Ferric Uptake Regulator and Its Role in the Pathogenesis of Nontypeable Haemophilus influenzae

Bacterial strains and culture media used.Strains and plasmids used are listed in Table 1. NTHi strain 86-028NP, recovered from the nasopharynx of a child with chronic OM, has been well characterized both in vitro (22, 23) and in chinchilla models of OM (24, 25). The genome sequence has been published (20).

For routine culturing, strains were grown on chocolate II agar plates (Fisher Scientific, Pittsburgh, PA). The 86-028NPΔfur::Tn903(pTS-fur) strain in which fur could be overexpressed was cultured on chocolate agar plates containing 200 μg of spectinomycin/ml. For routine liquid culture, all cells were grown in BHI supplemented with 2 μg of NAD/ml and 2 μg of heme/ml (sBHI). Alternatively, cells were grown in an iron-depleted defined iron source (DIS) medium, a modification of a defined medium developed by Hasan et al. and Coleman et al. (26, 27) and previously used for culture of strain 86-028NP (28). As needed, DIS medium was supplemented with human hemoglobin at a concentration of 10 μg/ml of medium. Cells were iron depleted by growing at 37°C for 16 h in DIS medium. Iron-depleted cells were then transferred to DIS medium supplemented with 10 μg of human hemoglobin/ml and grown at 37°C with shaking at 50 rpm.

To analyze transcriptional responses to changes in iron concentration, bacteria were grown to mid-log phase in DIS medium supplemented with 10 μg of human hemoglobin/ml, as previously outlined. Cultures were then split into two and 2,2′-bipyridine added, at a final concentration of 500 μM, to one aliquot (29). After 15 min at 37°C with shaking at 50 rpm, the cells were removed from both aliquots for RNA purification. Cells were also removed from both aliquots for plating to determine that chelation was not affecting cell viability.

Construction of a strain 86-028NPΔfur::Tn903 mutant.The chloramphenicol resistance gene in pKD3 (30) was excised with SwaI and PvuII and replaced with the HincII fragment from Tn903, which contains a kanamycin resistance gene, to generate pKD3kan. Bipartite PCR primers were designed. Primer 1 contained 19 bp of homology to the P1 binding site adjacent to the FLP recognition target (FRT) site 3′ of the kanamycin resistance gene and a second region of 51 bp, homologous to the intergenic region upstream of fur, as well as the 3′ end of the upstream gene, fldA. Primer 2 had 17 bp of homology to the P2 binding site adjacent to the FRT site 5′ of the kanamycin resistance gene, and a second region of 52 bp, homologous to the 3′ 24 bp of fur, as well as the adjacent 3′ intergenic region downstream of fur. Using these two primers, the kanamycin resistance gene from pKD3kan was amplified. The resulting PCR product and a cosmid-size clone of strain 86-028NP DNA that contained the fur gene were simultaneously transformed into the recombineering Escherichia coli strain DY380, previously incubated at 42°C to inactivate the temperature-sensitive repressor cI857. This allowed expression of the recombination genes exo, bet, and gam to facilitate recombination of the kanamycin gene-containing PCR product into the cosmid-sized clone. The fur gene was thus deleted. The construct was then purified from strain DY380, and the region surrounding the deleted fur gene was amplified by PCR. The amplicon was transformed into strain 86-028NP using the MIV method (31). fur mutants were selected on a gradient of heme derived by pouring a slope of GC agar supplemented with 2 μg of heme/ml, 2 μg of β-NAD/ml, and 20 μg of kanamycin/ml, overlaid with a flat surface of GC agar supplemented with 2 μg of β-NAD/ml and 20 μg of kanamycin/ml. The identity of the fur mutant was confirmed by PCR and sequencing.

Construction of marked strains.To distinguish strains in a competition model of experimental OM, 86-028NP and 86-028NPΔfur::Tn903 were marked. 86-028NP was transformed with pGZRS-39A, a Haemophilus-Actinobacillus pleuropneumoniae shuttle vector that contains the kanamycin resistance gene from Tn903 (32). 86-028NPΔfur::Tn903 was transformed with pSPEC1, a variant of pGZRS-39A, in which the kanamycin resistance gene was replaced by a spectinomycin resistance gene (28). The number of parental cells and the number of fur mutant cells could thus be enumerated in a mix of strains by plating an aliquot of cells on chocolate agar plates which contained 200 μg of spectinomycin/ml and a second aliquot on chocolate plates which contained 20 μg of kanamycin/ml. The numbers of kanamycin-resistant cells and the numbers of spectinomycin-resistant cells were approximately equal to the total number of viable cells within each effusion. Plasmids could also be purified from cells isolated from effusions (data not shown). The plasmids were thus maintained in vivo. These observations confirmed data that showed pGZRS-39A is stably maintained by strain 86-028NP during experimentally induced OM in chinchillas (25, 33).

Overexpression of fur in the 86-028NPΔfur::Tn903 mutant.fur was overexpressed from pTS, a tetracycline-inducible E. coli-Haemophilus shuttle vector, which was constructed as follows. A 788-bp region containing the tet repressor/promoter of pRSM2947 (34) was amplified with primers that contained an engineered SpeI restriction site at the 5′ end of the tet repressor/promoter. A 303-bp region of the multiple cloning site (MCS) from the pET-30b expression vector (EMD Chemicals, San Diego, CA) was excised with XbaI and BlpI. The tet repressor/promoter fragment and the MCS-containing fragment were then ligated via the XbaI/SpeI compatible ends. The ends of the resulting construct were blunted and phosphorylated with an End-It DNA End-Repair kit (Epicentre, Madison, WI), cloned into pSMART-LCKan (Lucigen, Middleton, WI), and transformed into the E. cloni strain of E. coli (Lucigen). To maintain a unique NdeI site in the MCS, site-directed mutagenesis was used to mutate the NdeI site at the 3′ end of tetR. This resulted in a single point mutation, which changed the fourth A of the NdeI site to T and did not alter the encoded amino acid sequence. The 1.2-kb fragment containing the spectinomycin gene was excised from pSpecR by BamHI restriction digest and cloned into the BamHI site of the MCS. The region of DNA, which included the tet promoter/repressor, MCS, and spectinomycin resistance gene flanked by terminators from within the pSMART-LCKan vector backbone, was PCR amplified. The PCR amplicon was then ligated to a fragment of pLS88 (35, 36), a broad-range vector that replicates in Haemophilus, as follows. The origin of replication and kanamycin resistance gene of pLS88, was PCR amplified with primers designed with 5′ BamHI restriction sites. The pLS88 amplicon was digested with BamHI (Fermentas, Glen Burnie, MD), and the ends were blunted and phosphorylated with the End-It DNA End-Repair kit and ligated with the blunt-ended PCR product containing the tet promoter and spectinomycin resistance gene from the pSMART-based construct. The resultant construct was transformed into H. influenzae Rd KW20. Plasmids were prepared from transformants that exhibited both kanamycin and spectinomycin resistance and characterized by restriction analysis. The spectinomycin resistance gene was removed from the construct by BamHI digestion and the construct was then self-ligated. Finally, site-directed mutagenesis was used to remove a NotI site in the pLS88 backbone. The resultant plasmid was named pT. To generate pT-fur, the coding sequence of fur was PCR amplified with primers that generated an NdeI site 5′ of fur and a BamHI site 3′ of fur. fur was then cloned between NdeI and BamHI in pT. In the course of these studies, it was then necessary to reintroduce the spectinomycin resistance gene back into pT. PCR primers containing BglII restriction sites were used to amplify pT-fur without the Tn903 cassette. The spectinomycin resistance gene from pSPECR (37) was then excised using BglII and ligated to the pTΔTn903-fur fragment. The ligation product was transformed into electrocompetent DH10B (Invitrogen), and transformants selected on LB agar containing 50 μg of spectinomycin/ml. The resulting pTS-fur construct was then transformed into the 86-028NPΔfur::Tn903 mutant by electroporation, and transformants were selected on chocolate agar containing 200 μg of spectinomycin/ml.

RNA isolation.Strains were grown statically for 16 h at 37°C in DIS medium. Cells were then diluted into DIS medium supplemented with 10 μg of human hemoglobin/ml and grown for four generations at 37°C in air with shaking at 50 rpm. Ten-milliliter aliquots of cells were removed and pelleted at 4°C for 15 min at 3,220 × g and RNA purified using TRIzol reagent (Life Technologies, Grand Island, NY) as outlined in Mason et al. (25). The RNA concentration was determined using a NanoDrop spectrophotometer (NanoDrop Products, Wilmington, DE), and each sample was carefully analyzed for integrity using an Agilent 2100 bioanalyzer (Agilent Technologies, Santa Clara, CA) to confirm the absence of degraded RNA. Only samples that passed this initial quality control were used in subsequent analyses.

Array manufacturing.Expression analyses were performed using custom arrays printed commercially by Agilent Technologies. Agilent's SurePrint technology uses phosphoramadite chemistry in combination with high-performance Hewlett-Packard inkjet technology for in situ synthesis of 60-mer oligonucleotides. Using Agilent eArray, a strain 86-028NP custom microarray was designed that contained probes for every known open reading frame (ORF). Each ORF was represented by three probes to ensure complete coverage of each transcript. The SurePrint technology prints eight arrays, with 15,000 elements per array, on a single slide, ensuring low technical variability across samples in a given experiment.

Hybridization procedures, data processing, and analysis.Array hybridizations were performed by the Biomedical Genomics Core facility, The Research Institute at Nationwide Children's Hospital (http://genomics.nchresearch.org/). Four sets of biological replicates were generated for microarray analyses. Agilent's One-Color microarray-based gene expression analysis labeling protocol (Agilent Technologies) was used to label samples with Cy3 using a random primed amino-allyl reverse transcription approach to generate labeled cDNA. After purification, the labeled test (fur mutant) and control (parent) samples were denatured and hybridized to the array overnight. Microarray slides were then washed and scanned with an Agilent G2505C microarray scanner at a 2-μm resolution. Images were analyzed with Feature Extraction 10.9 (Agilent Technologies). Median foreground intensities were obtained for each spot and imported into the mathematical software package “R”. “R” was used for all data input, diagnostic plots, normalization, and quality checking steps of the analysis process using scripts developed by the genomics core. Briefly, the data set was filtered to remove positive control elements and elements which were flagged as bad. Using the negative controls on the arrays, the background threshold was determined, and all values less than this value were flagged; local background correction was not performed since this has been shown to only introduce noise. The data were then normalized by the LOESS method using the LIMMA (linear models for microarray data) package in “R” as described previously (38). Finally, the median value for the replicate probes was used as the expression measurement for that given ORF. Complete statistical analysis was then performed in “R” using both the LIMMA and the Siggenes Bioconductor packages. The SAM (significance analysis of microarrays) statistical analysis package was used to identify genes differentially expressed due to the loss of Fur (http://www-stat.stanford.edu/∼tibs/SAM/ [39]). With SAM, the false discovery rate (FDR) and q-value method were used to measure the expected proportion of false positives in the set of all differential expression calls (40). A two-class unpaired analysis with a FDR of 10% was used to maximize sensitivity without significantly impacting accuracy. The q-values (FDR) for each gene were determined and then combined with a fold change cutoff (i.e., less than or equal to −2-fold downregulation or greater than or equal to 2-fold upregulation) to identify the final list of differentially expressed transcripts. Of the different statistical approaches for analysis of array data, this approach is considered the most robust. All microarray data are MIAME compliant and the transcriptional data has been deposited into the MIAME compliant database, GEO, under accession number GSE39874 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE39874).

qRT-PCR.Quantitative reverse transcription-PCR (qRT-PCR) was used to confirm the relative expression of a subset of genes identified by microarray analyses. This utilized RNA samples obtained from strain 86-026NP, strain 86-028NPΔfur::Tn903, the overexpressing strain 86-028NPΔfur::Tn903(pTS-fur), and strain 86-028NP with or without iron chelation. qRT-PCR was performed with a one-step QuantiTect SYBR green RT-PCR kit (Qiagen, Valencia, CA). Three biological replicates and three technical replicates were performed for each gene analyzed. Fold changes were calculated. All threshold cycle (C_T) values were normalized to the endogenous control gyrA. Relative quantitation was calculated from the median C_T value using ΔΔC_T, and statistical significance was determined using the Student two-tailed t test. A fold change in gene expression of >2-fold, with a P value of <0.05, was assessed as being significant.

Proteomic analyses of the Fur regulon in strain 86-028NP.Strain 86-026NP and strain 86-028NPΔfur::Tn903 were grown statically for 16 h at 37°C in DIS medium supplemented with 5 μg of PPIX/ml. Cells were then diluted into either 250 ml of DIS medium supplemented with 10 μg of human hemoglobin/ml or 250 ml of BHI supplemented with 2 μg of NAD/ml and 2 μg of heme/ml. In each case, cells were grown for four generations at 37°C in air with shaking at 50 rpm. The cells were harvested by centrifugation at 3,220 × g and fractionated into cytoplasmic, sarcosyl-soluble (inner membrane-enriched) and sarcosyl-insoluble (outer membrane-enriched) fractions (41, 42). Fractions were resolved on 7.0% sodium dodecyl sulfate (SDS) polyacrylamide gels, fixed, and stained with Coomassie brilliant blue R250. Protein bands to be analyzed were excised and stored in 5% acetic acid. Proteins were digested in-gel with trypsin and the peptides analyzed by capillary-liquid chromatography-nanospray tandem mass spectrometry using a Thermo Finnigan LTQ mass spectrometer equipped with a nanospray source operated in positive ion mode (Thermo Scientific, Waltham, MA). Protein identifications were made using Mascot Daemon (Matrix Science, Boston, MA) using the guidelines determined by Carr et al. (43).

Fur binding assay.Strains 86-028NPΔfur::Tn903 and 86-028NPΔfur::Tn903(pTS-fur) were grown in sBHI. When the cells were in mid-log phase, 200 ng of anhydrotetracycline/ml of medium was added to induce fur expression. After 1 h of Fur induction, the cells were harvested by centrifugation at 5,700 × g for 15 min and resuspended in Fur-binding buffer (10 mM Tris-HCl [pH 7.5], 1 mM MgCl₂, 0.5 mM dithiothreitol, 50 mM KCl, 100 μM MnCl₂, 5% glycerol). The cells were mechanically disrupted by using Lysing B Matrix (MP Biomedicals, Solon, OH), the lysing beads were removed by pelleting at 3,220 × g for 5 min, and the membrane-containing fraction was removed by centrifugation at 147,000 × g for 60 min. Using biotinylated primers, the promoter regions of Fur-regulated genes that contained predicted Fur boxes were amplified by PCR. To generate the hfeA promoter region containing a scrambled Fur box sequence, the predicted Fur box sequence was scrambled using Shuffle DNA (http://bioinformatics.org/sms/). Virtual Footprint (http://prodoric.tu-bs.de/vfp/) was then used to ensure the scrambled Fur box sequence no longer contained a predicted Fur box. A biotinylated hfeA promoter region containing a scrambled Fur box was then constructed using an adaptation of the gene splicing method developed by Horton et al. (44). Briefly, the promoter regions immediately 5′ and 3′ of the hfeA Fur box were amplified, with the primers proximal to the Fur box having 3′ overhangs with complementary sequences that contained the scrambled Fur box sequence. The resulting PCR products were column purified and then mixed. These products acted as primers for each other in a third PCR and thus amplified the hfeA promoter region containing a scrambled Fur box. The resultant biotinylated PCR product was gel purified. One hundred nanograms of all biotinylated PCR products was gently mixed with 100 μg of the cytoplasmic enriched supernatants in 1 ml of Fur-binding buffer containing 50 μg of sheared salmon sperm DNA/ml. After mixing for 60 min at room temperature, 50 μl of streptavidin-coated agarose beads (Life Technologies) was added to the lysate-PCR product mix, followed by mixing for 30 min at room temperature. The beads were then pelleted by centrifugation at 20,800 × g for 1 min and washed three times with 1 ml of Fur-binding buffer containing 50 μg of sheared salmon sperm DNA/ml and then two times with 1 ml of Fur-binding buffer. Washed beads were boiled in Laemmli sample buffer, and the proteins were resolved on 4.0 to 20.0% SDS-polyacrylamide gradient gels and transferred to a nitrocellulose membrane. The membrane was probed with an anti-Pseudomonas Fur polyclonal antibody (a generous gift from Michael Vasil).

Animal infection model.To determine whether fur expression impacts the course of experimental OM, strain 86-028NPΔfur::Tn903 and the parent strain, 86-028NP, were marked and then monitored in competition for survival in adult chinchillas (Chinchilla lanigera; Rauscher's Chinchilla Ranch, LaRue, OH) with no previous evidence of middle ear disease as determined by video otoscopy and tympanometry. Briefly, NTHi strains were grown on chocolate agar containing 200 μg of spectinomycin/ml or 20 μg of kanamycin/ml and then incubated overnight at 37°C in 5% CO₂. Bacteria were suspended in pyrogen-free sterile saline (Hospira, Inc., Lake Forest, IL) to an optical density at 600 nm of 0.25. Cultures were then serially diluted in pyrogen-free saline, and approximately 300 bacteria were introduced via the transbullar route, bilaterally into the middle ears of the chinchillas. At each time point, chinchillas were anesthetized, and middle ear fluids removed by epitympanic tap of the inferior bullae. Tympanic fluids were serially diluted in sterile saline and plated on chocolate agar, as well as on chocolate agar containing 200 μg of spectinomycin/ml or 20 μg of kanamycin/ml. Animal care and all procedures were performed in concordance with institutional and federal guidelines and were conducted under an approved protocol (AR08-00027).