- Research article
- Open Access
Multiple controls affect arsenite oxidase gene expression in Herminiimonas arsenicoxydans
BMC Microbiology volume 10, Article number: 53 (2010)
The Erratum to this article has been published in BMC Microbiology 2017 17:74
Both the speciation and toxicity of arsenic are affected by bacterial transformations, i.e. oxidation, reduction or methylation. These transformations have a major impact on environmental contamination and more particularly on arsenic contamination of drinking water. Herminiimonas arsenicoxydans has been isolated from an arsenic- contaminated environment and has developed various mechanisms for coping with arsenic, including the oxidation of As(III) to As(V) as a detoxification mechanism.
In the present study, a differential transcriptome analysis was used to identify genes, including arsenite oxidase encoding genes, involved in the response of H. arsenicoxydans to As(III). To get insight into the molecular mechanisms of this enzyme activity, a Tn5 transposon mutagenesis was performed. Transposon insertions resulting in a lack of arsenite oxidase activity disrupted aoxR and aoxS genes, showing that the aox operon transcription is regulated by the AoxRS two-component system. Remarkably, transposon insertions were also identified in rpoN coding for the alternative N sigma factor (σ54) of RNA polymerase and in dnaJ coding for the Hsp70 co-chaperone. Western blotting with anti-AoxB antibodies and quantitative RT-PCR experiments allowed us to demonstrate that the rpoN and dnaJ gene products are involved in the control of arsenite oxidase gene expression. Finally, the transcriptional start site of the aoxAB operon was determined using rapid amplification of cDNA ends (RACE) and a putative -12/-24 σ54-dependent promoter motif was identified upstream of aoxAB coding sequences.
These results reveal the existence of novel molecular regulatory processes governing arsenite oxidase expression in H. arsenicoxydans. These data are summarized in a model that functionally integrates arsenite oxidation in the adaptive response to As(III) in this microorganism.
Arsenic is present in various environments, released from either anthropogenic or natural sources. This element is toxic for living organisms and known to be a human carcinogen . Its toxicological effects depend, at least in part, on its oxidation state and its chemical forms, inorganic species being considered as more toxic . The contamination of drinking water by the two inorganic forms, arsenite As(III) and arsenate As(V), has been reported in different parts of the world  and constitutes a major threat of public health. Microorganisms are known to take part in the transformation, i.e oxidation, reduction or methylation of the metalloid, having a deep impact on arsenic contamination in environment.
Several bacteria and prokaryotes have developed adaptation, resistance and colonization mechanisms, which allow them to live in hostile arsenic contaminated environments. H. arsenicoxydans is a Gram-negative β-proteobacterium isolated from an industrial activated sludge plant and exhibiting a remarkable set of arsenic resistance determinants . The H. arsenicoxydans adaptive response to arsenic is organized in a complex and sophisticated network. In particular, differential proteome studies have recently demonstrated the synthesis of several proteins encoded by the three ars resistance operons, e.g. arsenate reductase ArsC, flavoprotein ArsH and regulator ArsR [5, 6] and the induction of oxidative stress protein encoding genes, e.g. catalase (katA), superoxide dismutase (sodB) and alkyl hydroperoxide reductase (ahpC) .
One of the most noticeable response to arsenic in H. arsenicoxydans is the ability of this bacterium to oxidize As(III) to As(V), a less toxic and less mobile form, via an arsenite oxidase activity. The two genes coding for this heterodimeric enzyme are organized in an operonic structure, and have been named aoxA and aoxB for the small and the large subunit, respectively [6, 8, 9]. Homologous genes have been since identified in various microorganisms [6, 10–13]. In Agrobacterium tumefaciens, a complex transcriptional regulation has been recently suggested, involving As(III) sensing, two-component signal transduction by an AoxS sensor kinase and an AoxR regulator, and quorum sensing . Nevertheless, the molecular mechanisms involved in the control of arsenite oxidase expression remain largely unknown.
To extend our view of arsenic response , in particular with respect to the arsenic oxidation process, we conducted a comprehensive transcriptomic, genetic and molecular analysis of H. arsenicoxydans following exposure to As(III). These approaches allowed us to identify major determinants involved in the control of arsenite oxidation.
Gene expression profiling in response to arsenic
The response to As(III) was analyzed in exponentially growing cells using microarrays. The data from three biological replicates were combined after normalization and statistical analysis carried out using the R software and packages http://www.r-project.org. The set of genes was further refined to include only those genes that showed a valid p-value and whose expression was altered by a factor of 2 or more when compared to the level measured in the absence of arsenic.
This experiment led to the identification of 293 genes showing an arsenic-induced expression change (> 2 fold (log2 = 1)). Among these genes, 133 (45%) were up-regulated and the remaining part, i.e. 160 genes, was down-regulated. The relative changes in gene expression ranged from a 11-fold down-regulation (rpsN gene encoding a ribosomal protein) to a 126-fold up-regulation (putative gene involved in phosphate transport). A list of those genes is shown in Additional file 1, Table S1. The corresponding HEAR gene numbers are available in the Arsenoscope relational database http://www.genoscope.cns.fr/agc/mage/arsenoscopevia the MaGe web interface .
The 293 genes differentially expressed were clustered according to the function of the corresponding encoded proteins. Among the 133 genes that were induced by at least a 2-fold factor, about 11% are involved in metabolism, 26% in transport and binding protein, 26% in cellular processes and 31% have no assigned function. The high percentage of genes with unknown function is in accordance with the proportion of unknown function proteins identified in the genome of H. arsenicoxydans[6, 7]. In agreement with our previous results, genes involved in arsenic resistance, phosphate transport and flagellar biosynthesis were induced in the presence of As(III) (see Additional file 1, Table S1), further supporting the relationship between these physiological processes [6, 7]. Interestingly, only one methyl-accepting chemotaxis protein (MCP) gene was induced in our microarray data, suggesting a role for this protein in the sensing of arsenic. This mechanism is currently under investigation. Genes encoding the putative nitroreductase AoxC and the cytochrome c552 precursor AoxD as well as the response regulator AoxRS were found to be induced by As(III) (see Additional file 1, Table S1). AoxR has been proposed to act as a positive regulator of the aox operon upon phosphorylation by AoxS in A. tumefaciens. Our transcriptomic data suggest that the regulation machinery is, at least in part, similar in H. arsenicoxydans. Futhermore, genes coding for the arsenite oxidase AoxAB subunits were found to be among the most induced genes in the presence of As(III). However, even though these results extended our knowledge of the arsenic response in H. arsenicoxydans, they did not led to a better understanding of the molecular mechanisms involved in the control of arsenite oxidation. This prompted us to perform a transposon mutagenesis experiment.
Identification of arsenite oxidase accessory genes by screening an Aox activity deficient mutant library
To identify genes possibly involved in the control of arsenite oxidation in H. arsenicoxydans, a library of 10,000 kanamycin resistant mutants was constructed by transposon mutagenesis, as previously described . These clones were tested by silver nitrate staining  for arsenate production on As(III)-supplemented CDM agar plates. As compared to the wild-type strain, whose arsenite oxidase activity was revealed by a brownish precipitate, 10 mutants with a lack of As(III) oxidase activity were obtained. These strains showed no precipitate (Figure 1A), as observed for the M1 and M2 strains used as negative controls. Indeed, these strains carry a mutation in aoxA or aoxB genes coding for the small and the large subunit of arsenite oxidase, respectively . Genes disrupted by transposon insertions were identified in these 10 new mutants. As expected, four of the 10 mutants showed insertions in the aoxAB operon (Figure 2A). More interestingly, six mutants carried a transposon insertion outside the aoxAB operon. Two mutants were found to be affected in the aoxRS two-component signal transduction system (mutants Ha482 and Ha483, respectively) located upstream of the aoxAB operon in H. arsenicoxydans (Figure 2A). These results further support our transcriptomic data suggesting that these two genes play a role in arsenic response. Two transposon insertions were shown to disrupt genes of the modEABC operon coding for a molybdenum high-affinity transport system , i.e. modC encoding an ATP-binding cassette transport protein (mutant Ha3437) and modB encoding a molybdenum transport system permease (mutant Ha3438) (Figure 2B). Remarkably, transposon insertions were also located in dnaJ encoding a heat shock protein (Hsp40), (mutant Ha2646) (Figure 2C) and in rpoN encoding the alternative nitrogen sigma factor (sigma 54) of RNA polymerase (mutant Ha3109) (Figure 2D).
Two additional experiments were performed to complete the physiological characterization of these mutants with respect to arsenite oxidation. First, arsenic species were quantified by HPLC-ICP-AES on filtered culture supernatants. H. arsenicoxydans was grown in liquid medium supplemented with 1.33 mM arsenite and showed 100% transformation of As(III) into As(V) after 48 h, whereas M1 (aoxA) and M2 (aoxB) mutants used as controls were not able to transform As(III) into As(V). The same loss of arsenite oxidase activity was measured in Ha482 (aoxS), Ha483 (aoxR), Ha2646 (dnaJ) and Ha3109 (rpoN) mutants. In contrast to the results obtained on agar plates, Ha3437 (modC) and Ha3438 (modB) strains showed 100% transformation of arsenite (Table 1, Figure 1A). Previous studies have demonstrated that the bioavailability of metals or trace elements considerably varies according to the type of matrix used for microbial growth . We therefore assumed that Mo might be partly sequestred on CDM agar medium, resulting in a lack of arsenite oxidase activity on plate. To test this hypothesis, As(III) oxidase tests were performed on CDM agar plates supplemented with 50 μM Mo. The addition of Mo to the solid medium restored As(III) oxidase activity in both Ha3437 (modC) and Ha3438 (modB) mutants while it had no effect on other mutant strains (Figure 1B).
Second, we have previously demonstrated that the polar flagellum-dependent motility of H. arsenicoxydans is increased in the presence of As(III), suggesting that arsenite oxidation may result in a gain of energy . The motility of mutant strains was therefore tested on plates containing different concentrations of As(III), i. e. 0.66 mM, 1.33 mM and 2 mM. The diameter of the swarming rings was measured after 72 h. As shown in Figure 3, the disruption of aoxA, aoxB, aoxR, aoxS or rpoN genes abolished the improvement of swarming performances in the presence of As(III). Unlike those mutants, a disruption in dnaJ completely abolished the motility of H. arsenicoxydans in the presence or the absence of As(III). DnaJ is known to be essential for the expression of the flhDC flagellar master operon in Escherichia coli. The lack of motility observed in the dnaJ mutant suggests the existence of a similar flhDC-dependent regulation of flagellar genes in H. arsenicoxydans. More importantly, the dnaJ mutant was affected in both the motility and arsenite oxidation, suggesting that these two mechanisms are co-regulated.
Effect of AoxR, AoxS, RpoN and DnaJ on arsenite oxidase synthesis
To get further insight into the involvement of AoxR, AoxS, RpoN and DnaJ in arsenite oxidase activity, Western immunoblotting experiments were performed using antibodies raised against AoxB. The abundance of this protein was evaluated from total protein extracts of H. arsenicoxydans wild-type and mutant strains grown in the presence or not of As(III). AoxB was detected as a single band corresponding to a molecular mass of 92 kDa in As(III)-challenged H. arsenicoxydans strain (Figure 4). This single band was not observed in the various mutant strains. Furthermore, arsenite oxidase activity on native gel was only detected in As(III)-challenged wild type total extract (data not shown). Taken together these results suggest that the lack of activity in the mutant strains is due to the absence of AoxB protein, which may result from an effect of AoxR, AoxS, RpoN and DnaJ on aoxAB expression.
Effect of AoxR, AoxS, RpoN and DnaJ on the control of arsenite oxidase operon expression
To determine the involvement of aoxR, aoxS, dnaJ and rpoN on aoxAB transcription, we performed quantitative RT-PCR experiments. For each strain, changes in aoxB transcript abundance were compared to two internal controls, i.e. the putative RNA methyltransferase gene and the peptide deformylase gene, in cultures challenged or not by As(III). The expression of aoxB mRNA was increased by a 9.4 fold factor after As(III) exposure in the H. arsenicoxydans wild-type strain. In contrast, aoxB expression was not increased in Ha482 (aoxS), Ha483 (aoxR), Ha3109 (rpoN) and Ha2646 (dnaJ) mutant strains, suggesting that the corresponding proteins play a crucial role in aoxAB operon expression (Table 2).
Molecular mechanisms of arsenite oxidase transcription
The aoxR and aoxS genes encode a two-component system while rpoN encodes a sigma factor which recognizes a particular promoter with a specific -12/-24 binding site. These three proteins may therefore play a role in the initiation of aoxAB transcription. To get further insight into the molecular interactions between those regulators and the aoxAB promoter, we mapped the transcriptional start site of this operon by the amplification of aoxAB cDNA ends and 5'RACE. Messenger RNAs were extracted from induced (1.33 mM As(III)) and non induced H. arsenicoxydans wild-type strain cultures. A single transcriptional start site was identified from induced cells at -26 bp relative to the translation start codon, while no transcriptional start site was identified from non induced cells. In agreement with this, a TGGCACG CAGTTTGC putative -12/-24 σ54-dependent promoter motif was identified upstream of the aoxAB transcriptional start site (Figure 5). In addition, multiple alignment of aoxAB promoter sequences present in databases revealed a similarity to promoters recognized by σ54 in A. tumefaciens, Thiomonas sp., Rhizobium sp. NT-26, Achromobacter sp., Rhodoferax ferrireducens, Ochrobactrum tritici (Figure 5A). In contrast, no such σ54-dependent promoter motif was found in several strains containing the aoxAB operon but lacking the two-component transduction system aoxRS operon, such as Chloroflexus aurantiacus, Chlorobium limicola, Thermus thermophilus, Burkholderia multivorans, Roseobacter litoralis, Pseudomonas sp.TS44, Chlorobium phaeobacteroides and Chloroflexus aggregans (Figure 5B).
Remarkably, a multiple alignment of amino acid sequences revealed that AoxR shares significant homology with a number of σ54 RNA polymerase transcriptional activators, i.e. 35.96% identity with ZraR and 35.26% identity with AtoC from E. coli K12. AoxR contains three conserved domains shared by most Enhancer Binding Proteins (EBP), namely a N-terminal response regulator receiver domain (amino acids 18-130), a central σ54 interaction domain (amino acids 147-368) common to all σ54 dependent EBPs (Pfam E-value 10-116; http://www.sanger.ac.uk/cgi-bin/Pfam/getacc?PF00158) and a C-terminal DNA binding helix-turn-helix (HTF) domain (amino acids 421-463) enable to bind to specific upstream activation sequences . AoxR shares similarities with several EBPs of σ54 essential for the formation of an open complex formation during σ54-dependent transcriptional initiation, in particular the σ54 activator sequence GAFTGA loop 1 which directly binds to σ54 conserved region III (Figure 6) . Taken together, these observations strongly suggest that AoxR interacts directly with RpoN to initiate the transcription of aoxAB operon in H. arsenicoxydans.
Despite many works devoted to arsenic metabolism in microorganisms, little is known about the regulation of arsenite oxidase activity. In the present study, the combination of transcriptomic, genetic and molecular data provided a comprehensive view of the role of various proteins in the control of arsenite oxidation in H. arsenicoxydans (Figure 7). We showed that some proteins play an indirect role in this process, as their presence is not essential for AoxAB synthesis. In this respect, the AoxB large subunit contains a Mo site required in arsenite oxidase enzymatic activity . Ha3437 (modC) and Ha3438 (modB) mutations were located in the molybdenum high-affinity transport system operon, which further support the key role of this element in enzyme activity. In addition, the recovery of As(III) oxidase activity in these two mutants in the presence of an excess molybdenum suggests that Mo may also be transported through an alternative uptake system in mod mutants, e.g. a low-affinity uptake system involving non specific permeases such as HEAR0069, HEAR0154, HEAR1749 or HEAR2391 or a sulfate transport system, as described in E. coli mod mutants .
More importantly, our results suggest that AoxR and RpoN constitute a transcriptional complex that play a major role in the initiation of aoxAB operon transcription. Three mutants, i.e. Ha482 (aoxS), Ha483 (aoxR) and Ha3109 (rpoN), were affected in this process. The amino acids sequence analysis of H. arsenicoxydans AoxR and AoxS revealed the existence in these proteins of structural features common to partners of two-component signal transduction systems, which are composed of a sensor kinase and a response regulator . Moreover, the comparison of AoxS and AoxR protein sequences with those of A. tumefaciens revealed similarities. Indeed, the AoxS protein sequence contains short blocks of conserved motifs that are consistent with a role of sensor histidine kinase, e.g. the "H" (amino acids 279 to 287: LAHEVNNPL), the "G2" (amino acids 435 to 441: GRIGLGL) and the "N" (amino acids 380 to 391: VRQIVLNLVLNA) domains. In addition, four highly invariant residues playing a central role in phosphorylation correspond to Asp9, Asp10, Asp57 and Lys107 in the H. arsenicoxydans AoxR protein. The three asparagine residues are known to interact together to form an "acid pocket", characteristic of orthodox receivers, into which the side chain of Lys109 protrudes . Finally, a transmembrane region and a 17 amino acids residue cluster possibly exposed to the periplasm are present in AoxS and could serve as a signal receptor in the presence of As(III) in the medium. The detection of As(III) would then lead to AoxS autophosphorylation at a histidine residue via ATP hydrolysis and phosphotransfer to an aspartate residue in the response regulator AoxR, as recently proposed in A. tumefaciens.
Remarkably, our results demonstrated for the first time that the alternative N sigma factor (σ54) is essential for the initiation of arsenite oxidase transcription. Indeed, a mutation in the corresponding gene led to a complete loss of As(III) oxidation and aoxB transcription in Ha3109 (rpoN). σ54 is one of the alternative sigma subunits of RNA polymerase responsible for specific binding to DNA. The core RNA polymerase complexed with σ54 is usually associated with nitrogen assimilation and fixation, but is also known to play a role in various physiological processes, e.g. flagellar synthesis, carbon source utilization or bacterial virulence . To date, only one report has shown that σ54 participates in the transcription of genes possibly involved in metal tolerance, i.e. the zraR/S genes that code for a zinc and lead responsive two-component regulatory system in E. coli. RNA polymerase together with σ54 binds to a specific promoter site, with the consensus DNA sequence YTGGCACGNNNNTTGCWNNw , forming a transcriptionally inactive closed complex. Such a characteristic -12/-24 σ54-dependent promoter motif, i.e. TGGCACG CAGTTTGC, was identified 26 pb upstream of the transcriptional initiation codon of aoxAB with respect to the +1 transcriptional start site (Figure 5), which confirmed the need for RpoN in the initiation of aoxAB transcription. Changes in the conformation of σ54-RNA polymerase are nucleotide dependent. Indeed, the DNA melting step absolutely requires the interaction with a transcriptional activator protein. Most of these σ54-dependent activators share three domains found in AoxR, i.e. a C-terminal DNA binding domain that binds to upstream activation sequences, a conserved central domain belonging to the AAA+ (ATPases associated with various cellular activities) protein family to proceed with initiation of transcription and a N-terminal receiver domain that regulates its own AAA+ domain [20, 28, 29]. A multiple alignment of the central domain revealed a conservation of a common architecture between AoxR and σ54 EBPs. Indeed, seven highly conserved sequence motifs corresponding to a σ54 interaction domain of AoxR further support the direct interaction of AoxR with RpoN to stimulate the transcription of aoxAB operon in H. arsenicoxydans (Figure 6). This central σ54 interaction domain has been already used to identify new σ54 EBPs [30–37]. It contains the highly conserved signature sequence GAFTGA loop (conserved region CIII, Figure 6) proved to be required for direct interaction with σ54[21, 38, 39], the Walker A (conserved region I, Figure 6) and B (conserved region CIV, Figure 6) ATPase motifs necessary for the open complex formation.
Both multiple sequences alignment and genetic environment analysis of aox promoters suggest the existence of a wide diversity in the transcriptional control of the aox operon. Indeed, as in H. arsenicoxydans, a σ54-dependent promoter signature was identified in bacteria possessing a two-component transduction system AoxRS operon downstream of the aoxAB operon, e.g. A. tumefaciens and O. tritici (Figure 5A). In contrast, no σ54-dependent promoter motif and no aoxR homologous gene were found in other bacteria, e.g. C. aurantiacus or C. aggregans (Figure 5B). These observations suggest that the transcription of the aox operon in these bacteria may involve other regulatory proteins and that AoxR may represent a specific co-activator of RpoN in the initiation of the aox operon transcription. Finally, our results provide evidence that the DnaJ co-chaperone is required for As(III) oxidation. DnaJ is part of the DnaK-DnaJ-GrpE Hsp70 machinery. Hsp70 chaperones represent one of the most potent defence cellular mechanism against environmental insults as DnaK-DnaJ-GrpE are known to assist protein folding [40, 41] or to be involved in mRNA stability . In the present study we showed that there is no induction of aoxAB transcription in the dnaJ mutant, resulting in a loss of AoxAB synthesis.
Several possible mechanisms involving DnaJ in the regulation of arsenite oxidase can be hypothesized. DnaJ may be required for the proper folding or activity of the AoxR regulator. Such a function has been demonstrated for the positive regulator CRP in a dnaJ deletion mutant in E. coli. Similarly, a post-transcriptional regulation of the arsenite oxidase itself can not be excluded. Moreover, a Tat (Twin-Arginine Translocation) signal has been detected in the AoxA sequence of H. arsenicoxydans. Proteins secreted to the periplasm via a Tat protein export pathway are known to require a folding by Hsp70 chaperones before their secretion. DnaJ could be one of these chaperones [44, 45]. Another possible target of DnaJ may be the RpoN sigma factor, as this chaperone has been demonstrated to play a role in the regulation of σS in various species . Alternatively, several mechanisms are known to be involved in the stability of messenger RNA. For example, in E. coli, a long 5' untranslated region (UTR) has been observed upstream of the transcriptional start site of the flhDC flagellum master operon. This region plays a crucial role in the stability of the mRNA controlled by CsrA . In the present report, the aoxAB transcriptional start site was located 26 bp upstream of the translational start codon, providing evidence that such a long 5'UTR does not exist upstream of the aox operon. However, this does not rule out a role of DnaJ on the stability of aoxAB mRNA, as HSP proteins are known to function as mRNA stabilizers and to protect them against nuclease degradation . Further experiments will therefore be required to fully elucidate the molecular mechanisms of arsenite oxidase regulation in H. arsenicoxydans.
Taken together, our observations provide evidence that multiple proteins play a role in the control of arsenite oxidation in H. arsenicoxydans. The following regulatory model is proposed: AoxS responds to the presence of As(III) in the environment and autophosphorylates. The phosphate is then transferred to AoxR, which acts as a positive regulator of the aox operon and activates the initiation of the transcription in association with RpoN. In addition, DnaJ acts on the expression or the stability of both arsenite oxidation and motility genes, demonstrating that these two functions are strongly linked. Our results include the role of RpoN and DnaJ in arsenite oxidase synthesis, which provide further insight into the molecular mechanisms used by H. arsenicoxydans to cope with the most toxic form of arsenic in its environment.
Bacterial strains and growth media
Bacterial strains used in this study are listed in Table 3. H. arsenicoxydans ULPAs1 was grown in a chemically defined medium (CDM), supplemented by 2% agar when required . Escherichia coli S17-1 strain  was cultivated in LB medium (MP Biochemicals). Matings were performed on CDM to which 10% (wt/vol) LB medium was added, as previously described . Tryptone swarm plates containing CDM supplemented with 1% Bacto-Tryptone and 0.25% agar were used to assess bacterial motility.
The mini-Tn5::lacZ2 transposon  was delivered by mobilization of the suicide vector pUT/mini-Tn5::lacZ2 from E. coli S17-1 (λ-pyr) to H. arsenicoxydans. Conjugation was performed and transformants were selected as previously described .
Selection of arsenite oxidase mutants
Mutants were screened for arsenite oxidase activity as previously described . Agar plates were flooded with a 0.1 M AgNO3 solution to visualize arsenite oxidation . Mutants affected in molybdenum metabolism were also tested on CDM agar plates supplemented with 50 μM Na2MoO4, 2H2O and 1.33 mM As(III).
DNA manipulation and insertion mapping
DNA manipulations were carried out according to standard protocols, as described by Sambrook et al. . Total DNA was isolated from mutant strains with the Wizard Genomic DNA purification kit (Promega). Transposon insertion sites were mapped as previously described . Briefly, total DNA was digested and self-ligated. This ligation mixture was used as a template for PCR amplification using mini-transposon specific primers. The PCR products obtained were purified with a PCR purification kit (Qiagen) and sequenced on an Applied Biosystems ABI prism 3130×l capillary sequencer. The resulting sequences were compared to the H. arsenicoxydans genome sequence  to identify disrupted CDS. Finally, insertion sites and transposon orientations were precisely mapped by sequencing PCR products obtained with two primers hybridizing upstream and downstream, respectively, of the insertion site of each disrupted gene (see Additional file 2, Table S2).
Arsenic speciation determination
H. arsenicoxydans wild type and mutants were grown for 48 hours in CDM medium supplemented with 1.33 mM As(III). Culture supernatants were filtered through sterile 0,22 μm pore size filters (VWR). Arsenic species were separated by high-performance liquid chromatography (HPLC) and quantified by inductively coupled plasma-atomic emission spectrometry (ICP-AES), as previously described .
Strains were grown at 25°C for 24 h (OD = 0,15) and cultures were induced by addition of 0.66 mM or 1.33 mM As(III) for 8 hours before extraction. Samples were harvested and stored at -80°C. RNA was extracted as previously described . After extraction procedure, RNA integrity was checked by electrophoregram analysis on a BioAnalyser (Agilent) and total RNA concentration was determined spectrophotometrically with a Nanodrop.
Microarrays and data analysis
Microarrays containing 60-mer oligonucleotides for all predicted H. arsenicoxydans genes http://www.genoscope.cns.fr/agc/mage/arsenoscope were used, as previously described . Briefly, total RNA (5 μg) was reverse transcribed and indirectly labelled according to manufacturer's instructions with some modifications . The quality and concentration determination as well as hybridization and scanning were performed as previously described  Three distinct biological RNA samples were prepared from in each growth condition (with and without As(III) induction) and labelled either by Cy3 or Cy5 in a dye-swap design. Microarray data were deposited in ArrayExpress (accession E-MEXP-2199 and A-MEXP-1594). Data normalization and statistical analysis were performed as previously described . Briefly, data were acquired and analyzed by Genepix Pro 6.0 (Axon Instrument). The experiment design included three biological replicates. For each of them, induced and non-induced cells were compared in dye swap experiments. The resulting arrays were analyzed using the R software http://www.r-project.org. A slide by slide Loess normalization was performed using the limma package . Valid log2 expression ratios from replicated spots were averaged on each array so as to get statistically independent ratios for each oligonucleotide included in the array design. For the same reason, dye swap arrays were also averaged. Oligos that had no valid expression ratios on the ten arrays were excluded from the data set for further analysis, which was carried out using the varmixt package and the VM option . The resulting raw p-values were adjusted according to a Benjamini and Yekutieli procedure . Genes showing a valid p-value and a more than two-fold decreased or increased expression were considered as differentially expressed between the two conditions and were retained for further study.
Quantitative real time PCR
Quantitative PCR experiments were performed with RNA prepared as described for microarrays. RNA aliquots were purified with the RNeasy Plus mini kit (Qiagen) to ensure the elimination of genomic DNA. Total RNA concentration was determined spectrophotometrically using a Nanodrop and RNA integrity was electrophoretically verified. Total RNA (1,9 μg) was reverse transcribed with SuperScript III first-strand synthesis system for RT-PCR (Invitrogen) using random hexamers. Real time quantitative PCR was carried out with a MyiQ single-color Real-time PCR detection system. The reaction mixture contained 12,5 μl of MESA Blue qPCR MasterMix Plus for SYBR Assay with fluorescein (Eurogentec), 5 μl of cDNA and 300 nM of each primer in a total volume of 25 μl. Thermocycling conditions were as follow: 5 min at 95°c and 40 cycles of 15 s at 95°C, 15 s at 61°c and 1 min at 72°C. The PCR efficiency of the genes of interest and internal control genes were optimized to be similar enough by adjusting the primer concentrations to 300 nM each (data not shown). For each quantitative PCR run, non-template controls were performed to identify false positives and negative controls without reverse transcriptase were performed for each cDNA synthesis reaction and verified in real time PCR to determine the presence of contaminating genomic DNA. Two biological replicates (independent cultures) and two quantitative PCR replicates were performed for each experience. Amplification products were designed to be less than 175 bp in size. The pairs of primers used are listed in Additional file 2, Table S2. Two housekeeping genes, i.e. HEAR2922 coding for a putative RNA methyltransferase and HEAR0118 coding for a peptide deformylase, were used as standards to obtain normalized aoxB (HEAR0478) gene ratio  in the As(III) induced sample compared to the non-induced sample. These two housekeeping genes showed a stable expression between the two analyzed conditions (without As(III) and after an 8 hours As(III) exposure) when observing the microarrays data. The data were analyzed with the Relative Expression Software Tool . Statistical significance was defined as a p-value of ≤ 0.05.
The transcriptional start site of aoxAB operon was determined using the 5'RACE system for rapid amplification of cDNA ends (Invitrogen). Total RNA was obtained as described before. Reverse transcriptase reactions were performed using 5 μg total RNA and a gene specific reverse primer (5'-CATGGGCACTTGAATGTCTTG-3'). Reactions were heated at 70°C for 10 min and immediately prewarmed at 50°C before addition of Super-Script II reverse transcriptase. Reverse transcription was conducted at 50°C for 50 min and stopped at 70°c for 15 min. Purification and tailing of cDNA were performed according to manufacturer's instructions. The resulting cDNA was amplified by PCR using the provided Abridged Anchor Primer and a gene specific primer (5'-ATGCTGTGCGCGACGATATCG-3') located upstream of the original cDNA primer.
Preparation of protein extracts, SDS-PAGE and PAGE separation
Western immunoblotting were performed from late exponential phase wild-type and mutant strains grown in 1 liter CDM (with and without the presence of 100 mg/liter As(III)). The cultures were harvested by centrifugation for 10 min at 9,000 × g. Cell pellets were resuspended in distilled water and sonicated at 100 A (15 times 1 min with 1 min interval on ice, 80% duty cycle). Cell debris were removed by centrifugation (15 min at 13,000 × g). The supernatant was collected (total extract) and stored at -20°C. The protein concentration of each sample was measured with a Bio-Rad protein assay kit. First, fifty micrograms of each protein extract was loaded onto an 11% polyacrylamide-SDS gel. Second, fifty micrograms of each protein extract were loaded onto a polyacrylamide gel (native gel). The assay of arsenite oxidase activity followed the transfer of reducing equivalents from arsenite to 2,4-dichlorophenolindophenol (DCIP) as described by Anderson et al. . Briefly, the reduction of DCIP (60 μM) was monitored in the presence of 200 μM sodium arsenite in 50 μM MES, pH 6.0, at 25°C.
Preparation of antibodies and Western blot analysis
Monoclonal antibodies raised against an AoxB peptide were obtained from Proteogenix. Briefly, a hexadecapeptide with the SKNRDRVALPPVNAQK sequence was synthesized. This peptide corresponds to the N-terminal 16 amino acids of the arsenite oxidase large subunit of H. arsenicoxydans. The peptide was then coupled to keyhole limpet haemocyanin (KLH). Two rabbits were injected at multiple subcutaneaous sites with peptide-KLH at 14 days intervals. Animals were prebled at day 0, bled at day 49 (from an ear vein) and totally bled at day 90. Antibodies were partially purified on an affinity column substituted with the peptide.
After SDS-PAGE electrophoresis, the proteins were electrotransfered to a nitrocellulose membrane (Schleicher and Schuell, BA-85) using a Trans-Blot system (Bio-Rad) at 100 V, 4°C for 1 h. The membranes were washed twice in Tris buffered saline (TBS: 10 mM Tris-HCl pH7.5, 150 mM NaCl) and blocked in TBS with 0,3% bovine serum albumin (BSA). The membrane was then washed three times in Tris-buffered saline with TritonX100 and Tween 20 (TBS-T: 20 mM TrisHCl pH7.5, 500 mM NaCl, 0,2% Triton X-100, 0,05% Tween20), and incubated for 1 h with the AoxB antisera (1:800 dilution) in TBS-T with 0,3% BSA. Excess antibodies were removed by repeated washing with TBS-T. After 30 min incubation in TBS-T containing the secondary antibody (1:800 dilution of goat IgG against rabbit IgG, Sigma) conjugated with alkaline phosphatase, the membrane was washed twice with TBS-T and revealed by NBT/BCIP color reagent using standard procedures.
Mead MN: Arsenic: In search of an antidote to a global poison. Environ Health Perspect. 2005, 113: A378-A386.
Rosen BP: Biochemistry of arsenic detoxification. FEBS Lett. 2002, 529: 86-92. 10.1016/S0014-5793(02)03186-1.
Smith AH, Lingas EO, Rahman M: Contamination of drinking-water by arsenic in Bangladesh: A public health emergency. Bull World Health Organ. 2000, 78: 1093-1103.
Muller D, Simeonova DD, Riegel P, Mangenot S, Koechler S, Lièvremont D, Bertin PN, Lett MC: Herminiimonas arsenicoxydans sp. nov., a metalloresistant bacterium. Int J Syst Evol Microbiol. 2006, 56: 1765-1769. 10.1099/ijs.0.64308-0.
Carapito C, Muller D, Turlin E, Koechler S, Danchin A, Van Dorsselaer A, Leize-Wagner E, Bertin PN, Lett MC: Identification of genes and proteins involved in the pleiotropic response to arsenic stress in Caenibacter arsenoxydans, a metalloresistant beta-proteobacterium with an unsequenced genome. Biochimie. 2006, 88: 595-606. 10.1016/j.biochi.2005.11.004.
Muller D, Medigue C, Koechler S, Barbe V, Barakat M, Talla E, Bonnefoy V, Krin E, Arsene-Ploetze F, Carapito C: A tale of two oxidation states: bacterial colonization of arsenic-rich environments. PLoS genetics. 2007, 3 (4): e53-10.1371/journal.pgen.0030053.
Weiss S, Carapito C, Cleiss J, Koechler S, Turlin E, Coppee JY, Heymann M, Kugler V, Stauffert M, Cruveiller S: Enhanced structural and functional genome elucidation of the arsenite-oxidizing strain Herminiimonas arsenicoxydans by proteomics data. Biochimie. 2009, 91: 192-203. 10.1016/j.biochi.2008.07.013.
Alvarez-Martinez CE, Lourenço RF, Baldini RL, Laub MT, Gomes SL: The ECF sigma factor sT is involved in osmotic and oxidative stress responses in Caulobacter crescentus. Mol Microbiol. 2007, 66: 1240-1255. 10.1111/j.1365-2958.2007.06005.x.
Muller D, Lièvremont D, Simeonova DD, Hubert JC, Lett MC: Arsenite oxidase aox genes from a metal-resistant beta-proteobacterium. J Bacteriol. 2003, 185: 135-141. 10.1128/JB.185.1.135-141.2003.
Santini JM, Hoven Vanden RN: Molybdenum-containing arsenite oxidase of the chemolithoautotrophic arsenite oxidizer NT-26. J Bacteriol. 2004, 186: 1614-1619. 10.1128/JB.186.6.1614-1619.2004.
Quéméneur M, Heinrich-Salmeron A, Muller D, Lièvremont D, Jauzein M, Bertin PN, Garrido F, Joulian C: Diversity surveys and evolutionary relationships of aoxB genes in aerobic arsenite-oxidizing bacteria. App Environ Microbiol. 2008, 74: 4567-4573. 10.1128/AEM.02851-07.
Cai L, Rensing C, Li X, Wang G: Novel gene clusters involved in arsenite oxidation and resistance in two arsenite oxidizers: Achromobacter sp. SY8 and Pseudomonas sp. TS44. App Microbiol Biotechnol. 2009, 83 (4): 715-25. 10.1007/s00253-009-1929-4.
Clingenpeel SR, D'Imperio S, Oduro H, Druschel GK, McDermott TR: Cloning and in situ expression studies of the Hydrogenobaculum arsenite oxidase genes. App Environ Microbiol. 2009, 75: 3365-3365. 10.1128/AEM.00336-09.
Kashyap DR, Botero LM, Franck WL, Hassett DJ, McDermott TR: Complex regulation of arsenite oxidation in Agrobacterium tumefaciens. J Bacteriol. 2006, 188: 1081-1088. 10.1128/JB.188.3.1081-1088.2006.
Vallenet D, Labarre L, Rouy Z, Barbe V, Bocs S, Cruveiller S, Lajus A, Pascal G, Scarpelli C, Médigue C: MaGe: A microbial genome annotation system supported by synteny results. Nucleic Acids Res. 2006, 34: 53-65. 10.1093/nar/gkj406.
Lett M-C, Paknikar K, Lièvremont D: A simple and rapid method for arsenite and arsenate speciation. Biohydrometallurgy - Fundamentals, Technology and Sustainable Development, Part B. Edited by: Jr VSTCaOG. 2001, Amsterdam: Elsevier Science, 541-546. (1348 pp)
Mouncey NJ, Mitchenall LA, Pau RN: Mutational analysis of genes of the mod locus involved in molybdenum transport, homeostasis, and processing in Azotobacter vinelandii. J Bacteriol. 1995, 177: 5294-5302.
Peijnenburg WJGM, Jager T: Monitoring approaches to assess bioaccessibility and bioavailability of metals: Matrix issues. Ecotoxicol Environ Saf. 2003, 56: 63-77. 10.1016/S0147-6513(03)00051-4.
Soutourina OA, Bertin PN: Regulation cascade of flagellar expression in Gram-negative bacteria. FEMS Microbiol Rev. 2003, 27: 505-523. 10.1016/S0168-6445(03)00064-0.
Studholme DJ, Dixon R: Domain architectures of σ54-dependent transcriptional activators. J Bacteriol. 2003, 185: 1757-1767. 10.1128/JB.185.6.1757-1767.2003.
Rappas M, Schumacher J, Beuron F, Niwa H, Bordes P, Wigneshweraraj S, Keetch CA, Robinson CV, Buck M, Zhang X: Structural insights into the activity of enhancer-binding proteins. Science. 2005, 307: 1972-1975. 10.1126/science.1105932.
Ellis PJ, Conrads T, Hille R, Kuhn P: Crystal structure of the 100 kDa arsenite oxidase from Alcaligenes faecalis in two crystal forms at 1.64 Å and 2.03 Å. Structure. 2001, 9: 125-132. 10.1016/S0969-2126(01)00566-4.
Grunden AM, Shanmugam KT: Molybdate transport and regulation in bacteria. Arch Microbiol. 1997, 168: 345-354. 10.1007/s002030050508.
Parkinson JS, Kofoid EC: Communication modules in bacterial signaling proteins. Annu Rev Genet. 1992, 26: 71-112. 10.1146/annurev.ge.26.120192.000443.
Kustu S, Santero E, Keener J, Popham D, Weiss D: Expression of sigma 54 (ntrA)-dependent genes is probably united by a common mechanism. Microbiol Rev. 1989, 53: 367-376.
Leonhartsberger S, Huber A, Lottspeich F, Böck A: The hydH/G genes from Escherichia coli code for a zinc and lead responsive two-component regulatory system. J Mol Biol. 2001, 307: 93-105. 10.1006/jmbi.2000.4451.
Barrios H, Valderrama B, Morett E: Compilation and analysis of σ54-dependent promoter sequences. Nucleic Acids Res. 1999, 27: 4305-4313. 10.1093/nar/27.22.4305.
Schumacher J, Joly N, Rappas M, Zhang X, Buck M: Structures and organisation of AAA+ enhancer binding proteins in transcriptional activation. J Struct Biol. 2006, 156: 190-199.
Zhang X, Chaney M, Wigneshweraraj SR, Schumacher J, Bordes P, Cannon W, Buck M: Mechanochemical ATPases and transcriptional activation. Mol Microbiol. 2002, 45: 895-903. 10.1046/j.1365-2958.2002.03065.x.
Yang XF, Alani SM, Norgard MV: The response regulator Rrp2 is essential for the expression of major membrane lipoproteins in Borrelia burgdorferi. Proc Natl Acad Sci Unit States Am. 2003, 100: 11001-11006. 10.1073/pnas.1834315100.
Stafford GP, Scanlan J, McDonald IR, Murell JC: rpoN, mmoR and mmoG, genes involved in regulating the expression of soluble methane monooxygenase in Methylosinus trichosporium OB3b. Microbiology. 2003, 149: 1771-1784. 10.1099/mic.0.26060-0.
Zhu L, Peng Q, Song F, Jiang Y, Sun C, Zhang J, Huang D: Structure and regulation of the gab gne cluster, involved in the γ-aminobutyric acid shunt, are controlled by a σ54 factor in Bacillus thuringiensis. J Bacteriol. 2010, 192: 346-355. 10.1128/JB.01038-09.
Debarbouille M, Gardan R, Arnaud M, Rapoport G: Role of bkdR, a transcriptional activator of the SigL-dependent isoleucine and valine degradation pathway in Bacillus subtilis. J Bacteriol. 1999, 181: 2059-2066.
Dombrecht B, Marchal K, Vanderleyden J, Michiels J: Prediction and overview of the RpoN-regulon in closely related species of the Rhizobiales. Genome Biol. 2002, 3 (12): RESEARCH0076-10.1186/gb-2002-3-12-research0076.
Cases I, Ussery DW, De Lorenzo V: The σ54 regulon (sigmulon) of Pseudomonas putida. Environ Microbiol. 2003, 5: 1281-1293. 10.1111/j.1462-2920.2003.00528.x.
Endoh T, Habe H, Yoshida T, Nojiri H, Omori T: A CysB-regulated and σ54-dependent regulator, SfnR, is essential for dimethyl sulfone metabolism of Pseudomonas putida strain DS1. Microbiology. 2003, 149: 991-1000. 10.1099/mic.0.26031-0.
Grigoroudis AI, Panagiotidis CA, Lioliou EE, Vlassi M, Kyriakidis DA: Molecular modeling and functional analysis of the AtoS-AtoC two-component signal transduction system of Escherichia coli . Biochim Biophys Acta Gen Subj. 2007, 1770: 1248-1258. 10.1016/j.bbagen.2007.04.004.
Bordes P, Wigneshweraraj SR, Schumacher J, Zhang X, Chaney M, Buck M: The ATP hydrolyzing transcription activator phage shock protein F of Escherichia coli : Identifying a surface that binds σ54. Proc Natl Acad Sci Unit States Am. 2003, 100: 2278-2283. 10.1073/pnas.0537525100.
Dago AE, Wigneshweraraj SR, Buck M, Morett E: A role for the conserved GAFTGA motif of AAA+ transcription activators in sensing promoter DNA conformation. J Biol Chem. 2007, 282: 1087-1097. 10.1074/jbc.M608715200.
Hartl FU: Molecular chaperones in cellular protein folding. Nature. 1996, 381: 571-580. 10.1038/381571a0.
Mayer MP, Bukau B: Hsp70 chaperone systems: Diversity of cellular functions and mechanism of action. Biol Chem. 1998, 379: 261-268.
Yoon H, Hong J, Ryu S: Effects of chaperones on mRNA stability and gene expression in Escherichia coli. J Microbiol Biotechnol. 2008, 18: 228-233.
Ohki R, Kawamata T, Katoh Y, Hosoda F, Ohki M: Escherichia coli dnaJ deletion mutation results in loss of stability of a positive regulator, CRP. J Biol Chem. 1992, 267: 13180-13184.
Berks BC, Sargent F, Palmer T: The Tat protein export pathway. Mol Microbiol. 2000, 35: 260-274. 10.1046/j.1365-2958.2000.01719.x.
Pérez-Rodriguez R, Fisher AC, Perlmutter JD, Hicks MG, Chanal A, Santini CL, Wu LF, Palmer T, DeLisa MP: An Essential Role for the DnaK Molecular Chaperone in Stabilizing Over-expressed Substrate Proteins of the Bacterial Twin-arginine Translocation Pathway. J Mol Biol. 2007, 367: 715-730. 10.1016/j.jmb.2007.01.027.
Rodriguez F, Arsène-Ploetze F, Rist W, Rüdiger S, Schneider-Mergener J, Mayer MP, Bukau B: Molecular Basis for Regulation of the Heat Shock Transcription Factor sigma32 by the DnaK and DnaJ Chaperones. Mol Cell. 2008, 32: 347-358. 10.1016/j.molcel.2008.09.016.
De Lorenzo V, Herrero M, Jakubzik U, Timmis KN: Mini-Tn5 transposon derivatives for insertion mutagenesis, promoter probing, and chromosomal insertion of cloned DNA in gram-negative eubacteria. J Bacteriol. 1990, 172: 6568-6572.
Sambrook J, Fritsch EF, Maniatis T: Molecular cloning: a laboratory manual. 1989, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 2
Smyth GK, Speed TP: Normalization of cDNA microarray data. Methods. 2003, 31: 265-273. 10.1016/S1046-2023(03)00155-5.
Delmar P, Robin S, Daudin JJ: VarMixt : Efficient variance modeling for the differential analysis of replicated gene expression data. Bioinformatics. 2004, 21 (4): 502-8. 10.1093/bioinformatics/bti023.
Benjamini Y, Yekutieli D: The control of the false discovery rate in multiple hypothesis testing under dependency. Ann Stat. 2001, 29: 1165-1188. 10.1214/aos/1013699998.
Pfaffl MW: A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 2001, 29 (9): e45-10.1093/nar/29.9.e45.
Pfaffl MW, Horgan GW, Dempfle L: Relative expression software tool (REST) for group-wise comparison and statistical analysis of relative expression results in real-time PCR. Nucleic Acids Res. 2002, 30 (9): e36-10.1093/nar/30.9.e36.
Anderson GL, Williams J, Hille R: The purification and characterization of arsenite oxydase from Alcaligenes faecalis, a molybdenum-containing hydroxylase. J Biol Chem. 1992, 267: 23674-23682.
JCA was supported by a grant from the French Ministry of Education and Research. Financial support came from the Centre National de la Recherche Scientifique, the Agence Nationale de la Recherche (ANR 07-BLAN-0118 project) and the Université de Strasbourg. This work was done in the frame of the Groupement de Recherche (GDR2909-CNRS): « Métabolisme de l'Arsenic chez les Micro-organismes».
SK and JCA wrote the manuscript and performed the genetic experiments. SK carried out the quantitative PCR and 5' RACE experiments. JCA performed the Western immunoblotting analysis. CP, OS, MAD and JYC conceived and performed the transcriptomic experiments and the data analyses. DL performed the chemical experiments. FGC, FH, FAP and PB helped to analyze the data and critically revised the manuscript. PB coordinated and conceived the study. All authors read and approved the final manuscript.
Sandrine Koechler, Jessica Cleiss-Arnold contributed equally to this work.
An erratum for this article can be found at http://dx.doi.org/10.1186/s12866-017-0976-8
Electronic supplementary material
Authors’ original submitted files for images
Below are the links to the authors’ original submitted files for images.