstrated to have a complex genetic response to changing iron levels, including the iron-activated transcription of a number of genes (9, 23). Furthermore, our laboratory has recently generated N. gonorrhoeae fur mutant and complemented fur mutant strains and demonstrated that a number of genes are activated in a Fur-dependent manner in N. gonorrhoeae (unpublished data). We therefore sought to determine whether nrrF is conserved in N. gonorrhoeae. The N. gonorrhoeae FA 1090 genome was examined for a nrrF orthologue by using BLAST analysis (1). As shown in Fig. 5A, nrrF is conserved with more than 93% identity in the FA 1090 genome, where it is located in the IG region between the NGO2002 and NGO2003 ORFs. This conservation extends downstream to five bases before the end of the predicted terminator tail and upstream beyond the predicted transcriptional start site of nrrF to encompass the Fur-binding site (lined region, Fig. 5A), as well as the −35 and −10 promoter elements (boxed regions, Fig. 5A). Transcription of the N. gonorrhoeae nrrF homologue was examined in a N. gonorrhoeae F62 wt strain, as well as in an isogenic fur mutant and complemented fur mutant by RT-PCR and was shown to be identical to that of N. meningitidis (data not shown). No evidence of sequence conservation in other organisms beyond Neisseria was found when examined by BLAST (data not shown). In addition, the conservation of NrrF's secondary structure was investigated by using the program QRNA (22), but no orthologue candidates could be detected (data not shown). Lastly, the NrrF sequence was examined by using the MFold RNA structural prediction software (32). This analysis generated a structure with a structured loop and a terminator hairpin tail connected by an AU-rich stretch of nucleotides (Fig. 5B). This structure closely resembles the predicted structure or the RyhB sRNA of E. coli and regions of the NrrF sRNA predicted to interact with the sdhA and sdhC mRNAs are highlighted by boldface, capital letters.

• Open in new tab
• Download powerpoint

FIG. 5.

In silico analysis of nrrF. (A) BLAST alignment of the IG region containing the nrrF locus shows a 189/201-bp region with 94% identity to a region of the N. gonorrhoeae FA 1090 genome. This includes all but the last five residues of the predicted terminator tail, as well as the upstream promoter elements containing the proposed Fur-binding site (indicated by the line) and the −35 and −10 elements (boxed). (B) Predicted secondary structure of NrrF from MFold. Bases predicted to base pair with the sdhA and sdhC mRNAs are indicated in boldface uppercase letters.