Microbiology (2009), 155, 1901–1911 DOI 10.1099/mic.0.026062-0 Correspondence Paul W. O’Toole pwotoole@ucc.ie Effect of FliK mutation on the transcriptional activity of the s 54 sigma factor RpoN in Helicobacter pylori Francois P. Douillard,1 Kieran A. Ryan,1 Jason Hinds2 and Paul W. O’Toole1 1Department of Microbiology and Alimentary Pharmabiotic Centre, University College Cork, Western Road, Cork, Ireland 2Bacterial Microarray Group, Division of Cellular and Molecular Medicine, St George’s University of London, Cranmer Terrace, London SW17 0RE, UK Helicobacter pylori is a motile Gram-negative bacterium that colonizes and persists in the human gastric mucosa. The flagellum gene regulatory circuitry of H. pylori is unique in many aspects compared with the Salmonella/Escherichia coli paradigms, and some regulatory checkpoints remain unclear. FliK controls the hook length during flagellar assembly. Microarray analysis of a fliK-null mutant revealed increased transcription of genes under the control of the s 54 sigma factor RpoN. This sigma factor has been shown to be responsible for transcription of the class II flagellar genes, including flgE and flaB. No genes higher in the flagellar hierarchy had altered expression, Received 21 November 2008 suggesting specific and localized FliK-dependent feedback on the RpoN regulon. FliK thus Revised 5 March 2009 appears to be involved in three processes: hook-length control, export substrate specificity and Accepted 9 March 2009 control of RpoN transcriptional activity. INTRODUCTION Helicobacter pylori infection is responsible for gastrointestinal disorders such as peptic and duodenal ulcers (Veldhuyzen van Zanten & Sherman, 1994), and is a predisposing factor for gastric adenocarcinoma (EUROGAST, 1993) and B-cell MALT lymphoma (Parsonnet et al., 1994). Epidemiological studies show large disparities in H. pylori infection rates between geographical regions (Kikuchi & Dore, 2005). Developing countries are the most severely infected by H. pylori, due to poor living conditions and limited access to therapies (Frenck & Clemens, 2003). The continuing emergence of new antibiotic-resistant strains underlines the urgent need to expand our understanding of the biology of H. pylori,in order to facilitate the development of new treatments. Motility is a key feature of H. pylori and is required for colonization and persistence (Eaton et al., 1991, 1992, 1996). In motile bacteria, flagella contribute to motility, Abbreviations: CGH, comparative genome hybridization; qRT-PCR, quantitative real-time PCR. The microarray design is available in BmG@Sbase (accession number ABUGS- 18; http://bugs.sgul.ac.uk/A-BUGS-18) and also ArrayExpress (accession number A-BUGS-18). Fully annotated microarray data have been deposited in BmG@Sbase (accession number E-BUGS-78; http:// bugs.sgul.ac.uk/E-BUGS-78) and also ArrayExpress (accession number E-BUGS-78). A supplementary table, listing oligonucleotide primers used in this study, is available with the online version of this paper. adhesion and inflammatory response by the host cells (Eaton et al., 1992; Galkin et al., 2008). Three RNA 8054 28 polymerase sigma factors, s , s and s , control the transcription of H. pylori flagellar genes (Niehus et al., 2004; Scarlato et al., 2001). s 80 modulates the transcription of the early flagellar genes (class I). Middle flagellar structural genes (class II) are under the control of RpoN (s 54 ). RpoN transcriptional activity is tightly regulated by the FlgR/FlgS activation system and HP0958, an RpoN chaperone (Brahmachary et al., 2004; Pereira & Hoover, 2005; Ryan et al., 2005a). The class II regulon includes HP0906/fliK, which encodes the hook-length-control protein. Transcription of class III genes (late flagellar genes) is controlled by FliA (s 28) and its anti-sigma factor FlgM (Colland et al., 2001; Josenhans et al., 2002). The initial annotations of H. pylori genome sequences did not identify all flagellar genes expected by comparison with the Salmonella/Escherichia coli paradigm (Alm et al., 1999; Tomb et al., 1997). Thus, the anti-sigma 28 factor FlgM was only identified subsequently, by rigorously searching for low-level conservation (Colland et al., 2001; Josenhans et al., 2002). Similarly, the hook-length protein FliK has only been recently identified based on bioinformatic analyses (Ryan et al., 2005b). The HP0906 gene product has been shown to be FliK in H. pylori (Ryan et al., 2005b). FliK in Salmonella controls hook length (Ferris & Minamino, 2006). Ablation of the H. pylori fliK gene impairs motility, and similar to a fliK-null mutant in Salmonella, the cells harbour polyhook structures (Ryan 026062 G 2009 SGM Printed in Great Britain F. P. Douillard and others et al., 2005b). A dramatic reduction in flagellin production and overproduction of the FlgE hook protein are also observed, and flgE and flaB transcription levels are significantly increased in the fliK mutant (Ryan et al., 2005b). These genes have been shown to be under the control of RpoN, the s 54 sigma factor of H. pylori (Beier & Frank, 2000; Brahmachary et al., 2004; Niehus et al., 2004; Scarlato et al., 2001). Three genes belonging to the intermediate gene class do not show altered transcription levels according to targeted quantitative real-time PCR (qRT-PCR) analysis (Ryan et al., 2005b). We thus suggested that the FliK protein is required to turn off the RpoN regulon during flagellar assembly. However, the transcriptional profile of the whole flagellar regulon was not investigated, leaving the extent of FliK feedback on flagellar gene expression unclear. In the present study, we used a pan-H. pylori array, based on the genomes of strains NCTC26695 and J99. Array comparative genomic hybridization (CGH) was performed to identify specific chromosomal regions that were missing in CCUG17874, the motile strain used in our laboratory for flagellum genetics. Global transcript analysis of a CCUG17874 mutant lacking the fliK gene was performed to further investigate the role of FliK in flagellar biogenesis in H. pylori. We integrated these microarray data into the currently known flagellar regulatory model. METHODS Bacterial strains, media and growth conditions. H. pylori type strain CCUG 17874 (Culture Collection, University of Gothenburg, Sweden) was cultured as previously described (Ryan et al., 2005b). H. pylori mutants defective in the fliK gene (Ryan et al., 2005a, b) and flgE gene (O’Toole et al., 1994) have been previously described. The mutants used in this study were cultured on Columbia agar base (CBA) plates containing the appropriate antibiotic: 10 mg chloramphenicol ml21 (Sigma) or 15 mg kanamycin ml21 (Sigma). Extraction of genomic DNA from H. pylori. Genomic DNA from 2 day-old H. pylori plate cultures was isolated using the DNeasy tissue kit (Qiagen). The genomic DNA was then quantified using a Nanodrop ND-1000 spectrophotometer. RNA isolation from H. pylori. Total RNA was isolated from 20 h H. pylori liquid cultures using the Qiagen RNeasy Mini kit. H. pylori cells were harvested and centrifuged for 15 s at ¢10 000 g. Cell pellets were then resuspended in 750 ml Bacteria RNA Protect reagent (Qiagen). The remainder of the protocol was performed as per the manufacturer’s instructions (RNeasy Mini kit; Qiagen). RNA quality was assessed using a Bioanalyser 2100 Instrument (Agilent Technologies) and quantified by NanoDrop ND-1000 spectrophotometer. Prior to further array experiments, total RNA samples were DNase-treated using a Turbo DNA-free kit (Ambion). H. pylori microarray design and construction. Design and construction of the H. pylori microarray (BmG@S HPv1.0.0; Bacterial Microarray Group at St George’s, University of London) was completed using the approaches described by Hinds et al. (2002a, b). In brief, PCR products were designed to represent all 1576 ORFs in H. pylori NCTC26695 and all 1495 ORFs in H. pylori J99 (Alm et al., 1999; Tomb et al., 1997). The microarrays were constructed by robotic spotting of PCR products in duplicate on UltraGaps amino- silane-coated glass slides (Corning) using a MicroGrid II automated microarrayer (BioRobotics) and post-print-processed according to the slide manufacturer’s instructions. CGH. Because of its history of genetic analysis and its motility, H. pylori strain CCUG17874 (equivalent to the type strain NCTC11637) was used to study global transcription patterns of different flagellar mutants. The genome of this strain has not been sequenced. The macrodiversity of CCUG17874 was therefore investigated using CGH. Nucleic acid labelling was undertaken using a modified protocol described by Hinds et al. (2002a). Briefly, 5 mg wild-type CCUG17874 genomic DNA was labelled with dCTP Cy3. A 41.5 ml mixture containing 5 mg genomic DNA and 3 mg random primers (Promega) was heated at 95 uC for 5 min, snap-cooled and quickly centrifuged at 10 000 g.A5 ml volume of 106 REact2 buffer (Invitrogen), 1 ml dNTP (5 mM dDTP, 2 mM dCTP), 1 ml DNA polymerase I large Klenow fragment (Invitrogen) and 1.5 ml Cy3 dCTP Fluorolink (GE Healthcare) was then added. Similarly, wild-type NCTC26695 genomic DNA was labelled with dCTP Cy5. The mixture was incubated at 37 uC for 90 min in the dark. Subsequent steps, including purification of labelled cDNA, hybridization, washing and scanning, were performed as described below. Array hybridizations for the wild-type and for the mutant were performed in duplicate. Following quantile normalization (Bolstad et al., 2003), a dynamic cut-off-based analysis was performed as described in Kim et al. (2002). The output files give three lists: divergent genes, uncertain genes and present genes. Confirmatory analysis of selected CGH data by PCR. PCRs were performed under standard conditions to confirm selected results obtained by CGH. Seven genes were selected from the CGH data. Primer pair sequences were designed using the Primer3 online software (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi) and are listed in Supplementary Table S1. Type II microarray analysis. To compare the transcriptional profiles of the wild-type and HP0906 mutant strains, the H. pylori whole genome microarray was used in a common reference or type II experimental design whereby Cy5-labelled cDNA from each strain was co-hybridized to an array with a Cy3-labelled genomic DNA reference. Nucleic acid labelling and microarray hybridizations were undertaken according to BmG@S standard protocols (Hinds et al., 2002a). In brief, for the common reference, 5 mg wild-type CCUG17874 genomic DNA was labelled with dCTP Cy3 using random primers (Promega) and DNA polymerase I large Klenow fragment (Invitrogen). Cy5-labelled cDNA was generated from 4 mg total RNA during first-stand synthesis using random primers (Promega) and Superscript II reverse transcriptase (Invitrogen). The Cy3-and Cy5-labelled nucleic acid mixtures were then co-purified using the MinElute PCR Purification kit (Qiagen), mixed in a hybridization solution of 46 SSC and 0.3 % SDS, and hybridized under a LifterSlip (Erie Scientific) for 18 h at 65 uC. Microarray slides were washed once in 16 SSC, 0.06 % SDS at 65 uC for 2 min and twice in 0.066 SSC for 2 min, dried by centrifugation, and scanned using a dual-laser Affymetrix 428 scanner. The images and data were analysed using the GeneDirector software package, which includes ImaGene and GeneSight v2.0 (Biodiscovery). Genes designated as missing or uncertain in CCUG17874 were filtered out. Array hybridizations for the wild-type and for the mutant were performed in triplicate. Ratios for wild-type DNA versus wild-type RNA and wild-type DNA versus mutant RNA were exported to Excel (Microsoft). Following the within-slide normalization in GeneSight, the genes containing empty values were discarded. Quantile normalization was then performed to make the entire distribution of the values identical between each array slide (Bolstad et al., 2003). The triplicates for the wild-type and for the mutant were averaged 1902 Microbiology 155 Modulation of RpoN activity by FliK and final ratios (log2) were calculated. One-way ANOVA was used to calculate statistical confidences. Genes with a P value less than 0.05 and a fold-change greater than 2.00 were designated differentially expressed in the mutant. Quantitative analysis of transcription by real-time PCR. qRT- PCR was performed as a confirmatory test on four flagellar genes following global transcript analysis by microarray. Real-time PCR primers were designed using the Primer3 software package (Rozen & Skaletsky, 2000) and are listed in Supplementary Table S1. RNA (500 ng) was reverse-transcribed using Improm-II reverse trancriptase (Promega) and 500 ng random hexamers, as described in the manufacturer’s manual. qRT-PCR was performed on flagellar genes using primers listed in Supplementary Table S1. The reaction mixture was prepared as described in the manufacturer’s protocol. Briefly, the amplification by qRT-PCR was performed in a final volume of 12.5 ml including 1 ml cDNA, 50 nM of each primer, 6.25 ml26 Master Mix (Biogen) and 1 : 60 000 Sybr Green I (Bio/Gene). qRT-PCRs were run and monitored using an ABI 7000 Thermo cycler and ABI Prism 7000 SDS software (both from Applied Biosystems). Reactions were performed in triplicate (technical replicates) from at least two independent RNA preparations (biological replicates). Relative fold- changes of expression were calculated as described by Pfaffl (2001). The era gene was used as a housekeeping gene (Sebert et al., 2002). Each transcript abundance was therefore calculated relative to the era gene transcript abundance. Reverse transcription and amplification of intergenic regions by PCR. RNA (500 ng) was reverse-transcribed using Improm-II reverse trancriptase (Promega) and 500 ng random hexamers, as described in the manufacturer’s manual. Primer pairs were designed to amplify the intergenic region between hp0114 and hp0115 and an internal region of the hp0115 gene (Supplementary Table S1). Next, the two regions were amplified by PCR using cDNA preparations. Bioinformatics analysis. Predictions of stress-induced DNA duplex destabilization were performed using the SIDD online server developed by the C. Benham laboratory (Bi & Benham, 2004). Predictions of transcriptional terminators were performed using TransTermHP (Kingsford et al., 2007). RESULTS CGH of H. pylori strain CCUG17874 We and others have performed genetic and microbiological analysis of H. pylori motility in strain CCUG17874, whose genome has not been sequenced. To investigate its genome content, we used an H. pylori DNA array representing all the ORFs from strains NCTC26695 and J99. To analyse the level of genetic conservation between CCUG17874 and the sequenced strains, and to improve the subsequent type II array experiment analyses, CGH was performed. Based on the trinary output analysis method described by Kim et al. (2002), genes were classified into three groups: absent (highly divergent), uncertain and absolutely present. The genes designated uncertain have values that are in the transition region, where they cannot be confidently assigned to one or the other group. The uncertain group consisted of 130 genes. Seven genes annotated as highly divergent, uncertain or absolutely present were analysed by PCR using genomic DNA from CCUG17874 (test strain) and NCTC26695 (reference strain) (Fig. 1). Fig. 1. Confirmatory PCR for CGH analysis. Seven genes were amplified using the genomic DNA of CCUG17874 and NCTC26695. omp27 and flgI were assigned as divergent. groES, flaG and omp13 were classified as uncertain. cheA and omp22 were classified as present. The PCR data confirmed the CGH results, with one exception. The HP1192 gene was shown to be present in CCUG17874 using a PCR approach, whereas CGH data indicated its absence (data not shown). Sequence divergence is correlated with hybridization values (Kim et al., 2002). It has been reported that the Cy5 : Cy3 ratio decreases twofold if the corresponding gene has 85 % identity with the gene used for the array design (van Hijum et al., 2008). Thus, some of the divergent genes in CCUG17874 may be present but their sequences may be poorly conserved compared with the sequenced reference strain. The confirmatory PCRs provided empirical data to allow rational adjustment of the cut-off value for spot intensity in the array data, and thus to narrow the breakpoint between genes classified as present and divergent (Fig. 1). The uncertain gene list was thus reduced from 130 to seven genes following the PCR screening. Out of the seven, five genes encode hypothetical proteins. The two remaining genes encode transketolase and geranyltranstransferase. We classified these genes as divergent for the data analysis of further type II array experiments. flgI, encoding the flagellar basal-body P-ring protein, is an essential structural component of the flagellar superstructure. CGH data indicated that flgI was absent from the genome of strain CCUG17874. However, the standard deviation of hybridization values for this gene was very high, suggesting that the CGH output value for this gene was not reliable. PCR investigation confirmed the presence of flgI, which is consistent with the motile phenotype of the CCUG17874 strain used in this study (Fig. 1). Table 1 lists genes present in NCTC26695 and J99 but missing in CCUG17874. Significantly, 14.25 % of H. pylori NCTC26695 genes were absent in the CCUG17874 genome. Most of the missing genes encode hypothetical proteins and appear to be strain-and H. pylori-specific (Table 1). Some genes that encode restriction enzymes, outer-membrane proteins, insertion elements (possibly in http://mic.sgmjournals.org F. P. Douillard and others Table 1. Genes absent from the genome of H. pylori strain CCUG17874, relative to two sequenced genomes, as determined by CGH Strain Number of divergent genes Cellular role category NCTC26695 CCUG17874 147 14 7 5 4 5 2 3 16 24 2 1 Unknown function DNA metabolism (restriction enzyme) Cell envelope Pseudogene Energy metabolism Mobile and extrachromosomal element function Chemotaxis and motility Cellular processes: pathogenesis Others Unknown function DNA metabolism (topoisomerase) Other (DNA transfer protein) reduced copy number in the CCUG17874 genome), virulence-associated proteins and transposases (possibly in reduced copy number in the CCUG17874 genome) were also absent. Transposable elements are often present in multiple copies in the H. pylori genome (Logan & Berg, 1996; Tomb et al., 1997) and may be associated with disruption of operons and large regions in the CCUG17874 genome. Consideration of the location of genes identified as missing by the CGH analysis revealed that five large regions have been disrupted in the genome of CCUG 17874 relative to the NCTC26695 genome (Fig. 2). These regions consist of genes encoding hypothetical proteins and/or mobile and extrachromosomal elements, such as transposases and recombinases. These genes designated as divergent in CCUG17874 were filtered out in further type II array experiments using that H. pylori strain. Transcription analysis of the fliK mutant A previous transcriptional analysis showed that FliK is under the control of the s 54 sigma factor (Niehus et al., Fig. 2. Overview of the divergence of the genome of H. pylori CCUG17874 relative to NCTC26695. Outer rings show strand location of genes, with genes missing in CCUG17874 (relative to NCTC26695) in red, and genes present in green. The middle ring shows average mol% G+C content. The inner ring is GC skew. 1904 Microbiology 155 Modulation of RpoN activity by FliK 2004). To provide the first genome-wide analysis of FliKrelated regulatory circuitry in H. pylori, we performed global transcript analysis of the HP0906 insertional mutant. Fifty-four genes were differentially transcribed in the HP0906 mutant, including 34 hypothetical or putative proteins. Differentially expressed genes with functional annotations are listed in Table 2. Seven genes encoding ribosomal proteins were significantly upregulated in the FliK mutant. The microarray data also showed upregulation of stress-related proteins, such as ferredoxin and thioredoxin, in the FliK mutant. HP0923, encoding an outer-membrane protein (Omp22) that is highly immunoreactive (Kim et al., 2000), was downregulated. Three RpoN-dependent genes, flaB, flgE and HP1076, were upregulated (Table 2). Because of its likely role in the flagellar regulon, we carefully examined expression fold-changes of the flagellar genes in the HP0906 mutant (Table 3). Only one class I gene (fliP/HP0684) was significantly downregulated. The genes for RpoN, the regulator FlhA, FlgR (HP0703) and FlgS (HP0244) were transcribed at wild-type levels in the fliK mutant. All class III genes were also transcribed at wild-type levels. The FliA sigma factor and its anti-sigma factor FlgM were not significantly differentially expressed. Only one gene in the intermediate class was significantly upregulated, hp0367. The response regulator ompR was upregulated 1.9-fold and this upregulation was confirmed by qRT-PCR, which indicated a fold upregulation of 2.25±0.54-fold. fliK mutation affected the transcription of almost all class II genes in the flagellar regulon (Table 3), with only three exceptions. HP0114, part of the RpoN regulon and located downstream of flaB/HP0115, was not upregulated compared with other RpoN-dependent genes in the fliK mutant (Fig. 3, Table 3). HP0114 also has a different transcript profile compared with the other RpoN-dependent genes in an HP0958 mutant (Douillard et al., 2008). We therefore performed bioinformatic analysis to identify DNA regulatory regions, such as promoters, terminations and intrinsic termination motifs (rho-independent transcription terminators) (Bi & Benham, 2004; Kingsford et al., 2007). The algorithm developed by Bi & Benham (2004) identifies regions where the DNA duplex is destabilized, i.e. promoters and terminators. The results suggested that HP0114 and flaB are not co-transcribed (Fig. 3a), which is consistent with our data. In addition, a transcription terminator prediction algorithm (Kingsford et al., 2007) identified a potential terminator GGGC GTTA GCCC located downstream of flaB (confidence score 88 out of 100), followed immediately by a U-rich region. Although reverse-transcription investigations showed a weak band corresponding to the intergenic region between HP0114 and flaB, it appears that these two genes are not completely co-transcribed. The weak PCR band obtained could actually correspond to low-level read-through from the flaB transcript. Table 2. Selected genes differentially expressed in the HP0906 mutant Only genes with fold-change ¢2.00, P ¡0.05 and known function are listed in this Table. Fold-changes and P values were calculated based on three independent biological replicates, as described in Methods. Genes discussed in this paper are shown in bold type. TIGR ORF number Annotation Fold-change P value Downregulated genes Hp26695-0683 Hp26695-1406 Hp26695-1206 Hp26695-0923 Hp26695-1480 Hp26695-0684 Hp26695-0662 Hp26695-1208 Upregulated genes Hp26695-1315 Hp26695-1246 Hp26695-1306 Hp26695-0084 Hp26695-1310 Hp26695-0870 Hp26695-0115 Hp26695-1314 Hp26695-0824 Hp26695-0125 Hp26695-0277 Hp26695-1076 UDP-N-acetylglucosamine pyrophosphorylase (glmU) Biotin synthetase (bioB) Multidrug resistance protein (hetA) Outer-membrane protein (omp22) Seryl-tRNA synthetase (serS) Flagellar biosynthesis protein (fliP) RNase III (rnc) Ulcer-associated adenine-specific DNA methyltransferase Ribosomal protein S19 (rps19) Ribosomal protein S6 (rps6) Ribosomal protein S14 (rps14) Ribosomal protein L13 (rpl13) Ribosomal protein S17 (rps17) Flagellar hook (flgE) Flagellin B (flaB) Ribosomal protein L22 (rpl22) Thioredoxin (trxA) Ribosomal protein L35 (rpl35) Ferredoxin Hypothetical protein 0.237 0.386 0.402 0.410 0.425 0.432 0.440 0.451 2.053 2.073 2.082 2.167 2.273 2.407 2.432 2.464 2.466 2.514 2.550 3.374 0.009 0.004 0.009 0.039 0.001 0.000 0.013 0.003 0.047 0.002 0.030 0.035 0.047 0.008 0.007 0.009 0.002 0.000 0.008 0.007 http://mic.sgmjournals.org F. P. Douillard and others Table 3. Differential expression ratios of all known flagellar genes in the HP0906 mutant relative to the wild-type Fold-changes and P values were calculated based upon three independent biological replicates, as described in Methods. ORFs and gene annotations are based on the TIGR database (Tomb et al., 1997). The genes were assigned to previously proposed flagellar classes (Niehus et al., 2004). Genes with fold-changes in expression with P ¡0.05 are shown in bold type. Dashes indicate values excluded during array data analysis due to variation or technical problems with array features. Proposed class TIGR ORF no. Putative gene product (gene) Dhp0906 P value Class I Class II Class III Intermediate HP0019 HP0082 HP0099 HP0103 HP0173 HP0244 HP0246 HP0325 HP0326 HP0327 HP0351 HP0352 HP0391 HP0392 HP0393 HP0584 HP0599 HP0616 HP0684 HP0685 HP0703 HP0714 HP0770 HP0815 HP0816 HP0840 HP1041 HP1067 HP1092 HP1286 HP1419 HP1420 HP1462 HP1575 HP1585 HP0114 HP0115 HP0295 HP0869 HP0870 HP0906 HP1076 HP1119 HP1120 HP1154 HP1155 HP1233 HP0472 HP0601 HP1051 HP1052 HP0165 Chemotaxis protein (cheV) Methyl-accepting chemotaxis transducer (tlpC) Methyl-accepting chemotaxis protein (tlpA) Methyl-accepting chemotaxis protein (tlpB) Flagellar biosynthetic protein (fliR) Signal-transducing protein, histidine kinase (atoS) Flagellar basal-body P-ring protein (flgI) Flagellar basal-body L-ring protein (flgH) CMP-N-acetylneuraminic acid synthetase (neuA) Flagellar protein G (flaG) Basal-body M-ring protein (fliF) Flagellar motor switch protein (fliG) Purine-binding chemotaxis protein (cheW) Histidine kinase (cheA) Chemotaxis protein (cheV) Flagellar motor switch protein (fliN) Haemolysin secretion protein precursor (hylB) Chemotaxis protein (cheV) Flagellar biosynthesis protein (fliP) Flagellar biosynthesis protein (fliP) Response regulator RNA polymerase sigma 54 factor (rpoN) Flagellar biosynthesis protein (flhB) Flagellar motor rotation protein (motA) Flagellar motor rotation protein (motB) flaA1 protein Flagellar biosynthesis protein (flhA) Chemotaxis protein (cheY) Flagellar basal-body rod protein (flgG) Conserved hypothetical secreted protein (fliZ) Flagellar biosynthesis protein (fliQ) Flagellar export protein ATP synthase (fliI) Secreted protein involved in flagellar motility Homologue of FlhB protein (flhB2) Flagellar basal-body rod protein (flgG) Hypothetical protein Flagellin B (flaB) Flagellin B homologue (fla) Hydrogenase expression/formation protein (hypA) Flagellar hook (flgE) Hook-length-control regulator (fliK) Hypothetical protein Flagellar hook-associated protein 1 (HAP1) (flgK) Hypothetical protein Hypothetical protein (operon with murG) Transferase, peptidoglycan synthesis (murG) Putative flagellar muraminidase (flgJ) Outer-membrane protein (omp11) Flagellin A (flaA) Hypothetical protein UDP-3-O-acyl N-acetylglucosamine deacetylase (envA) Hypothetical protein 1.175 0.952 1.051 1.592 0.978 0.756 – 1.340 0.889 0.684 1.266 1.293 1.780 1.513 1.249 1.451 1.025 1.147 0.432 1.176D 0.855 0.706* 0.649 1.405 1.009D 1.211 0.983 1.347 1.593 1.309 0.791 0.774 1.324 1.122 0.748 1.199 2.432* 1.459 1.808 2.407* Not valid 3.374 1.442 1.772 1.614 1.668 1.396 1.142 1.204 1.083 1.013 1.422 0.16 0.50 0.77 0.01 0.79 0.12 – 0.01 0.20 0.16 0.05 0.02 0.02 0.02 0.20 0.05 0.84 0.44 0.00 0.44 0.46 0.12 0.19 0.06 0.48 0.26 0.79 0.31 0.03 0.66 0.15 0.12 0.00 0.15 0.09 0.43 0.01 0.02 0.00 0.01 Not valid 0.01 0.04 0.03 0.09 0.06 0.00 0.47 0.47 0.59 0.94 0.44 1906 Microbiology 155 Modulation of RpoN activity by FliK Table 3. cont. Proposed class TIGR ORF no. Putative gene product (gene) Dhp0906 P value Not assigned HP0166 HP0366 HP0367 HP0488 HP0907 HP0908 HP1028 HP1029 HP1030 HP1031 HP1032 HP1033 HP1034 HP1035 HP1122 HP1440 HP1557 HP1558 HP1559 HP0751 HP0752 HP0753 HP0754 HP0410 HP0492 HP0797 Response regulator (ompR) Spore coat polysaccharide biosynthesis protein C Hypothetical protein Hypothetical protein Hook-assembly protein, flagella (flgD) Flagellar hook (flgE) Hypothetical protein Hypothetical protein FliY protein (fliY) Flagellar motor switch protein (fliM) Alternative transcription initiation factor, sigma 28 (fliA) Hypothetical protein ATP-binding protein (ylxH) Flagellar biosynthesis protein (flhF) Anti-sigma 28 factor (flgM) Hypothetical protein Flagellar basal-body protein (fliE) Flagellar basal-body rod protein (flgC) (proximal rod protein) Flagellar basal-body rod protein (flgB) (proximal rod protein) Polar flagellin (flaG2) Flagellar cap protein (fliD) Flagellar chaperone (fliS) Flagellar chaperone (fliT) Flagellar sheath-associated protein (hpaA2) Flagellar sheath-associated protein (hpaA3) Flagellar sheath-associated protein (hpaA) 1.923* 0.941 1.790 0.789 0.909 0.802 1.271 1.035 1.094 1.018 1.120 0.921 0.921 0.999 1.392 0.323 1.254 1.055 1.716 1.130 0.831 1.118 – 0.944 1.161 1.420 0.06 0.77 0.02 0.06 0.40 0.06 0.17 0.57 0.54 0.82 0.18 0.44 0.55 0.96 0.03 0.00 0.08 0.71 0.01 0.27 0.19 0.39 – 0.52 0.18 0.11 *Confirmatory analysis by qRT-PCR was performed for these genes. DThe values for these genes were missing in the microarray analysis and were then investigated by qRT-PCR. Four genes, flaB/HP0115, flgE/HP0870, motB/HP0816 and rpoN/HP0714, were subsequently analysed by qRT-PCR to confirm the differential expression indicated by array data (Fig. 4). RpoN controls transcription of flaB and flgE. motB/HP0816 is a class I gene that is expressed at wild-type levels in the HP0906 mutant. The fold-changes obtained were in good agreement with the microarray data (Fig. 4). RpoN-dependent gene transcription in a flgE mutant The flgE gene of H. pylori flagellar hook protein is controlled by the RpoN sigma factor. Completion of the hook is an important morphogenetic check-point in H. pylori flagellum biogenesis (Niehus et al., 2004; O’Toole et al., 1994). Targeted analysis of RpoN-dependent transcription was carried out by performing qRT-PCR for the rpoN and flaB genes (Fig. 4). Mutation of flgE caused a significant increase in rpoN and flaB transcription. DISCUSSION In the present study, we investigated the effect of fliK mutation on RpoN activity and flagellin synthesis in H. pylori. The FliK protein HP0906 controls hook length. In the model bacterium Salmonella enterica serovar Typhimurium, FliK is known to be involved in the switch of substrate export specificity (Minamino et al., 1999). Previous analyses of flagellum regulatory mechanisms have catered for strain-specific effects by performing array analyses on multiple strains harbouring the same mutation (Niehus et al., 2004). We approached this problem by first establishing the flagellar gene complement of the model strain CCUG17874, relative to two H. pylori genome sequences, by CGH. This analysis indicated that all known H. pylori flagellar genes were present in CCUG17874, consistent with the motile phenotype of that strain. This had the additional value of definitively identifying, for the first time in this commonly employed strain, the presence of several variable areas of the genome, including the plasticity zone, the cag pathogenicity island and the DNA modification/restriction genes. These loci have DNA with a low mol% G+C content and are known to be unstable (Alm & Trust, 1999). Most of the divergent genes in CCUG17874 were located in regions with low mol% G+C content and had no attributed function. Some low mol% G+C-content regions in the NCTC26695 and J99 genomes have also been found in self-replicating plasmids, highhttp:// mic.sgmjournals.org F. P. Douillard and others Fig. 3. Transcription analysis of the flaB gene region. (a) Probability profile of stress-induced DNA duplex destabilization of the flaB gene and linked genes, and expression ratio of the relevant genes in the HP0906 mutant. (b) Results from RT-PCR analysis of the flaB operon structure. Primer pairs were designed to amplify the intergenic region between the genes hp0114 and flaB (region 1) and a fragment of the flaB transcript (region 2). lighting that the diversity between strains is also the result of plasmid integration and horizontal transfer (Alm et al., 1999). The majority of genes identified as divergent in CCUG17874 relative to the NCTC26695 and J99 reference Fig. 4. qRT-PCR analysis of transcription of selected flagellar genes in H. pylori fliK and flgE mutants. Fold-changes and SDs (error bars) were calculated relative to abundance of the era transcript. qRT-PCRs were performed on at least two biological replicates. genomes encode DNA restriction/modification enzymes. Interestingly, CCUG17874 also lacked some genes encoding proteins associated with the cell envelope and pathogenesis that are likely to be involved in interaction with the host cells. It is noteworthy that the CCUG17874 genome may also contain additional genes that are strain- specific. Global transcript analysis of the fliK mutant indicated that insertional inactivation of this gene increased the expression of most RpoN-dependent genes, and it did not directly modulate any other key regulators of the flagellar regulon at a transcriptional level. The class I genes encoding RpoN and the FlgR/FlgS system were transcribed at wild-type levels in the HP0906 mutant. The most affected genes in the RpoN regulon were two essential structural genes, HP0870/flgE (flagellar hook) and HP0115/ flaB (minor flagellin), and also HP1076, encoding a hypothetical protein. Confirmation of the inclusion of HP1076 in the H. pylori RpoN regulon highlights the motivation for functional investigation of the role of this gene in motility. HP0114 is an essential gene for motility (Schirm et al., 2003). It was initially assigned to class II (RpoN-dependent genes) (Niehus et al., 2004). Our array analysis suggested that HP0114 and flaB are not co-transcribed. Using an algorithm to predict stress-induced DNA duplex destabilization, we identified a potential regulatory region located downstream of the flaB gene, further supporting the suggestion that the two genes are not in fact co-transcribed. Transcription termination predictions also identified a potential intrinsic terminator downstream of flaB.We concluded that HP0114 is required in the flagellar regulon, but that it is not in class II (RpoN-dependent genes), as suggested in an earlier study (Niehus et al., 2004), and is more likely to be in the intermediate class. The upregulation of the RpoN-dependent genes in the fliK mutant suggests a transcriptional activation mechanism for RpoN-dependent genes which is possibly triggered by the FlgS/FlgR system in response to the completion of the MS ring and/or export apparatus. Presumably, this activation is normally transient in wild-type cells. After completion of the hook and progression from RpoN-dependent to FliAdependent transcription, a downregulation or termination mechanism should also exist to terminate the activation of the RpoN regulon. The lack of such a mechanism in the fliK mutant apparently results in a sustained upregulation of the RpoN-dependent genes [exemplified phenotypically by the polyhook structure in a fliK mutant (Ryan et al., 2005b)]. We hypothesize that the signal inducing the FlgR/ FlgS activation system stops upon completion of the rod and the hook. In a wild-type cell, expression of the RpoNdependent genes, including the hook component, is thus prevented from occurring at a later stage of the flagellar assembly. The coupling between hook assembly completion and turning off RpoN-dependent gene transcription is mecha 1908 Microbiology 155 Modulation of RpoN activity by FliK nistically unclear. It is unlikely that the turn-off of the RpoN regulon is controlled by a FliA-dependent gene. In a Campylobacter jejuni fliA mutant, a poly-hook phenotype has not been reported, in contrast to that previously observed in FliK mutants (Jagannathan et al., 2001). In theory, the FliK protein could be directly responsible for switching. In S. enterica, the FliK protein is secreted during hook polymerization (Minamino et al., 1999). We tentatively suggest that the accumulation of FliK upon completion of the hook affects RpoN transcription activity by interacting directly or indirectly with the FlgR/FlgS system or the RpoN sigma factor. This is indicated in the model in Fig. 5, which builds on our previous work (Douillard et al., 2008). However, no protein–protein interactions between FlgR or FlgS and FliK have so far been identified (Rain et al., 2001). This suggests that the signal controlling the transcription of the class II genes does not directly involve FliK itself, but more likely another component of the flagellar apparatus, such as the flagellar protein FlhB (not shown in the model). This protein controls the switch of substrate specificity export from rod–hook–basal body class genes to filament genes in S. enterica serovar Typhimurium (Williams et al., 1996). The fragment FlhBC resulting from the autocatalytic cleavage of FlhB shares similarities with a novel flagellar protein, FlhX (Wand et al., 2006), which may also be implicated. Upon completion of the hook, the RpoN regulon would be turned off by a FliK-dependent mechanism, illustrated by dotted lines in Fig. 5. However, the mechanisms for this proposed onwards relay of a FliK–FlhB-transmitted signal to RpoN are not clear. The elevation of RpoN-dependent gene expression in the flgE mutant is compatible with this model, since the transition to FliA-mediated gene expression cannot occur in the absence of a completed hook- basal body. We have recently shown that the HP0958 protein controls the translation of the flaA mRNA (Douillard et al., 2008), as well as being an RpoN chaperone (Pereira & Hoover, 2005). Significantly, HP0958 also binds to FliH (Rain et al., 2001), the inhibitor of the flagellum export ATPase FliI (Lane et al., 2006). In the termination stages of this model (Fig. 5), the HP0958 protein becomes detached from its association with the sigma factor RpoN. In parallel, the anti-sigma factor FlgM is secreted from the cell. Thus, the fliA-dependent genes, including flaA, are then transcribed at high levels, consistent with our previously reported repression of flaA transcription and translation in a fliK mutant (Ryan et al., 2005b). [Curiously, however, mutation of flgE does not abolish FlaA protein production (O’Toole et al., 1994), though it should be noted that flaA transcription was not analysed in that study.] As has been Fig. 5. Proposed and expanded model of action of FliK (HP0906) and HP0958 in H. pylori flagellar biogenesis. Known events or interactions are indicated by unbroken arrows, proposed or unclear events by dotted arrows. The potential roles of the membrane-bound FlhB protein, or FlhX, are omitted for clarity. The lower boxed section indicates late-stage events after flagellin gene transcription has commenced. http://mic.sgmjournals.org F. P. Douillard and others previously proposed, HP0958 acts as a post-transcriptional regulator on flagellin synthesis by targeting flaA mRNA to the export apparatus and promoting coupled FlaA translation/secretion (Douillard et al., 2008). Experiments are in progress to test these hypotheses. ACKNOWLEDGEMENTS H. pylori flagellum research in P. W. O’T.’s lab was supported by a Science Foundation Ireland grant from the Research Frontiers Programme. We acknowledge the Wellcome Trust for supporting BmG@S (Bacterial Microarray Group at St George’s, University of London). We thank A. Zomer for valuable discussions on the manuscript and M. J. Claesson for the generation of the circular genome map. REFERENCES Alm, R. A. & Trust, T. J. (1999). Analysis of the genetic diversity of Helicobacter pylori: the tale of two genomes. J Mol Med 77, 834–846. Alm, R. A., Ling, L.-S., Moir, D. T., King, B. L., Brown, E. D., Doig, P. C., Smith, D. R., Noonan, B., Guild, B. C. & other authors (1999). Genomic-sequence comparison of two unrelated isolates of the human gastric pathogen Helicobacter pylori. Nature 397, 176–180. Beier, D. & Frank, R. (2000). Molecular characterization of two- component systems of Helicobacter pylori. J Bacteriol 182, 2068–2076. Bi, C. & Benham, C. J. (2004). WebSIDD: server for predicting stress- induced duplex destabilized (SIDD) sites in superhelical DNA. Bioinformatics 20, 1477–1479. Bolstad, B. M., Irizarry, R. A., Astrand, M. & Speed, T. P. (2003). A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics 19, 185– 193. Brahmachary, P., Dashti, M. G., Olson, J. W. & Hoover, T. R. (2004). Helicobacter pylori FlgR is an enhancer-independent activator of s 54RNA polymerase holoenzyme. J Bacteriol 186, 4535–4542. Colland, F., Rain, J.-C., Gounon, P., Labigne, A., Legrain, P. & De Reuse, H. (2001). Identification of the Helicobacter pylori anti-s 28 factor. Mol Microbiol 41, 477–487. Douillard, F. P., Ryan, K. A., Caly, D. L., Hinds, J., Witney, A. A., Husain, S. E. & O’Toole, P. W. (2008). Posttranscriptional regulation of flagellin synthesis in Helicobacter pylori by the RpoN chaperone HP0958. J Bacteriol 190, 7975–7984. Eaton, K. A., Brooks, C. L., Morgan, D. R. & Krakowka, S. (1991). Essential role of urease in pathogenesis of gastritis induced by Helicobacter pylori in gnotobiotic piglets. Infect Immun 59, 2470– 2475. Eaton, K. A., Morgan, D. R. & Krakowka, S. (1992). Motility as a factor in the colonisation of gnotobiotic piglets by Helicobacter pylori. J Med Microbiol 37, 123–127. Eaton, K. A., Suerbaum, S., Josenhans, C. & Krakowka, S. (1996). Colonization of gnotobiotic piglets by Helicobacter pylori deficient in two flagellin genes. Infect Immun 64, 2445–2448. EUROGAST (1993). An international association between Helicobacter pylori infection and gastric cancer. Lancet 341, 1359–1362. Ferris, H. U. & Minamino, T. (2006). Flipping the switch: bringing order to flagellar assembly. Trends Microbiol 14, 519–526. Frenck, R. W., Jr & Clemens, J. (2003). Helicobacter in the developing world. Microbes Infect 5, 705–713. Galkin, V. E., Yu, X., Bielnicki, J., Heuser, J., Ewing, C. P., Guerry, P. & Egelman, E. H. (2008). Divergence of quaternary structures among bacterial flagellar filaments. Science 320, 382–385. Hinds, J., Laing, K. G., Mangan, J. A. & Butcher, P. D. (2002a). Glass slide microarray for bacterial genomes. In Methods in Microbiology: Functional Microbial Genomics, pp. 83–99. Edited by B. W. Wren & N. Dorell. London: Academic Press. Hinds, J., Witney, A. A. & Vass, J. K. (2002b). Microarray design for bacterial genomes. In Methods in Microbiology: Functional Microbial Genomics, pp. 67–82. Edited by B. W. Wren & N. Dorell. London: Academic Press. Jagannathan, A., Constantinidou, C. & Penn, C. W. (2001). Roles of rpoN, fliA, and flgR in expression of flagella in Campylobacter jejuni. J Bacteriol 183, 2937–2942. Josenhans, C., Niehus, E., Amersbach, S., Horster, A., Betz, C., Drescher, B., Hughes, K. T. & Suerbaum, S. (2002). Functional characterization of the antagonistic flagellar late regulators FliA and FlgM of Helicobacter pylori and their effects on the H. pylori transcriptome. Mol Microbiol 43, 307–322. Kikuchi, S. & Dore, M. P. (2005). Epidemiology of Helicobacter pylori infection. Helicobacter 10, 1–4. Kim, J. S., Chang, J. H., Seo, W. Y., Yu, G. J., Chung, S. I. & Yum, J. S. (2000). Cloning and characterization of a 22 kDa outer-membrane protein (Omp22) from Helicobacter pylori. Mol Cells 10, 633–641. Kim, C. C., Joyce, E. A., Chan, K. & Falkow, S. (2002). Improved analytical methods for microarray-based genome-composition analysis Genome Biol 3, research0065. Kingsford, C. L., Ayanbule, K. & Salzberg, S. L. (2007). Rapid, accurate, computational discovery of Rho-independent transcription terminators illuminates their relationship to DNA uptake. Genome Biol 8, R22. Lane, M. C., O’Toole, P. W. & Moore, S. A. (2006). Molecular basis of the interaction between the flagellar export proteins FliI and FliH from Helicobacter pylori. J Biol Chem 281, 508–517. Logan, R. P. & Berg, D. E. (1996). Genetic diversity of Helicobacter pylori. Lancet 348, 1462–1463. Minamino, T., Gonzalez-Pedrajo, B., Yamaguchi, K., Aizawa, S. I. & Macnab, R. M. (1999). FliK, the protein responsible for flagellar hook length control in Salmonella, is exported during hook assembly. Mol Microbiol 34, 295–304. Niehus, E., Gressmann, H., Ye, F., Schlapbach, R., Dehio, M., Dehio, C., Stack, A., Meyer, T. F., Suerbaum, S. & Josenhans, C. (2004). Genome-wide analysis of transcriptional hierarchy and feedback regulation in the flagellar system of Helicobacter pylori. Mol Microbiol 52, 947–961. O’Toole, P. W., Kostrzynska, M. & Trust, T. J. (1994). Non-motile mutants of Helicobacter pylori and Helicobacter mustelae defective in flagellar hook production. Mol Microbiol 14, 691–703. Parsonnet, J., Hansen, S., Rodriguez, L., Gelb, A. B., Warnke, R. A., Jellum, E., Orentreich, N., Vogelman, J. H. & Friedman, G. D. (1994). Helicobacter pylori infection and gastric lymphoma. N Engl J Med 330, 1267–1271. Pereira, L. & Hoover, T. R. (2005). Stable accumulation of s in Helicobacter pylori requires the novel protein HP0958. J Bacteriol 187, 4463–4469. Pfaffl, M. W. (2001). A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 29, e45. Rain, J. C., Selig, L., De Reuse, H., Battaglia, V., Reverdy, C., Simon, S., Lenzen, G., Petel, F., Wojcik, J. & other authors (2001). The protein–protein interaction map of Helicobacter pylori. Nature 409, 211–215. 1910 Microbiology 155 Modulation of RpoN activity by FliK Rozen, S. & Skaletsky, H. (2000). Primer3 on the WWW for general users and for biologist programmers. Methods Mol Biol 132, 365–386. Ryan, K. A., Karim, N., Worku, M., Moore, S. A., Penn, C. W. & O’Toole, P. W. (2005a). HP0958 is an essential motility gene in Helicobacter pylori. FEMS Microbiol Lett 248, 47–55. Ryan, K. A., Karim, N., Worku, M., Penn, C. W. & O’Toole, P. W. (2005b). Helicobacter pylori flagellar hook-filament transition is controlled by a FliK functional homolog encoded by the gene HP0906. J Bacteriol 187, 5742–5750. Scarlato, V., Delany, I., Spohn, G. & Beier, D. (2001). Regulation of transcription in Helicobacter pylori: simple systems or complex circuits? Int J Med Microbiol 291, 107–117. Schirm, M., Soo, E. C., Aubry, A. J., Austin, J., Thibault, P. & Logan, S. M. (2003). Structural, genetic and functional characterization of the flagellin glycosylation process in Helicobacter pylori. Mol Microbiol 48, 1579–1592. Sebert, M. E., Palmer, L. M., Rosenberg, M. & Weiser, J. N. (2002). Microarray-based identification of htrA,a Streptococcus pneumoniae gene that is regulated by the CiaRH two-component system and contributes to nasopharyngeal colonization. Infect Immun 70, 4059– 4067. Tomb, J.-F., White, O., Kerlavage, A. R., Clayton, R. A., Sutton, G. G., Fleischmann, R. D., Ketchum, K. A., Klenk, H. P., Gill, S. & other authors (1997). The complete genome sequence of the gastric pathogen Helicobacter pylori. Nature 388, 539–547. van Hijum, S. A., Baerends, R. J., Zomer, A. L., Karsens, H. A., Martin- Requena, V., Trelles, O., Kok, J. & Kuipers, O. P. (2008). Supervised Lowess normalization of comparative genome hybridization data – application to lactococcal strain comparisons. BMC Bioinformatics 9,93. Veldhuyzen van Zanten, S. J. & Sherman, P. M. (1994). Helicobacter pylori infection as a cause of gastritis, duodenal ulcer, gastric cancer and nonulcer dyspepsia: a systematic overview. CMAJ 150, 177–185. Wand, M. E., Sockett, R. E., Evans, K. J., Doherty, N., Sharp, P. M., Hardie, K. R. & Winzer, K. (2006). Helicobacter pylori FlhB function: the FlhB C-terminal homologue HP1575 acts as a ‘‘spare part’’ to permit flagellar export when the HP0770 FlhBCC domain is deleted. J Bacteriol 188, 7531–7541. Williams, A. W., Yamaguchi, S., Togashi, F., Aizawa, S.-I., Kawagishi, I. & Macnab, R. M. (1996). Mutations in fliK and flhB affecting flagellar hook and filament assembly in Salmonella typhimurium. J Bacteriol 178, 2960–2970. Edited by: R. J. Maier http://mic.sgmjournals.org