A periplasmic arsenite-binding protein involved in regulating arsenite oxidation Authors: Guanghui Liu, Mengyao Liu, Eun-Hae Kim, Walid S. Maaty, Brian Bothner, Benfang Lei, Christopher Rensing, Gejiao Wang, and Timothy R. McDermott This is the peer reviewed version of the following article: [Liu, Guanghui, Mengyao Liu, Eun-Hae Kim, Walid S. Maaty, Brian Bothner, Benfang Lei, Christopher Rensing, Gejiao Wang, and Timothy R. McDermott. “A Periplasmic Arsenite-Binding Protein Involved in Regulating Arsenite Oxidation.” Environmental Microbiology 14, no. 7 (December 19, 2011): 1624–1634], which has been published in final form at http://dx.doi.org/10.1111/j.1462-2920.2011.02672.x. This article may be used for non-commercial purposes in accordance with Wiley Terms and Conditions for Self-Archiving. Liu, Guanghui, Mengyao Liu, Eun-Hae Kim, Walid S. Maaty, Brian Bothner, Benfang Lei, Christopher Rensing, Gejiao Wang, and Timothy R. McDermott. “A Periplasmic Arsenite-Binding Protein Involved in Regulating Arsenite Oxidation.” Environmental Microbiology 14, no. 7 (December 19, 2011): 1624–1634. doi:10.1111/j.1462-2920.2011.02672.x. Made available through Montana State University’s ScholarWorks scholarworks.montana.edu A periplasmic arsenite-binding protein involved in regulating arsenite oxidation Guanghui Liu,1,2 Mengyao Liu,3 Eun-Hae Kim,5 Walid S. Maaty,4 Brian Bothner,4 Benfang Lei,3 Christopher Rensing,5 Gejiao Wang1** and Timothy R. McDermott2* 1State Key Laboratory of Agricultural Microbiology, College of Life Science and Technology, Huazhong Agricultural University, Wuhan 430070, China. Departments of 2Land Resources and Environmental Sciences, 3Immunology and Infectious Disease, and 4Chemistry and Biochemistry, Montana State University, Bozeman, MT 59717, USA. 5Department of Soil, Water, and Environmental Science, The University of Arizona, Tucson, AZ 85721, USA. Summary Arsenic (As) is the most common toxic element in the environment, ranking first on the Superfund List of Hazardous Substances. Microbial redox transforma-tions are the principal drivers of As chemical specia-tion, which in turn dictates As mobility and toxicity. Consequently, in order to manage or remediate envi-ronmental As, land managers need to understand how and why microorganisms react to As. Studies have demonstrated a two-component signal trans-duction system comprised of AioS (sensor kinase) and AioR (response regulator) is involved in regulat-ing microbial AsIII oxidation, with the AsIII oxidase structural genes aioB and aioA being upregulated by AsIII. However, it is not known whether AsIII is first detected directly by AioS or by an intermediate. Herein we demonstrate the essential role of a peri-plasmic AsIII-binding protein encoded by aioX, which is upregulated by AsIII. An DaioX mutant is defective for upregulation of the aioBA genes and conse-quently AsIII oxidation. Purified AioX expressed without its TAT-type signal peptide behaves as a monomer (MW 32 kDa), and Western blots show AioX to be exclusively associated with the cytoplasmic membrane. AioX binds AsIII with a KD of 2.4 mM AsIII; however, mutating a conserved Cys108 to either alanine or serine resulted in lack of AsIII binding, lack of aioBA induction, and correlated with a negative AsIII oxidation phenotype. The discovery and charac-terization of AioX illustrates a novel AsIII sensing mechanism that appears to be used in a range of bacteria and also provides one of the first examples of a bacterial signal anchor protein. Introduction Transport and bioavailability of arsenic (As) in the envi-ronment is dependent on chemical speciation; hence, the abiotic and biotic processes that regulate arsenite (AsIII) oxidation and arsenate (AsV) reduction have important implications for watershed quality management in As-impacted environments. The various abiotic and biotic factors that control As fate and transport are not mutually exclusive; however, it is now understood that microbial As redox transformations are an important (if not the princi-pal) force controlling As speciation in most environments (Cullen and Reimer, 1989; Pontius et al., 1994; Inskeep et al., 2001; Oremland and Stolz, 2005; Stolz et al., 2006). Thus, in order to better understand microbe–As interac-tions in nature and to more effectively strategize bioreme- diation efforts, it is critical that there be a more formal and foundational understanding of how microbes sense and react to As. At present, a fairly detailed model explaining the genet-ics, regulation and function of detoxification-based AsV reduction is in place. At the minimum, this involves pro-teins encoded by arsRBC: ArsR is a repressor controlling the expression of arsRBC, ArsB extrudes AsIII from the cell, and ArsC is an AsV reductase that converts AsV to AsIII, which is the substrate for ArsB (Bhattacharjee and Rosen, 2007). In addition, AsV reductase enzymes involved in anaerobic AsV respiration have been charac-terized from three organisms (Kraft and Macy, 1998; Afkar et al., 2003; Saltikov and Newman, 2003; Malasarn et al., 2007) and the encoding genes (arrAB) have been char-acterized (Saltikov and Newman, 2003). Studies on the genetics and physiology of AsIII oxida-tion are at an initial stage. Early accomplishments include the characterization of one of the two identified AsIII oxi-dases (Phillips and Taylor, 1976; Anderson et al., 1992; Ellis et al., 2001). The genes coding for AsIII oxidase have been cloned (Muller et al., 2003; Santini and vanden Hoven, 2004), and more recently a phylogenetically dis- tinct AsIII oxidase from Alkalilimnicola ehrlichii was cloned and characterized (Zargar et al., 2010). Note that in this report we are installing modified gene symbol nomencla- ture resulting from recent international discussions designed to unify the arsenite oxidase literature with respect to gene symbols and to eliminate confusion with other proteins. To that end, aox/aro/aso are all now des- ignated aio (arenite oxidase), and the arsenite oxidase large subunit is designated as A and the small subunit as B, e.g. aoxAB is now aioBA (see Lett et al., 2011). Our previous efforts with Agrobacterium tumefaciens identified a two-component signal transduction pair, aioSR (previously aoxSR), as being essential to AsIII oxi- dation (Kashyap et al., 2006a). In addition, Kashyap and colleagues (2006b) found that a molybdate transporter and a Na+/H+ antiporter are also essential for AsIII oxida- tion. Later, Koechler and colleagues (2010) using a similar transposon mutation approach also identified the aioSR two-component pair and molybdate transporter as being essential for AsIII oxidation, and in addition provided evi- dence that RpoN (alternative sigma factor, s54) and DnaJ (heat shock protein J) are also required for this process. The experiments summarized in the current study take the next step in characterizing important regulatory ele- ments that control bacterial AsIII oxidation. Specifically, we describe a gene and its encoded protein that is essen- tial for the upregulation of aioAB and for AsIII oxidation. We present the initial characterization of the encoded protein, which behaves in a manner consistent with it being a periplasmic AsIII-binding protein. Results Identification and expression analysis of aioX All mutants, plasmids and genetic constructs generated for this study are shown in Table 1. Previous work identi- fied an open reading frame upstream of the arsenite oxidase regulatory locus in A. tumefaciens strain 5A (Kashyap et al., 2006a). To understand its potential role in AsIII oxidation, the full coding sequence (921 bp) was determined by TAIL-PCR and sequencing, and is referred to as aioX as it is a homologue of the aioX gene referred to as aoxX by Cai and colleagues (2009) and encodes a 306-amino-acid protein that shares significant identity and similarity with a number of variously annotated solute- binding proteins and can be found associated with aio genes in a range of AsIII oxidizing microbes (Figs S1 and S2). Because genes involved in AsIII oxidation are typically induced by AsIII, aioX expression was examined by quantitative reverse transcriptase (qRT)-PCR. Levels of aioX transcript in early log-phase cells in the presence of 100 mM AsIII were 3615  411 copies per nanogram of Table 1. Bacterial strains and plasmids used in this study. Strain/plasmid Relevant properties or derivation Source or reference Strains Agrobacterium tumefaciens 5A Wild type, soil isolate, As(III) oxidizing Macur et al. (2004) M53 aioX deletion mutant This study Escherichia coli DH5a supE44 lacU169(j80lacZM15) hRDR17 recA1 endA1 gyrA96 thi-1 relA1 Hanahan (1983) HB101 supE44 hsdS20 (rB- mB-) recA13 ara-14 proA2 lacY1 galK2 rpsL20 xyl-5 metl-1 Boyer and Roulland-Dussoix (1969) JM110 dam and dcm deficient, cloning and expression host Stratagene TOP10 High-competency cloning host Invitrogen BL21Star™(DE3)pLysS F- ompT hsdSB (rB- mB-) gal dcm me131 (DE3) pLysS (CamR) Invitrogen Plasmids pGEM-T PCR TA cloning vector; AmpR Promega pRK2013 Conjugation helper plasmid; KanR Figurski and Helinski (1979) pJQ200SK sacB sacR SucS; GentR Quandt and Hynes (1993) pCPP30- Broad host range; TetR Michael Kahn, Washington State University pCPP30::aioX pCPP30 with 1412 bp fragment PCR cloned from strain 5A containing 304 bp upstream sequence of aioX, whole aioX coding region and 191 bp partial aioS; TetR This study pCPP30::Cys108Ser Site-directed mutagenesis of Cys-108 codon to Ser codon in aioX gene of pCPP aioX; TetR This study pET-52b(+) T7 RNA polymerase-based expression vector; AmpR Novagen pETaioX3 804 bp NcoI–SacI fragment of aioX gene without signal sequence and stop codon cloned into the multiple sites of pET-52b(+); AmpR This study pETaioX3C108A Site-directed mutagenesis of Cys-108 codon to Ala codon in aioX gene of pET-52b(+); AmpR This study pETaioX3C108S Site-directed mutagenesis of Cys-108 codon to Ser codon in aioX gene of pET-52b(+); AmpR This study total RNA, which was sevenfold that of AsIII-naïve cells (516 106 copies per nanogram of total RNA). Requirement of aioX for AsIII oxidation To investigate the role of aioX in AsIII oxidation, a deletion mutation (nucleotides 12–861) was created in aioX (Fig. S3). Loss of AsIII oxidation in the DaioX mutant strain M53 was quantitatively demonstrated by HPLC-ICP-MS analysis of culture supernatants (Fig. 1). After 7 h growth in 100 mM AsIII, growth profiles were similar for the differ- ent strains (Fig. 1A), although no AsIII oxidation was detected in mutant M53 or M53 carrying the control plasmid pCPP30 (Fig. 1B). In contrast, AsIII oxidation was observed with the wild-type strain and with M53 contain- ing pCPP30::aioX (Fig. 1B). Association of AioX with the cytoplasmic membrane SignalP 3.0 software (http://www.cbs.dtu.dk/services/ SignalP) predicted the AioX N-terminus contains a 39-amino-acid signal peptide with the twin-arginine trans- location (TAT) motif SRRMAIG (Fig. S2). The signal peptide exhibits features consistent with it being an uncleaved signal anchor, with hidden Markov modelling predicting amino acids 1–20 to be cytoplasmic (n-region), amino acids 21–43 to be membrane spanning (h-region), and the balance of the protein (aa 44–306) to be periplas- mic (c-region). Western blot analysis of cytoplasmic, peri- plasmic and cytoplasmic membrane-partitioned proteins with anti-AioX antibodies prepared against purified AioX (Fig. 2A) showed that AioX to be strictly associated with the cytoplasmic membrane (Fig. 2B). This is consistent with AioX being anchored to the cytoplasmic membrane in an N-in/C-out orientation, i.e. a type II membrane protein (von Heijne, 1988). Note also that the apparent molecular weight of the monomer released from the solubilized membrane (MW = 34.2 kDa; Fig. 2B) is consistent with AioX retaining the signal peptide. To verify the cell frac- tionation procedure accurately reflects the targeted cell components, the above fractions were assayed for alka- line phosphatase (phosphate stressed cells), a known periplasmic marker protein expressed under phosphate- stress conditions (Wanner, 1996). The periplasm fraction contained 77% of total alkaline phosphatase activity, illus- trating that the periplasm fraction was enriched with peri- plasm proteins, although apparently lacking AioX. AioX is involved in regulation Since solute-binding proteins are often involved in regu- lation, the potential for AioX to be involved in AsIII-based regulation was then examined by quantitative RT-PCR analysis of expression of the AsIII oxidase structural genes aioBA (Fig. 3). In AsIII naïve cells, aioBA was not detected in either the wild-type or mutant strains (Fig. 3A), whereas in AsIII-treated cells the presence of aioBA mRNA was readily evident in the wild-type strain but absent in mutant M53 (Fig. 3A). The presence of pCPP30::aioX converted the mutant back to wild type with respect to AsIII-based regulatory control of aioBA (Fig. 3B). Fig. 1. AsIII oxidation properties of the wild-type strain 5A and the DaioX mutant. Example of reproducible experiments illustrating: (A) growth profiles and (B) AsIII oxidation profiles in the presence of 100 mM AsIII. () wild-type strain 5A; () D aioX mutant M53; () M53 (pCPP30::aioX); and () M53 (pCPP30). Fig. 2. Purification and Western blot analysis of AioX. A. Relative purity of the wild-type (lane 1) and Cys108Ser mutant (lane 2) versions of AioX purified without the signal peptide and with a His6 tag; 2.5 mg of total protein was loaded in each well. B. Western blots of AsIII induced strain 5A total periplasm (lane 1), cytoplasm (lane 2) and cytoplasmic membrane (lane 3) extracts, and of purified AioX (lane 4, separate gel). Each cell extract fraction was loaded as 2.5 mg total protein, whereas 0.25 mg of purified AioX was used as positive control. Discussion This study describes a significantly novel development in our understanding of how microbe-arsenic interactions are regulated. Heretofore, only the repressor ArsR and the AsIII metallochaperone ArsD proteins involved in regulating arsenic detoxification have been shown to specifically interact with AsIII via distinct cysteine resi- dues (Bhattacharjee and Rosen, 2007). The frequent occurrence of aioX in known AsIII-oxidizing microbes (Fig. S1) and its close physical proximity to other genes known to be essential to AsIII oxidation (Fig. S1) sug- gested aioX encodes a protein important for AsIII oxida- tion. The present study confirmed the essential nature of this gene (Fig. 1B, Fig. S4), and that it encodes a protein that is critical for normal regulatory control of genes also known to be essential to AsIII oxidation (Figs 1 and 3). Annotation, protein modelling exercises and amino acid alignments of AioX with its homologues suggested this protein is located to the periplasm, being translo- cated via a TAT signal peptide (Fig. S2). However, Western blot evidence showed that AioX was exclusively associated with the cytoplasmic membrane and electro- phoresed as a 34.1 kDa peptide (Fig. 2), arguing that the signal peptide remains uncleaved. As such, this provides experimental support to suggest that AioX is a Gram- negative signal anchor protein, for which there are very few examples (Nielsen and Krogh, 1998; Brunak et al., 2007) and thus another novel observation deriving from this investigation. When expressed and characterized without the signal peptide (to afford purification), AioX was found to behave as a monomer (Fig. 4) and thus is consistent with descrip- tions of other characterized periplasmic solute-binding proteins (Lever, 1972; De Pina et al., 1995; He et al., 2009). Evidence of AsIII binding is threefold: (i) in size exclusion chromatography experiments, AioX eluted as a monomer, although pre-incubation with AsIII altered AioX behaviour in the gel matrix, indicating that AsIII binding effects a conformational change in AioX (Fig. 4), (ii) this conformation change is consistent with reproducible shifts in the intrinsic fluorescence profile with incremental increases in added AsIII, yielding a KD estimate of 2.4 mM AsIII (Fig. 5B), and (iii) direct binding assays illustrated AsIII co-eluting with AioX and separate from the AsIII breakthrough (Fig. 5A). Substitution of the highly conserved Cys108 resulted in complete loss of AsIII oxidation (Fig. S4), lack of aioBA induction (Fig. 3B), and collapse of AsIII binding (Fig. 5A and B). An AsIII : AioX stoichiometry less than unity is perhaps explained by AsIII disassociating from AioX during transit in the column, although we draw attention to the similar stoichiometry for the sulfhydryl-specific dye used in the ligand binding competition experiments that demonstrated the potential for AsIII to disassociate from AioX (Fig. 6). The exact mechanism of AsIII binding pre- sumably involves Cys108, although we note this substan- tially deviates from the current AsIII-binding paradigm that always involve proteins (ArsR and ArsD) employing mul- tiple cysteines and typically as homodimers (Bhattachar- jee and Rosen, 2007). Potentially, AioX may actually function as a multimer while anchored to the membrane. Current efforts are underway to further elucidate AioX form and function of AsIII binding. Our previous efforts yielded a fairly simple model for explaining AsIII oxidation regulation (Kashyap et al., 2006a). AsIII was viewed to be detected by the sensor kinase AioS, which then phosphorylates its cognate response regulator AioR that then actuates the upregula- tion of essential genes such as aioBA (Kashyap et al., 2006a). A recent study by Sardiwal and colleagues (2010) demonstrated AioR being phosphorylated by a compo- nent of AioS and thus supports this model. However, the experiments reported herein illustrate that this above model is too simple. The DaioX mutant was totally devoid of aioBA expression (Fig. 3), implying AioX plays an essential role, likely either in sensing or in transducing the AsIII signal. Given the combined evidence showing AioX to be a membrane bound exported protein (Fig. S2, Fig. 2) capable of binding AsIII (Figs 4–6), it is reasonable at this juncture to conclude that AioX is a periplasm located AsIII-binding protein. The role of periplasmic solute-binding proteins in regu- lating gene expression via two-component systems is well supported by the literature (reviewed by Tam and Saier, 1993; Mascher et al., 2006). This class of proteins is sometimes involved in initializing solute transport, such as PstS in high-affinity phosphate transport (Luecke and Quiocho, 1990). In preliminary experi- ments, we as yet have been unable to show that AioX is required for AsIII uptake. Given the regulatory phenotype of the DaioX mutant (Fig. 3) and the properties of AioX (Figs 2–5 and Fig. S2), our current working model is that AioX actually functions as the AsIII sensor, with AsIII binding resulting in a conformational change that then facilitates intramolecular communication with the sensor kinase AioS. This would be similar to the A. tumefaciens glucose-binding protein, ChvE, which mediates the sugar-induced virulence response governed by the VirA/ VirG two-component signal transduction system (He et al., 2009). In summary, the combination of experiments summa- rized herein support an updated model that now includes AioX as part of an apparent three component system that involves AioX binding AsIII, and that then initiates signal transduction by either AsIII being transferred to AioS or an AioX conformational change causing AioX to then specifi- cally interact with AioS to facilitate the same signalling pathway. Accordingly, molecular and genetic regulatory models for microbial AsIII oxidation should now reflect a more sophisticated system. Experimental procedures Bacterial strains and growth conditions Bacterial strains and plasmids used in this study are listed in Table 1. The A. tumefaciens strains were grown at 30°C in a defined minimal mannitol ammonium (MMN) medium as described previously (Somerville and Kahn, 1983), but modified to contain 50 mM phosphorus. Escherichia coli strains were cultured at 37°C in Luria–Bertani medium (Sambrook et al., 1989). Kanamycin (25 mg ml-1), gentamicin (25 mg ml-1), tetracycline (10 mg ml-1), chloramphenicol (34 mg ml-1) or ampicillin (100 mg ml-1) were added as needed. Isopropyl b-D-1-thiogalactopyranoside (IPTG, 0.4 mM) and 4-chloro-3-indoyl-b-D-galactoside (X-Gal, 40 mg per litre of agar medium) were added as required for detect- ing PCR amplicons cloned into the pGEM-T vector. Identification of the aioX gene Thermal asymmetric interlaced polymerase chain reaction (TAIL-PCR) was performed to identify the complete aioX gene in strain 5A as described previously by Liu and Whittier (1995). Three specific primers (5ALB1, 5′-CTTTCC CGCTGTCGTG-3′, 5ALB2, 5′-GAGAGCAGTTCCGGTTC TG-3′ and 5ALB3, 5′-GGCCAGATAGGTCTTCGTGAC-3′) complementary to the sequences flanking the aioSRBA locus (GenBank Accession No. DQ151549) were used in combina- tion with short arbitrary, degenerate primers (Liu and Whittier, 1995). After TAIL-PCR, all products were gel-purified, sub- cloned into pGEM-T and sequenced. Multiple inferred amino acid sequence alignments of putative AioX homologues obtained from NCBI database were performed with CLUSTALX 2.0 (Thompson et al., 1997). Creation of aioX mutants and complementation A deletion mutation was introduced into the aioX coding region using cross-over PCR. Two separate PCRs were per- formed with primer pairs: DaioX1 [GGATCCGCGACAGGT GCGGATGA] with DaioX2 [CCCATCGATTAAACTTAAACA CGATTTGACACCGACGACCTCCCTC] and DaioX3 [TGT TTAAGTTTAATCGATGGGCCACCCGAAAGCTACGA] with DaioX4 [GGATCC CGAAAGAGGCGTGTATGGTCCCGAT] (BamHI restriction sites in bold) to generate fragments upstream and downstream of the DNA targeted for deletion. The resulting products were mixed and used as template for cross-over PCR with primers DaioX1 and DaioX4. The two fragments were capable of annealing to each other by the 21 bp complementary tag sequence (underlined in the primers DaioX2 and DaioX3). The generated fusion frag- ment of ~ 1400 bp was subcloned into BamHI-digested pJQ200SK, resulting in pJQ200SKDaioX, which was then mobilized into strain 5A by the conjugative E. coli HB101 (pRK2013). The wild-type aioX allele was then replaced by the DaioX gene using levansucrase selection as we previ- ously described (McDermott and Kahn, 1992). GmR merodip- loid transconjugants were selected on minimal mannitol agar and then transconjugants were subcultured onto minimal mannitol-15% sucrose agar. SucroseR GmSen transconjugants were then screened using diagnostic PCR to identify a double recombinant, followed by sequencing of the amplicons to verify that the correct mutation had been introduced. The resulting DaioX mutant is referred to as strain M53. A Cys108Ala mutation and a Cys108Ser mutation were introduced using a QuikChange II XL Site-Directed Mutagen- esis Kit (Stratagene). The Cys108 residue was changed to Ala (C108A) using the primers (EHKC108AF, GGCCGCCT GGATTGCGGGCTACCCGTTCATG and EHKC108AR, CA TGAACGGGTAGCCCGCAATCCAGGCGGCC). The change to serine used primers EHKC108SF (5′-GGCCGCCTGGAT TTCTGGCTACCCGTTCATG-3′) and EHKC108SR (5″-CA TGAACGGGTAGCCAGAAATCCAGGCGGCC-3′). Plasmid DNA was isolated with the QIAprep® Spin Miniprep Kit (Qiagen). Restriction endonuclease digestion, DNA purifica- tion, ligation and transformation were performed with the manufacturer’s standard protocols. All putative mutations were confirmed by DNA sequencing. PCR and DNA sequencing For mutant complementation, a 1412 bp fragment containing the complete aioX coding region along with 304 bp upstream sequence and 191 bp downstream DNA (including part of downstream adjacent aioS) was PCR-amplified using primers [CGGCCGGGTGGTAGCGAGCGAAAT, PstI site in bold and GAATTCGGGATAAGAGCGGTAGACAA, EcoRI in bold], and subcloned into PstI+EcoRI-digested pCPP30. The resulting plasmid was transferred into the mutant M53 by conjugation (as above). Gene expression analyses Reverse transcriptase PCR (RT-PCR) and quantitative (q)RT-PCR were applied to assess transcription of the genes described in this study. Total RNA was extracted from strains 5A or M53 grown with and without 100 mM AsIII using RNeasy® Mini Kit (Qiagen), then treated with DNase using Turbo DNA-free (Ambion, Austin, TX) and purified using the Ambion MEGAclear kit (Ambion, Austin, TX) following the manufacturer’s instructions. RNA preparations were verified to be free of DNA by PCR. Primers aioXRTF (5′-TCATA CCTCATCGTCGGTCA-3′) and aioXRTR (5′-GAGCGCGT TTCTTATTCTGG-3′) were designed for routine RT-PCR monitoring of aioX. Primers P4 and P5 (described in Kashyap et al., 2006a) were used for detecting expression of aioBA. The RT-PCRs were performed using the Access Quick RT-PCR system (Promega) following the manufacturer’s rec- ommended protocol. The annealing temperature for the aioX RT-PCRs was 53°C. For qRT-PCR, primer XrealF (5′-TG GATACGTCTGGGAAGTCATG-3′) paired with XrealR (5′- GCGTTTCTTATTCTGGCAACC-3′) were used. For RT- PCRs involving the 16S rRNA, primers 8F and 1392R were used as we have previously described (Kashyap et al., 2006a). For qRT-PCR, 10 ng of total RNA was first reverse transcribed by M-MLV Reverse Transcriptase (Ambion, Austin, TX). The resulting cDNA was then used as template for qPCR with the GoTaq® qPCR Master Mix (Promega). The standard curves in quantitative RT-PCR were generated using plasmid pGEM-T containing cloned target sequences. Plasmid construction, expression and purification of AioX proteins The AioX wild-type and mutant proteins were expressed in E. coli BL21 Star™(DE3)pLysS with aioX genes on vector pET-52b(+). The aioX coding region without the 117 bp signal peptide sequence was PCR cloned by Pfx50™ DNA Polymerase (Invitrogen) with primers 5′-CCATGGGCGA GTTGCTGTCCGTG-3′ (NcoI site in bold) and 5′-GAGCTC CCCTAGCCTCCGAACAC-3′ (SacI site in bold), and sub- cloned into NcoI+SacI-digested sites of pET-52b(+), resulting in pETaioX3 or pETaioX3Cys108Ser, each with a C-terminal his tag. Cells were grown at 37°C overnight in 50 ml of M9ZB medium (Studier, 1991) containing the required antibiotics and transferred into 3 l of fresh media with the same antibi- otics. Cells were induced at an OD600 of ~ 0.5 by adding IPTG, harvested by centrifugation (7000 g for 10 min at 4°C) after induction (overnight for the wild-type protein and 6 h induc- tion for mutant protein), and suspended in 20 mM Tris-HCl, pH 8.0. After washing with Tris-HCl, the pellets were lysed via sonication on ice for 10 min. Unbroken cells were removed by centrifugation at 12 000 g for 20 min. The soluble superna- tant was mixed with l ml of pre-equilibrated TALON® Metal Affinity Resins (Clontech) and gently agitated at 4°C for 30 min on a platform shaker to allow the polyhistidine-tagged protein to bind the resin. The resin was transferred to a 2 ml gravity-flow column and washed by 10 ml of Tris-HCl, then eluted with a linear gradient of 0.005–0.5 M imidazole in Tris-HCl, pH 8.0. Fractions were collected, analysed by SDS-PAGE and verified to be AioX by Matrix Assisted Laser Desorption/Ionization Time-of-Flight (MALDI-TOF) Mass Spectrometry. Fractions containing a single band corre- sponding to the 32 kDa polyhistidine-tagged protein was observed with > 95% purity and dialysed against 3 l of 20 mM Tris-HCl (pH 8.0) overnight at 4°C. An additional clean-up step was included to remove some non-target proteins that bound to the resin. Proteins were loaded into DEAE Sepharose Fast Flow Column (GE Healthcare), which was pre-equilibrated with Tris-HCl, pH 8.0 and eluted with a linear gradient of 0–0.5 M NaCl. The AioX was pooled and stored at -80°C until use. Protein concentration was determined using Pierce® BCA Protein Assay Kit (Thermo Scientific). Measurement of arsenite binding: (i) AsIII–protein association Purified wild type and the Cys108Ser mutant AioX were incu- bated with AsIII (AsIII : AioX molar ratio of 2:1) at room temperature for 1 h. The protein-AsIII mixture was passed through a Sephadex® G-25 (fine) desalting column (GE Healthcare) that had been pre-equilibrated with 20 mM Tris- HCl, pH 8.0. Eluted fractions were analysed for protein via the Bradford assay and for AsIII by inductively coupled mass spectrometry (ICP-MS) (Agilent 7500ce). Measurement of arsenite binding: (ii) fluorescence spectroscopy Fluorescence measurements were performed with a Cary Eclipse Fluorescence Spectrophotometer (Varian Scientific Instruments, Mulgrave, Australia) at room temperature. Tryp- tophan fluorescence was monitored with an excitation wave- length of 280 nm and the emission scans were conducted between 300 and 500 nm. The wild-type and mutant proteins were 1.0 mM in 20 mM Tris-HCl (pH 8.0), unless otherwise noted. The fluorescence of Tris-HCl buffer alone or with AsIII was performed as blank control between scanning of dupli- cate samples for each AsIII concentration used. Size exclusion chromatography Size exclusion chromatography was performed on an ÄKTA FPLC system (Amersham Pharmacia Biotech) using a Agilent Bio SEC-3 Column (3 mm, 300 Å, 4.6 mm ¥ 300 mm). The mobile phase contained 20 mM Tris-HCl (pH 7.5) with 150 mM NaCl and was run at a flow rate of 0.2 ml min-1. The protein concentration was measured using UV absorbance at 280 nm. Two standard proteins (conalbumin, 75 kDa and myoglobin, 16.7 kDa) were used to estimate molecular mass based on elution volume, primarily to distinguish between the monomer and dimmer forms of AioX. Cell extract preparation The periplasmic, cytoplasmic and membrane proteins were fractionated from AsIII-exposed cells based on the procedure of De Maagd and Lugtenberg (1986) with modification. In brief, in addition to adding 100 mM AsIII, the cells were starved for phosphate in order to induce alkaline phos- phatase, which was the periplasmic marker enzyme. The cells were pelleted by centrifugation, resuspended in 5 ml of suspension buffer (200 mM Tris-HCl, pH 8.0, 0.5 M sucrose, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride) supple- mented with 0.2 mg of lysozyme per ml, and incubated at room temperature for 30 min. The cells were pelleted by centrifugation at 4°C and the supernatant was saved as the periplasmic fraction. The resulting pellet contained sphero- plasts. To isolate the cytoplasmic proteins, the spheroplasts were lysed by suspension in 5 ml of chilled double-distilled water, followed by brief sonication. Unspheroplasted cells were removed by low-speed centrifugation. The supernatant was reserved as the cytoplasmic + cytoplasmic membrane fractions which were further fractionated by ultra- centrifugation (262 000 g for 1 h at 4°C) to pellet the mem- branes. The supernatant was reserved as the cytoplasmic fraction. The membrane pellet was then dissolved in SDS sample buffer and kept as the membrane fraction. Proteins of the supernatant (periplasmic and cytoplasmic fractions) were concentrated by acetone precipitation. Alkaline phosphatase (AP) activity was measured by recording the hydrolysis of p-nitrophenyl phosphate (Bessey et al., 1946). Briefly, 10 ml of the different fractions were mixed well with 280 ml of buffer A (300 mM 2-amino-2-methyl-1, 3-propanediol, 2 mM MgCl2, pH 10.25) and 10 ml of substrate buffer (400 mM 4-nitrophenyl phosphate). The hydrolysis of p-nitrophenyl phosphate was measured spectrophotometrically at 405 nm and 37°C with a SpectraMax Plus384 UV/Vis spectro- photometer (Molecular Devices, Sunnyvale, CA). One unit of the AP activity was defined as catalysing 1 micromole of p-nitrophenyl phosphate per minute per milligram of protein at 37°C. Mouse immunization and Western immunoblot analysis Two female BALB/c mice (4 weeks old; from National Cancer Institute, Frederick Animal Production Area, Frederick, MD) were immunized subcutaneously and then boosted with 30 mg of purified recombinant AioX suspended in 160 ml of aluminium hydroxide gel (Sigma) on days 1 and 14. Immune antiserum was collected 2 weeks after the second boost. Western immunoblot analysis was performed to detect the location of the AioX as described (Lei et al., 2004). Equal amounts of fractions and purified recombinant proteins were resolved by SDS-PAGE and transferred to nitrocellulose membranes (BioTrace™ NT, Pall Corporation) with Towbin transfer buffer using a Trans-Blot SD semidry transfer cell (Bio-Rad) at 15 V for 45 min. The membrane was blocked with 1:20 Amersham Liquid Block in Tween buffer (0.1% Tween 20 in PBS) for 1 h and subsequently incubated for 1 h with anti-AioX mouse antiserum added to the block solution (1:500 dilution). The membrane was then rinsed twice and washed three times for 15 min each with 0.1% Tween 20 in PBS. The membrane was incubated for 1 h with goat anti- mouse IgG-HRP (1:2000 dilution, Santa Cruz Biotechnology) in the block solution and rinsed and washed as described above. Antibody–antigen interaction was visualized by enhanced chemiluminescence. Binding-competition assays The thiol-reactive dye BODIPY® 577/618 maleimide (Invitro- gen) was used to demonstrate the reversible binding between AioX and AsIII. AioX and AsIII (2:1 molar ratio) were incubated for 1 h incubation at room temperature, followed by dye addition (1:1, AsIII : dye ratio). The binding was stopped by adding 200 mM DTT immediately. Subsequently, the pro- teins were analysed by SDS-PAGE and gels were scanned at excitation/emission maxima of ~ 577/618 nm using the Typhoon Trio Variable Mode Imager (GE Healthcare). All gels were quantified using TotalLab Quant image analysis soft- ware (http://www.totallab.com/products/totallabquant). Acknowledgements This research was supported by a Major International Joint Research Project of Chinese National Natural Science Foun- dation (31010103903) to G.W. (Co-PIs, T.R.M. and C.R.) and by the US National Science Foundation to T.R.M. and B.B. (MCB 0817170). G.L. was supported by the PhD student exchange scholarship of the Ministry of Education, China. The authors thank B.P. Rosen for stimulating discussion. References Afkar, E., Lisak, J., Saltikov, C., Basu, P., Oremland, R.S., and Stolz, J.F. (2003) The respiratory arsenate reductase from Bacillus selenitireducens strain MLS10. FEMS Micro- biol Lett 226: 107–112. Anderson, C.L., Williams, J., and Hille, R. (1992) The purifi- cation and characterization of arsenite oxidase from Alcali- genes faecalis, a molybdenum-containing hydroxylase. J Biol Chem 267: 23674–23682. Bessey, O.A., Lowry, O.H., and Brock, M.J. (1946) A method for the rapid determination of alkaline phosphatase with five cubic millimeters of serum. J Biol Chem 164: 321–329. Bhattacharjee, H., and Rosen, B.P. (2007) Arsenic metabo- lism in prokaryotic and eukaryotic microbes. In Molecular Microbiology of Heavy Metals. Nies, D.H., and Silver, S. (eds). Heidelberg, Germany: Springer, pp. 371–406. Boyer, H.W., and Roulland-Dussoix, D. (1969) A complemen- tation analysis of the restriction and modification of DNA in Escherichia coli. Mol Biol 41: 459–472. Brunak, S., Emanuelsson, O., Nielsen, H., and von Heijne, G. (2007) Locating proteins in the cell using TargetP, SignalP and related tools. Nat Protoc 2: 953–971. Cai, L., Rensing, C., Li, X., and Wang, G. (2009) Novel gene clusters involved in arsenite oxidation and resistance in two arsenite oxidizers: Achromobacter sp. SY8 and Pseudomonas sp. TS44. Appl Microbiol Biotechnol 83: 715–725. Cullen, W.R., and Reimer, K.J. (1989) Arsenic speciation in the environment. Chem Rev 89: 713–764. De Maagd, R.A., and Lugtenberg, B. (1986) Fractionation of Rhizobium leguminosarum cells into outer membrane, cytoplasmic membrane, periplasmic, and cytoplasmic components. J Bacteriol 167: 1083–1085. De Pina, K., Navarro, C., McWalter, L., Boxer, D.H., Price, N.C., Kelly, S.M., et al. (1995) Purification and character- ization of the periplasmic nickel-binding protein NikA of Escherichia coli K12. Eur J Biochem 227: 857–865. Ellis, P.J., Conrads, T., Hille, R., and Kuhn, P. (2001) Crystal structure of the 100 kDa arsenite oxidase from Alcalin- genes faecalis in two crystal forms at 1.64 Å and 2.03 Å. Structure 9: 125–132. Figurski, D.H., and Helinski, D.R. (1979) Replication of an origin-containing derivative of plasmid RK2 dependent on a plasmid function provided in trans. Proc Natl Acad Sci USA 76: 1648–1652. Hanahan, D. (1983) Studies on transformation of Escherichia coli with plasmids. J Mol Biol 166: 557–580. He, F., Nair, G.N., Soto, S.C., Chang, Y., Hsu, L., DeGrado, W.F., and Binns, A.N. (2009) Molecular basis of ChvE function in sugar binding, sugar utilization, and virulence in Agrobacterium tumefaciens. J Bacteriol 191: 5802– 5813. von Heijne, G. (1988) Transcending the impenetrable: how proteins come to terms with membranes. Biochim Biophys Acta 947: 307–333. Inskeep, W.P., McDermott, T.R., and Fendorf, S.E. (2001) Arsenic (V)/(III) cycling in soils and natural waters: chemi- cal and microbiological processes. In Environmental Chemistry of Arsenic. Frankenberger, W.F., and Macy, J.M. (eds). New York, USA: Marcell Dekker, pp. 183–215. Kashyap, D.R., Botero, L.M., Franck, W.L., Hassett, D.J., and McDermott, T.R. (2006a) Complex regulation of arsenite oxidation in Agrobacterium tumefaciens. J Bacteriol 188: 1081–1088. Kashyap, D.R., Botero, L.M., Lehr, C., Hasset, D.J., and McDermott, T.R. (2006b) A Na+:H+ antiporter and a molybdate transporter are essential for arsenite oxidation in Agrobacterium tumefaciens. J Bacteriol 188: 1577– 1584. Koechler, S., Cleiss-Arnold, J., Proux, C., Sismeiro, O., Dillies, M.A., Goulhen-Chollet, F., et al. (2010) Multiple controls affect arsenite oxidase gene expression in Hermi- niimonas arsenicoxydans. BMC Microbiol 10: 53. Kraft, T., and Macy, J.M. (1998) Purification and character- ization of the respiratory arsenate reductase of Chysi- ogenes arsenatis. Eur J Biochem 255: 647–653. Lei, B., Liu, M., Chesney, G.L., and Musser, J.M. (2004) Identification of new candidate vaccine antigens made by Streptococcus pyogenes. J Infect Dis 189: 79–89. Lett, M.C., Muller, D., Lièvremont, D., Silver, S., and Santini, J. (2011) Unified nomenclature for genes involved in prokaryotic aerobic arsenite oxidation. J Bacteriol doi:10.1128/JB.06391-11. Lever, J.E. (1972) Purification and properties of a component of histidine transport in Salmonella typhimurium. J Biol Chem 247: 4317–4326. Lin, Y.F., Yang, J., and Rosen, B.P. (2007) ArsD residues Cys12, Cys13 and Cys18 form an As(III) binding site required for arsenic metallochaperone activity. J Biol Chem 282: 16783–16791. Liu, Y.G., and Whittier, R.F. (1995) Thermal asymmetric inter- laced PCR: automatable amplification and sequencing of insert end fragments from P1 and YAC clones for chromo- some walking. Genomics 25: 674–681. Luecke, H., and Quiocho, F.A. (1990) High specificity of a phosphate transport protein determined by hydrogen bonds. Nature 347: 402–406. McDermott, T.R., and Kahn, M.L. (1992) Cloning and mutagenesis of the Rhizobium meliloti isocitrate dehydro- genase gene. J Bacteriol 174: 4790–4797. Macur, R.E., Jackson, C.R., Botero, L.M., McDermott, T.R., and Inskeep, W.P. (2004) Bacterial populations associated with the oxidation and reduction of arsenic in an unsatur- ated soil. Environ Sci Technol 38: 104–111. Malasarn, D., Keeffe, J.R., and Newman, D.K. (2007) Char- acterization of the arsenate respiratory reductase from Shewanella sp. strain ANA-3. J Bacteriol 190: 135–142. Mascher, T., Helmann, J.D., and Unden, G. (2006) Stimulus perception in bacterial signal-transduction histidine kinases. Microbiol Mol Biol Rev 70: 910–938. Muller, D., Lievremont, D., Simeonova, D.D., Hubert, J.-C., and Lett, M.-C. (2003) Arsenite oxidase aox genes from a metal-resistant b-proteobacterium. J Bacteriol 185: 135– 141. Nielsen, H., and Krogh, A. (1998) Prediction of signal pep- tides and signal anchors by a hidden Markov model. In Proceedings of the Sixth International Conference on Intel- ligent Systems for Molecular Biology. Glasgow, J., Little- john, T., Major, F., Lathrop, R., Sankoff, D., and Sensen, C. (eds). Menlo Park, CA, USA: AAAI Press Proceedings, pp. 122–130. Oremland, R.S., and Stolz, J.F. (2005) Arsenic, microbes and contaminated aquifers. Trends Microbiol 13: 45–49. Phillips, S.E., and Taylor, M.L. (1976) Oxidation of arsenite to arsenate by Alcaligenes faecalis. Appl Environ Microbiol 32: 392–399. Pontius, F.W., Brown, K.G., and Chen, C.J. (1994) Health implications of arsenic drinking water. J Am Water Works Assoc 86: 52–63. Quandt, J., and Hynes, M.F. (1993) Versatile suicide vectors which allow direct selection for gene replacement in gram- negative bacteria. Gene 127: 15–21. Ruan, X., Bhattacharjee, H., and Rosen, B.P. (2006) Cys-113 and Cys-422 form a high affinity metalloid binding site in the ArsA ATPase. J Biol Chem 281: 9925–9934. Saltikov, C., and Newman, D.K. (2003) Genetic identification of a respiratory arsenate reductase. Proc Natl Acad Sci USA 100: 10983–10988. Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd edn. Cold Spring Harbor, NY, USA: Cold Spring Harbor Laboratory Press. Santini, G.M., and vanden Hoven, R.N. (2004) Molybdenum- containing arsenite oxidase of the chemolithoautotrophic arsenite oxidizer NT-26. J Bacteriol 186: 1614–1619. Sardiwal, S., Santini, J.M., Osborne, T.H., and Djordjevic, S. (2010) Characterization of a two-component signal trans- duction system that controls arsenite oxidation in the chemolithoautotroph NT-26. FEMS Microbiol Lett 313: 20–28. Shi, W., Wu, J., and Rosen, B.P. (1994) Identification of a putative metal binding site in a new family of metalloregu- latory proteins. J Biol Chem 269: 19826–19829. Somerville, J.E., and Kahn, M.L. (1983) Cloning of the glutamine synthetase I gene from Rhizobium meliloti. J Bacteriol 156: 168–176. Stolz, J.F., Basu, P., Santini, J.M., and Oremland, R.S. (2006) Arsenic and selenium in microbial metabolism. Annu Rev Microbiol 60: 107–130. Studier, F.W. (1991) Use of bacteriophage T7 lysozyme to improve an inducible T7 expression system. J Mol Biol 219: 37–44. Tam, R., and Saier, M.H., Jr (1993) Structural, functional, and evolutionary relationships among extracellular solute-binding receptors of bacteria. Microbiol Rev 57: 320–346. Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., and Higgins, D.G. (1997) The CLUSTAL_X windows inter- face: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res 25: 4876–4882. Wanner, B.L. (1996) Phosphorus assimilation and control of the phosphate regulon. In Escherichia coli and Salmonella: Cellular and Molecular Biology, 2nd edn. Neidhardt, F.C., Curtis, R., III, Ingraham, J.L., Lin, E.C.C., Low, K.B., Magasanik, B., et al. (eds). Washington, DC, USA: ASM Press, pp. 1357–1381. Zargar, K., Hoeft, S., Oremland, R.S., and Saltikov, C. (2010) Identification of a novel arsenite oxidase gene, arxA, in the haloalkaliphilic, arsenite-oxidizing bacterium Alkalilimnicola ehrlichii strain MLHE-1. J Bacteriol 192: 3755–3762. Supporting information Additional Supporting Information may be found in the online version of this article: Fig. S1. Clusters of the aio genes in Agrobacterium tumefa- ciens strain 5A and various arsenite-oxidizing bacteria. Representative aio gene clusters are from Xanthobacter autotrophicus Py2 (NC_009720), Roseovarius sp. 217 (NZ_AAMV01000002), Starkeya novella DSM 506 (NC_014217), Rhizobium sp. NT-26 (AY345225), Ochrobac- trum tritici strain SCII24 (FJ465505), Alcaligenes faecalis strain NCIB 8687 (AY297781), Rhodoferax ferrireducens T118 (NC_007908), Herminiimonas arsenicoxydans (NC_009138), Thiomonas intermedia K12 (NC_014153), Achromobacter sp. SY8 (EF523515), Alkalilimnicola ehrlichii MLHE-1 (NC_008340), Pseudomonas sp. TS44 (EU311944), Burkholderia multivorans ATCC 17616 (NC_010801), Candi- datus Nitrospira defluvii (NC_014355) and Thermus thermo- philus HB8 plasmid pTT27 (NC_006462). Homologues were marked as same colour. All aioX homologues were indicated as red colour. Note: all aox/aro/aso gene symbols are changed to aio. Fig. S2. Amino acid alignments of various AioX proteins annotated in different bacteria. The 5A AioX TAT signal peptide is as indicated, with the twin arginines shown as double vertical arrows. The conserved cysteine (aa 108) that was mutated to either an alanine or a serine is also highlighted. Fig. S3. Diagnostic PCR to verify aioX deletion. A total of 850 bp of nucleotides (grey filled area) is deleted in strain M53. Primer pairs P95/P2218 and P946/P1522 were used to confirm deletion (P95, AGACCCAACACGGAGCG and P2218, CCAGCATTCGTCGCAAGA; P946, GACCTGGAAG TGCTGGACG and P1522, GCTGGCTTTCCCGCTGT). Fig. S4. Qualitative As(III) oxidation phenotype confirmed by AgNO3 staining. The presence of As(V) is indicated by dark brown colour associated with the agar as seen when inocu- lated with WT and M53 carrying pCPP30::aioX. Loss of As(III) oxidation phenotype resulting from an DaioX mutation is also shown in strain M53 (DaioX mutant) and M53 carrying pCPP30::Cys108Ser. Regions of the agar plate were spot inoculated as indicated.