A CDC25 homologue from rice functions as an arsenate reductase


  • Gui-Lan Duan,

    1. Department of Soil Environmental Sciences, Research Center for Eco-environmental Sciences, Chinese Academy of Sciences, 18 Shuangqing RD, Beijing 100085, People's Republic of China;
    Search for more papers by this author
  • Yao Zhou,

    1. Department of Biochemistry and Molecular Biology, Wayne State University, School of Medicine, Detroit, MI 48201, USA;
    Search for more papers by this author
  • Yi-Ping Tong,

    1. Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, Beijing 100083, China
    Search for more papers by this author
  • Rita Mukhopadhyay,

    1. Department of Biochemistry and Molecular Biology, Wayne State University, School of Medicine, Detroit, MI 48201, USA;
    Search for more papers by this author
  • Barry P. Rosen,

    1. Department of Biochemistry and Molecular Biology, Wayne State University, School of Medicine, Detroit, MI 48201, USA;
    Search for more papers by this author
  • Yong-Guan Zhu

    1. Department of Soil Environmental Sciences, Research Center for Eco-environmental Sciences, Chinese Academy of Sciences, 18 Shuangqing RD, Beijing 100085, People's Republic of China;
    Search for more papers by this author

Authors for correspondence: Yong-Guan Zhu Tel: +86 10 629 36940 Fax: +86 10 629 23563 Email: ygzhu@rcees.ac.cnBarry P. Rosen Tel: +1 313 577 1512 Fax: +1 313 577 2765 Email: brosen@med.wayne.edu


  • • Enzymatic reduction of arsenate to arsenite is the first step in arsenate metabolism in all organisms studied. The rice genome contains two ACR2-like genes, OsACR2.1 and OsACR2.2, which may be involved in regulating arsenic metabolism in rice.
  • • Here, we cloned both OsACR2 genes and expressed them in an Escherichia coli strain in which the arsC gene was deleted and in a yeast (Saccharomyces cerevisiae) strain with a disrupted ACR2 gene. OsACR2.1 complemented the arsenate hypersensitive phenotype of E. coli and yeast. OsACR2.2 showed much less ability to complement.
  • • The gene products were purified and demonstrated to reduce arsenate to arsenite in vitro, and both exhibited phosphatase activity. In agreement with the complementation results, OsACR2.1 exhibited higher reductase activity than OsACR2.2. Mutagenesis of cysteine residues in the putative active site HC(X)5R motif led to nearly complete loss of both phosphatase and arsenate reductase activities.
  • • In planta expression of OsACR2.1 increased dramatically after exposure to arsenate. OsACR2.2 was observed only in roots following arsenate exposure, and its expression was less than OsACR2.1.


Arsenic ranks first on the US Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) Priority List of Hazardous Substances (http://www.atsdr.cdc.gov/cercla/05list.html), and the US Environmental Protection Agency (EPA) has classified arsenic as a group A human carcinogen (http://www.epa.gov/ttn/atw/hlthef/arsenic.html). Arsenic is a ubiquitous metalloid that is introduced into the environment from both anthropogenic and geochemical sources (Smith et al., 1998). In Bangladesh, arsenic contamination of groundwater is believed to cause arsenic-related disorders in 80% of the population (Alam et al., 2002; Das et al., 2004). Moreover, arsenic-contaminated groundwater is used not only as drinking water by tens of millions of people but also for rice cultivation, particularly during the dry season. This has led to an arsenic build-up in the soil of paddy fields and a 10-fold elevation of arsenic concentration in rice grains and straw (Meharg & Rahman, 2003). In China, in addition to arsenic contamination in groundwater, metal mining has caused severe arsenic contamination of arable land (paddy soils) and rice. For example, in Chenzhou, in the Hunan province, arsenic in rice was found to range from 0.5 to 7.5 µg g−1 (Liao et al., 2005). Rice is the staple food for more than half the world's population, and arsenic in rice eventually works its way up the food chain to humans, causing serious health problems (Smith et al., 2002; Meharg, 2004; Williams et al., 2006). For Asian populations, arsenic ingestion through rice consumption is one of the major sources of exposure. In a typical Bangladeshi diet, it is estimated that inorganic arsenic intake from rice is equivalent to the intake from drinking water (Williams et al., 2005, 2006). Understanding the mechanisms of arsenic metabolism in rice may help to improve the safety of rice grown in arsenic-contaminated soils and to minimize the health impact of arsenic contamination of soil and water.

Arsenic metabolism has been extensively studied in microorganisms (Rosen, 2002). In all organisms studied, enzymatic reduction is the first step for arsenate metabolism. Reduction is a prerequisite not only for the fate of arsenic in the organism, but also for its toxicity (Ghosh et al., 1999; Radabaugh & Aposhian, 2000; Mukhopadhyay & Rosen, 2002; Mukhopadhyay et al., 2002). Disruption of arsenate reductase genes leads to arsenate hypersensitivity in both E. coli and Saccharomyces cerevisiae (Mukhopadhyay & Rosen, 1998, 2002). Reduction of arsenate to arsenite also occurs in plants and animals. In mammals, purine nucleoside phosphorylase has been reported to reduce arsenate (Gregus & Nemeti, 2002; Radabaugh et al., 2002). However, the physiological relevance of this finding has been questioned (Nemeti et al., 2003). More recently, the glycolytic enzyme glyceraldehyde-3-phosphate dehydrogenase has been shown to reduce arsenate in vitro and has been proposed to have this function in vitro (Gregus & Nemeti, 2005). Dhankher et al. (2006) reported that silencing of the arsenate reductase gene in Arabidopsis thaliana (AtACR2) resulted in hypersensitivity to high concentrations of arsenate. Duan et al. (2005) demonstrated that reduction of arsenate is an important step in arsenic metabolism in the arsenic hyperaccumulator, Pteris vittata. Recently, Pickering et al. (2006) examined the distribution of arsenic species in vivo and arsenate reduction in fronds of P. vittata by X-ray absorption spectroscopy (XAS) and XAS imaging. In rice grains, arsenite and DMAs are the primary arsenic species (D’Amato et al., 2004; Williams et al., 2005). Transformation of arsenic to a variety of species in rice indicates that arsenate reduction and subsequent methylation are crucial steps in the biotransformation of inorganic arsenic.

Arsenate reductases have evolved independently at least three times (Rosen, 2002). In bacteria, two families of ArsC arsenate reductases have been characterized and a third family of ACR arsenate reductases has been identified in eukaryotes (Mukhopadhyay et al., 2002). Two ACR-type arsenate reductases have been characterized: ScAcr2p from S. cerevisiae (Mukhopadhyay et al., 2000) and LmACR2 from Leishmania major (Zhou et al., 2004). Both are believed to have evolved from protein-tyrosine phosphatases (PTPases), which include the CDC25 (cell division cycle) cell cycle dual-specificity tyrosine phosphatases. These phosphatases and reductases share a common HC(X)5R catalytic motif, the signature sequence of phosphatases. L. major LmACR2 exhibits phosphatase activity (Zhou et al., 2006). The yeast ScAcr2p does not exhibit phosphatase activity but can be converted into a phosphatase by substitution of just three residues (Mukhopadhyay et al., 2003).

Plant arsenate reductases belong to the ACR family. A. thaliana has a homologue of the catalytic domain of CDC25 that has been shown to reduce arsenate (Bleeker et al., 2006). It has also been shown to have phosphatase activity and induces short cell length when overexpressed in fission yeast, which suggest that it has a CDC25-like function (Landrieu et al., 2004a,b; Sorrell et al., 2005). This enzyme, called variously AtACR2, AtASR and AtCDC25, has been identified and characterized as an arsenate reductase both in microorganisms and in planta (Bleeker et al., 2006; Dhankher et al., 2006). Bleeker et al. (2006) cloned a homologue of ScAcr2p from Holcus lanatus and showed that it has arsenate reductase activity. Ellis et al. (2006) cloned the gene for a homologue of ScAcr2p from P. vittata (PvACR2). Unlike AtACR2, PvACR2 does not exhibit phosphatase activity. Rathinasabapathi et al. (2006) reported that the PV4-8 TPI (cytosolic triosephosphate isomerases) from P. vittata increased arsenate resistance in an E. coli strain in which arsC was deleted and suggested that this enzyme might function as an arsenate reductase.

Although arsenate reductase genes have been cloned and characterized in A. thaliana and P. vittata (Bleeker et al., 2006; Dhankher et al., 2006; Ellis et al., 2006), the goal of those studies focused on phytoremediation to maximize arsenic uptake and increase translocation from root to shoot. For plants destined for human consumption, the objective is precisely the opposite of that for phytoremediation, namely to minimize uptake and limit translocation to edible tissues. In view of the dramatic health problems caused by consumption of arsenic-contaminated food crops, particularly rice, knowledge about the mechanisms of arsenic metabolism in rice is essential.

In this paper we report the purification and characterization of two small CDC25 homologues from O. sativa L., OsACR2.1 and OsACR2.2. OsACR2.1 and, to a lesser extent, OsACR2.2 exhibit bifunctional activities as phosphatases and arsenate reductases. In addition, we examined expression of these genes in rice plants grown under normal conditions and in the presence of arsenate and/or phosphate deficiency, and in planta, OsACR2.1, and again to a lesser extent OsACR2.2, appears to be an arsenate inducible protein. The results suggest that OsACR2.1 is the major arsenate reductase of O. sativa.

Materials and Methods

Strains, plasmids, and media

Strains and plasmids used in this study are described in Table 1. Escherichia coli strains were grown at 37°C in a low phosphate medium or Luria-Bertani medium (Sambrook et al., 1989) supplemented with appropriate antibiotics and inducers. Saccharomyces cerevisiae strains were grown at 30°C in complete YPD (yeast extract-peptone-dextrose) medium (Adams et al., 1998) (for phenotypic analysis) or minimal medium with 2% glucose or 2% galactose plus auxotrophic requirements. S. cerevisiae strains Y10000 (wild-type) and Y15615 (Δacr2) were purchased from Euroscarf (http://web.uni-frankfurt.de/fb15/mikro/euroscarf/).

Table 1.  Strains and plasmids
Strain/PlasmidGenotypeReference or source
Bacterial strains
 JM109recA1 supE44 endA1 hsdR17 gyrA96 relA1 thi (lac-proAB)F′[traD36 proAB+ lac1q lacZ M15]Sambrook et al. (1989)
 BL21(DE3)hsdS galcIts857 ind1 sam7 nin5 lacUV5-T7 gene 1)Sambrook et al. (1989)
 WC3110K12 FIN(rrnD-rrnE) ΔarsCMukhopadhyay et al. (2000)
 W3110K12 FIN(rrnD-rrnE)(Bachmann, 1987)
Yeast strains
 Y10000MATα his3Δ1 leu2Δ0 lys2Δ0 ura3Δ0Euroscarf
 Y15615BY4742 Matα his3Δ1leu2Δ0 lys2Δ0 ura3Δ0 YPR200c::kanMX4Euroscarf
 pET-28aE. coli cloning and expression vector, KmrNovagen
 pYES2Multicopy E. coli– yeast shuttle vector, Apr, URA3, gal1Invitrogen
 pET-28a-ACR2OsACR2 genes from O. sativa cloned into the NdeI and HindIII site of pET-28aThis study
 pYES2-ACR2OsACR2 genes from O. sativa cloned into the HindIII-XbaI sites of pYES2This study

Plant culture and treatment

Rice (Oryza sativa L. cv. Jiahua 1) seeds were sterilized in 10% H2O2 solution for 15 min followed by thorough washing with deionized water. The seeds were germinated in moist perlite. After 2 wk, uniform seedlings were selected and transplanted to PVC pots (15 cm diameter and 14 cm height, six plants per pot) containing 1000 ml one-third-strength nutrient solution. The composition of the full-strength nutrient solution was as follows: 5 mm NH4NO3, 2 mm K2SO4, 4 mm CaCl2, 1.5 mm MgSO4, 50 µm Fe(II)-EDTA, 10 µm H3BO4, 1 µm ZnSO4, 1 µm CuSO4, 5 µm MnSO4, 0.5 µm Na2MoO4, 0.2 µm CoSO4 and 1.3 mm KH2PO4. The nutrient solution was changed twice a week and the pH of the solution was adjusted to 5.5 using HCl or NaOH (Hewitt, 1966). All plants were cultivated in a growth room at a 14-h light period (260–350 µmol m−2 s−1). The temperature was regulated to 28°C day and 20°C night and the relative humidity was maintained at c. 60%.

After 1 wk, –P-treated plants were starved of phosphate (1.3 mm KH2PO4 was replaced with 1.3 mm KCl in nutrient solution); +P-treated plants were grown with normal phosphate concentration in the nutrient solution (1.3 mm KH2PO4).

One week after phosphate treatment, plants were treated with arsenate. For cDNA synthesis, the plants were exposed to 55 µm sodium arsenate (approx. 8 ppm) with normal phosphate nutrient solution (+P) for 8 h before harvesting. For semiquantitative RT-polymerase chain reaction (PCR), plants were grown with normal phosphate nutrient solution (+P) or with phosphate starvation nutrient solution (–P) and/or different times of exposure to 55 µm sodium arsenate (0, 1, 2, 4, 8, 12 and 24 h). For arsenate reductase activity measurement in extracts of rice roots and shoots, rice plants were treated with or without 25 µm of sodium arsenate and/or with or without phosphate for 1 wk.

RNA extraction and cDNA synthesis

Total RNA was isolated from 4-wk-old rice plants using the phenol/chloroform/isoamylalcohol technique (Marco et al., 1990). cDNA was synthesized in 20 µl reactions from total RNA after DNase treatment (Invitrogen, Carlsbad, CA, USA), using 200 U of MMLV reverse transcriptase (Invitrogen) and oligo-(dT) as a primer (Sambrook et al., 1989).

Expression recombinants construction

The ACR2 genes from O. sativa cDNA were amplified by PCR. PCR was performed in a Peltier Thermal Cycler (PTC 200, MJ Research, Waltham, MA, USA). DNA sequencing was performed by using a CEQ2000 DNA sequencer (Beckman Coulter, Fullerton, CA, USA). For construction of the E. coli expression vector pET28a-OsACR2.1, the forward primer was 5′-GCGGGATCATATGGCGCGGAGCGTGTCGTACGT-3′ and the reverse primer was 5′-GTGATAAGCTTTTACAACTCAGGTTCTTCAGGTG-3′. For construction of pET28a-OsACR2.2, the forward primer was 5′-GCGGGATCATATGGCGAGGGGCGTCTCCTACGT-3′ and the reverse primer was 5′-GTGATAAGCTTTCAAGAGCACACACCCTTGCAAG-3′. PCR (94°C for 30 s, 58°C for 30 s and 72°C for 45 s, 30 cycles) was run with cDNA. NdeI and HindIII sites (underlined) were created at the 5′ and 3′ ends of the gene, respectively. For construction of the S. cerevisiae expression vector pYES2-OsACR2.1, the forward primer was 5′-GTGATAAGCTTATGGCGCGGAGCGTGTCGTACGT-3′ and the reverse primer was 5′-CCTAGTCTAGATTACAACTCAGGTTCTTCAGGTG-3′. For construction of vector pYES2-OsACR2.2, the forward primer was 5′-GTGATAAGCTTATGGCGAGGGGCGTCTCCTACGT-3′ and the reverse primer was 5′-CCTAGTCTAGATCAAGAGCACACACCCTTGCAAG-3′. HindIII and XbaI sites (underlined) were created at the 5′ and 3′ end of the gene, respectively.

Oligonucleotide-directed mutagenesis

Mutations in OsACR2s were introduced by site-directed mutagenesis using the QuikChange™ site-directed mutagenesis procedure (Stratagene, La Jolla, CA, USA). The two pET28a-ACR2 plasmids were used as templates for creating single cysteine-to-serine mutations. The mutagenic oligonucleotides used for both strands with changes introduced (underlined) were as follows: OsACR2.1 C71S, 5′-CCGTCGTCTTCCACTCCGCCCTCAGCAAGGTG-3′ and 5′-CACCTTGCTGAGGGCGGAGTGGAAGACGACGG-3′; OsACR2.2 C70S, 5′-CTCGTCTTCCACTCTGCCCTCAGCAAG-3′ and 5′-CTTGCTGAGGGCAGAGTGGAAGACGAG-3′. The mutations were confirmed by sequencing both genes.

Metalloid sensitivity assays

The arsenate resistance phenotype of cells expressing ACR2 genes was determined in both bacteria and yeast. In E. coli, strains W3110 (wild-type) or WC3110 (DE3) (ΔarsC) were used for arsenate sensitivity assays (Shi et al., 1999). Cells of E. coli were grown overnight in a low-phosphate medium at 37°C supplemented with 40 µg ml−1 kanamycin. Overnight cultures were diluted 100-fold in low-phosphate medium containing various concentrations of sodium arsenate and 0.3 mm isopropyl-β-D-thiogalactopyranoside (IPTG). Growth was estimated from the absorption at 600 nm after 15 h of growth at 37°C with shaking. S. cerevisiae strains Y10000 (wild-type) and Y15615 (ACR2Δ) bearing plasmids pYES2.0 and OsACR2s-pYES were grown overnight at 30°C in minimal medium containing 2% galactose supplemented with 0.2 mg ml−1 uracil. The overnight cultures were then diluted to an A600 of 0.1 into the same medium containing varying amounts sodium arsenate and allowed to grow for an additional 24 h.

Purification of OsACR2s, glutaredoxin (Grx2), thioredoxin (Trx1p) and thioredoxin reductase (Trr1p)

Escherichia coli Grx2 and yeast Trx1p, and Trr1p were purified as described previously (Mukhopadhyay et al., 2000). OsACR2s were purified from E. coli strain Bl21 (DE3). A single colony bearing a plasmid pET28a-OsACR2.1 or pET28a-OsACR2.2 was grown overnight in 200 ml LB medium containing 40 µg ml−1 kanamycin with shaking at 37°C, diluted to 2 l of the same medium. At an A600 nm of 0.5, IPTG was added as an inducer to a final concentration of 0.3 mm, and the culture was grown for an additional 4 h at 37°C. The cells were harvested, washed once with a buffer consisting of 50 mm MOPS, pH 7.5, containing 20 mm imidazole, 0.5 m NaCl, 10 mmβ-mercaptoethanol, and 20% glycerol (buffer A), suspended in buffer A at a ratio of 5 ml of buffer per g of wet cells, and lysed by a single passage through a French pressure cell at 20 000 psi. Diisopropyl fluorophosphate (2.5 µl g−1 of wet cells) was added to the lysate immediately after lysis. The lysate was centrifuged at 100 000 g for 60 min at 4°C, and the supernatant solution was loaded at a flow rate of 0.5 ml min−1 onto a Ni2+-nitrilotriacetic acid column pre-equilibrated with buffer A. The column was then washed with 250 ml of buffer A followed by elution with 125 ml of buffer A containing 0.2 m imidazole. OsACR2s were identified by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). Fractions containing purified OsACR2s were pooled and concentrated. All purified proteins were stored at −70°C until use. Protein concentrations were determined from the absorbance at 280 nm by using the following extinction coefficients (Gill & von Hippel, 1989): OsACR2.1, 10 250 m−1 cm−1; OsACR2.2, 12 810 m−1 cm−1; Trx1p, 9770 m−1 cm−1; Trr1p, 23 380 m−1 cm−1; Grx2, 21 860 m−1 cm−1; and Grx1p, 5360 m−1 cm−1.

Enzyme assays

Arsenate reductase activity was measured at 37°C using a coupled assay (Shi et al., 1999; Mukhopadhyay et al., 2000). The assay buffer contained 50 mm MOPS, 50 mm MES, pH 6.5, 0.1 mg ml−1 bovine serum albumin, 0.25 mm NADPH, 5 nm yeast glutathione reductase (Calbiochem, Darmstadt, Germany), 1 mm GSH, and 50–100 µm OsACR2. Sodium arsenate and Grx2 were added as indicated. Reductase activity was monitored at 340 nm and expressed as nanomoles of NADPH oxidized per mg of OsACR2 by using a molar extinction coefficient of 6200 m−1 cm−1 for NADPH. Each assay was repeated at least three times with two separate batches of purified protein. From rice roots and shoots, enzyme extractions and activity determination were manipulated using the coupled enzymatic reaction described by Duan et al. (2005).

Phosphatase activity was assayed at 37°C with 5 µm wild-type or mutant OsACR2 proteins with the indicated amounts of p-nitrophenyl phosphate (pNPP) in 0.1 m MOPS/MES buffer, pH 6.5 (Zhou et al., 2006). The assay was initiated by the addition of pNPP, and the rate of hydrolysis was estimated from the increase in absorption at 405 nm. Each value was corrected for nonenzymatic pNPP hydrolysis. The data were analyzed with SigmaPlot 9.0 using an extinction coefficient for nitrophenol of 18 000 m−1 cm−1.

Semiquantitative RT-PCR

Total RNA was isolated from shoots and roots of the rice plants grown with different exposure times to 55 µm and normal phosphate (+P) or absent phosphate (–P). cDNA was synthesized from c. 5 µg of total RNA, as described above. To evaluate cDNA content, a rice actin gene was used as a constitutive internal standard. The tubulin forward primer was GGAACTGGTATGGTCAAGGC, and the reverse primer was AGTCTCATGGATAACCGCAG. PCR (94°C for 30 s, 56°C for 30 s and 72°C for 45 s; 35 cycles) was run with cDNA; the PCR product length was 591 bp. Evaluation of the expression of the two OsACR2s was performed by specific PCR, using OsACR2 primers. PCR (94°C for 30 s, 58°C for 30 s and 72°C for 45 s; 35 cycles) was run with the same amount of cDNA of different samples. The PCR products were electrophoresed in 1% agarose gels, stained with ethidium bromide, imaged using a Gel Documentation workstation and Image AutoAnalysis Software (Gel Doc 2000™ and Quantity One™, Bio-Rad Co., Hercules, CA, USA). PCR reactions were performed on cDNAs obtained from two different RNA extractions performed on samples from two independent experiments and repeated at least three times for each cDNA.


Phylogenetic analysis of plant ACR2 homologs

The DNA sequences of ScACR2 and AtACR2 were used to search for homologues in the rice dbEST database, and two rice sequences (CF333258, CF293892) were identified. RT-PCR was performed to clone the two putative ACR2 genes from a rice root cDNA library with gene-specific primers, and single fragments were cloned and sequenced. A 414 bp gene was designated as OsACR2.1 (GenBank accession number AY860059), and a 393 bp gene was designated as OsACR2.2 (accession number AY860058). OsACR2.1 and OsACR2.2 have 137 and 130 residues, with predicted molecular masses of 14 963 and 14 330 Da, respectively. Both OsACR2s are similar to ScAcr2p (130 residues, 14 883 Da) (21.4 and 25%, respectively) and AtACR2 (146 residues, 16 447 Da) (51 and 55.5%, respectively), and they each have an HC(X)5R active site motif (Fig. 1a).

Figure 1.

(a) Alignment of predicted plants ACR2 compared with some identified CDC25s and ACR2. Os, Oryza sativa; Sp, Schizosaccharomyces pombe; Sc, Saccharomyces cerevisiae; At, Arabidopsis thaliana; Mt, Medicago truncatula; Zm, Zea mays; Hv, Hordeum vulgare; Ta, Triticum aestivum; Hl, Holcus lanatus; Cr, Chlamydomonas reinhardtii; Lm, Leishmania major; Pv, Pteris vittata. (b) Phylogenetic tree.

The DNA sequences of OsACR2s were used to search for homologues in other higher plants (Fig. 1a). These plant ACR2 homologues share a high amino acid identity (68.8%), and each predicted protein contains the HC(X)5R signature sequence of the catalytic domains of protein tyrosine phosphatases, as well as ScAcr2p and LmACR2. Phylogenetic analysis CDC25s and ACR2s indicate that the plant ACR2s are equally related to both but resemble ACR2s more closely in size than the much larger CDC25s (Fig. 1b).

OsACR2.1 complements arsenate-sensitive phenotype in both E. coli and yeast

The ability of the two OsACR2 genes to complement the arsenate-sensitive phenotype of an E. coli Δars strain, WC3110, was examined. OsACR2.1 restored arsenate resistance at 2 mm sodium arsenate to that of the wild-type strain W3110 (Fig. 2a). OsACR2.2 complemented only at low arsenate concentrations (< 0.8 mm), but this was still higher than cells of WC3110 with the pET-28a vector alone, which was sensitive to 0.4 mm. Both proteins were expressed equally well, as determined by immunoblotting with antihistag antibody (Fig. 2c). Thus both rice genes could complement the arsC deletion, although to varying degrees. The ability of the two OsACR2 genes to complement the arsenate-sensitive phenotype of the S. cerevisiae strain Y15615 (acr2Δ) was also examined. Plasmids pYES2-OsACR2.1 and pYES2-OsACR2.2 carrying rice ACR2 genes under the control of the GAL1 promoter were transformed into strain Y15615. In the presence of 2% galactose, cells bearing plasmid pYES2-OsACR2.1 exhibited an arsenate-resistant phenotype comparable to that of a wild-type control, strain Y10000 (Fig. 2b). Induction was required: there was no complementation when galactose was replaced by glucose (data not shown). On the other hand, strain Y15615 bearing pYES2-OsACR2.2 exhibited only a slight increase in arsenate resistance. These results clearly show that OsACR2.1 encodes a functional arsenate reductase. On the other hand, the poor complementation by OsACR2.2 suggests that this gene may not encode a physiologically functioning arsenate reductase.

Figure 2.

Function complementation in Escherichia coli and Saccharomyces cerevisiae. (a) OsACR2s confer arsenate resistance in E. coli. OsACR2 was cloned behind the T7 promoter in vector pET28a with an N-terminal 6-histidine tag and expressed in E. coli arsC deletion strain WC3110 (DE3). OD600, cell density; As(V), sodium arsenate. (b) OsACR2s confer arsenate resistance in S. cerevisiae. OsACR2 was cloned behind the GAL1 promoter in vector pYES2 and expressed in S. cerevisiae ACR2 disruption strain Y15615. (c) Expression of OsACR2s in whole cell, cytosol and inclusion bodies of E. coli strain WC3110.

Both OsACR2s exhibit arsenate reductase activity

In E. coli, both OsACR2.1 and OsACR2.2 were expressed as soluble proteins in the cytosol (Fig. 2c). The two proteins were purified by nickel-nitrilotriacetic acid chromatography, and their ability to reduce arsenate to arsenite was examined (Fig. 3a). The ArsC enzyme of the E. coli plasmid R773 uses glutaredoxin and GSH as electron donors, as do members of the ACR2 family, including yeast ScAcr2p and Leishmania LmACR2. In the presence of GSH, either yeast glutaredoxin (YGrx1) or E. coli glutaredoxin-2 (Grx2) could serve as an electron donor for OsACR2.1 and OsACR2.2 activity. Members of the second family of bacterial arsenate reductases use thioredoxin as electron donor. However, neither OsACR2.1 nor OsACR2.2 was able to use yeast thioredoxin (Trx1) together with thioredoxin reductase as an electron donor for the reduction of arsenate (Fig. 3b).

Figure 3.

OsACR2s exhibit arsenate reductase activity in vitro. (a) Arsenate reductase activity of wild-type and cysteine mutants of the two OsACR2s. Arsenate reductase activity was assayed with 15 µm protein, 50 mm sodium arsenate, 0.5 µm glutaredoxin (Grx2). Other assay conditions were as described in the Materials and Methods section. (b) Electron donor requirements for OsACR2s arsenate reductase activity. Arsenate reductase activity was assayed with 15 µm protein, 50 mm sodium arsenate, 0.5 µm Grx2 or 0.5 µm yeast glutaredoxin (YGrx1) or 0.5 µm thioredoxin (Trx1). Other assay conditions were as described in the Materials and Methods section.

Arsenate reduction by both OsACR2s exhibited Michaelis-Menten kinetics. Their kinetic parameters (Table 2) are similar to LmACR2 and ScAcr2p, the first two characterized eukaryote arsenate reductases. Purified OsACR2.1 has a higher turnover number than OsACR2.2, and this higher enzyme activity may, in part, explain its ability to confer higher resistance to arsenate (Fig. 2a,b).

Table 2.  Kinetic parameters of purified OsACR2.1s
ProteinSubstrateKm(mm)Vmax (Nmol mg−1 min−1)Kcat (s−1)Kcat/Km (m−1 s−1)
  1. Km, concentration of substrate that leads to half-maximal velocity; Vmax, maximum velocity as substrate concentrations increase; Kcat, catalytic rate constant.

OsACR2.1Arsenate12.2 (P = 0.0005)1170.3326.7
Grx2 0.1 (P < 0.0001) 920.26 2.6 × 106
pNPP15.4 (P < 0.0001)1800.5032.4
C71SArsenate24.4 (P = 0.0003) 170.05 1.9
OsACR2.2Arsenate 9.2 (P < 0.0001) 550.1416.0
Grx2 0.2 (P < 0.0001) 690.19 0.7 × 106
pNPP15.0 (P = 0.0001) 580.1610.4
C70SArsenate 9.9 (P = 0.0005) 120.03 3.2

Both purified OsACR2s exhibit phosphatase activity

Leishmania major LmACR2 exhibits protein phosphotyrosine phosphatase activity (Zhou et al., 2006), and the yeast ScAcr2p can be converted into a phosphatase by substitution of just three residues (Mukhopadhyay et al., 2003). The phosphatase activity of OsACR2.1 and OsACR2.2 was assayed by their ability to hydrolyze the artificial substrate pNPP (Fig. 4). The phosphatase activity of both OsACR2s obeyed Michaelis-Menten kinetics (Table 2). These parameters are similar to LmACR2 (Zhou et al., 2006), which also has both arsenate reductase and phosphatase activities. Again, OsACR2.1 exhibited a higher maximal rate of phosphatase activity than OsACR2.2.

Figure 4.

OsACR2s exhibit phosphatase activity in vitro. Phosphatase activity of wild-type OsACR2s, cysteine mutant OsACR2s and arsenic inhibition. Phosphatase activity was assayed with 15 µm protein and 20 mm p-nitrophenyl phosphate (pNPP). For inhibition assays, enzymes were preincubated with arsenate or arsenite for 5 min at 37°C before initiation of the reaction. Other assay conditions were as described in the Materials and Methods section.

Arsenate inhibited the phosphatase activity, while arsenite had no effect (Fig. 4). When the protein was incubated with 15 mm arsenate for 5 min, OsACR2.1 + As(V), the protein lost phosphate activity completely. While when the protein was incubated with 15 mm arsenite for 5 min, OsACR2.1 + As(III), the phosphatase activity was not affected.

The cysteine residue of the HC(X)5R motif is involved in OsACR2 arsenate reductase and phosphatase activities. To analyze the effect of substitution of the cysteine residues in the HCX5R motif (Cys-71 in OsACR2.1 and Cys-70 in OcACR2.2) on enzymatic activity, the cysteine residues were altered to serine residues by mutagenesis. The two mutant genes showed reduced ability to complement the arsenate-sensitive phenotype of the E. coli strain WC3110 at low arsenic (V) concentrations (Fig. 2a). The purified mutant proteins exhibited low arsenate reductase activity (Figs 3a, 4), with a nearly 10-fold reduction in Vmax (Table 2). The mutant proteins lost phosphatase activity. These results are consistent with the cysteine residue of the HCX5R motif of OsACR2.1 and OsACR2.2 being involved in both phosphatase and reductase activities.

Transcriptional patterns and enzyme activity of OsACR2s in response to phosphorus nutrition and arsenate exposure in planta

Quantification of OsACR2s expression in rice was performed using a semiquantitative PCR method in the linear range (Fig. 5). The OsACR2 genes expressed weakly in both roots and shoots of rice before addition of arsenate to the nutrient solution. OsACR2.2 was only weakly expressed in shoots under all conditions examined. Expression of OsACR2.1 increased dramatically in both roots and shoots after addition of arsenate for up to 24 h. OsACR2.2 was observed only in roots following arsenate exposure, and its expression was lower than OsACR2.1. Expression of OsACR2.1 under conditions of phosphate starvation was faster than in the presence of phosphate (+P), reaching its maximum expression at 4 h under phosphate-limited conditions and 8 h under phosphate-replete conditions, respectively. In addition, the expression was higher and persisted longer under phosphate-limited conditions (no decrease from 4 h to 24 h) than under phosphate-replete conditions (substantial decreased after 8 h). The difference in OsACR2.1 expression in the presence and absence of phosphate was more apparent in roots than in shoots. Expression of OsACR2.2 was hardly affected by the phosphate nutritional status. Both arsenate and phosphate had more of an effect on expression of OsACR2.1 than of OsACR2.2. Expression of OsACR2.1 was higher than OsACR2.2 in both the presence and absence of phosphate and in both roots and shoots. These results indicate that the physiological function of OsACR2.1 may be more related to arsenic metabolism than that of OsACR2.2.

Figure 5.

Transcriptional patterns of OsACR2 under different phosphorus nutrition status and different times of arsenate exposure. Quantification of OsACR2 expression was performed using the semiquantitative PCR method in the linear range. Total RNA was reverse-transcribed and PCR products were blotted and photographed by Gel Doc 2000™ and Quantity One™ (Bio-Rad Co.). Analyses were repeated at least three times for each cDNA obtained from two different RNA extractions.

Arsenate reductase activity was examined in extracts of rice roots and shoots treated with or without 25 µm of arsenate and/or with or without phosphate for 1 wk (Fig. 6). The data clearly show that arsenate reductase activity responds to the arsenate and phosphorus concentrations in the culture medium. When plants were grown with normal phosphate concentration in the nutrient solution, reductase activity in rice treated with arsenate was twofold higher than rice without arsenate. These results indicate that the arsenate reductase activity of rice, although present constitutively at moderate levels, is arsenate-inducible. When the plants were induced by arsenate, arsenate reductase activity under conditions of phosphate deficiency was higher than under phosphate sufficiency, but this difference was not as significant as arsenate exposure. Even though the two sets of experiments were on different timescales, they are consistent with each other: arsenate exposure and phosphate deprivation both increase expression of the OsACR genes and arsenate reductase activity.

Figure 6.

Arsenate reductase activity in rice (Oryza sativa). Arsenate reductase activity was examined in extracts of rice roots (grey bars) and shoots (black bars) treated with or without 25 µm of arsenate (As) and/or with or without phosphate (P) for 1 wk (± SD, n = 4).


Paddy soils must produce rice for the dense population of Southeast Asia, and remediation of the large areas contaminated with low-to-moderate concentrations of arsenic is almost impossible. Insights into the biochemical and molecular mechanisms of arsenic metabolism in rice plants may present opportunities to improve crop safety in arsenic-contaminated areas. The significant transformation of arsenic species observed in rice plants suggests that there is one or more arsenate reductases involved in the biotransformation of inorganic arsenic. Characterization of arsenate reductases in rice plants is crucial for formulation of strategies to reduce arsenic accumulation and/or decrease the proportion of inorganic arsenic in rice grains.

In this study we characterized two O. sativa homologues, OsACR2.1 and OsACR2.2, of yeast ScAcr2p. By heterologous expression in E. coli and yeast, both OsACR2s genes complemented the arsenate-sensitive phenotype of an E. coli Δars strain (WC3110) and S. cerevisiae acr2Δ strain (Y15615) (Fig. 2a,b). Both purified OsACR2s exhibit arsenate reductase activity, in the presence of GSH and yeast glutaredoxin (ScGrx1p) or E. coli glutaredoxin-2 (Grx2), but not yeast thioredoxin (Trx1) (Fig. 3a,b). However, from both complementation and enzyme activity, it appears that OsACR2.1 is more active than OsACR2.2. In addition, the expression of OsACR2.1 was higher than OsACR2.2 in planta (Fig. 5). These results suggest that the physiological function of OsACR2.1 may be as an arsenate reductase, while OsACR2.2 may have a different function, for example as a CDC25-like regulator of cell cycle.

Members of the ACR2 family share a HC(X)5R motif with members of the superfamily of phosphotyrosine protein phosphatases. The cysteine residue in this motif is believed to be essential for phosphatase activity of PTPases. A cysteine mutant of AtACR2 lost phosphatase activity (Landrieu et al., 2004a). Similarly, cysteine-to-serine substitutions resulted in complete loss of ability of both OsACR2s to hydrolyze pNPP (Fig. 4). Mutants in the conserved arginine and cysteine residues of the HC(X)5R sequence of both ScAcr2p and LmACR2 lost arsenate reductase activity (Mukhopadhyay et al., 2000; Zhou et al., 2004). While the two OsACR2 cysteine mutants lost nearly 90% of wild-type levels of arsenate reductase activity, they are not inactive. The OsACR2s have a number of other cysteine residues, including a cluster of cysteines at their C-termini. Whether one or more of these cysteine residues can substitute in part for the loss of the HCX5R cysteine is a topic for future investigation.

Expression of arsenate reductases has a significant impact on the translocation of arsenic from roots to shoots, and translocation is a key process in arsenic accumulation in the rice grain. Dhankher et al. (2006) reported that, in A. thaliana, AtACR2 knockdown resulted in a 10- to 16-fold increase in arsenic accumulation in shoots (350–500 ppm) and a decrease in arsenic accumulation in roots compared with wild-type plants grown in the presence of less than 8 ppm arsenate. To reduce arsenic translocation and/or decrease the proportion of inorganic arsenic in rice grains, it is important to understand the factors that regulate arsenate reductase activity. Arsenate is an analogue of phosphate and competes for the same absorption sites in the root apoplast and for the same uptake carriers in the root plasma membrane (Meharg & Macnair, 1992; Abedin et al., 2002a,b; Wang et al., 2002). Once inside plant cells, arsenate disrupts cellular energy metabolism by replacing phosphate in ATP to form the unstable ADP-arsenate (Kenney & Kaplan, 1988). In this study, we investigated the interactions between phosphate and arsenate in terms of the transcriptional patterns of the OsACR2s and arsenate reductase activity (Figs 5, 6). The data clearly show that OsACR2 transcription levels and arsenate reductase activities correlated with the presence or absence of arsenate and phosphate in the culture medium. Both gene expression and arsenate reductase activities were induced by the presence of arsenate and the absence of phosphate, but the impact of phosphate was not as significant as arsenate. However, reductase activity is present at relatively high levels constitutively, raising the question of the physiological role of the two OsACR2s. Arsenate reductase activity is likely an adventitious activity for these enzymes. Aside from these two proteins, there are no other CDC25 homologues in O. sativa. Since both OsACR2s are phosphatases, it is reasonable to propose that one or both serve a role as PTPases in the regulation of the cell cycle. Their ability to detoxify arsenate may be a secondary role for which OsACR2.1 is better suited than OsACR2.2.


This work was supported by Natural Science Foundation of China (40225002) and Chinese Ministry of Science and Technology (2002CB410808) grants to YGZ, and by NIH grant GM52216 to BPR. We appreciate helpful discussion with Dr H Schat and Professor Andrew Smith.