Expression of the arsenite oxidation regulatory operon in Rhizobium sp. str. NT‐26 is under the control of two promoters that respond to different environmental cues

Abstract Rhizobium sp. str. NT‐26 is a Gram‐negative facultative chemolithoautotrophic arsenite oxidizer that has been used as a model organism to study various aspects of arsenite oxidation including the regulation of arsenite oxidation. The three regulatory genes, aioX, aioS, and aioR, are cotranscribed when NT‐26 was grown in the presence or absence of arsenite. The aioXSR operon is upregulated in stationary phase but not by the presence of arsenite in the growth medium. The two transcription start sites upstream of aioX were determined which led to the identification of two promoters, the housekeeping promoter RpoD and the growth‐phase‐dependent promoter RpoE2. Promoter–lacZ fusions confirmed their constitutive and stationary phase expressions. The involvement of the NT‐26 sigma factor RpoE2 in acting on the NT‐26 RpoE2 promoter was confirmed in vivo in Escherichia coli, which lacks a rpoE2 homolog, using a strain carrying both the promoter–lacZ fusion and the NT‐26 rpoE2 gene. An in silico approach was used to search for other RpoE2 promoters and AioR‐binding motifs and led to the identification of other genes that could be regulated by these proteins including those involved in quorum sensing, chemotaxis, and motility expanding the signaling networks important for the microbial metabolism of arsenite.

In NT-26, As III is oxidized to As V in the periplasm by the As III oxidase (Aio), which is a bioenergetic enzyme that contains a large catalytic subunit (AioA) with a molybdopterin guanine dinucleotide at its active site and a 3Fe-4S cluster, and a small (AioB) Rieske cluster (Santini & vanden Hoven, 2004;Warelow et al., 2013). Homologs of the aioB and aioA genes have been identified in many phylogenetically diverse prokaryotes including members of the Bacteria and Archaea (van Lis et al., 2013). In many cases, the aioB and aioA genes are either upstream or downstream of three regulatory genes, aioX, aioS, and aioR (Slyemi, Moinier, Talla, & Bonnefoy, 2013). In NT-26, the aioB and aioA genes are in an operon with cytC and moeA1, downstream of a RpoN promoter (σ 54 ), and operon expression is induced by As III (Figure 1a) (Santini et al., 2007). The regulatory genes, aioX, aioS, and aioR, are in a separate operon upstream of aioB (Sardiwal, Santini, Osborne, & Djordjevic, 2010), which has been shown to be constitutively expressed (this study). The proposed regulation of aioB and aioA involves As III sensing by the periplasmic protein AioX and the AioX-As III complex presumably acts as a ligand for the sensor histidine kinase, AioS, which autophosphorylates and then phosphorylates the transcriptional regulator, AioR, which binds upstream of the RpoN promoter (TGGCACAACGATTGCA) switching on transcription (Andres et al., 2013;Kang, Bothner, Rensing, & McDermott, 2012;Liu et al., 2012;Sardiwal et al., 2010).
AioR appears to play a wider role in regulating gene expression in arsenite oxidizers. Recently, it has also been shown to positively regulate gene expression of the chemotaxis gene mcp in Agrobacterium tumefaciens GW4 by binding to the mcp regulatory region (Shi et al., 2017). The AioR-binding consensus sequence was also found upstream of the mcp gene in NT-26 and Herminiimonas arsenicoxydans ULPAs1 (Shi et al., 2017).
The overall aim of this work was to better understand the physiological roles of AioX, AioS, and AioR in regulating gene expression in NT-26. To do this, we studied the expression of the aioX, aioS, and aioR genes under different growth conditions using quantitative reverse transcription PCR (qRT-PCR). We found that the three genes were cotranscribed and that there was an increase in gene expression in stationary phase. Two transcription start sites were identified upstream of aioX which resulted in the discovery of two promoters, RpoD (σ 70 ) and RpoE2 (σ 24 ), that operate under different growth conditions. Promoter functional studies confirmed the differences observed in aioX, aioS, and aioR gene expression under different growth conditions. In silico analyses also implicate the sigma factor RpoE2 in regulating quorum sensing and motility.

| Bacterial culture
A rifampicin-resistant (Rif R ) spontaneous mutant of NT-26 (Santini & vanden Hoven, 2004) was grown in McCartney bottles containing 10 ml minimal salts medium (MSM) containing 0.04% yeast exact (YE) (Oxoid™) with and without 5 mmol/L As III (Santini et al., 2000). Routine transfers were done using a 5% (v/v) inoculum of NT-26 grown overnight in the respective medium. All cultures were incubated at 28°C under aerobic conditions with shaking at 150 rpm. For the qRT-PCR and promoter activity experiments, the cells were grown until late-log (OD 600 from 0.100 to 0.140) and stationary phases (OD 600 from 0.170 to 0.24) (Santini et al., 2000). Escherichia coli was routinely cultured in lysogeny broth (LB).

| Nucleic acid isolation
NT-26 genomic DNA (gDNA) was isolated using the Wizard ® Genomic DNA purification kit (Promega) according to the manufacturer's instructions.
Total RNA was isolated from NT-26 using the SV Total RNA Isolation System (Promega) following the manufacturer's instructions.
To avoid DNA contamination, the DNA-free Kit™ (Ambion) was used according to the manufacturer's instructions and the RNA stored at −80°C. RNA was isolated from five biological replicates for each of the conditions tested by qPCR.
The plasmid pPHU234 (Hübner et al., 1991) and recombinant plasmids were isolated from E. coli using the QIAprep Spin Miniprep Kit (Qiagen) according to the manufacturer's instructions.
F I G U R E 1 Organization of the NT-26 aio gene cluster. (a) aioX encodes periplasmic As III -binding protein; aioS encodes sensor histidine kinase; aioR encodes transcriptional regulator; aioB encodes small subunit of the As III oxidase; aioA encodes the large catalytic subunit of the As III oxidase; cytC encodes a cytochrome c; moeA1 encodes a molybdenum cofactor biosynthesis protein. (b) RT-PCR analysis of the cotranscription of aioX-aioS, and aioS-aioR using aioXF-aioSR and aioSF-aioRR primers (Table S3) and RNA isolated from NT-26 grown to late-log phase with and without As III

| RT-PCR
The Access RT-PCR system kit (Promega) was used to confirm the cotranscription of aioX, aioS, and aioR in accordance with the manufacturer's instructions. To confirm that the samples were free of DNA contamination, the RT step was removed and only DNA polymerase was used in the reaction; no PCR products were obtained in these reactions. The primers used in the RT-PCR are listed in Table S3.

| Real time PCR
qPCR were performed using the PikoReal 96 Real-Time PCR System (Thermo Scientific) using the DyNAmo™ ColorFlash SYBR ® Green qPCR Kit (Thermo Scientific). cDNA was synthesized using the RevertAid Premium First Strand cDNA Synthesis kit (Thermo Scientific) and the quantitative PCR (qPCR) First Strand cDNA Synthesis protocol was performed using random primers provided and according to the manufacturer's specifications. The amount of total RNA used to synthesize cDNA was 1 μg and in the qPCR a final concentration of 2 ng/μl of cDNA was used.
The baseline and quantification cycle (Cq) of each reaction was automatically determined using PikoReal Software version 2.1 (Thermo Scientific). The software qBase+ (Hellemans, Mortier, De Paepe, Speleman, & Vandesompele, 2007) was used to analyze the qPCR data to normalize expression levels of the target genes based on the expression of the reference genes glnA, gyrB, and gltA. qBase+ software was also used to calculate the expression level of each gene in the conditions tested, to plot the results using 95% confidence interval, and to perform the analyses of variance (ANOVAs). When comparing two different conditions, the expression of a gene was considered significantly different when p < .05 and not significantly different when p > .05 based on the ANOVA.

| Promoter identification
The SMARTer RACE 5′ kit from Clonetech (Takara Bio Inc., Shiga, Japan) was used to determine the TSS upstream of aioX following the manufacturer's specifications. All the reagents, cells, enzymes, and vectors used were provided in the kit aside from the gene-specific primers designed for aioX (Table S3). DNA samples were sequenced by GATC Biotech (Germany) using the LIGHTrun™ Sanger Technology (GATC Biotech AG). MEGA 6.0 (Tamura, Stecher, Peterson, Filipski, & Kumar, 2013) was used to analyze the sequence chromatograms.
To identify the TSS, sequences obtained from the different cloned fragments (four for each TSS) were aligned to the NT-26 aioX sequence obtained in the MAGE interface (ID: NT26v4_p10026) (Vallenet et al., 2006) using ClustalW (Thompson, Higgins, & Gibson, 1994).

| Promoter activity
To test whether the putative promoters upstream of aioX were functional, the P aioX1 and P aioX2 fragments were PCR amplified using the primers aioX p1 and aioX p2, forward and reverse, containing BamHI and PstI restriction sites and cloned upstream of a promoterless lacZ gene in the plasmid pPHU234 at the BamHI/PstI sites (Hübner et al., 1991). The plasmids were transferred into NT-26 Rif R by conjugation as described previously (Santini & vanden Hoven, 2004). The promoter assays were performed in E. coli strain S17 λ pir. To quantify the promoter function, β-galactosidase activity was measured as described previously (Zhang & Bremer, 1995), and this was done with three biological replicates (Table S4).

| In silico search for RpoD-and RpoE2regulated genes
For the region identified as P aioX1 , the sequences at −35 and −10 were used to build the pattern [TGGACA-(N)16-TACAGT]. For the region identified as P aioX2 , the RpoE2-binding motif described previously for Sinorhizobium meliloti (Sauviac et al., 2007) was identified by eye and used to construct the pattern [GGAAC-(N)18-TT-(N)8-G]. The two motifs were used as input for the "Find Individual Motif Occurrences" tool (FIMO) (Grant, Bailey, & Noble, 2011), which matched the motifs against a library of upstream regions (up to 400 nucleotides in length from the starting ATG) for every gene in NT-26. The resulting alignments were used to generate a summary motif with MEME/MAST (Bailey & Elkan, 1994).

| Cotranscription of the aioX, aioS, and aioR genes in NT-26
To determine whether the aioX, aioS, and aioR genes were cotranscribed in NT-26, and therefore, part of the same operon, RT-PCR was performed using the RNA isolated from NT-26 grown heterotrophically either in the presence or absence of As III (Figure 1b).
Two sets of primers were used, one to amplify the 3′ end of aioX and the 5′ end of aioS and, and the second to amplify the 3′ end of aioS and the 5′ end of aioR (Figure 1a, see arrows). The results demonstrate that aioX, aioS, and aioR are cotranscribed under both conditions.

| The effect of As III and growth phase on aioX, aioS, or aioR expression
As determined by qRT-PCR, As III had no effect on the expression of aioX, aioS, and aioR (Figure 2a). The qPCR was normalized using the reference genes that encode glutamine synthetase (glnA), citrate synthetase (gltA), and DNA gyrase subunit B (gyrB). These reference genes were selected based on their expression stability in NT-26 when grown heterotrophically in the presence and absence of As III in late-log and stationary phases.
In NT-26, the aioX, aioS, and aioR genes were upregulated in stationary phase of growth when compared to late-log phase. The increase in expression was statistically significant (with p < .05) with increases of 3.7-fold for aioX, twofold for aioS, and threefold for aioR ( Figure 2b and Table S1).

| Identification of two transcription start sites and the associated promoters upstream of aioX
To determine whether the aioXSR operon is under the control of one or more promoters, the transcription start site(s) (TSS) upstream of aioX was determined using 5′ RACE when NT-26 was grown heterotrophically until late-log or stationary phase. Two different TSS were identified: (1) the proximal one named TSS1 and (2) the distal one named TSS2 (Figure 3a). The untranslated region (UTR) was identified and the TSS corresponds to the first 5′ nucleotide of the UTR. These were named UTR1 and TSS1 for the proximal TSS (highlighted in red), and UTR2 and TSS2 for the distal one (highlighted in blue). The presence of two different TSS suggests that this operon is regulated by two different promoters.
To identify the promoters, the two regions upstream of both TSSs were subjected to visual inspection and in silico analysis. For the con- which has been previously shown to be a consensus sequence for a RpoD promoter (Harley & Reynolds, 1987). This consensus sequence was found upstream of 176 different genes in NT-26 (Table S2) (Table S2); such a high number of hits suggest the motif was too generic. Nevertheless, the NT-26 RpoE2 promoter motif was found upstream of the rpoE2 gene, which is also the case in R. meliloti (Sauviac et al., 2007). In addition, the RpoE2binding motif was found upstream of putative genes involved in chemotaxis and motility, qseB and fliG, respectively.

| Functional analysis of the RpoD and RpoE2 promoters in NT-26 using reporter gene fusion
To verify the function of the RpoD and RpoE2 promoters in regulating the aioXSR operon in NT-26, the region upstream of the TSS1 and TSS2, designated P aioX1 and P aioX2 , were cloned upstream of a promoterless lacZ gene in the plasmid pPHU234 (Hübner et al., 1991). The plasmid was transferred into NT-26 by conjugation  (Figure 4). The activity of the P aioX2 -lacZ gene fusion displayed increased β-galactosidase activity over time with the highest activity detected in stationary phase. As expected the P aioX1 -lacZ fusion was constitutively expressed with no significant increase in activity over time.

| The NT-26 RpoE2 is required for activity of the P aiox2 promoter in Escherichia coli
To confirm the involvement of the NT-26 sigma factor RpoE2 in regulating the aioXSR operon, E. coli, which does not naturally contain a rpoE2 homolog, was used as a host for in vivo experiments. E. coli containing P aioX2 alone showed no detectable β-galactosidase activity ( Figure 5), however, when a plasmid containing the NT-26 rpoE2 gene was also provided in trans β-galactosidase activity was detected ( Figure 5).

| DISCUSSION
It has been previously shown that the aioX, aioS, and aioR genes are essential for As III oxidation and expression of the arsenite F I G U R E 3 Identified promoters upstream of aioX. (a) The two identified transcription start sites (TSSs) upstream of aioX are underlined, the TSS1 is underlined in red and the TSS2 is underlined in blue. The six nucleotides flanking the −10 and −35 regions upstream of TSS1 correspond to the RpoD conserved region are in red and in bold. The predicted AioR-binding site is shown italicized and underlined. The conserved regions for the RpoE2 promoter are shown in blue and in bold. (b) The RpoD promoter conserved site motif built with MEME/MAST using conserved nucleotides in NT-26. (c) The RpoE2 promoter conserved site motif built with MEME/MAST using conserved nucleotides in NT-26 oxidase genes in NT-26 (Andres et al., 2013;Sardiwal et al., 2010).
The AioX, AioS, and AioR proteins are thought to be involved in a three-component system involved in the regulation of the aioB and aioA genes in the presence of As III in the growth medium (Andres et al., 2013;Sardiwal et al., 2010). In this study, we have shown that the aioX, aioS, and aioR genes are cotranscribed and that there is no effect on expression of these genes when NT-26 was grown in the presence of As III . Similar results have also been reported for Thiomonas arsenitoxydans 3As, where the presence or absence of As III had no effect on the expression of aioX, aioS, or aioR (Slyemi et al., 2013). However, in A. tumefaciens 5A, the presence of As III induces the expression of aioX (Liu et al., 2012), and in H. arsenicoxydans ULPAS-1, aioX, aioS, and aioR are all upregulated after 8 hr exposure to As III (Cleiss-Arnold et al., 2010).
In NT-26, a growth-phase-dependent effect on expression was observed where the aioX, aioS, and aioR genes were upregulated in stationary phase (Figure 2). These results can be explained by the detection of two different TSSs and the identification of two promoters, RpoD and RpoE2. The TSS upstream of aioX in T. arsenitoxydans 3As was also determined and a RpoD consensus sequence identified (Moinier et al., 2014). Given the constitutive expression of the aioXSR operon, the identification of a RpoD promoter was expected. Perhaps surprisingly, the aioXSR operon was upregulated in stationary phase resulting in the identification of a RpoE2 promoter.
In R. meliloti, the RpoE2 sigma factor is involved in the general stress and starvation response (Sauviac et al., 2007), and this may also be the case in NT-26.
The RpoE2-binding consensus sequence was also found upstream of the genes rpoe2, kat, and qseB (refer to Table S2), the latter two of which encode putative proteins involved in the response to oxidative stress and flagella regulation, respectively. In NT-26, the kat gene was also found to be upregulated by As III (Andres et al., 2013), and in Sinorhizobium meliloti, it is known to be regulated by RpoE2 (Sauviac et al., 2007). In E. coli, the qseB gene encodes a putative regulatory protein involved in quorum sensing and flagella gene expression (Sperandio, Torres, & Kaper, 2002), and was also shown in NT-26 to be upregulated by As III (Andres et al., 2013). NT-26 was also shown to be more motile when grown in the presence of As III , reinforcing the link presented here between the regulation of qseB and the aioXSR operon regulation by RpoE2 (Andres et al., 2013). We hypothesize that in stationary phase when the As III concentration is reduced (Santini et al., 2000), a greater abundance of AioX, AioS, and AioR would allow NT-26 to sense and respond to lower concentrations of As III . This together with the As III -induced regulation of chemotaxis would facilitate its motility toward As III .
In NT-26, the consensus sequences for the predicted AioR-binding site was found upstream of the putative chemoreceptor-encoding gene, mcp (Shi et al., 2017). Mcp is a chemoreceptor described in A. tumefaciens GW4 to bind As III and it has been shown to be important for chemotaxis toward As III in this organism (Shi et al., 2017). In NT-26, it F I G U R E 4 β-Galactosidase activity determined at different growth stages for the lacZ-promoter in NT-26 grown heterotrophically (with 0.04% yeast extract, YE) with and without As III . The negative control samples (i.e., the promoterless pPHU234 plasmid) are shown in black. The β-galactosidase activity for NT-26 containing the RpoE2 promoter (P aioX1 ) in trans is shown in blue and NT-26 containing the RpoD promoter (P aioX2 ) in trans is shown in red. The data plotted correspond to the average of three independent experiments F I G U R E 5 β-Galactosidase activity determined for the lacZ-promoter fusions also harboring the NT-26 rpoE2 gene in Escherichia coli. C, negative control corresponds to the activity of the promoterless plasmid pPHU234; PaioX1, negative control of the RpoE2 promoter alone; P aioX1 + RpoE2, RpoE2 promoter and also a plasmid harboring the NT-26 rpoE2 gene. The β-galactosidase activity plotted is the average of three independent experiments seems that Mcp may also be involved in chemotaxis toward As III and that mcp expression is under the control of AioR, further strengthening the link between As III sensing, As III oxidation, and As III -induced chemotaxis. The chemotaxis genes, involved in regulating the activity and direction of the flagella (Wadhams & Armitage, 2004), cheY, cheR, cheW, and cheD also contain a putative RpoE2-binding site in their promoter regions (Table S2) (no RpoD promoters were identified).
The presence of As III in the growth medium has no effect on the expression of the aioXSR operon, however, the results presented herein introduces possible links between the regulation of As III oxidation, motility, and quorum sensing, and opens up the possibility that the regulatory proteins AioX, AioS, and AioR may have other roles other than regulating As III oxidation in NT-26. The results presented here also suggest the involvement of the sigma factor RpoE2 in the regulation of As III oxidation and the link to chemotaxis and confirms the involvement of RpoD in regulating expression of the aioXSR operon as first suggested for T. arsenitoxydans 3As (Moinier et al., 2014).