- Email: [email protected]
Global responses of Aliivibrio salmonicida to hydrogen peroxide as revealed by microarray analysis
Marine Genomics 3 (2010) 193–200
Contents lists available at ScienceDirect
Marine Genomics j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / m a r g e n
Global responses of Aliivibrio salmonicida to hydrogen peroxide as revealed by microarray analysis Hege L. Pedersen a, Erik Hjerde a, Steinar M. Paulsen a,1, Hilde Hansen a, Lotte Olsen a,b, Sunniva K. Thode a, Marcos T. Dos Santos a, Ruth H. Paulssen b, Nils-Peder Willassen a,c, Peik Haugen a,c,⁎ a b c
Department of Chemistry, University of Tromsø, N-9037 Tromsø, Norway Institute of Clinical Medicine, University of Tromsø, N-9037 Tromsø, Norway The Norwegian Structural Biology Centre, University of Tromsø, N-9037 Tromsø, Norway
a r t i c l e
i n f o
Article history: Received 29 January 2010 Received in revised form 4 October 2010 Accepted 5 October 2010 Keywords: Aliivibrio salmonicida Hydrogen peroxide Oxidative stress Microarray Hitra disease
a b s t r a c t Aliivibrio salmonicida causes “cold-water vibriosis” (or “Hitra disease”) in fish, including marine-reared Atlantic salmon. During development of the disease the bacterium will encounter macrophages with antibacterial activities such as production of damaging reactive oxygen species (ROS). To defend itself the bacterium will presumably start producing detoxifying enzymes, reducing agents, and proteins involved in DNA and protein repair systems. Even though responses to oxidative stress are well studied for a few model bacteria, little work has been done in general to explain how important groups of pathogens, like members of the Vibrionaceae family, can survive at high levels of ROS. We have used bioinformatic tools and microarray to study how A. salmonicida responds to hydrogen peroxide (H2O2). First, we used the recently published genome sequence to predict potential binding sites for OxyR (H2O2 response regulator). The computer-based search identified OxyR sites associated with 20 single genes and 8 operons, and these predictions were compared to experimental data from Northern blot analysis and microarray analysis. In general, OxyR binding site predictions and experimental results are in agreement. Up- and down regulated genes are distributed among all functional gene categories, but a striking number of ≥ 2 fold up regulated genes encode proteins involved in detoxification and DNA repair, are part of reduction systems, or are involved in carbon metabolism and regeneration of NADPH. Our predictions and –omics data corroborates well with findings from other model bacteria, but also suggest species-specific gene regulation. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Damaging reactive oxygen species (ROS) form inside cells as byproducts from normal cellular processes, they may be produced actively by macrophages as part of the innate immune system, and they are sometimes produced by bacteria in the competition for survival (Imlay, 2008). Bacteria respond to ROS by producing elevated levels of detoxifying enzymes, reduction systems, and proteins involved in protection and repair of DNA. In Escherichia coli and other Gramnegative bacteria the presence of hydrogen peroxide (H2O2) is for example counteracted by the production of catalases (Storz and Tartaglia, 1992), i.e. enzymes that catalyses the decomposition of ⁎ Corresponding author. Department of Chemistry, University of Tromsø, Norway. Tel.: +47 776 5288; fax: +47 776 5350. E-mail addresses: [email protected] (H.L. Pedersen), [email protected] (E. Hjerde), [email protected] (S.M. Paulsen), [email protected] (H. Hansen), [email protected] (L. Olsen), [email protected] (S.K. Thode), [email protected] (M.T.D. Santos), [email protected] (R.H. Paulssen), [email protected] (N.-P. Willassen), [email protected] (P. Haugen). 1 Current address: MABCENT-SFI, Centre on Marine bioactives and drug discovery, University of Tromsø, Tromsø, Norway. 1874-7787/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.margen.2010.10.002
H2O2 to water and oxygen. The total response to H2O2 is however more complicated, and for E. coli and a few other well-studied bacteria, global changes in the transcriptome and proteome have been investigated by microarray (Zheng et al., 2001; Li et al., 2004; Palma et al., 2004; Chang et al., 2005; Stohl et al., 2005; Zeller et al., 2005, 2007; Diaz et al., 2006; Sund et al., 2008) and 2D gel electrophoresis (Morgan et al., 1986; VanBogelen et al., 1987; reviewed by Han and Lee, 2006), respectively. These and other analyses show that the global regulator OxyR has a key role in conferring resistance to H2O2 (Zheng et al., 1998, 2001; Aslund et al., 1999), although other systems such as PerR (Horsburgh et al., 2001a,b; Helmann et al., 2003), σs (Loewen et al., 1998; Gort et al., 1999), Fur (Zheng et al., 1999), Fnr (Gort et al., 1999), MarAB (Jair et al., 1995) and ArcA (Lu et al., 2002; Wong et al., 2007) may also be involved. OxyR knock-out mutants are significantly more sensitive to H2O2 exposure and are less likely to cause infections (Sund et al., 2008; Lau et al., 2005). In the presence of H2O2 the E. coli OxyR is activated by formation of an intra-molecular disulfide bridge between cysteines 199 and 208 (Zheng et al., 1998; Choi et al., 2001). The E. coli OxyR dimers (or tetramers) bind to promoters at degenerated binding sites (ATAG/CTAT boxes) located in four adjacent major grooves (Toledano et al., 1994), and typically activates transcription of the associated genes (Zheng
194
H.L. Pedersen et al. / Marine Genomics 3 (2010) 193–200
et al., 2001). Some of the best studied examples of OxyR up regulated genes in E. coli are dps, which encodes a nonspecific DNA binding protein, katG, which encodes a catalase and oxyS which encodes a regulatory small RNA. The Gram-negative bacterium Aliivibrio salmonicida causes disease in farmed Atlantic salmon (Salmo salar), Atlantic cod (Gadus morhua) and rainbow trout (Oncorhynchus mykiss). During the 80s, the bacterium was directly responsible for severe economic losses in the fish-farming industry, but is today kept under control by vaccinating juvenile fish prior to their release into marine fish farms. The complete genome of A. salmonicida strain LFI1238 was recently published (Hjerde et al., 2008): it consists of one large and one small circular chromosome of 3.3 and 1.2 Mb, respectively, (typical for species that belong to the Vibrionaceae family) and four plasmids. A striking feature of the genome is the mosaic composition that apparently originates from high numbers of intra-chromosomal rearrangements, deletions and duplications, and mobile genetic elements (i.e., insertion sequences, genomic islands and prophages). Interestingly, A. salmonicida lacks several genes that play important roles in oxidative stress responses in E. coli (e.g., soxS, katG, katE, ahpF, and dps). This observation inspired us to investigate the global responses of A. salmonicida to H2O2. A computational approach was used to predict OxyR binding sites, and microarray was subsequently used to study changes in the transcriptome. Time-dependent responses were measured by harvesting samples 15, 30 and 60 min after addition of H2O2. Our results reveal responses that are well-known from the literature from studies using other bacteria, but also show new solutions to gene regulation in oxidative stress responses. 2. Materials and methods 2.1. Bacterial strains, culture conditions and sampling A. salmonicida strain LFI1238 was grown in Luria-Bertani (LB) medium (Sambrook et al., 1989) with 2.5% NaCl at 15 °C and 220 rpm. For stimulation experiments, cells were grown to mid-exponential phase (OD600nm = 0.5), split in two halves and 100 μM or 1-5 mM H2O2 (Sigma Aldrich) was added to one of the cultures (see main text). Cells were harvested at 5, 30 and 60 min for the initial Northern blots analysis, and at 15, 30 and 60 min for the microarray analysis. Three independent experiments were done for the microarray analysis. RNA extraction is described below. 2.2. Northern blot analysis The Northern blot analysis was performed essentially as previously described (Ahmad et al., 2009). Briefly, total RNA was isolated from bacteria using the Trizol protocol. Ten μg RNA was separated on 1.2% denaturing formamide agarose gels, and run for 2 min at 20 V, 10 min at 40 V and finally at 60 V until the dye reached the end of the gel. RNA was next transferred to a Hybond-N + nylon membrane (Amersham) by capillary transfer. For Northern analysis [α-32P]-labeled dsDNA was used as probes according to the Northern Max TM instruction manual (Ambion). Hybridisations were performed over-night at 42 °C and signals were acquired on phosphoimaging screens (Fujifilm) and scanned using a BAS-5000 phosphoimager (Fujifilm). Quantification of signals was done using the ImageGauge software v4.0 (Fujifilm). 2.3. Microarray analysis For microarray analysis RNA was extracted from cells using the RNA-isol reagent (Fischer scientific), and DNA was removed using the DNA-free kit (Ambion). Any traces of DNase were removed with the RNeasy Minelute Cleanup kit (Qiagen). RNA quality was checked with a Bioanalyzer, Nanodrop, and by incubating the total RNA for 1 h at 37 °C with 500 ng plasmid DNA and then running the sample on a 1%
agarose gel (i.e., to check for any remaining DNase activity). cDNA was generated and labelled by using 15 μg RNA, the Aminoallyl cDNA labeling kit (Ambion), and the CyDye™ Post-Labeling Reactive Dye Pack (GE Healthcare). Cy3 and Cy5 were used as dyes. Samples were hybridised to “Vibrio salmonicida V1.0.1 AROS” slides at 42 °C on a TECAN HS4800 hybridisation station, and microarray slides were subsequently washed, once in 0.1 × SSC/0.1% SDS for 5 min at 42 °C, then once in 0.1 × SSC/0.1% SDS for 10 min at room temperature, and finally four times in 0.1 × SSC for 1 min at room temperature. Slides were scanned using a GenePix 4000B scanner (Axon Instruments Inc.). Images were explored using the GenePix Pro v6.1 software and analysis of expression data was done using J-Express Pro v2.7 (Dysvik and Jonassen, 2001). Background was subtracted after quantification using a global background subtraction method. The absolute pixel values for the spots of each gene were averaged. The ratio between the average intensity signals from three H2O2 treated samples and three untreated samples represented the fold change in gene expression. All experiments were mean normalised based on expression ratio data. Microarray data has been uploaded to The NCBI Gene Expression Omnibus (GEO) database, and is available through accession number GSE20082. 2.4. Computer based prediction of OxyR binding sites To predict OxyR binding sites in the A. salmonicida genome we first compiled sequences of E. coli promoters that contain experimentally verified OxyR boxes, or are assumed to contain OxyR boxes (i.e. gene expression is OxyR-dependent). The extracted sequences were 250 nt upstream of the translation start codon. Extracted promoter regions were associated with the following genes: ahpC, dps, dsbG, fhuF, flu, fur, gor, grxA, hemH, katG, oxyR, oxyS, rcsC, trxC, sufA, uxuA, ybjC, yhjA and ygaQ. Sequences were extracted from the evolutionary related gamma-proteobacteria Vibrio and Aliivibrio genomes (204 sequences). The complete genome of nine Vibrio and three Aliivibrio species, and genome draft sequences of seven Vibrio species were extracted from the NCBI ftp site (ftp://ftp.ncbi.nlm.nih.gov). Draft sequences were automatically annotated using the GenDB pipeline (Meyer et al., 2003). The sequence dataset and the motif discovery tool MEME (Bailey and Elkan, 1994) were used to search for conserved motifs using the parameters set to search for motifs 30 nt or more in length. Manual inspection of the result revealed 41 nt OxyR box-like motifs in 99 sequences (Supplemental material 1). Based on the 99 sequences, an alignment matrix (Supplemental material 2) which describes the frequency of each base at each of the 41 positions, was built using BioEdit v.7.0.9 (Hall, 1999). In addition, an OxyR box consensus was made with WebLogo (Crooks et al., 2004). Using the Patser program (Hertz and Stormo, 1999) and the Vibrio/Aliivibrio adjusted OxyR box alignment matrix, the genome of A. salmonicida was searched for potential OxyR binding sites. OxyR box searches were performed within the region -250 to + 1 relative to the translational start codon of annotated genes. The lower Patser cut-off score was set at 7.0 whereas all other Patser parameters were set to default. 3. Results 3.1. Computer based prediction of OxyR binding sites in A. salmonicida In this study we set out to globally predict OxyR binding sites in the recently published A. salmonicida genome (Hjerde et al., 2008). E. coli and A. salmonicida OxyR proteins share overall 65% identity and 88% similarity, and 77% identity and 100% similarity in the conserved DNA-binding domain. We therefore found it reasonable to believe that OxyR–DNA interactions are conserved in the two gammaproteobacteria. Based on our assumption, we searched the literature for E. coli genes with OxyR-dependent expression patterns and
H.L. Pedersen et al. / Marine Genomics 3 (2010) 193–200
195
Fig. 1. OxyR-box consensus sequence based on experimentally verified and proposed OxyR-binding sites of known OxyR-regulated genes in E. coli and homologous genes in Vibrio and Aliivibrio species. (A) WebLogo was used to build a consensus sequence logo, in which the height of individual letters within a stack of letters represents the relative frequency of that letter at a given position, and the overall height of the stack represents the degree of conservation at that position. (B) Model of oxidised OxyR binding to DNA based on the study of Toledano et al. (1994).
experimentally verified OxyR binding sites. A total of 19 genes were selected for further analysis (see Materials and methods), and 250 nt from the corresponding promoters were extracted from GenBank and aligned. Homologous promoters from 16 Vibrio and three Aliivibrio species were added to the alignment. The dataset, which contained 222 sequences, was next used as input in MEME analysis to search for conserved motifs. MEME identified a 41 nt conserved motif with a distinct periodicity of 4 nt (ATAG or CTAT) separated by seven degenerate nt in 99 of the 222 sequences. Seven of the 19 OxyR boxes from E. coli promoters were recovered in the final dataset, which show that even this relatively relaxed motif can be recovered by our search method. Finally, we generated an alignment matrix based on the 99 OxyR motifs, and used this to generate a WebLogo (Fig. 1), which describes the frequency of each base at each position. This motif is in agreement with the OxyR recognition motif (5’-ATAGntnnnanCTATnnnnnnnATAGntnnnanCTAT) suggested by Toledano et al. (1994). Also, the alignment matrix was used as input in Patser to search the A. salmonicida genome for OxyR binding sites. OxyR binding sites were predicted in front of 20 protein coding sequences (CDSs) and 8 operons in A. salmonicida (provided in Supplementary material 3). OxyR boxes with Patser scores above 10 (high confidence values) were found upstream of a phosphorylase (VSAL_I1055), two conserved hypothetical proteins (VSAL_I1295 and VSAL_2547), a ribosomal protein (VSAL_I2837), and a membrane protein (VSAL_II0074), and in front of oxyR, ahpC and grxA, which were included in the initial alignment to construct the alignment matrix. Also, high confidence OxyR boxes were found associated with three operons, i.e., an operon containing two CDSs annotated as hybrid peroxiredoxin and dihydrolipoyl dehydrogenase (VSAL_I2713-12), an operon containing three CDSs involved in the degradation of carbohydrates (VSAL_II0687-85), and an operon containing two CDSs encoding a patatin-like phospholipase and catalase (VSAL_II0214-13). Three of the putative target genes were either pseudogenes or gene remnants and were not included in the results. For known OxyR-activated genes, OxyR binds to the DNA upstream of the -35 promoter of target genes (Toledano et al., 1994). In A. salmonicida all OxyR boxes were found in the proximity of the -35 promoter with a distance of between 11 and 201 nt from the translational start codon of the gene.
exponential phase, 1–5 mM H2O2 was added and growth was subsequently monitored for 300 min. The result is presented in Fig. 2A and shows that at these concentrations cellular growth was inhibited, but not completely stopped. Five mM H2O2 was chosen for further gene expression analyses. To evaluate if relevant genes had been activated, bacteria were sampled 5, 30 and 60 min after stimulation, and total RNA from cells was subjected to Northern blot analyses. Expression of ahpC was monitored because it is associated
3.2. Expression of sodC and ahpC increases in response to H2O2 To identify nonlethal and appropriate doses of H2O2 for studying oxidative stress responses in A. salmonicida, the bacterium was cultivated in the presence of increasing concentrations of H2O2. We wanted to find conditions that would trigger relevant cellular H2O2 responses, but at the same time would not over-stimulate the cells (over-stimulation activates/represses non-relevant genes). We considered H2O2 concentrations with minor inhibitory effects on cell growth as desired stress conditions. The bacterium was grown to
Fig. 2. Gene expression profiles in response to hydrogen peroxide. Growth curve of A. salmonicida cultured to mid-exponential phase (OD600 = 0.6), split and added H2O2 (A). H2O2 was added at different concentrations (A, enlarged), except for the control, which was unchanged. The concentrations of H2O2 were 1 mM, 5 mM, 8 mM and 10 mM. Northern blots show the expression of different mRNA transcripts in response to H2O2 (B). All the CDSs were normalised to the 16 S ribosomal RNA signal (endogenous control) and compared with the control. Samples were collected 0, 5, 30 and 60 min after treatment with 5 mM H2O2.
196
H.L. Pedersen et al. / Marine Genomics 3 (2010) 193–200
with an OxyR box in our OxyR predictions, and is activated by OxyR upon H2O2 stimulation in E. coli, Pseudomonas putida and Pseudomonas aeruginosa (Christman et al., 1985; Ochsner et al., 2000; Hishinuma et al., 2006). In addition, we examined the expression of sodC, sodB, fur and rpoS, which are also implicated in oxidative stress response in E. coli (Hopkin et al., 1992; Compan and Touati, 1993; Steinman et al., 1994; Gort et al., 1999; Ruiz-Laguna et al., 2000). Of these, only sodC is associated with an OxyR box in A. salmonicida. Five min after addition of H2O2 (Fig. 2B) the levels of ahpC and sodC mRNAs are significantly induced, whereas the levels of sodB and fur mRNAs apparently are reduced. In summary, we conclude that 5 mM H2O2 slightly reduces growth of A. salmonicida and increases expression of sodC and ahpC. This indicates that the growth conditions used in our study are adequate for studying H2O2 responses. 3.3. Microarray analysis reveals H2O2-dependent gene expression of defence molecules, dehydrogenases and reductases Next, microarray analysis was used to investigate global changes in the A. salmonicida transcriptome following exposure to H2O2. A new batch of H2O2 was used for this experiment, and therefore the added concentration of H2O2 was re-calibrated by repeating the experiment described in Fig. 2A. In this experiment, 0.1 mM H2O2 induced appropriate reduction in cellular growth. H2O2 is unstable and spontaneously decomposes to water and oxygen gas, which explains why 5 mM H2O2 resulted in a similar reduction in growth in the initial experiment (Fig. 2A) as described above. In three independent experiments, A. salmonicida was grown in LB medium with 2.5%
NaCl to OD600nm = 0.6 (i.e., approx. mid-exponential growth phase) and then subjected to 0.1 mM H2O2. Samples were collected 15, 30 and 60 min after treatment. Untreated cells collected at same time points were used as control. Total RNA from the three cultures (biological replicates) were pooled and subjected to cDNA synthesis (described in Materials and methods), and run on three 70-mer-based whole genome microarray slides (technical replicates). Results were analysed with GenePix and J-Express. First we wanted to investigate how the differentially expressed genes were distributed into functional classes, and for this purpose we took into consideration genes with fold change of ≥1.5 or ≤0.66 (i.e., 0.66 corresponds to 1.5 fold down regulation). Fig. 3 shows a graphical presentation of the functional classes and the number of differentially expressed genes 15 and 60 min after H2O2 treatment. In general, the distribution of genes in 30 min samples was similar to that of 15 min, and was omitted from the figure for simplicity. Two main conclusions can be drawn from the analysis. First, affected genes are distributed into all classes of genes. Second, significantly more genes are differentially expressed at 60 min than at 15 min. Specifically, 61 and 120 genes are up regulated at 15 and 60 min, respectively, and 75 and 151 genes are down regulated at 15 and 60 min, respectively. This finding is expected as more genes will be affected late due to secondary effects. Table 1 lists genes with fold change ≥2 at one or more time points (i.e., 15, 30 or 60 min), and similarly Table 2 list genes with fold change ≤0.5 (i.e., equal to or greater than 2 fold down regulated). A complete list with all expressed genes is available in Supplemental material 4. The chromosomal organisation of genes and operons of particular interest, and the position of associated OxyR boxes are shown in Fig. 4. The
Fig. 3. Functional distribution of CDSs with changes in expression levels above or below 1.5 fold between H2O2 treated samples and untreated samples. The number in parenthesis represents the percentage of the total number of CDSs within the genome in each functional class.
H.L. Pedersen et al. / Marine Genomics 3 (2010) 193–200 Table 1 Genes showing ≥2 fold up regulated mRNA expression compared to control after treatment with H2O2. The CDSs are divided into classes based on their function.
Table 1 (continued) Fold change CDS
Fold change CDS
Gene Product
Biosynthesis of cofactors, carriers VSAL_I0136 iucC Siderophore biosynthesis protein a Hybrid peroxiredoxin VSAL_I2713 (thioredoxin reductase) grxA Glutaredoxin 1 VSAL_I2030a VSAL_I1845 trxC Thioredoxin 2 VSAL_I0804 ybbN Thioredoxin Energy metabolism, carbon VSAL_I2201 fpr Ferredoxin–NADP reductase VSAL_II1079 nfsA NADPH-flavin oxidoreductase VSAL_II0687a zwf Glucose-6-phosphate 1dehydrogenase VSAL_II0686 pgl 6-phosphogluconolactonase VSAL_II0685 gnd 6-phosphogluconate dehydrogenase Protection responses VSAL_I2199 norV VSAL_II0511a VSAL_II0215
sodC katA
Anaerobic nitric oxide reductase flavorubredoxin Superoxide dismutase [Cu-Zn] Catalase
Laterally acquired elements VSAL_II0502 Putative bacteriophage integrase (fragment) VSAL_I2066 Group II intron reverse transcriptase/maturase Macromolecule degradation VSAL_I2740 hslV ATP-dependent protease HslV (heat shock protein HslV) VSAL_I2475 recN DNA repair protein RecN Membrane/exported/lipoproteins VSAL_II1068 Putative exported protein VSAL_II0512 Putative exported protein VSAL_I1830 Putative exported protein Not classified VSALI_2712
Dihydrolipoyl dehydrogenase (dihydrolipoamide dehydrogenase) Putative DNA-damageinducible protein
VSAL_p320_11
Transport/binding proteins VSAL_I0137 TonB-dependent ironsiderophore receptor VSAL_I2802 secB Protein-export protein SecB Unknown function VSAL_I2064a
15 min 30 min 60 min
Central intermediary metabolism VSAL_I2822 glmS Glucosamine–fructose-6phosphate aminotransferase
15 min 30 min 60 min 3.0
1.5
1.0
2.8
1.7
1.3
2.9
1.7
12.4
4.4
2.5
3.0 0.8 1.0
2.4 1.6 2.0
2.4 2.3 2.3
a An OxyR box has been predicted in the promoter region: VSAL_I2713: 22.0, VSAL_I2030: 21.3, VSAL_II0511: 7.9, VSAL_II0687: 16.3, VSAL_I2064: 13.0, VSAL_II0214: 13.6.
6.4 0.9 2.9
2.7 1.7 1.7
1.5 2.1 1.2
3.1 3.0
1.7 1.9
1.2 1.4
3.8
2.5
1.8
2.2 9.4
2.2 2.9
1.6 1.7
2.3
2.3
1.9
3.1
1.3
1.1
1.0
1.9
2.4
2.1
2.3
2.2
2.1 2.0 1.0
0.8 2.3 1.7
1.2 1.7 2.3
13.3
4.4
2.0
2.0
2.4
2.2
numbers of ≥2 fold up regulated genes are 28, 20 and 15 at time points 15, 30 and 60 min after H2O2 treatment, respectively, whereas the same numbers are 31, 13 and 72 for ≤0.5 fold down regulated genes. All significantly differentially expressed genes at 30 min, except one (i.e., VSAL_I2202 encoding a hypothetical protein), are also significantly differentially expressed after 15 min. In contrast, only four genes with fold change of ≥2 are common between 15 and 60 min samples. These genes include the recN (recombination and DNA repair), pVS320_0013 (DNA damage), VSAL_I2713 (thioredoxin reductase) and grxA (glutaredoxin 1). The tables contain genes that are known to be induced (e.g., grxA, trxC, ahpC, recN and secB) (Zheng et al., 2001; Chang et al., 2005; Stohl et al., 2005) or repressed (glpT) (Chang et al., 2005) in response to H2O2 in other bacteria, but they also contain genes that suggest a H2O2dependent regulation which is different from that in E. coli. Specifically, the gene zwf is regulated by SoxS in E. coli. The soxS gene is, however, not present in A. salmonicida, and instead the zwf promoter contains a strong OxyR box. Thus, zwf is likely under the direct regulation of OxyR. Moreover, the two most strongly H2O2 induced genes (i.e., VSAL_I2712 and VSAL_I2713) apparently constitute a two-gene operon, which is located adjacent to oxyR. The intergenic region between VSAL_I2713 and oxyR contains a strong OxyR box (21.04 Patser score) 134 nt upstream of the VSAL_I2713 start codon (see Fig. 4). It should also be mentioned that both katA and sodC are induced by H2O2 (9.4 and 2.2 fold at 15 min), and are both associated with putative OxyR boxes (Patser scores of 13.62 and 7.86, respectively) (see Fig. 4). Four genes with predicted OxyR binding sites were not detected on the microarray (probably due to technical reasons). Finally, further examination of Table 1 reveals that the majority of ≥2 fold up regulated genes encode proteins involved in detoxification or DNA protection and repair, are part of one of two reduction systems, or are involved in carbon metabolism and important for regenerating NADPH. These results are further discussed below.
3.4
1.8
1.3
1.0
1.5
2.1
4.1
2.1
1.3
1.2
1.9
2.2
2.5 –
1.4 2.2
1.2 1.9
Defined families, DeoR VSAL_I2823 srlR
Putative glucitol operon repressor
4. Discussion 4.1. Genes involved in detoxification and DNA protection
VSAL_I2065 VSAL_I2202 Fatty acid biosynthesis VSAL_II0214a
Patatin-like phospholipase
7.1
2.7
1.2
Adaptation VSAL_I2961
Small heat shock protein
1.1
1.0
2.2
2.7
1.2
1.2
0.9
1.7
2.1
ibpA
Gene Product
7.6
Conserved hypothetical protein Conserved hypothetical protein Hypothetical protein Hypothetical protein
VSAL_I0549
197
Amino acid biosynthesis VSAL_I0134 L-2,4-diaminobutyrate decarboxylase Cell processes, chaperones VSAL_I0018 groS 10 kDa chaperonin 1
(continued on next page)
The genome of A. salmonicida harbours genes that are important for detoxification of superoxide (sodB and sodC), hydrogen peroxide (katA) and alkyl hydroperoxides (ahpC). The A. salmonicida sodB and sodC probably encode cytoplasmic and periplasmic superoxide dismutases, respectively. Our computer prediction suggests that sodC and katA are associated with OxyR boxes (see Fig. 4), and this finding is supported by our microarray data which shows that these two genes are significantly up regulated 15 and 30 min after H2O2 treatment. This makes sense as sodB probably maintains a relatively stable level of superoxide dismutase activity inside the cell, whereas sodC might be responsible for responding to elevated levels of superoxide by neutralising O-2 in the periplasmic space. It is documented that H2O2 can induce damages to DNA (Miller and Britigan, 1997). A. salmonicida encodes a number of genes involved in DNA protection, recombination and repair, such as recA (multifunctional protein in recombination and SOS response), recBCD (recombination
198
H.L. Pedersen et al. / Marine Genomics 3 (2010) 193–200
Table 2 Genes showing ≥2 fold down regulated (≤0.5) mRNA expression compared to control after treatment with H2O2. The CDSs are divided into classes based on their function. Fold change CDS
Gene Product
Transport/binding protein VSAL_I0640 oadB Oxaloacetate decarboxylase, beta subunit VSAL_I0743 yhaO Inner membrane transport protein (pseudogene) VSAL_I1713 glpT Glycerol-3-phosphate transporter VSAL_I2279 ccmC Cytochrome c-type biogenesis protein CcmC VSAL_II0338 MFS transporter VSAL_II0384 glpF Glycerol uptake facilitator protein VSAL_II0817 putP Proline permease VSAL_II1002 Sodium/proton-dependent alanine carrier protein
15 min 30 min 60 min 0.9
0.7
0.5
0.4
–
0.8
0.4
0.5
0.5
0.7
0.7
0.5
0.4 0.3
0.5 0.3
0.4 0.3
0.5 1.0
0.7 0.8
0.5 0.5
0.7
0.7
0.5
0.4 1.0
0.9 0.7
0.9 0.5
0.9
0.8
0.5
0.3
0.4
0.4
0.4
0.5
0.6
0.3 0.5
0.3 0.7
0.4 0.8
Macromolecule degradation VSAL_II0039 malQ 4-alpha-glucanotransferase VSAL_I2667 degQ Exported serine protease
0.5 0.5
0.9 0.7
0.7 1.4
Membrane/exported/lipoprotein VSAL_I2545 yggE Putative exported protein VSAL_II0323 Putative lipoprotein
0.5 0.7
0.8 0.6
1.0 0.4
0.7
0.6
0.4
1.0
0.8
0.5
1.3
1.0
0.5
1.7
1.1
0.5
Energy metabolism, carbon VSAL_I2595 acnB Aconitate hydratase 2 (citrate hydro-lyase 2) VSAL_II0135 Putative cytochrome b561 VSAL_II0161 cyoC Cytochrome o ubiquinol oxidase subunit III VSAL_II0163 cyoA Cytochrome o ubiquinol oxidase subunit II Central intermediary metabolism VSAL_I1712 glpQ Glycerophosphoryl diester phosphodiesterase precursor VSAL_II0337 ugpQ Glycerophosphoryl diester phosphodiesterase VSAL_II0385 glpK Glycerol kinase VSAL_I0419 cpdB 2',3'-cyclic-nucleotide 2'phosphodiesterase
Nucleotide biosynthesis VSAL_I0738 guaA GMP synthase [glutaminehydrolyzing] VSAL_I0585 carA Carbamoyl-phosphate synthase small chain Sulphur metabolism VSAL_I0404 cysH VSAL_I0420
cysD
Phosphoadenosine phosphosulfate reductase Sulfate adenylyltransferase subunit 2
Biosynthesis of cofactors VSAL_II0859 bioF 8-amino-7-oxononanoate synthase
0.4
0.4
0.4
0.5
0.7
0.7
RNA polymerase, sigma factor VSAL_I2531 rpoE RNA polymerase sigma-E factor
0.4
0.8
0.9
Threonine biosynthesis VSAL_I2555 thrB Homoserine kinase
1.0
0.8
0.5
Degradation of DNA VSAL_I1364
Putative exported nuclease
and DNA repair), recFOR (recombination, DNA protection and repair), recG (helicase activity in recombination and repair), recN (recombination and DNA repair) and rmuC (recombination). Based on our computer predictions none of these genes is associated with OxyR boxes, and
Table 1 shows that only recN has mRNA levels ≥2 fold increase after H2O2 treatment. A majority of the remaining genes are up regulated typically between 1.7 and 1.2 folds. Despite the relatively modest increase in gene expression our microarray results indicate a general response to protect and repair DNA. 4.2. Genes in reducing systems Interestingly, a set of H2O2-induced genes in our microarray experiment belong to one of two reducing systems, i.e., the glutathione- and thioredoxin-dependent reduction systems. Both systems are responsible for maintaining a reduced cytoplasmic environment in E. coli by reducing disulfide bonds, and by converting alkyl peroxides to corresponding alcohols (Ritz and Beckwith, 2001; Carmel-Harel and Storz, 2000). In A. salmonicida, several genes that are part of these two reducing systems are found: gshA (glutamatecysteine ligase), gshB (subunit of glutathione synthetase), gor (subunit of glutathione reductase), grxA, grxB and grxC (glutaredoxins), trxA and trxC (thioredoxins), trxB (thioredoxin reductase), ahpC (subunit of alkylhydroperoxide reductase) and bcp (thioredoxin reductase). The earlier described OxyR box predictions suggest that at least ahpC and grxA are directly regulated by OxyR. Remarkably, these are indeed among the most up regulated genes 15 min after addition of H2O2 (Table 1). trxC, which is induced by OxyR in E. coli (Ritz et al., 2000) is also significantly up regulated, but only 60 min after treatment with H2O2. 4.3. Genes and proteins involved in regeneration of NADPH, and carbon metabolism Our microarray data shows that three significantly up regulated genes 15 min after H2O2 treatment encode the whole set of enzymes involved in the oxidative (and NADPH generating) stage of the pentose phosphate pathway (PPP). These genes are pgl (encodes 6phosphogluconolactonase), gnd (encodes 6-phosphogluconate dehydrogenase) and zwf (encodes glucose-6-phosphate 1-dehydrogenase). PPP's two main functions in the cell are to generate pentose phosphate for synthesis of nucleotides, and to supply the cell with NADPH. Furthermore, this pathway plays an important role in defence during oxidative stress (Pandolfi et al., 1995; Puskas et al., 2000; He et al., 2007; Lundberg et al., 1999). In the A. salmonicida genome zwf, pgl and gnd are organised in the same operon on chromosome II, and are associated with a strong OxyR box (Fig. 4). 4.4. Concluding remarks Pathogenic bacteria that invade their hosts will be challenged with reactive oxygen species originating from the immune system of the host. Our computational and experimental study shows that A. salmonicida can respond to such challenges by up regulating expression of molecules that can neutralize reactive oxygen species (catalase, peroxidase, glutaredoxin and thioredoxins). In total, OxyR boxes were predicted in front of 20 single genes and 8 operons, and six of the 28 genes and operons were recovered on the microarray list of most up regulated CDSs. This list contains genes that are known to be induced by H2O2 in other bacteria, but also suggests H2O2-dependent regulation that is different from that of e.g., E. coli. For example, zwf, which is regulated by SoxS in E. coli is apparently regulated by OxyR in A. salmonicida (the soxS gene is not present in A. salmonicida). By studying the global responses of this pathogen to stress conditions that are similar to those it faces within its host, it is our goal to gain increasingly more knowledge on the molecular mechanisms underlying its pathogenic lifestyle. Supplementary materials related to this article can be found online at doi:10.1016/j.margen.2010.10.002.
H.L. Pedersen et al. / Marine Genomics 3 (2010) 193–200
199
Fig. 4. Selected mRNA transcripts differentially regulated after treatment with H2O2, based on the microarray data analysis. Schematic drawing that shows the organisation of neighbouring genes in the A. salmonicida genome. Each arrow represents a CDS or an OxyR box and shows the direction of the gene in the genome. Blue represents genes significantly up regulated, while orange represents genes significantly down regulated when compared to control. In addition, yellow represents transcriptional regulators and black represents a putative OxyR box. The CDSs are numbered according to the genome annotation of A. salmonicida and the gene name is indicated when known. In bold are the Patser score of the OxyR box and the distance between the OxyR box and the transcription start site of the downstream gene.
Acknowledgement This work was supported by The University of Tromsø, the Norwegian Research Council and The National Programme for Research and
Functional Genomics in Norway (FUGE). We are grateful to The Microarray Resource Centre in Tromsø (MRCT) for offering facilities and equipments to run the microarray experiments. We wish to thank Christopher G. Fenton for help with initial microarray data analysis.
200
H.L. Pedersen et al. / Marine Genomics 3 (2010) 193–200
References Ahmad, R., Hjerde, E., Hansen, G.A., Haugen, P., Willassen, N.P., 2009. Prediction and experimental testing of ferric uptake regulator regulons in Vibrios. J. Mol. Microbiol. Biotechnol. 16, 159–168. Aslund, F., Zheng, M., Beckwith, J., Storz, G., 1999. Regulation of the OxyR transcription factor by hydrogen peroxide and the cellular thiol-disulfide status. Proc. Natl Acad. Sci. U. S. A. 96, 6161–6165. Bailey, T.L., Elkan, C., 1994. Fitting a mixture model by expectation maximization to discover motifs in biopolymers. Proc. Int. Conf. Intell. Syst. Mol. Biol. 2, 28–36. Carmel-Harel, O., Storz, G., 2000. Roles of the glutathione- and thioredoxin-dependent reduction systems in the Escherichia coli and Saccharomyces cerevisiae responses to oxidative stress. Annu. Rev. Microbiol. 54, 439–461. Chang, W., Small, D.A., Toghrol, F., Bentley, W.E., 2005. Microarray analysis of Pseudomonas aeruginosa reveals induction of pyocin genes in response to hydrogen peroxide. BMC Genomics 6, 115. Choi, H., Kim, S., Mukhopadhyay, P., Cho, S., Woo, J., Storz, G., Ryu, S., 2001. Structural basis of the redox switch in the OxyR transcription factor. Cell 105, 103–113. Christman, M.F., Morgan, R.W., Jacobson, F.S., Ames, B.N., 1985. Positive control of a regulon for defenses against oxidative stress and some heat-shock proteins in Salmonella typhimurium. Cell 41, 753–762. Compan, I., Touati, D., 1993. Interaction of six global transcription regulators in expression of manganese superoxide dismutase in Escherichia coli K-12. J. Bacteriol. 175, 1687–1696. Crooks, G.E., Hon, G., Chandonia, J.M., Brenner, S.E., 2004. WebLogo: a sequence logo generator. Genome Res. 14, 1188–1190. Diaz, P.I., Slakeski, N., Reynolds, E.C., Morona, R., Rogers, A.H., Kolenbrander, P.E., 2006. Role of oxyR in the oral anaerobe Porphyromonas gingivalis. J. Bacteriol. 188, 2454–2462. Dysvik, B., Jonassen, I., 2001. J-Express: exploring gene expression data using Java. Bioinformatics 17, 369–370. Gort, A.S., Ferber, D.M., Imlay, J.A., 1999. The regulation and role of the periplasmic copper, zinc superoxide dismutase of Escherichia coli. Mol. Microbiol. 32, 179–191. Hall, T.A., 1999. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symp. Ser. 41, 95–98. Han, M.J., Lee, S.Y., 2006. The Escherichia coli proteome: past, present, and future prospects. Microbiol. Mol. Biol. Rev. 70, 362–439. He, W., Wang, Y., Liu, W., Zhou, C.Z., 2007. Crystal structure of Saccharomyces cerevisiae 6-phosphogluconate dehydrogenase Gnd1. BMC Struct. Biol. 7, 38. Helmann, J.D., Wu, M.F., Gaballa, A., Kobel, P.A., Morshedi, M.M., Fawcett, P., Paddon, C., 2003. The global transcriptional response of Bacillus subtilis to peroxide stress is coordinated by three transcription factors. J. Bacteriol. 185, 243–253. Hertz, G.Z., Stormo, G.D., 1999. Identifying DNA and protein patterns with statistically significant alignments of multiple sequences. Bioinformatics 15, 563–577. Hishinuma, S., Yuki, M., Fujimura, M., Fukumori, F., 2006. OxyR regulated the expression of two major catalases, KatA and KatB, along with peroxiredoxin, AhpC in Pseudomonas putida. Environ. Microbiol. 8, 2115–2124. Hjerde, E., Lorentzen, M.S., Holden, M.T., Seeger, K., Paulsen, S., Bason, N., Churcher, C., Harris, D., Norbertczak, H., Quail, M.A., Sanders, S., Thurston, S., Parkhill, J., Willassen, N.P., Thomson, N.R., 2008. The genome sequence of the fish pathogen Aliivibrio salmonicida strain LFI1238 shows extensive evidence of gene decay. BMC Genomics 9, 616. Hopkin, K.A., Papazian, M.A., Steinman, H.M., 1992. Functional differences between manganese and iron superoxide dismutases in Escherichia coli K-12. J. Biol. Chem. 267, 24253–24258. Horsburgh, M.J., Clements, M.O., Crossley, H., Ingham, E., Foster, S.J., 2001a. PerR controls oxidative stress resistance and iron storage proteins and is required for virulence in Staphylococcus aureus. Infect. Immun. 69, 3744–3754. Horsburgh, M.J., Ingham, E., Foster, S.J., 2001b. In Staphylococcus aureus, fur is an interactive regulator with PerR, contributes to virulence, and Is necessary for oxidative stress resistance through positive regulation of catalase and iron homeostasis. J. Bacteriol. 183, 468–475. Imlay, J.A., 2008. Cellular defenses against superoxide and hydrogen peroxide. Annu. Rev. Biochem. 77, 755–776. Jair, K.W., Martin, R.G., Rosner, J.L., Fujita, N., Ishihama, A., Wolf Jr., R.E., 1995. Purification and regulatory properties of MarA protein, a transcriptional activator of Escherichia coli multiple antibiotic and superoxide resistance promoters. J. Bacteriol. 177, 7100–7104. Lau, G.W., Britigan, B.E., Hassett, D.J., 2005. Pseudomonas aeruginosa OxyR is required for full virulence in rodent and insect models of infection and for resistance to human neutrophils. Infect. Immun. 73, 2550–2553. Li, H., Singh, A.K., McIntyre, L.M., Sherman, L.A., 2004. Differential gene expression in response to hydrogen peroxide and the putative PerR regulon of Synechocystis sp. strain PCC 6803. J. Bacteriol. 186, 3331–3345.
Loewen, P.C., Hu, B., Strutinsky, J., Sparling, R., 1998. Regulation in the rpoS regulon of Escherichia coli. Can. J. Microbiol. 44, 707–717. Lu, S., Killoran, P.B., Fang, F.C., Riley, L.W., 2002. The global regulator ArcA controls resistance to reactive nitrogen and oxygen intermediates in Salmonella enterica serovar Enteritidis. Infect. Immun. 70, 451–461. Lundberg, B.E., Wolf Jr., R.E., Dinauer, M.C., Xu, Y., Fang, F.C., 1999. Glucose 6-phosphate dehydrogenase is required for Salmonella typhimurium virulence and resistance to reactive oxygen and nitrogen intermediates. Infect. Immun. 67, 436–438. Meyer, F., Goesmann, A., McHardy, A.C., Bartels, D., Bekel, T., Clausen, J., Kalinowski, J., Linke, B., Rupp, O., Giegerich, R., Pühler, A., 2003. GenDB — an open source genome annotation system for prokaryote genomes. Nucleic Acids Res. 31, 2187–2195. Miller, R.A., Britigan, B.E., 1997. Role of oxidants in microbial pathophysiology. Clin. Microbiol. Rev. 10, 1–18. Morgan, R.W., Christman, M.F., Jacobson, F.S., Storz, G., Ames, B.N., 1986. Hydrogen peroxide-inducible proteins in Salmonella typhimurium overlap with heat shock and other stress proteins. Proc. Natl Acad. Sci. U. S. A. 83, 8059–8063. Ochsner, U.A., Vasil, M.L., Alsabbagh, E., Parvatiyar, K., Hassett, D.J., 2000. Role of the Pseudomonas aeruginosa oxyR-recG operon in oxidative stress defense and DNA repair: OxyR-dependent regulation of katB-ankB, ahpB, and ahpC-ahpF. J. Bacteriol. 182, 4533–4544. Palma, M., DeLuca, D., Worgall, S., Quadri, L.E., 2004. Transcriptome analysis of the response of Pseudomonas aeruginosa to hydrogen peroxide. J. Bacteriol. 186, 248–252. Pandolfi, P.P., Sonati, F., Rivi, R., Mason, P., Grosveld, F., Luzzatto, L., 1995. Targeted disruption of the housekeeping gene encoding glucose 6-phosphate dehydrogenase (G6PD): G6PD is dispensable for pentose synthesis but essential for defense against oxidative stress. EMBO J. 14, 5209–5215. Puskas, F., Gergely Jr., P., Banki, K., Perl, A., 2000. Stimulation of the pentose phosphate pathway and glutathione levels by dehydroascorbate, the oxidized form of vitamin C. FASEB J. 14, 1352–1361. Ritz, D., Beckwith, J., 2001. Roles of thiol-redox pathways in bacteria. Annu. Rev. Microbiol. 55, 21–48. Ritz, D., Patel, H., Doan, B., Zheng, M., Aslund, F., Storz, G., Beckwith, J., 2000. Thioredoxin 2 is involved in the oxidative stress response in Escherichia coli. J. Biol. Chem. 275, 2505–2512. Ruiz-Laguna, J., Prieto-Alamo, M.J., Pueyo, C., 2000. Oxidative mutagenesis in Escherichia coli strains lacking ROS-scavenging enzymes and/or 8-oxoguanine defenses. Environ. Mol. Mutagen. 35, 22–30. Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Steinman, H.M., Weinstein, L., Brenowitz, M., 1994. The manganese superoxide dismutase of Escherichia coli K-12 associates with DNA. J. Biol. Chem. 269, 28629–28634. Stohl, E.A., Criss, A.K., Seifert, H.S., 2005. The transcriptome response of Neisseria gonorrhoeae to hydrogen peroxide reveals genes with previously uncharacterized roles in oxidative damage protection. Mol. Microbiol. 58, 520–532. Storz, G., Tartaglia, L.A., 1992. OxyR: a regulator of antioxidant genes. J. Nutr. 122, 627–630. Sund, C.J., Rocha, E.R., Tzianabos, A.O., Wells, W.G., Gee, J.M., Reott, M.A., O'Rourke, D.P., Smith, C.J., 2008. The Bacteroides fragilis transcriptome response to oxygen and H2O2: the role of OxyR and its effect on survival and virulence. Mol. Microbiol. 67, 129–142. Toledano, M.B., Kullik, I., Trinh, F., Baird, P.T., Schneider, T.D., Storz, G., 1994. Redoxdependent shift of OxyR–DNA contacts along an extended DNA-binding site: a mechanism for differential promoter selection. Cell 78, 897–909. VanBogelen, R.A., Kelley, P.M., Neidhardt, F.C., 1987. Differential induction of heat shock, SOS, and oxidation stress regulons and accumulation of nucleotides in Escherichia coli. J. Bacteriol. 169, 26–32. Wong, S.M., Alugupalli, K.R., Ram, S., Akerley, B.J., 2007. The ArcA regulon and oxidative stress resistance in Haemophilus influenzae. Mol. Microbiol. 64, 1375–1390. Zeller, T., Moskvin, O.V., Li, K., Klug, G., Gomelsky, M., 2005. Transcriptome and physiological responses to hydrogen peroxide of the facultatively phototrophic bacterium Rhodobacter sphaeroides. J. Bacteriol. 187, 7232–7242. Zeller, T., Mraheil, M.A., Moskvin, O.V., Li, K., Gomelsky, M., Klug, G., 2007. Regulation of hydrogen peroxide-dependent gene expression in Rhodobacter sphaeroides: regulatory functions of OxyR. J. Bacteriol. 189, 3784–3792. Zheng, M., Aslund, F., Storz, G., 1998. Activation of the OxyR transcription factor by reversible disulfide bond formation. Science 279, 1718–1721. Zheng, M., Doan, B., Schneider, T.D., Storz, G., 1999. OxyR and SoxRS regulation of fur. J. Bacteriol. 181, 4639–4643. Zheng, M., Wang, X., Templeton, L.J., Smulski, D.R., LaRossa, R.A., Storz, G., 2001. DNA microarray-mediated transcriptional profiling of the Escherichia coli response to hydrogen peroxide. J. Bacteriol. 183, 4562–4570.
Contents lists available at ScienceDirect
Marine Genomics j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / m a r g e n
Global responses of Aliivibrio salmonicida to hydrogen peroxide as revealed by microarray analysis Hege L. Pedersen a, Erik Hjerde a, Steinar M. Paulsen a,1, Hilde Hansen a, Lotte Olsen a,b, Sunniva K. Thode a, Marcos T. Dos Santos a, Ruth H. Paulssen b, Nils-Peder Willassen a,c, Peik Haugen a,c,⁎ a b c
Department of Chemistry, University of Tromsø, N-9037 Tromsø, Norway Institute of Clinical Medicine, University of Tromsø, N-9037 Tromsø, Norway The Norwegian Structural Biology Centre, University of Tromsø, N-9037 Tromsø, Norway
a r t i c l e
i n f o
Article history: Received 29 January 2010 Received in revised form 4 October 2010 Accepted 5 October 2010 Keywords: Aliivibrio salmonicida Hydrogen peroxide Oxidative stress Microarray Hitra disease
a b s t r a c t Aliivibrio salmonicida causes “cold-water vibriosis” (or “Hitra disease”) in fish, including marine-reared Atlantic salmon. During development of the disease the bacterium will encounter macrophages with antibacterial activities such as production of damaging reactive oxygen species (ROS). To defend itself the bacterium will presumably start producing detoxifying enzymes, reducing agents, and proteins involved in DNA and protein repair systems. Even though responses to oxidative stress are well studied for a few model bacteria, little work has been done in general to explain how important groups of pathogens, like members of the Vibrionaceae family, can survive at high levels of ROS. We have used bioinformatic tools and microarray to study how A. salmonicida responds to hydrogen peroxide (H2O2). First, we used the recently published genome sequence to predict potential binding sites for OxyR (H2O2 response regulator). The computer-based search identified OxyR sites associated with 20 single genes and 8 operons, and these predictions were compared to experimental data from Northern blot analysis and microarray analysis. In general, OxyR binding site predictions and experimental results are in agreement. Up- and down regulated genes are distributed among all functional gene categories, but a striking number of ≥ 2 fold up regulated genes encode proteins involved in detoxification and DNA repair, are part of reduction systems, or are involved in carbon metabolism and regeneration of NADPH. Our predictions and –omics data corroborates well with findings from other model bacteria, but also suggest species-specific gene regulation. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Damaging reactive oxygen species (ROS) form inside cells as byproducts from normal cellular processes, they may be produced actively by macrophages as part of the innate immune system, and they are sometimes produced by bacteria in the competition for survival (Imlay, 2008). Bacteria respond to ROS by producing elevated levels of detoxifying enzymes, reduction systems, and proteins involved in protection and repair of DNA. In Escherichia coli and other Gramnegative bacteria the presence of hydrogen peroxide (H2O2) is for example counteracted by the production of catalases (Storz and Tartaglia, 1992), i.e. enzymes that catalyses the decomposition of ⁎ Corresponding author. Department of Chemistry, University of Tromsø, Norway. Tel.: +47 776 5288; fax: +47 776 5350. E-mail addresses: [email protected] (H.L. Pedersen), [email protected] (E. Hjerde), [email protected] (S.M. Paulsen), [email protected] (H. Hansen), [email protected] (L. Olsen), [email protected] (S.K. Thode), [email protected] (M.T.D. Santos), [email protected] (R.H. Paulssen), [email protected] (N.-P. Willassen), [email protected] (P. Haugen). 1 Current address: MABCENT-SFI, Centre on Marine bioactives and drug discovery, University of Tromsø, Tromsø, Norway. 1874-7787/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.margen.2010.10.002
H2O2 to water and oxygen. The total response to H2O2 is however more complicated, and for E. coli and a few other well-studied bacteria, global changes in the transcriptome and proteome have been investigated by microarray (Zheng et al., 2001; Li et al., 2004; Palma et al., 2004; Chang et al., 2005; Stohl et al., 2005; Zeller et al., 2005, 2007; Diaz et al., 2006; Sund et al., 2008) and 2D gel electrophoresis (Morgan et al., 1986; VanBogelen et al., 1987; reviewed by Han and Lee, 2006), respectively. These and other analyses show that the global regulator OxyR has a key role in conferring resistance to H2O2 (Zheng et al., 1998, 2001; Aslund et al., 1999), although other systems such as PerR (Horsburgh et al., 2001a,b; Helmann et al., 2003), σs (Loewen et al., 1998; Gort et al., 1999), Fur (Zheng et al., 1999), Fnr (Gort et al., 1999), MarAB (Jair et al., 1995) and ArcA (Lu et al., 2002; Wong et al., 2007) may also be involved. OxyR knock-out mutants are significantly more sensitive to H2O2 exposure and are less likely to cause infections (Sund et al., 2008; Lau et al., 2005). In the presence of H2O2 the E. coli OxyR is activated by formation of an intra-molecular disulfide bridge between cysteines 199 and 208 (Zheng et al., 1998; Choi et al., 2001). The E. coli OxyR dimers (or tetramers) bind to promoters at degenerated binding sites (ATAG/CTAT boxes) located in four adjacent major grooves (Toledano et al., 1994), and typically activates transcription of the associated genes (Zheng
194
H.L. Pedersen et al. / Marine Genomics 3 (2010) 193–200
et al., 2001). Some of the best studied examples of OxyR up regulated genes in E. coli are dps, which encodes a nonspecific DNA binding protein, katG, which encodes a catalase and oxyS which encodes a regulatory small RNA. The Gram-negative bacterium Aliivibrio salmonicida causes disease in farmed Atlantic salmon (Salmo salar), Atlantic cod (Gadus morhua) and rainbow trout (Oncorhynchus mykiss). During the 80s, the bacterium was directly responsible for severe economic losses in the fish-farming industry, but is today kept under control by vaccinating juvenile fish prior to their release into marine fish farms. The complete genome of A. salmonicida strain LFI1238 was recently published (Hjerde et al., 2008): it consists of one large and one small circular chromosome of 3.3 and 1.2 Mb, respectively, (typical for species that belong to the Vibrionaceae family) and four plasmids. A striking feature of the genome is the mosaic composition that apparently originates from high numbers of intra-chromosomal rearrangements, deletions and duplications, and mobile genetic elements (i.e., insertion sequences, genomic islands and prophages). Interestingly, A. salmonicida lacks several genes that play important roles in oxidative stress responses in E. coli (e.g., soxS, katG, katE, ahpF, and dps). This observation inspired us to investigate the global responses of A. salmonicida to H2O2. A computational approach was used to predict OxyR binding sites, and microarray was subsequently used to study changes in the transcriptome. Time-dependent responses were measured by harvesting samples 15, 30 and 60 min after addition of H2O2. Our results reveal responses that are well-known from the literature from studies using other bacteria, but also show new solutions to gene regulation in oxidative stress responses. 2. Materials and methods 2.1. Bacterial strains, culture conditions and sampling A. salmonicida strain LFI1238 was grown in Luria-Bertani (LB) medium (Sambrook et al., 1989) with 2.5% NaCl at 15 °C and 220 rpm. For stimulation experiments, cells were grown to mid-exponential phase (OD600nm = 0.5), split in two halves and 100 μM or 1-5 mM H2O2 (Sigma Aldrich) was added to one of the cultures (see main text). Cells were harvested at 5, 30 and 60 min for the initial Northern blots analysis, and at 15, 30 and 60 min for the microarray analysis. Three independent experiments were done for the microarray analysis. RNA extraction is described below. 2.2. Northern blot analysis The Northern blot analysis was performed essentially as previously described (Ahmad et al., 2009). Briefly, total RNA was isolated from bacteria using the Trizol protocol. Ten μg RNA was separated on 1.2% denaturing formamide agarose gels, and run for 2 min at 20 V, 10 min at 40 V and finally at 60 V until the dye reached the end of the gel. RNA was next transferred to a Hybond-N + nylon membrane (Amersham) by capillary transfer. For Northern analysis [α-32P]-labeled dsDNA was used as probes according to the Northern Max TM instruction manual (Ambion). Hybridisations were performed over-night at 42 °C and signals were acquired on phosphoimaging screens (Fujifilm) and scanned using a BAS-5000 phosphoimager (Fujifilm). Quantification of signals was done using the ImageGauge software v4.0 (Fujifilm). 2.3. Microarray analysis For microarray analysis RNA was extracted from cells using the RNA-isol reagent (Fischer scientific), and DNA was removed using the DNA-free kit (Ambion). Any traces of DNase were removed with the RNeasy Minelute Cleanup kit (Qiagen). RNA quality was checked with a Bioanalyzer, Nanodrop, and by incubating the total RNA for 1 h at 37 °C with 500 ng plasmid DNA and then running the sample on a 1%
agarose gel (i.e., to check for any remaining DNase activity). cDNA was generated and labelled by using 15 μg RNA, the Aminoallyl cDNA labeling kit (Ambion), and the CyDye™ Post-Labeling Reactive Dye Pack (GE Healthcare). Cy3 and Cy5 were used as dyes. Samples were hybridised to “Vibrio salmonicida V1.0.1 AROS” slides at 42 °C on a TECAN HS4800 hybridisation station, and microarray slides were subsequently washed, once in 0.1 × SSC/0.1% SDS for 5 min at 42 °C, then once in 0.1 × SSC/0.1% SDS for 10 min at room temperature, and finally four times in 0.1 × SSC for 1 min at room temperature. Slides were scanned using a GenePix 4000B scanner (Axon Instruments Inc.). Images were explored using the GenePix Pro v6.1 software and analysis of expression data was done using J-Express Pro v2.7 (Dysvik and Jonassen, 2001). Background was subtracted after quantification using a global background subtraction method. The absolute pixel values for the spots of each gene were averaged. The ratio between the average intensity signals from three H2O2 treated samples and three untreated samples represented the fold change in gene expression. All experiments were mean normalised based on expression ratio data. Microarray data has been uploaded to The NCBI Gene Expression Omnibus (GEO) database, and is available through accession number GSE20082. 2.4. Computer based prediction of OxyR binding sites To predict OxyR binding sites in the A. salmonicida genome we first compiled sequences of E. coli promoters that contain experimentally verified OxyR boxes, or are assumed to contain OxyR boxes (i.e. gene expression is OxyR-dependent). The extracted sequences were 250 nt upstream of the translation start codon. Extracted promoter regions were associated with the following genes: ahpC, dps, dsbG, fhuF, flu, fur, gor, grxA, hemH, katG, oxyR, oxyS, rcsC, trxC, sufA, uxuA, ybjC, yhjA and ygaQ. Sequences were extracted from the evolutionary related gamma-proteobacteria Vibrio and Aliivibrio genomes (204 sequences). The complete genome of nine Vibrio and three Aliivibrio species, and genome draft sequences of seven Vibrio species were extracted from the NCBI ftp site (ftp://ftp.ncbi.nlm.nih.gov). Draft sequences were automatically annotated using the GenDB pipeline (Meyer et al., 2003). The sequence dataset and the motif discovery tool MEME (Bailey and Elkan, 1994) were used to search for conserved motifs using the parameters set to search for motifs 30 nt or more in length. Manual inspection of the result revealed 41 nt OxyR box-like motifs in 99 sequences (Supplemental material 1). Based on the 99 sequences, an alignment matrix (Supplemental material 2) which describes the frequency of each base at each of the 41 positions, was built using BioEdit v.7.0.9 (Hall, 1999). In addition, an OxyR box consensus was made with WebLogo (Crooks et al., 2004). Using the Patser program (Hertz and Stormo, 1999) and the Vibrio/Aliivibrio adjusted OxyR box alignment matrix, the genome of A. salmonicida was searched for potential OxyR binding sites. OxyR box searches were performed within the region -250 to + 1 relative to the translational start codon of annotated genes. The lower Patser cut-off score was set at 7.0 whereas all other Patser parameters were set to default. 3. Results 3.1. Computer based prediction of OxyR binding sites in A. salmonicida In this study we set out to globally predict OxyR binding sites in the recently published A. salmonicida genome (Hjerde et al., 2008). E. coli and A. salmonicida OxyR proteins share overall 65% identity and 88% similarity, and 77% identity and 100% similarity in the conserved DNA-binding domain. We therefore found it reasonable to believe that OxyR–DNA interactions are conserved in the two gammaproteobacteria. Based on our assumption, we searched the literature for E. coli genes with OxyR-dependent expression patterns and
H.L. Pedersen et al. / Marine Genomics 3 (2010) 193–200
195
Fig. 1. OxyR-box consensus sequence based on experimentally verified and proposed OxyR-binding sites of known OxyR-regulated genes in E. coli and homologous genes in Vibrio and Aliivibrio species. (A) WebLogo was used to build a consensus sequence logo, in which the height of individual letters within a stack of letters represents the relative frequency of that letter at a given position, and the overall height of the stack represents the degree of conservation at that position. (B) Model of oxidised OxyR binding to DNA based on the study of Toledano et al. (1994).
experimentally verified OxyR binding sites. A total of 19 genes were selected for further analysis (see Materials and methods), and 250 nt from the corresponding promoters were extracted from GenBank and aligned. Homologous promoters from 16 Vibrio and three Aliivibrio species were added to the alignment. The dataset, which contained 222 sequences, was next used as input in MEME analysis to search for conserved motifs. MEME identified a 41 nt conserved motif with a distinct periodicity of 4 nt (ATAG or CTAT) separated by seven degenerate nt in 99 of the 222 sequences. Seven of the 19 OxyR boxes from E. coli promoters were recovered in the final dataset, which show that even this relatively relaxed motif can be recovered by our search method. Finally, we generated an alignment matrix based on the 99 OxyR motifs, and used this to generate a WebLogo (Fig. 1), which describes the frequency of each base at each position. This motif is in agreement with the OxyR recognition motif (5’-ATAGntnnnanCTATnnnnnnnATAGntnnnanCTAT) suggested by Toledano et al. (1994). Also, the alignment matrix was used as input in Patser to search the A. salmonicida genome for OxyR binding sites. OxyR binding sites were predicted in front of 20 protein coding sequences (CDSs) and 8 operons in A. salmonicida (provided in Supplementary material 3). OxyR boxes with Patser scores above 10 (high confidence values) were found upstream of a phosphorylase (VSAL_I1055), two conserved hypothetical proteins (VSAL_I1295 and VSAL_2547), a ribosomal protein (VSAL_I2837), and a membrane protein (VSAL_II0074), and in front of oxyR, ahpC and grxA, which were included in the initial alignment to construct the alignment matrix. Also, high confidence OxyR boxes were found associated with three operons, i.e., an operon containing two CDSs annotated as hybrid peroxiredoxin and dihydrolipoyl dehydrogenase (VSAL_I2713-12), an operon containing three CDSs involved in the degradation of carbohydrates (VSAL_II0687-85), and an operon containing two CDSs encoding a patatin-like phospholipase and catalase (VSAL_II0214-13). Three of the putative target genes were either pseudogenes or gene remnants and were not included in the results. For known OxyR-activated genes, OxyR binds to the DNA upstream of the -35 promoter of target genes (Toledano et al., 1994). In A. salmonicida all OxyR boxes were found in the proximity of the -35 promoter with a distance of between 11 and 201 nt from the translational start codon of the gene.
exponential phase, 1–5 mM H2O2 was added and growth was subsequently monitored for 300 min. The result is presented in Fig. 2A and shows that at these concentrations cellular growth was inhibited, but not completely stopped. Five mM H2O2 was chosen for further gene expression analyses. To evaluate if relevant genes had been activated, bacteria were sampled 5, 30 and 60 min after stimulation, and total RNA from cells was subjected to Northern blot analyses. Expression of ahpC was monitored because it is associated
3.2. Expression of sodC and ahpC increases in response to H2O2 To identify nonlethal and appropriate doses of H2O2 for studying oxidative stress responses in A. salmonicida, the bacterium was cultivated in the presence of increasing concentrations of H2O2. We wanted to find conditions that would trigger relevant cellular H2O2 responses, but at the same time would not over-stimulate the cells (over-stimulation activates/represses non-relevant genes). We considered H2O2 concentrations with minor inhibitory effects on cell growth as desired stress conditions. The bacterium was grown to
Fig. 2. Gene expression profiles in response to hydrogen peroxide. Growth curve of A. salmonicida cultured to mid-exponential phase (OD600 = 0.6), split and added H2O2 (A). H2O2 was added at different concentrations (A, enlarged), except for the control, which was unchanged. The concentrations of H2O2 were 1 mM, 5 mM, 8 mM and 10 mM. Northern blots show the expression of different mRNA transcripts in response to H2O2 (B). All the CDSs were normalised to the 16 S ribosomal RNA signal (endogenous control) and compared with the control. Samples were collected 0, 5, 30 and 60 min after treatment with 5 mM H2O2.
196
H.L. Pedersen et al. / Marine Genomics 3 (2010) 193–200
with an OxyR box in our OxyR predictions, and is activated by OxyR upon H2O2 stimulation in E. coli, Pseudomonas putida and Pseudomonas aeruginosa (Christman et al., 1985; Ochsner et al., 2000; Hishinuma et al., 2006). In addition, we examined the expression of sodC, sodB, fur and rpoS, which are also implicated in oxidative stress response in E. coli (Hopkin et al., 1992; Compan and Touati, 1993; Steinman et al., 1994; Gort et al., 1999; Ruiz-Laguna et al., 2000). Of these, only sodC is associated with an OxyR box in A. salmonicida. Five min after addition of H2O2 (Fig. 2B) the levels of ahpC and sodC mRNAs are significantly induced, whereas the levels of sodB and fur mRNAs apparently are reduced. In summary, we conclude that 5 mM H2O2 slightly reduces growth of A. salmonicida and increases expression of sodC and ahpC. This indicates that the growth conditions used in our study are adequate for studying H2O2 responses. 3.3. Microarray analysis reveals H2O2-dependent gene expression of defence molecules, dehydrogenases and reductases Next, microarray analysis was used to investigate global changes in the A. salmonicida transcriptome following exposure to H2O2. A new batch of H2O2 was used for this experiment, and therefore the added concentration of H2O2 was re-calibrated by repeating the experiment described in Fig. 2A. In this experiment, 0.1 mM H2O2 induced appropriate reduction in cellular growth. H2O2 is unstable and spontaneously decomposes to water and oxygen gas, which explains why 5 mM H2O2 resulted in a similar reduction in growth in the initial experiment (Fig. 2A) as described above. In three independent experiments, A. salmonicida was grown in LB medium with 2.5%
NaCl to OD600nm = 0.6 (i.e., approx. mid-exponential growth phase) and then subjected to 0.1 mM H2O2. Samples were collected 15, 30 and 60 min after treatment. Untreated cells collected at same time points were used as control. Total RNA from the three cultures (biological replicates) were pooled and subjected to cDNA synthesis (described in Materials and methods), and run on three 70-mer-based whole genome microarray slides (technical replicates). Results were analysed with GenePix and J-Express. First we wanted to investigate how the differentially expressed genes were distributed into functional classes, and for this purpose we took into consideration genes with fold change of ≥1.5 or ≤0.66 (i.e., 0.66 corresponds to 1.5 fold down regulation). Fig. 3 shows a graphical presentation of the functional classes and the number of differentially expressed genes 15 and 60 min after H2O2 treatment. In general, the distribution of genes in 30 min samples was similar to that of 15 min, and was omitted from the figure for simplicity. Two main conclusions can be drawn from the analysis. First, affected genes are distributed into all classes of genes. Second, significantly more genes are differentially expressed at 60 min than at 15 min. Specifically, 61 and 120 genes are up regulated at 15 and 60 min, respectively, and 75 and 151 genes are down regulated at 15 and 60 min, respectively. This finding is expected as more genes will be affected late due to secondary effects. Table 1 lists genes with fold change ≥2 at one or more time points (i.e., 15, 30 or 60 min), and similarly Table 2 list genes with fold change ≤0.5 (i.e., equal to or greater than 2 fold down regulated). A complete list with all expressed genes is available in Supplemental material 4. The chromosomal organisation of genes and operons of particular interest, and the position of associated OxyR boxes are shown in Fig. 4. The
Fig. 3. Functional distribution of CDSs with changes in expression levels above or below 1.5 fold between H2O2 treated samples and untreated samples. The number in parenthesis represents the percentage of the total number of CDSs within the genome in each functional class.
H.L. Pedersen et al. / Marine Genomics 3 (2010) 193–200 Table 1 Genes showing ≥2 fold up regulated mRNA expression compared to control after treatment with H2O2. The CDSs are divided into classes based on their function.
Table 1 (continued) Fold change CDS
Fold change CDS
Gene Product
Biosynthesis of cofactors, carriers VSAL_I0136 iucC Siderophore biosynthesis protein a Hybrid peroxiredoxin VSAL_I2713 (thioredoxin reductase) grxA Glutaredoxin 1 VSAL_I2030a VSAL_I1845 trxC Thioredoxin 2 VSAL_I0804 ybbN Thioredoxin Energy metabolism, carbon VSAL_I2201 fpr Ferredoxin–NADP reductase VSAL_II1079 nfsA NADPH-flavin oxidoreductase VSAL_II0687a zwf Glucose-6-phosphate 1dehydrogenase VSAL_II0686 pgl 6-phosphogluconolactonase VSAL_II0685 gnd 6-phosphogluconate dehydrogenase Protection responses VSAL_I2199 norV VSAL_II0511a VSAL_II0215
sodC katA
Anaerobic nitric oxide reductase flavorubredoxin Superoxide dismutase [Cu-Zn] Catalase
Laterally acquired elements VSAL_II0502 Putative bacteriophage integrase (fragment) VSAL_I2066 Group II intron reverse transcriptase/maturase Macromolecule degradation VSAL_I2740 hslV ATP-dependent protease HslV (heat shock protein HslV) VSAL_I2475 recN DNA repair protein RecN Membrane/exported/lipoproteins VSAL_II1068 Putative exported protein VSAL_II0512 Putative exported protein VSAL_I1830 Putative exported protein Not classified VSALI_2712
Dihydrolipoyl dehydrogenase (dihydrolipoamide dehydrogenase) Putative DNA-damageinducible protein
VSAL_p320_11
Transport/binding proteins VSAL_I0137 TonB-dependent ironsiderophore receptor VSAL_I2802 secB Protein-export protein SecB Unknown function VSAL_I2064a
15 min 30 min 60 min
Central intermediary metabolism VSAL_I2822 glmS Glucosamine–fructose-6phosphate aminotransferase
15 min 30 min 60 min 3.0
1.5
1.0
2.8
1.7
1.3
2.9
1.7
12.4
4.4
2.5
3.0 0.8 1.0
2.4 1.6 2.0
2.4 2.3 2.3
a An OxyR box has been predicted in the promoter region: VSAL_I2713: 22.0, VSAL_I2030: 21.3, VSAL_II0511: 7.9, VSAL_II0687: 16.3, VSAL_I2064: 13.0, VSAL_II0214: 13.6.
6.4 0.9 2.9
2.7 1.7 1.7
1.5 2.1 1.2
3.1 3.0
1.7 1.9
1.2 1.4
3.8
2.5
1.8
2.2 9.4
2.2 2.9
1.6 1.7
2.3
2.3
1.9
3.1
1.3
1.1
1.0
1.9
2.4
2.1
2.3
2.2
2.1 2.0 1.0
0.8 2.3 1.7
1.2 1.7 2.3
13.3
4.4
2.0
2.0
2.4
2.2
numbers of ≥2 fold up regulated genes are 28, 20 and 15 at time points 15, 30 and 60 min after H2O2 treatment, respectively, whereas the same numbers are 31, 13 and 72 for ≤0.5 fold down regulated genes. All significantly differentially expressed genes at 30 min, except one (i.e., VSAL_I2202 encoding a hypothetical protein), are also significantly differentially expressed after 15 min. In contrast, only four genes with fold change of ≥2 are common between 15 and 60 min samples. These genes include the recN (recombination and DNA repair), pVS320_0013 (DNA damage), VSAL_I2713 (thioredoxin reductase) and grxA (glutaredoxin 1). The tables contain genes that are known to be induced (e.g., grxA, trxC, ahpC, recN and secB) (Zheng et al., 2001; Chang et al., 2005; Stohl et al., 2005) or repressed (glpT) (Chang et al., 2005) in response to H2O2 in other bacteria, but they also contain genes that suggest a H2O2dependent regulation which is different from that in E. coli. Specifically, the gene zwf is regulated by SoxS in E. coli. The soxS gene is, however, not present in A. salmonicida, and instead the zwf promoter contains a strong OxyR box. Thus, zwf is likely under the direct regulation of OxyR. Moreover, the two most strongly H2O2 induced genes (i.e., VSAL_I2712 and VSAL_I2713) apparently constitute a two-gene operon, which is located adjacent to oxyR. The intergenic region between VSAL_I2713 and oxyR contains a strong OxyR box (21.04 Patser score) 134 nt upstream of the VSAL_I2713 start codon (see Fig. 4). It should also be mentioned that both katA and sodC are induced by H2O2 (9.4 and 2.2 fold at 15 min), and are both associated with putative OxyR boxes (Patser scores of 13.62 and 7.86, respectively) (see Fig. 4). Four genes with predicted OxyR binding sites were not detected on the microarray (probably due to technical reasons). Finally, further examination of Table 1 reveals that the majority of ≥2 fold up regulated genes encode proteins involved in detoxification or DNA protection and repair, are part of one of two reduction systems, or are involved in carbon metabolism and important for regenerating NADPH. These results are further discussed below.
3.4
1.8
1.3
1.0
1.5
2.1
4.1
2.1
1.3
1.2
1.9
2.2
2.5 –
1.4 2.2
1.2 1.9
Defined families, DeoR VSAL_I2823 srlR
Putative glucitol operon repressor
4. Discussion 4.1. Genes involved in detoxification and DNA protection
VSAL_I2065 VSAL_I2202 Fatty acid biosynthesis VSAL_II0214a
Patatin-like phospholipase
7.1
2.7
1.2
Adaptation VSAL_I2961
Small heat shock protein
1.1
1.0
2.2
2.7
1.2
1.2
0.9
1.7
2.1
ibpA
Gene Product
7.6
Conserved hypothetical protein Conserved hypothetical protein Hypothetical protein Hypothetical protein
VSAL_I0549
197
Amino acid biosynthesis VSAL_I0134 L-2,4-diaminobutyrate decarboxylase Cell processes, chaperones VSAL_I0018 groS 10 kDa chaperonin 1
(continued on next page)
The genome of A. salmonicida harbours genes that are important for detoxification of superoxide (sodB and sodC), hydrogen peroxide (katA) and alkyl hydroperoxides (ahpC). The A. salmonicida sodB and sodC probably encode cytoplasmic and periplasmic superoxide dismutases, respectively. Our computer prediction suggests that sodC and katA are associated with OxyR boxes (see Fig. 4), and this finding is supported by our microarray data which shows that these two genes are significantly up regulated 15 and 30 min after H2O2 treatment. This makes sense as sodB probably maintains a relatively stable level of superoxide dismutase activity inside the cell, whereas sodC might be responsible for responding to elevated levels of superoxide by neutralising O-2 in the periplasmic space. It is documented that H2O2 can induce damages to DNA (Miller and Britigan, 1997). A. salmonicida encodes a number of genes involved in DNA protection, recombination and repair, such as recA (multifunctional protein in recombination and SOS response), recBCD (recombination
198
H.L. Pedersen et al. / Marine Genomics 3 (2010) 193–200
Table 2 Genes showing ≥2 fold down regulated (≤0.5) mRNA expression compared to control after treatment with H2O2. The CDSs are divided into classes based on their function. Fold change CDS
Gene Product
Transport/binding protein VSAL_I0640 oadB Oxaloacetate decarboxylase, beta subunit VSAL_I0743 yhaO Inner membrane transport protein (pseudogene) VSAL_I1713 glpT Glycerol-3-phosphate transporter VSAL_I2279 ccmC Cytochrome c-type biogenesis protein CcmC VSAL_II0338 MFS transporter VSAL_II0384 glpF Glycerol uptake facilitator protein VSAL_II0817 putP Proline permease VSAL_II1002 Sodium/proton-dependent alanine carrier protein
15 min 30 min 60 min 0.9
0.7
0.5
0.4
–
0.8
0.4
0.5
0.5
0.7
0.7
0.5
0.4 0.3
0.5 0.3
0.4 0.3
0.5 1.0
0.7 0.8
0.5 0.5
0.7
0.7
0.5
0.4 1.0
0.9 0.7
0.9 0.5
0.9
0.8
0.5
0.3
0.4
0.4
0.4
0.5
0.6
0.3 0.5
0.3 0.7
0.4 0.8
Macromolecule degradation VSAL_II0039 malQ 4-alpha-glucanotransferase VSAL_I2667 degQ Exported serine protease
0.5 0.5
0.9 0.7
0.7 1.4
Membrane/exported/lipoprotein VSAL_I2545 yggE Putative exported protein VSAL_II0323 Putative lipoprotein
0.5 0.7
0.8 0.6
1.0 0.4
0.7
0.6
0.4
1.0
0.8
0.5
1.3
1.0
0.5
1.7
1.1
0.5
Energy metabolism, carbon VSAL_I2595 acnB Aconitate hydratase 2 (citrate hydro-lyase 2) VSAL_II0135 Putative cytochrome b561 VSAL_II0161 cyoC Cytochrome o ubiquinol oxidase subunit III VSAL_II0163 cyoA Cytochrome o ubiquinol oxidase subunit II Central intermediary metabolism VSAL_I1712 glpQ Glycerophosphoryl diester phosphodiesterase precursor VSAL_II0337 ugpQ Glycerophosphoryl diester phosphodiesterase VSAL_II0385 glpK Glycerol kinase VSAL_I0419 cpdB 2',3'-cyclic-nucleotide 2'phosphodiesterase
Nucleotide biosynthesis VSAL_I0738 guaA GMP synthase [glutaminehydrolyzing] VSAL_I0585 carA Carbamoyl-phosphate synthase small chain Sulphur metabolism VSAL_I0404 cysH VSAL_I0420
cysD
Phosphoadenosine phosphosulfate reductase Sulfate adenylyltransferase subunit 2
Biosynthesis of cofactors VSAL_II0859 bioF 8-amino-7-oxononanoate synthase
0.4
0.4
0.4
0.5
0.7
0.7
RNA polymerase, sigma factor VSAL_I2531 rpoE RNA polymerase sigma-E factor
0.4
0.8
0.9
Threonine biosynthesis VSAL_I2555 thrB Homoserine kinase
1.0
0.8
0.5
Degradation of DNA VSAL_I1364
Putative exported nuclease
and DNA repair), recFOR (recombination, DNA protection and repair), recG (helicase activity in recombination and repair), recN (recombination and DNA repair) and rmuC (recombination). Based on our computer predictions none of these genes is associated with OxyR boxes, and
Table 1 shows that only recN has mRNA levels ≥2 fold increase after H2O2 treatment. A majority of the remaining genes are up regulated typically between 1.7 and 1.2 folds. Despite the relatively modest increase in gene expression our microarray results indicate a general response to protect and repair DNA. 4.2. Genes in reducing systems Interestingly, a set of H2O2-induced genes in our microarray experiment belong to one of two reducing systems, i.e., the glutathione- and thioredoxin-dependent reduction systems. Both systems are responsible for maintaining a reduced cytoplasmic environment in E. coli by reducing disulfide bonds, and by converting alkyl peroxides to corresponding alcohols (Ritz and Beckwith, 2001; Carmel-Harel and Storz, 2000). In A. salmonicida, several genes that are part of these two reducing systems are found: gshA (glutamatecysteine ligase), gshB (subunit of glutathione synthetase), gor (subunit of glutathione reductase), grxA, grxB and grxC (glutaredoxins), trxA and trxC (thioredoxins), trxB (thioredoxin reductase), ahpC (subunit of alkylhydroperoxide reductase) and bcp (thioredoxin reductase). The earlier described OxyR box predictions suggest that at least ahpC and grxA are directly regulated by OxyR. Remarkably, these are indeed among the most up regulated genes 15 min after addition of H2O2 (Table 1). trxC, which is induced by OxyR in E. coli (Ritz et al., 2000) is also significantly up regulated, but only 60 min after treatment with H2O2. 4.3. Genes and proteins involved in regeneration of NADPH, and carbon metabolism Our microarray data shows that three significantly up regulated genes 15 min after H2O2 treatment encode the whole set of enzymes involved in the oxidative (and NADPH generating) stage of the pentose phosphate pathway (PPP). These genes are pgl (encodes 6phosphogluconolactonase), gnd (encodes 6-phosphogluconate dehydrogenase) and zwf (encodes glucose-6-phosphate 1-dehydrogenase). PPP's two main functions in the cell are to generate pentose phosphate for synthesis of nucleotides, and to supply the cell with NADPH. Furthermore, this pathway plays an important role in defence during oxidative stress (Pandolfi et al., 1995; Puskas et al., 2000; He et al., 2007; Lundberg et al., 1999). In the A. salmonicida genome zwf, pgl and gnd are organised in the same operon on chromosome II, and are associated with a strong OxyR box (Fig. 4). 4.4. Concluding remarks Pathogenic bacteria that invade their hosts will be challenged with reactive oxygen species originating from the immune system of the host. Our computational and experimental study shows that A. salmonicida can respond to such challenges by up regulating expression of molecules that can neutralize reactive oxygen species (catalase, peroxidase, glutaredoxin and thioredoxins). In total, OxyR boxes were predicted in front of 20 single genes and 8 operons, and six of the 28 genes and operons were recovered on the microarray list of most up regulated CDSs. This list contains genes that are known to be induced by H2O2 in other bacteria, but also suggests H2O2-dependent regulation that is different from that of e.g., E. coli. For example, zwf, which is regulated by SoxS in E. coli is apparently regulated by OxyR in A. salmonicida (the soxS gene is not present in A. salmonicida). By studying the global responses of this pathogen to stress conditions that are similar to those it faces within its host, it is our goal to gain increasingly more knowledge on the molecular mechanisms underlying its pathogenic lifestyle. Supplementary materials related to this article can be found online at doi:10.1016/j.margen.2010.10.002.
H.L. Pedersen et al. / Marine Genomics 3 (2010) 193–200
199
Fig. 4. Selected mRNA transcripts differentially regulated after treatment with H2O2, based on the microarray data analysis. Schematic drawing that shows the organisation of neighbouring genes in the A. salmonicida genome. Each arrow represents a CDS or an OxyR box and shows the direction of the gene in the genome. Blue represents genes significantly up regulated, while orange represents genes significantly down regulated when compared to control. In addition, yellow represents transcriptional regulators and black represents a putative OxyR box. The CDSs are numbered according to the genome annotation of A. salmonicida and the gene name is indicated when known. In bold are the Patser score of the OxyR box and the distance between the OxyR box and the transcription start site of the downstream gene.
Acknowledgement This work was supported by The University of Tromsø, the Norwegian Research Council and The National Programme for Research and
Functional Genomics in Norway (FUGE). We are grateful to The Microarray Resource Centre in Tromsø (MRCT) for offering facilities and equipments to run the microarray experiments. We wish to thank Christopher G. Fenton for help with initial microarray data analysis.
200
H.L. Pedersen et al. / Marine Genomics 3 (2010) 193–200
References Ahmad, R., Hjerde, E., Hansen, G.A., Haugen, P., Willassen, N.P., 2009. Prediction and experimental testing of ferric uptake regulator regulons in Vibrios. J. Mol. Microbiol. Biotechnol. 16, 159–168. Aslund, F., Zheng, M., Beckwith, J., Storz, G., 1999. Regulation of the OxyR transcription factor by hydrogen peroxide and the cellular thiol-disulfide status. Proc. Natl Acad. Sci. U. S. A. 96, 6161–6165. Bailey, T.L., Elkan, C., 1994. Fitting a mixture model by expectation maximization to discover motifs in biopolymers. Proc. Int. Conf. Intell. Syst. Mol. Biol. 2, 28–36. Carmel-Harel, O., Storz, G., 2000. Roles of the glutathione- and thioredoxin-dependent reduction systems in the Escherichia coli and Saccharomyces cerevisiae responses to oxidative stress. Annu. Rev. Microbiol. 54, 439–461. Chang, W., Small, D.A., Toghrol, F., Bentley, W.E., 2005. Microarray analysis of Pseudomonas aeruginosa reveals induction of pyocin genes in response to hydrogen peroxide. BMC Genomics 6, 115. Choi, H., Kim, S., Mukhopadhyay, P., Cho, S., Woo, J., Storz, G., Ryu, S., 2001. Structural basis of the redox switch in the OxyR transcription factor. Cell 105, 103–113. Christman, M.F., Morgan, R.W., Jacobson, F.S., Ames, B.N., 1985. Positive control of a regulon for defenses against oxidative stress and some heat-shock proteins in Salmonella typhimurium. Cell 41, 753–762. Compan, I., Touati, D., 1993. Interaction of six global transcription regulators in expression of manganese superoxide dismutase in Escherichia coli K-12. J. Bacteriol. 175, 1687–1696. Crooks, G.E., Hon, G., Chandonia, J.M., Brenner, S.E., 2004. WebLogo: a sequence logo generator. Genome Res. 14, 1188–1190. Diaz, P.I., Slakeski, N., Reynolds, E.C., Morona, R., Rogers, A.H., Kolenbrander, P.E., 2006. Role of oxyR in the oral anaerobe Porphyromonas gingivalis. J. Bacteriol. 188, 2454–2462. Dysvik, B., Jonassen, I., 2001. J-Express: exploring gene expression data using Java. Bioinformatics 17, 369–370. Gort, A.S., Ferber, D.M., Imlay, J.A., 1999. The regulation and role of the periplasmic copper, zinc superoxide dismutase of Escherichia coli. Mol. Microbiol. 32, 179–191. Hall, T.A., 1999. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symp. Ser. 41, 95–98. Han, M.J., Lee, S.Y., 2006. The Escherichia coli proteome: past, present, and future prospects. Microbiol. Mol. Biol. Rev. 70, 362–439. He, W., Wang, Y., Liu, W., Zhou, C.Z., 2007. Crystal structure of Saccharomyces cerevisiae 6-phosphogluconate dehydrogenase Gnd1. BMC Struct. Biol. 7, 38. Helmann, J.D., Wu, M.F., Gaballa, A., Kobel, P.A., Morshedi, M.M., Fawcett, P., Paddon, C., 2003. The global transcriptional response of Bacillus subtilis to peroxide stress is coordinated by three transcription factors. J. Bacteriol. 185, 243–253. Hertz, G.Z., Stormo, G.D., 1999. Identifying DNA and protein patterns with statistically significant alignments of multiple sequences. Bioinformatics 15, 563–577. Hishinuma, S., Yuki, M., Fujimura, M., Fukumori, F., 2006. OxyR regulated the expression of two major catalases, KatA and KatB, along with peroxiredoxin, AhpC in Pseudomonas putida. Environ. Microbiol. 8, 2115–2124. Hjerde, E., Lorentzen, M.S., Holden, M.T., Seeger, K., Paulsen, S., Bason, N., Churcher, C., Harris, D., Norbertczak, H., Quail, M.A., Sanders, S., Thurston, S., Parkhill, J., Willassen, N.P., Thomson, N.R., 2008. The genome sequence of the fish pathogen Aliivibrio salmonicida strain LFI1238 shows extensive evidence of gene decay. BMC Genomics 9, 616. Hopkin, K.A., Papazian, M.A., Steinman, H.M., 1992. Functional differences between manganese and iron superoxide dismutases in Escherichia coli K-12. J. Biol. Chem. 267, 24253–24258. Horsburgh, M.J., Clements, M.O., Crossley, H., Ingham, E., Foster, S.J., 2001a. PerR controls oxidative stress resistance and iron storage proteins and is required for virulence in Staphylococcus aureus. Infect. Immun. 69, 3744–3754. Horsburgh, M.J., Ingham, E., Foster, S.J., 2001b. In Staphylococcus aureus, fur is an interactive regulator with PerR, contributes to virulence, and Is necessary for oxidative stress resistance through positive regulation of catalase and iron homeostasis. J. Bacteriol. 183, 468–475. Imlay, J.A., 2008. Cellular defenses against superoxide and hydrogen peroxide. Annu. Rev. Biochem. 77, 755–776. Jair, K.W., Martin, R.G., Rosner, J.L., Fujita, N., Ishihama, A., Wolf Jr., R.E., 1995. Purification and regulatory properties of MarA protein, a transcriptional activator of Escherichia coli multiple antibiotic and superoxide resistance promoters. J. Bacteriol. 177, 7100–7104. Lau, G.W., Britigan, B.E., Hassett, D.J., 2005. Pseudomonas aeruginosa OxyR is required for full virulence in rodent and insect models of infection and for resistance to human neutrophils. Infect. Immun. 73, 2550–2553. Li, H., Singh, A.K., McIntyre, L.M., Sherman, L.A., 2004. Differential gene expression in response to hydrogen peroxide and the putative PerR regulon of Synechocystis sp. strain PCC 6803. J. Bacteriol. 186, 3331–3345.
Loewen, P.C., Hu, B., Strutinsky, J., Sparling, R., 1998. Regulation in the rpoS regulon of Escherichia coli. Can. J. Microbiol. 44, 707–717. Lu, S., Killoran, P.B., Fang, F.C., Riley, L.W., 2002. The global regulator ArcA controls resistance to reactive nitrogen and oxygen intermediates in Salmonella enterica serovar Enteritidis. Infect. Immun. 70, 451–461. Lundberg, B.E., Wolf Jr., R.E., Dinauer, M.C., Xu, Y., Fang, F.C., 1999. Glucose 6-phosphate dehydrogenase is required for Salmonella typhimurium virulence and resistance to reactive oxygen and nitrogen intermediates. Infect. Immun. 67, 436–438. Meyer, F., Goesmann, A., McHardy, A.C., Bartels, D., Bekel, T., Clausen, J., Kalinowski, J., Linke, B., Rupp, O., Giegerich, R., Pühler, A., 2003. GenDB — an open source genome annotation system for prokaryote genomes. Nucleic Acids Res. 31, 2187–2195. Miller, R.A., Britigan, B.E., 1997. Role of oxidants in microbial pathophysiology. Clin. Microbiol. Rev. 10, 1–18. Morgan, R.W., Christman, M.F., Jacobson, F.S., Storz, G., Ames, B.N., 1986. Hydrogen peroxide-inducible proteins in Salmonella typhimurium overlap with heat shock and other stress proteins. Proc. Natl Acad. Sci. U. S. A. 83, 8059–8063. Ochsner, U.A., Vasil, M.L., Alsabbagh, E., Parvatiyar, K., Hassett, D.J., 2000. Role of the Pseudomonas aeruginosa oxyR-recG operon in oxidative stress defense and DNA repair: OxyR-dependent regulation of katB-ankB, ahpB, and ahpC-ahpF. J. Bacteriol. 182, 4533–4544. Palma, M., DeLuca, D., Worgall, S., Quadri, L.E., 2004. Transcriptome analysis of the response of Pseudomonas aeruginosa to hydrogen peroxide. J. Bacteriol. 186, 248–252. Pandolfi, P.P., Sonati, F., Rivi, R., Mason, P., Grosveld, F., Luzzatto, L., 1995. Targeted disruption of the housekeeping gene encoding glucose 6-phosphate dehydrogenase (G6PD): G6PD is dispensable for pentose synthesis but essential for defense against oxidative stress. EMBO J. 14, 5209–5215. Puskas, F., Gergely Jr., P., Banki, K., Perl, A., 2000. Stimulation of the pentose phosphate pathway and glutathione levels by dehydroascorbate, the oxidized form of vitamin C. FASEB J. 14, 1352–1361. Ritz, D., Beckwith, J., 2001. Roles of thiol-redox pathways in bacteria. Annu. Rev. Microbiol. 55, 21–48. Ritz, D., Patel, H., Doan, B., Zheng, M., Aslund, F., Storz, G., Beckwith, J., 2000. Thioredoxin 2 is involved in the oxidative stress response in Escherichia coli. J. Biol. Chem. 275, 2505–2512. Ruiz-Laguna, J., Prieto-Alamo, M.J., Pueyo, C., 2000. Oxidative mutagenesis in Escherichia coli strains lacking ROS-scavenging enzymes and/or 8-oxoguanine defenses. Environ. Mol. Mutagen. 35, 22–30. Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Steinman, H.M., Weinstein, L., Brenowitz, M., 1994. The manganese superoxide dismutase of Escherichia coli K-12 associates with DNA. J. Biol. Chem. 269, 28629–28634. Stohl, E.A., Criss, A.K., Seifert, H.S., 2005. The transcriptome response of Neisseria gonorrhoeae to hydrogen peroxide reveals genes with previously uncharacterized roles in oxidative damage protection. Mol. Microbiol. 58, 520–532. Storz, G., Tartaglia, L.A., 1992. OxyR: a regulator of antioxidant genes. J. Nutr. 122, 627–630. Sund, C.J., Rocha, E.R., Tzianabos, A.O., Wells, W.G., Gee, J.M., Reott, M.A., O'Rourke, D.P., Smith, C.J., 2008. The Bacteroides fragilis transcriptome response to oxygen and H2O2: the role of OxyR and its effect on survival and virulence. Mol. Microbiol. 67, 129–142. Toledano, M.B., Kullik, I., Trinh, F., Baird, P.T., Schneider, T.D., Storz, G., 1994. Redoxdependent shift of OxyR–DNA contacts along an extended DNA-binding site: a mechanism for differential promoter selection. Cell 78, 897–909. VanBogelen, R.A., Kelley, P.M., Neidhardt, F.C., 1987. Differential induction of heat shock, SOS, and oxidation stress regulons and accumulation of nucleotides in Escherichia coli. J. Bacteriol. 169, 26–32. Wong, S.M., Alugupalli, K.R., Ram, S., Akerley, B.J., 2007. The ArcA regulon and oxidative stress resistance in Haemophilus influenzae. Mol. Microbiol. 64, 1375–1390. Zeller, T., Moskvin, O.V., Li, K., Klug, G., Gomelsky, M., 2005. Transcriptome and physiological responses to hydrogen peroxide of the facultatively phototrophic bacterium Rhodobacter sphaeroides. J. Bacteriol. 187, 7232–7242. Zeller, T., Mraheil, M.A., Moskvin, O.V., Li, K., Gomelsky, M., Klug, G., 2007. Regulation of hydrogen peroxide-dependent gene expression in Rhodobacter sphaeroides: regulatory functions of OxyR. J. Bacteriol. 189, 3784–3792. Zheng, M., Aslund, F., Storz, G., 1998. Activation of the OxyR transcription factor by reversible disulfide bond formation. Science 279, 1718–1721. Zheng, M., Doan, B., Schneider, T.D., Storz, G., 1999. OxyR and SoxRS regulation of fur. J. Bacteriol. 181, 4639–4643. Zheng, M., Wang, X., Templeton, L.J., Smulski, D.R., LaRossa, R.A., Storz, G., 2001. DNA microarray-mediated transcriptional profiling of the Escherichia coli response to hydrogen peroxide. J. Bacteriol. 183, 4562–4570.