Bacterial strains and growth conditions
FA 1090 was grown from frozen stock cultures on GCB agar (Becton Dickinson) supplemented with 2% IsoVitaleX (Becton Dickinson) under 5% CO2 atmosphere at 37 °C. Fe-deplete (CDM-0) and Fe-replete medium (CDM-10), supplemented with 10 μm Fe (NO3)3 were prepared as previously reported [16, 52]. Briefly, organisms were inoculated to an OD600 = 0.100 from overnight growth on GCB agar plates. Cultures were grown to OD600 = 0.200 to deplete intracellular Fe pools. Half of the culture was transferred into a separate flask and CDM-0 was then added to both flasks to equal the original volume; Fe (NO3)3 was added to one of the flasks to a final concentration of 10 μM (CDM-10) . The cultures were then allowed to grow to stationary phase in a 37 °C shaking incubator (225 rpm), and an aliquot of cells from Fe deplete and Fe replete conditions were collected at 1,2,3,4, and 5 h for RNA isolation.
Total RNA was isolated using a hot phenol method previously described  and pooled from three separate biological replicates for each time and growth condition as described above. RNA samples were DNase treated [1U RQ1 RNase-Free DNase per 1 μg RNA (Promega)] followed by using the TriReagent protocol for clean-up (Molecular Research Center, Inc.). RNA quantitation was done using the NanoDrop ND-100 (Thermo Scientific) and RNA integrity assessed using the Agilent RNA 6000 Nano Kit on the Agilent 2100 Bioanalyzer. Purified RNA samples were stored at −80 °C until further use.
Transcript size selection
Total RNA was size selected for transcripts <500 bp following a protocol successfully used for sRNA sequencing in Pseudomonas aeruginosa . Briefly, 50 μg DNAse-treated total RNA was separated by electrophoresis on a denaturing 10 M urea/10% polyacrylamide gel (Jule Biotechnologies Inc). The region of the polyacrylamide gel containing RNA of (50–500 nt) was excised, followed by centrifugation in Gel Breaker tubes (IST Engineering). The RNA was eluted overnight in 400 μl 0.4 M NaCL and filtered through a 0.5 μm filter tube (IST Engineering). RNA was extracted from the filtrate by addition of Phenol:Chloroform:Isoamyl alcohol (25:24:1, v/v) (Ambion). The aqueous phase was collected after centrifugation in a Phase-Lock Gel (5 Prime), followed by ethanol precipitation (0.02 volume 5 mg/ml glycogen (Ambion); 0.1 volume 3 M sodium acetate pH 5.5 (Ambion); 2.5 volumes cold 100% ethanol).
Ribosomal RNA removal and TAP treatment
5 μg of size-selected RNA was depleted of 5S, 16S and 23S rRNA using the Ribo-Zero Magnetic Kit Gram Negative (Epicentre), followed by extraction in Phenol:Chloroform:Isoamyl alcohol and ethanol precipitation as described above. A portion of the rRNA depleted RNA samples were then treated with TAP (Epicentre) for 1 h @ 37 °C to convert 5′ phosphate-triphosphate to 5′-monophosphate, and then similarly purified and ethanol precipitated after the TAP reaction.
Library production and sequencing
Directional libraries were prepared at Iowa State University DNA Facility http://www.dna.iastate.edu/nextgensequencing.html using the Illumina TruSeq Small RNA kit. All 10 directional libraries were bar-coded and 5 libraries were run per one lane of the Illumina MiSeq v3, to obtain 150 bp single reads. An average of 1,455,012 reads was generated from each library; this yielded an overall 62% alignment of high quality reads to the FA 1090 genome and an average 24% of those reads aligning to unannotated regions.
Alignment and analysis of RNA-seq data
Reads were mapped to the Neisseria gonorrhoeae FA 1090 genome (NCBI GenBank Accession AE004969.1). Since the original FA 1090 annotation was over 10 years old, we re-annotated the FA 1090 genome sequence using the NCBI annotation pipeline on 6/22/2015. As an aid to the community, we developed an annotation conversion “look back” table to easily locate new locus tag numbers by querying old locus tag numbers using CD-Hit (Additional file 10: Table S5). RockHopper  was used for alignment and analysis of predicted sRNA.
Prediction of DNA regulatory signals
Potential sRNA reads between 40 and 500 bp were retained and nucleotide sequence was extracted 150 bp 5′ and 3′ to each putative sRNA. These regions were screened for the presence of putative promoter motif using Promoter Prediction by PBROM (Softberry)  and a rho independent terminator with ARNold (Erpin and RNAmotif programs). For prediction of FB motifs in FA 1090, we used a training set of 23 sequences previously shown to bind Fur in vitro by EMSA and/or in vivo in a FurTA assay to generate a consensus FB sequence in MEME v 4.9.1. The FB consensus sequence generated in (WebLogo)  was 18 bp representing a 7-1-7 inverted repeat (Additional file 3: Figure S1). In parallel, we built a hidden markov model (HMM) using hmmer v 1.8.5 and position weight matrix (PWM) model (MEME v 4.9.1). Both models were used to scan the FA 1090 genome. Leave-one-out cross validation (LOOCV)  was used to determine an empirical p-value cut-off of 0.008 for the HMM model and p-value cut-off of 0.00103 for the PWM model.
sRNA secondary structure and NrrF target predictions
RNAfold  was used to generate the secondary structure for NrrF and Nrs sRNAs on the ViennaRNA Web Server http://rna.tbi.univie.ac.at/. IntaRNA , a tool for prediction of RNA-RNA interactions was used to determine possible targets of NrrF in the Nrs RNAs. http://rna.informatik.uni-freiburg.de/IntaRNA/Input.jsp.
Rfam homology search to known regulatory ncRNAs
Homology to known regulatory ncRNAs was done by blasting the pool of candidate sRNAs to the Rfam 12.0 release database (http://rfam.xfam.org)  using the default scoring. Hits with an e-value less than 1 were reported.
Neisseriaceae homology search and heat map
Complete and draft genomes of 451 members of the Neisseriaceae family including both commensal and pathogenic organisms used in this study are listed in Additional file 8: Table S4. Blastn (NCBI Nucleotide v 2.2.28) of the 451 genomes was performed using the default p = 0.05, cut-off of ≥85% of sRNA query length and ≥85% sequence identity to Nrs and Nrf Fe regulated sRNAs. The blastn results were tabulated in a table of all genomes present in the blastn report vs. the blastn results for each of the 14 sRNAs. Scores of 0 or 1 were assigned to the following states; 0 for genomes not found for that specific sRNA and 1 for a hit indicating the presence of that sRNA. In-house scripts in R version 3.2.3 were written to reduce the number of draft genome duplicate contigs for presentation in a table and heat map and to bundle all contigs of same strain together into a single genome name. Clustering analysis for this heat map was generated using the default hclust function in the gplots package of R.
To estimate sRNA levels, quantitative real time PCR (qRT_PCR) was performed as previously described . Briefly, cDNA was generated from DNAse-treated total RNA using the High Capacity cDNA Archive Kit using random primers and including a negative control reaction lacking the reverse transcriptase enzyme (RT; Applied Biosystems). RT using gene specific primers (GSP) were set up as per the High Capacity cDNA Archive Kit with a few modifications. The reverse real time primers were used as the GSP for RT of antisense sRNAs. All GSP were at 0.2 pM final concentration with a 20 min annealing temperature at the respective primer Tm, replacing the 10 min room temperature annealing temperature as for RT using random primers. Amplification of all cDNA was performed on an ABI 7500 Fast Real-Time PCR system (Applied Biosystems) using SYBR green master mix (Applied Biosystems). The relative expression was calculated by the comparative threshold cycle (2-ΔΔCT) method, with fold changes calculated as a ratio of the qRT-PCR measurements of sRNA from the Fe-deplete/Fe-replete growth conditions. The relative concentration was reported in pg and quantified using the standard-curve method (user bulletin no. 2; Applied Biosystems). Real-time reactions were carried out in duplicate with porA as the endogenous reference. Primers and annealing temperatures are listed in Additional file 11: Table S6.
Fluorescent Primer Extension (FPE)
TSS was estimated using primer extensions employing fluorescent primers . Reactions were performed by addition of FAM-labeled primer (0.01 μM final concentration) to 20–30 μg of DNase treated total RNA (20 μl total reaction volume), heated for 5 min @ 70 ° C and placed on ice for 10 min. Samples were then incubated for 20 min at the specified primer annealing temperature and cooled to room temperature for 15 min. The reverse transcription reaction was performed as follows: annealed RNA/primer mixture, 400 U RT (SuperScript III Invitrogen), (1 mM dNTP and 0.01 M DTT final concentration) in a total reaction volume of 40 μl. The reaction was incubated for 2 h at 42 ° C followed by sodium acetate/ethanol precipitation. Fragment analysis was done on the ABI 3730 DNA analyzer using GeneScan 600-LIZ ladder at the Laboratory of Molecular Biology and Cytometry Research http://research.ouhsc.edu/CoreFacilities.aspx. A standard curve was generated in Peak Scanner (Applied Biosystems) software to calculate the size and intensity of the FAM labeled cDNA products. Primers and annealing temperatures are listed in Additional file 11: Table S6.
Northern blot analysis
Northern blots were performed by addition of 20ug of DNAse treated 5 h Fe replete RNA with an equal amount of 2× Gel loading buffer II (Life Technologies P/N: AM8547), denatured for 10 min at 95 °C, and loaded onto a denaturing 15% TBE-Urea polyacrylamide gel (Invitrogen). Gels were run for 3 h at 150 V and then transferred onto Hybond-N+ membrane (GE Healthcare) by wet blotting at 12 V for 2 h. RNA was then UV cross-linked twice to the membrane at 120 mJ/cm2 (Fisher Scientific). Probes were generated by labeling the 3′ end of the 5′ FAM-labeled oligonucleotides used for primer extensions (Additional file 11: Table S6) according to the manufacturer’s protocol (Roche cat. number 03353575910). Hybridization was performed overnight at 10 °C below the FAM-labeled oligonucleotide Tm in a ProBlot hybridization oven (Labnet International). Probe detection was performed using the DIG luminescent detection kit (Roche) according to the manufacturer’s protocol. Briefly, membranes were blocked in blocking buffer for 30 min at 25 °C, then incubated in antibody solution (anti-DIG antibody diluted in blocking buffer; 1:10,000) for 30 min at 25 °C, then washed twice in 1× washing buffer (0.1 M maleic acid, 0.15 M NaCl at pH 7.5, 0.3% Tween 20 [v/v]) for 15 min at 25 °C. Membranes were then equilibrated in detection buffer (0.1 M Tris, 0.1 M NaCl, pH9.5); incubated in 0.25 mM CDP-star at room temperature for 5 min and then exposed on the Kodak Gel Logic 1500.
Illumina RNA-seq data accession number
RNA_seq data are deposited at Gene Expression Omnibus (GEO) https://www.ncbi/nih.gov/geo under BioProject PRJNA356970.
Results were expressed as the standard error of the mean (SEM). Paired t test was used for all comparisons. A P value of <0.05 was considered significant.