Arbuscular mycorrhizal colonization reduces arsenate uptake in barley via downregulation of transporters in the direct epidermal phosphate uptake pathway

Authors


Author for correspondence:
H. M. Christophersen
Tel: +61 8 83037306
Email: helle.christophersen@adelaide.edu.au

Summary

  • Here, we used barley (Hordeum vulgare) grown in normal and compartmented pots to investigate sensitivity to arsenic (As) in the absence of a positive growth response to arbuscular mycorrhizas (AM).
  • We tested the hypothesis that upon inoculation with AM fungi downregulation of HvPht1;1 and HvPht1;2 genes (encoding high-affinity inorganic orthophosphate (Pi)-uptake systems in a direct pathway via root epidermis and root hairs) and upregulation of the AM-induced HvPht1;8 (encoding the Pi-uptake system responsible for transfer of Pi from the symbiotic interface to cortical cells) play a role in decreased As uptake and hence reduced As sensitivity in AM plants.
  • Barley did not respond, or responded negatively to colonization by Glomus intraradices in terms of growth. In terms of specific phosphorus (P) uptake (P uptake per g of root) barley was nonresponsive. There was a significant interaction between As treatment and colonization, resulting in a lower As concentration and uptake in AM compared with nonmycorrhizal (NM) plants.
  • The decreased uptake of As and higher P : As molar ratios in the AM barley can be explained by the operation of the AM pathway as indicated by induction of HvPht1;8 and by down-regulation of HvPht1;1 and HvPht1;2.

Introduction

Arsenic (As) is a nonessential element in biological processes and is toxic to plants, humans and animals. Arsenic contamination of water, air and soil from both geological and anthropogenic sources is a significant environmental health concern (Ng et al., 2003). Arsenic toxicity in humans is a widespread and increasingly recognized problem, mainly resulting from accumulation of As in rice and other grains from contaminated irrigation water (Meharg, 2004). Considerable current work is aimed at reducing As uptake and accumulation in plants and hence in human food chains, and understanding mechanisms is essential before effective mitigation strategies can be devised. The forms of arsenic available to plant roots in natural soil solutions are the inorganic species arsenite (As(III)) and arsenate (As(V)). These have different modes of action; As(III) toxicity results from its ability to react with protein sulfhydryl groups, thereby affecting their function (Leonard & Lauwerys, 1980), whereas As(V), being chemically similar to inorganic orthophosphate (Pi), is incorporated into compounds such as ATP and the As(V)-ATP analogue, and so deprives the cell of its energy source, ultimately leading to cell death (Meharg, 1994).

Both As(V) and As(III) are the primary forms of As in soils. In aerobic soils As(V) is the predominant species (Masscheleyn et al., 1991; Marin et al., 1993), so here we focus on uptake of As(V) by an important crop plant, Hordeum vulgare (barley). Because As(V) is a Pi analogue it is effectively transported across the plasma membrane of plants via Pi-transporters, apparently competing with Pi (Asher & Reay, 1979; Meharg et al., 1994).

Most plant species growing in aerobic soils, including major cereal crops, are colonized by arbuscular mycorrhizal (AM) fungi, forming symbioses that play significant roles in the uptake of Pi from soil (Smith & Read, 2008). The symbioses are integral to root function in natural soils, whereas corresponding nonmycorrhizal (NM) plants, frequently used for experiments, are not representative of the situation that would exist in the field. Understanding uptake of both Pi and As(V) in AM plants is therefore of considerable importance, particularly with respect to the possibility that the symbioses may exert a protective effect against As toxicity. Several investigations have shown that AM plants of a number of species are less sensitive to As toxicity, with growth increased compared with NM controls (Covey et al., 1981; Meharg et al., 1994; Ahmed et al., 2006; Chen et al., 2007; Pope et al., 2007; Ultra et al., 2007a,b; Xia et al., 2007; Xu et al., 2008). Because increased P nutrition in NM plants can also reduce As sensitivity (Christophersen et al., 2009) the AM effect may be linked to increases in Pi uptake and growth that occur when plant species inoculated with AM fungi show positive growth and phosphorus (P) responses. However, some plant species, including many cereals, do not show positive responses to AM fungi in terms of growth or total P uptake and may even take up less P in total and grow less well than NM controls (Johnson et al., 1997; Knudson et al., 2003; Zhu et al., 2003; Li et al., 2006; Grace et al., 2009a). It is important to investigate interactions between As toxicity and P nutrition in such plants.

Physiological and molecular evidence shows that AM plants have two pathways by which Pi can be absorbed (Smith et al., 2003). The direct pathway involves uptake via Pi transporters in the epidermis and root hairs, as also in NM plants. In barley, three Pi transporters are primarily expressed in roots and of these HvPht1;1 and HvPht1;2 are considered to play the most important role in direct uptake of Pi from soil (Smith et al., 1999; Rae et al., 2003;Schünmann et al., 2004a,b). These high-affinity transport systems are also involved in As(V) uptake, as shown by both physiological and molecular approaches (Asher & Reay, 1979; Meharg & MacNair, 1991; Meharg & MacNair, 1992; Shin et al., 2004). The mycorrhizal pathway involves uptake of Pi by transporters in external hyphae of the AM fungal symbiont (Harrison & van Buuren, 1995; Maldonado-Mendoza et al., 2002), translocation of P for long distances along the hyphae (see Ezawa et al., 2002) and transfer to the plant across symbiotic interfaces where AM-inducible plant Pi transporters are expressed in root cortical cells (Rausch et al., 2001; Harrison et al., 2002; Paszkowski et al., 2002; Glassop et al., 2005). Even where plants show no or negative growth responses to colonization, in terms of Pi uptake and growth, it has been shown that the mycorrhizal Pi uptake pathway is functional and that a considerable proportion of P can be taken up via the fungi (Zhu et al., 2003; Smith et al., 2004; Li et al., 2006; Grace et al., 2009a). This means that less Pi must pass through the direct Pi-uptake pathway, consistent with downregulation of the high-affinity Pi-transporter genes HvPht1;1 and HvPht1;2 in AM barley (Glassop et al., 2005). It has been proposed that low activity of the direct, high-affinity Pi-uptake pathway, in this case not associated with AM colonization, is linked to lower As(V) uptake in genotypes of Holcus lanatus that are tolerant to As (Meharg & MacNair, 1992). Furthermore, Gonzalez-Chavez et al. (2002) showed that low levels of AM colonization of nontolerant H. lanatus by an isolate of the AM fungus Glomus mosseae reduced As(V) uptake, suggesting lower activity of the direct Pi : As(V) uptake pathway.

Most previous studies investigating effects of AM symbiosis on As sensitivity in plants have used species that respond positively to AM colonization in regard to growth and P uptake (Chen et al., 2006, 2007; Dong et al., 2008; Xu et al., 2008). When AM plants are larger because of relief of P stress, concentrations of other nutrients (and As) may be lower than in NM plants because of ‘tissue dilution’. If AM plants are smaller than in the absence of As because of As toxicity, increased concentrations of nutrients (e.g. P) may follow, owing to ‘tissue concentration’ (Lambert et al., 1979). In consequence, elucidation of mechanisms underlying As sensitivity or toxicity based on tissue concentrations is difficult. In summary, the ability of AM fungi to alleviate As toxicity is well established in plants that respond positively to AM fungi but has not yet been clearly demonstrated for nonresponsive or negatively responsive plants. Furthermore, the mechanisms behind the lower As sensitivity in AM plants in general, whether or not they are positively or negatively responsive, have yet to be fully elucidated (Smith et al., 2009).

Here we used barley, in which the mycorrhizal Pi uptake pathway is known to operate (Zhu et al., 2003; Grace et al., 2009a), to investigate As sensitivity in the absence of a positive AM growth response. Previously we found that toxic effects of As on growth of NM barley were ameliorated by increased P in the plant. However increased soil P had no effect on the specific uptake (i.e. uptake per unit root weight) of As. We found severe reductions in growth of NM barley grown with 20 mg P kg−1 soil and 7 mg As kg−1 soil, and saw reduction of root hair development with increasing levels of As (Christophersen et al., 2009). Therefore, we reduced As additions here and included AM fungi to study their effect on As uptake and toxicity and improve understanding of the mechanisms involved in the decreased sensitivity to As in AM plants. We tested the hypothesis that downregulation of HvPht1;1 and HvPht1;2 (encoding high-affinity Pi-uptake systems in the direct pathway) upon inoculation with AM fungi and upregulation of the AM-induced HvPht1;8 (encoding the Pi-uptake system responsible for transfer of Pi from the symbiotic interface to cortical cells) play a role in decreased As uptake and hence the relatively low As sensitivity in AM plants. By using a compartmented pot system described by Drew et al. (2006) we were also able to assess whether As is transferred to the plant by the mycorrhizal Pi-uptake pathway.

Materials and Methods

Soil

A low P, fine sandy loam from an undisturbed and unfertilized site within the cemetery at Mallala, South Australia, was collected from the top 20 cm and passed through a 5 mm sieve. This site was chosen because it is adjacent to a former trial site used for studies of crop responses to P fertilization. Termination of the trial precluded ongoing use of the soil from the zero-P treatments. Soil and fine quartz sand were separately sterilized by autoclaving twice over 3 d at 121°C for 1 h. The Mallala loam contained 13.8 mg P kg−1 (Olsen P) (Olsen et al., 1954). In addition it contained 1.7 g kg−1 total nitrogen (N), 731 mg potassium (K) kg−1 and 10.3 mg sulphur (S) kg−1; the pH was 7.7 (CaCl2). Soil was mixed with sand in the ratio of 1 : 9 (soil : fine sand). This mix will be referred to as soil hereafter. Additional P was applied to the soil as KH2PO4 to provide additional 20 (Expt 1) or 10 (Expt 2) mg P kg−1 soil. For addition of As(V), KH2AsO4 was added to provide 0, 2.08 and 4.16 (Expt 1) or 2.5 (Expt 2) mg As kg−1 soil. These P and As concentrations were chosen based on our previous work showing that this concentration of P permits colonization by AM fungi and that this concentration of As slightly decreases growth of the NM plants, but that they otherwise remain apparently healthy (H. M. Christophersen, unpublished). Following addition of Pi and As(V) a basic nutrient solution for barley, minus P addition, modified from Murphy et al. (1997) was as added. Final nutrients added were as follows (mg kg−1 dry soil): NH4NO3, 320; Ca(NO3)2.4H2O, 330; KNO3, 283; K2SO4, 104.6; MgCl2.10H2O, 41.5; MnSO4.4H2O, 2.68; ZnSO4.7H2O, 3.4; CuSO4.5H2O, 3.7; CoSO4.7H2O, 1.7; H3BO3, 0.189; MnMoO4.2H2O, 0.36 and FeSO4.7H2O, 8.8). The soil was incubated for 4 d to allow the ions to equilibrate with the soil. In Expt 2, two soil samples from each soil treatment were taken for further analysis. The mean pH of the soil at the start of the experiment was 7.4 (CaCl2) and there were no significant differences between treatments. The amount of resin-extractable P and As in the soil before planting was 9.35 mg P kg−1 soil and 1.43 mg As kg−1 soil.

Inoculum of Glomus intraradices

Glomus intraradices Schenck and Smith DAOM181602 was used as the AM fungal symbiont. Inoculum was produced in pot cultures with clover (Trifolium subterraneum L.) in the same soil : sand mix as used for the main experiment. Pots received 10 ml wk−1 of half-strength Long Ashton solution minus P (Cavagnaro et al., 2001). Plants were grown for 8–12 wk and then dried until used. Clover grown under similar conditions with no added inoculum was used as NM mock inoculum.

The inoculum, a dry crude pot culture material containing soil, roots, spores and hyphae, was thoroughly mixed with the soil at a rate of 10% w : w. In Expt 2 inoculum was mixed with soil in the main plant compartment only. The NM pots were prepared as described for AM pots by adding dry mock inoculum.

Experimental design

Experiment 1 had a factorial design, with As treatment and inoculation as factors. Three As treatments (0, 2.08 and 4.16 mg As kg−1 soil) and two inoculation treatments (NM and AM) were included. Each treatment was replicated six times. This experiment was carried out mainly to determine the levels of As to apply in Expt 2. Experiment 2 was carried out using a pot system with three compartments for growth of the plants (Fig. 1) (Drew et al., 2006). The root hyphal compartment (RHC), buffer compartment (BC) and hyphal compartment (HC) were filled with 2.3, 0.6 and 1.8 kg of soil, respectively. The compartments were separated by a 30 μm mesh which allowed hyphae of AM fungi but not roots to enter the BC and HC. The experiment had a factorial design with As treatment and inoculation as factors. Four As treatments, depending on the placement of the As in the different compartments and two inoculation treatments (NM and AM) were included. Inoculum was placed only in the RHC. Addition of As in the different As treatments was as follows: (0–0) no As addition in RHC or HC, (2.5–0) addition of As in RHC only, (0–2.5) addition of As in HC only and (2.5–2.5) addition of As to both RHC and HC. Compartment BC never contained As and served as a buffer zone with the aim of preventing movement of As between the RHC and the HC. Each treatment was replicated three times. For both experiments pots were randomly rearranged on the bench after every watering to take into account variations in environmental conditions.

Figure 1.

 Diagram of the compartmented pot system, where Glomus intraradices was placed in the root hyphal compartment (RHC). Roots are represented by solid lines and hyphae by broken lines. The buffer compartment (BC) was separated from the RHC and the hyphal compartment (HC) by a 30-μm mesh. For further details, see the text.

Seed sterilization, plant growth and sample collection

Seeds of H. vulgare L. cv. Golden Promise were surface-sterilized, placed on wet sterilized filter paper for 2 d at 4°C and then germinated for 24 h at 25°C, in the dark. Two germinated seeds were planted per pot. In Expt 1 plants were grown for 3 wk in a semi-controlled glasshouse under natural light from late August to September. The glasshouse temperature was set at 22 ± 2°C during the day and at 14 ± 2°C during the night. In Expt 2 seedlings were thinned to one per pot after emergence and plants were grown for 5 wk in a growth chamber with a 16 h photoperiod (average of 500 μmol m−2 s−1 over the growth period) and a temperature range from 16 to 23°C. The soil was maintained at 10% gravimetric water content by watering to weight with RO water initially every second to third day, or more frequently as required in the later growth stages. In order to minimize movement of As from the HC to the RHC (Expt 2) and vice versa the pots were first watered in BC, followed by RHC and HC. Shoots and roots were harvested. Roots were washed in tap water and rinsed in 0.1 m HCl solution, followed by a thorough rinse in RO water (Marin et al., 1992) and patted dry with paper towels. Fresh weights were recorded before drying. Plant material was dried at 72°C for 48 h. Tissue P and As concentrations were determined by inductively coupled plasma atomic emission spectroscopy (ICP-AES). In Expt 1 a subsample of roots (apical 1–2 cm) was taken for RNA extractions and AM colonization. In Expt 2 the RHC was cored once on each side of the plant with a 2 cm diameter corer in order to collect a random root sample for extraction of RNA and determination of per cent AM colonization. This was done the day before harvesting. Roots collected for RNA extraction were fast frozen in liquid nitrogen and stored at −80°C. Roots for determination of colonization were stored in 50% ethanol until further examination.

Assessment of AM colonization

Roots in 50% ethanol were rinsed thoroughly with RO water and cleared in 10% KOH for 4–5 d at room temperature and then for 3 min at 90°C. They were stained with 5% Schaeffer black ink in 5% white vinegar for 3 min at 90°C and destained in acidified water (Vierheilig et al., 1998). Between each step the roots were placed on ice to cool down quickly and then rinsed in RO water. Roots were maintained in acidified water until they were mounted on slides in glycerol. The per cent colonization, including development of arbuscules and vesicles, was determined according to the magnified intersects method (McGonigle et al., 1990) and control pots were thoroughly checked for lack of colonization under a dissecting microscope.

Tissue digestion and analysis

Weighed samples of oven-dried plant tissue were digested with nitric/perchloric acid. Sample solutions were analysed by Radial circular optic system (CIROS) ICP-AES, by the Waite Analytical Services (University of Adelaide, South Australia). Some samples were analysed in duplicate to give an indication of the homogeneity of the samples. The per cent variation between the duplicate samples was less than 0.5% for both P and As.

Specific uptake of P and As was calculated as total P or As absorbed per g of root dry weight (RDW). The P : As ratios were calculated on a molar basis.

Soil analysis

Soil extractions (Expt 2 only) were done using an anion exchange membrane method (Kouno et al., 1995) omitting the steps related to fumigation-extraction with chloroform. The soil : water ratio used was 1 g per 15 ml. Extract solutions were diluted appropriately and analysed by Radial CIROS ICP-AES, by the Waite Analytical Services (University of Adelaide, South Australia).

Soil pH was measured in a 0.01 m CaCl2 solution. Two grams of soil were shaken in 10 ml CaCl2 solution in an end-over-end shaker for 1 h, the sample was left for soil to settle and pH was measured in the liquid phase.

RNA extractions

RNA was extracted using a Rneasy Plant Mini Kit (Qiagen) following the manufacturer’s instructions. The RNA extractions included the Qiagen ‘on the column’ RNAse-Free DNAse treatment. In addition extracted RNA was DNAse treated using the TURBO DNA-free kit (Applied Biosystems/Ambion) following the manufacturer’s instructions to remove any remains of genomic DNA. The RNA integrity was tested by running RNA on a 1.2% agarose gel and validation of the ribosomal RNA bands. RNA was quantified using the Thermo Scientific NanoDrop 1000 Spectrophometer (Analytical Technologies, Scoresby, Australia); OD 260/280 ratios ranged from 1.94 to 2.09.

Reverse transcription

For optimal use of 18S-ribosomal RNA (18S RNA) as the house-keeping gene (HKG), cDNA was synthesised using an mRNA and 18S-RNA co-application reverse transcription (Co-RT) method described by Zhu & Altmann (2005). The Affinity Script QPCR cDNA synthesis kit (Stratagene, La Jolla, CA, USA) was used for the RT reactions. Different concentrations (0.2, 1 or 5 μm) of the 18S-RNA universal primer, NS21 (Simon et al., 1992) and the oligo-(dT) primer (0.3 μg) were tested with 0.25 and 0.5 μg of total RNA as described by Zhu & Altmann (2005). The optimized combination of primers for the Co-RT reaction was 0.2 μm of NS21 and 0.3 μg of oligo-(dT) primer. Each 20 μl RT reaction included 0.5 μg of total RNA. The RT program was: 5 min at 25°C, 5 min at 42°C, 15 min at 55°C and 5 min at 95°C. cDNA was stored at −20°C. No reverse transcriptase (NRT) controls were included, in order to assess genomic DNA contamination in the downstream quantitative real-time PCR (qPCR) analysis.

Quantitative real-time PCR

Quantitative real-time PCR was performed using the Brilliant II QPCR master mix (Stratagene) containing SYBR green. Amplification and detection of PCR products were performed in a MX3000P instrument (Stratagene). Twenty microlitre QPCR reactions using 150 nmol of each primer and 2 μl templates were set up according to manufacturer’s instructions. High-pressure liquid chromatography (HPLC) -purified primers were synthesized by Operon Biotechnologies GmbH (https://www.operon-biotech.com/index.php). The cDNAs were diluted 1 : 10 for detection of the genes of interest (GOI) and 1 : 10 000 for detection of the HKG, Hv18S.

The QPCR cycling program was initiated with 10 min at 95°C to activate the polymerase, followed by 40 cycles of 95°C for 30 s, 60°C for 30 s and 72°C for 30 s for the GOI. For Hv18S the annealing temp was 61°C for 20 s. All reactions were performed with three biological replicates, each with three technical replications. The normalized relative quantity (NRQ) was determined using the software qbase (Hellemans et al., 2007); Hellemans et al. (2007) argue for the use of more than one HKG. We also tested the elongation factor 1 gene (primer design, Dr Neil Shirley, Australian Centre for Plant Functional Genomics (ACPFG), the University of Adelaide), and the cyclophilin and HSP 70 genes (Burton et al., 2004). The elongation factor 1 and cyclophilin genes showed the most stable expression. However, comparing NRQ when elongation factor and cyclophilin were used as HKG genes with NRQ when 18S-RNA was used gave similar results so here we only present data that was normalized using 18S-RNA. The relative quantity was normalized using one of the biological replicates in the NM As treatment 0–0 as calibrator. Primers used for detection of HvPht1;1, HvPht1;2 and HvPht1;8 transcripts were described by Glassop et al. (2005). The sequences for sense and antisense primers for detection of Hv18SrRNA were 5′-ACCATGGTGGTGACGGGTG-3′ and 5-′TTCCAATTACCAGACACTAACGC-3′, respectively.

Statistical analyses

jmpin (SAS Institute Inc, 1998/2000) was used for two-way analysis of variance with As treatment and inoculation as factors for growth, per cent colonization, P concentration, total P uptake and specific P uptake. The Shapiro–Wilk W test was performed to test for normal distribution of the data. Where residual values were not normally distributed with constant variance, a log transformation of raw data (RDW and shoot P concentration) was done before analysis. The Tukey–Kramer HSD test was used for pairwise comparisons of means.

The effects of the treatments on As concentration, total As uptake, specific As uptake, P : As molar ratios and NRQ of HvPht1;1, HvPht1;2 and HvPht1;8 were analysed individually using the permutational analysis of variance in permanova (Anderson, 2005) using Euclidean distances followed by the corresponding a posteriori pairwise comparison test (Anderson, 2001; McArdle & Anderson, 2001). Factors included in the analyses were: As treatment and inoculation and the interaction. Treatment effects were similarly analysed but with the combination of the two factors as single treatment (i.e. as a one-way analysis).

Results

Experiment 1

AM colonization and plant growth  Addition of As significantly decreased shoot dry weight (< 0.001), as did inoculation with G. intraradices (< 0.001); these effects were independent (Fig. 2). No colonization was found in NM roots. The percent colonisation in AM root samples was very low, i.e. ≤ 5% for all As addition levels (results not shown).

Figure 2.

 Shoot (above the line) and root (below the line) DW of nonmycorrhizal (open bars) and arbuscular mycorrhizal (closed bars) barley (Hordeum vulgare) grown in pots with no arsenic (As) or As added to the soil at the rate of 2.08 and 4.16 mg kg−1 soil. Data are presented as mean values ± SE (= 6). The effect of inoculation of both shoot DW and root DW was significant (< 0.0001) and independent on As addition. Same letters indicate no significantly difference between levels of As by the Tukey–Kramer HSD test at the 5% level.

Uptake of As and P  Concentrations of P in shoots ranged from c. 3.0–3.5 g kg−1. There were no effects of As addition, but concentrations in AM plants were significantly lower (by c. 10%) than in NM plants (Table 1a). The addition of As had an effect on root P concentration, which resulted in a significantly higher root P concentration at the highest As addition level compared with the two lower As addition levels. There was no effect of As or inoculation on specific P uptake (Table 1a).

Table 1.   Experiment 1: phosphorus (P) concentration, specific P uptake (a) arsenic (As) concentration and specific As uptake (b) of barley (Hordeum vulgare) grown in pots with As added to the soil at the rate of 2.08 and 4.16 mg kg−1 soil, and outputs of the corresponding anovas
(a)
As addition (mg As kg−1 soil)P concentration (g kg−1)Specific P uptake (mmol As g−1 RDW)
ShootRoot
NMAMNMAMNMAM
03.47 ± 0.113.15 ± 0.061.96 ± 0.132.11 ± 0.100.27 ± 0.02 0.27 ± 0.003
2.083.45 ± 0.133.23 ± 0.192.12 ± 0.172.25 ± 0.090.28 ± 0.010.28 ± 0.02
4.163.33 ± 0.163.00 ± 0.112.45 ± 0.162.40 ± 0.090.28 ± 0.020.25 ± 0.01
P > F
 As addition (As)0.2251<0.00010.5243
 Inoculation0.00510.20560.4095
 As × inoculation0.76640.31900.2901
(b)
As addition (mg As kg−1 soil)As concentration (mg kg−1)Specific As uptake (μmol As g−1 RDW)
ShootRoot
NMAMNMAMNMAM
  1. Data are presented as mean values ± SE (= 6). Same letters indicate no significantly difference between treatments at the 5% level.

  2. NM, nonmycorrhizal; AM arbuscular mycorrhizal; RDW, root dry weight.

  3. aBecause As was below the limit of detection in this treatment the concentrations and specific uptake could not be calculated.

0aaaaaa
2.086.64 ± 0.17 d5.44 ± 0.21 c192.25 ± 13.20 c155.53 ± 8.37 c2.78 ± 0.18 c2.25 ± 0.11 d
4.1612.49 ± 0.46 a8.68 ± 0.41 b653.33 ± 39.30 a490.00 ± 15.06 b9.16 ± 0.55 a6.84 ± 0.19 b
P > F
 As addition (As)0.03890.02200.0211
 Inoculation0.33490.32300.3122
 As × inoculation0.00010.00030.0002

There were effects of As addition and inoculation on both shoot and root As concentration and specific As uptake (Table 1b). This resulted in lower shoot As concentration and specific As uptake in AM compared with NM plants at both 2.08 and 4.16 mg As kg−1 soil. For roots, however, As concentration was lower in AM plants at the highest As addition level only. There were higher P : As ratios in AM compared with NM plants across the levels of As (Table 2).

Table 2.   Experiment 1: molar ratio of phosphorus (P) : arsenic (As) per plant for nonmycorrhizal (NM) and arbuscular mycorrhizal (AM) barley (Hordeum vulgare) grown with As added to the soil at the rate of 2.08 and 4.16 mg kg−1 soil, and outputs of the corresponding anovas
As addition (mg As kg−1 soil)P : As molar ratios
ShootRoot
NMAMNMAM
  1. Data are presented as mean values ± SE (= 6). Same letters indicate no significantly difference between treatments at the 5% level.

  2. aBecause As was below the limit of detection in this treatment the P : As ratios could not be calculated.

0aaaa
2.081202 ± 411387 ± 4425.4 ± 0.39 b33.6 ± 0.86 a
4.16624 ± 46803 ± 278.67 ± 0.21 d11.3 ± 0.29 c
P > F
 As addition (As)0.00340.0908
 Inoculation0.01240.3082
 As × inoculation0.94100.0003

Gene expression  The expression of the HvPht1;1 gene was not affected by either the addition of As (= 0.0844) or inoculation (= 0.8709), but there was a trend towards a higher expression of the HvPht1;1 gene at the highest As addition (4.16 mg As kg−1 soil) compared with the two lower As addition amounts (Fig. 3a). This effect was significant at the 10% significance level. No effect of As addition or inoculation was detected on the expression of the HvPht1;2 gene (Fig. 3b). The expression of the HvPht1;8 gene was, in general, low and no significant differences were found between treatments. However, there was a trend towards increased expression in AM plants (Fig. S1).

Figure 3.

 Normalized relative quantity (NRQ) of barley (Hordeum vulgare) inorganic orthophosphate (Pi)-transporter genes HvPht1;1 (a) and HvPht1;2 (b) in the direct uptake pathway in roots of nonmycorrhizal (NM) and arbuscular mycorrhizal (AM) barley plants. Arsenic (As) addition level: open bars, 0 mg As kg−1 soil; tinted bars, 2.08 mg As kg−1 soil; closed bars, 4.16 mg As kg−1 soil. The NRQ was determined using 18S-RNA as housekeeping gene and one of the biological replicates in the NM no As treatments as calibrator. Data are presented as mean values ± SE (= 6).

Experiment 2

AM colonization and plant growth  No colonization was found in NM roots. The per cent colonization by G. intraradices was between 32% and 50% (Table S1). No aspect of colonization in roots was affected by As treatment (= 0.3073).

Colonization resulted in a marginally significant reduction in shoot DW (= 0.0522) but no effect on root DW (= 0.4121; Fig. 4). However, the addition of As had significant effects on the DW of both shoots and roots (< 0.001 and = 0.011, respectively). Addition of As to the RHC (As treatments 2.5–0 and 2.5–2.5) resulted in a lower DW compared with the treatments where no As was added to the RHC (As treatments 0–0 and 0–2.5). Importantly, addition of As to the HC alone (As treatment 0–2.5) had no significant effect on plant growth compared with the control (As treatment 0–0) (Fig. 4).

Figure 4.

 Shoot (above the line) and root (below the line) DW of nonmycorrhizal (open bars) and arbuscular mycorrhizal (closed bars) barley (Hordeum vulgare) grown in compartmented pots with no arsenic (As; 0–0) or As added to the soil at the rate of 2.5 mg kg−1 soil in the root hyphal compartment (RHC) (2.5–0), hyphal compartment (HC) (0–2.5) or both (2.5–2.5). See text for full explanation. Data are presented as mean values ± SE (= 3). Same letters indicate no significantly difference between levels of As by the Tukey–Kramer HSD test at the 5% level.

Concentration of P and specific P uptake  There was no effect of mycorrhizal colonization on shoot and root P concentration or specific P uptake (Table 3a). However, As treatment had significant effects on both shoot and root P concentrations and specific P uptake. Shoot P concentration was higher when As was added to the RHC (As treatments 2.5–0 and 2.5–2.5) than when no As was added (As treatments 0–0 and 0–2.5). Root P concentration was higher when As was added to the RHC (As treatments 2.5–0 and 2.5–2.5) than when no As was added to any of the compartments (As treatment 0–0). Addition of As to the HC only (As treatment 0–2.5) gave mean values for P concentration in roots that were intermediate between those for As treatment 0–0 and those for As treatments 2.5–0 and 2.5–2.5. Specific P uptake was higher when As was added to the RHC (As treatments 2.5–0 and 2.5–2.5) than when no As was added (As treatments 0–0). Mean values for As treatment 0–2.5 were again intermediate between those for As treatment 0–0 and those for As treatments 2.5–0 and 2.5–2.5.

Table 3.   Experiment 2: phosphorus (P) concentration, specific P uptake (a) arsenic (As) concentration and specific As uptake (b) of barley (Hordeum vulgare) grown in compartmented pots with no As (0–0) or As added to the soil at the rate of 2.5 mg kg−1 soil in the root hyphal compartment (RHC, 2.5–0), hyphal compartment (HC, 0–2.5) or both (2.5–2.5), and outputs of the corresponding ANOVAs
(a)
As treatment (mg As kg−1 soil)P concentration (g kg−1)Specific P uptake (mmol P g−1 RDW)
ShootRoot
NMAMNMAMNMAM
0–01.74 ± 0.851.91 ± 1.200.81 ± 0.201.22 ± 0.300.16 ± 0.040.21 ± 0.05
2.5–02.40 ± 0.252.26 ± 0.341.59 ± 0.071.55 ± 0.320.30 ± 0.030.27 ± 0.09
0–2.51.78 ± 0.111.91 ± 0.111.17 ± 0.061.37 ± 0.120.22 ± 0.010.21 ± 0.02
2.5–2.52.23 ± 0.162.53 ± 0.161.47 ± 0.031.46 ± 0.160.28 ± 0.030.30 ± 0.01
P > F
 As treatment (As)<0.0010.01200.0164
 Inoculation0.23740.20860.1190
 As × inoculation0.48540.43860.3019
(b)
As treatment (mg As kg−1 soil)As concentration (mg kg−1)Specific As uptake (μmol As g−1 RDW)
ShootRoot
NMAMNMAMNMAM
  1. Data are presented as mean values ± SE (= 3). Same letters indicate no significantly difference between treatments at the 5% level.

  2. NM, nonmycorrhizal; AM, arbuscular mycorrhizal. See text for full explanation.

  3. aBecause As was below the limit of detection in this treatment the concentrations and specific uptake could not be calculated.

0–0aaaaaa
2.5–03.38 a3.01 a175.98 a119.51 ab2.50 a1.71 ab
0–2.5aa5.92 c16.43 c0.08 c0.22 c
2.5–2.52.81 a3.30 a145.84 a82.34 b2.07 a1.24 b
P > F
 As treatment (As)0.00010.00010.0001
 Inoculation0.69190.00290.004 
 As × inoculation0.01050.00680.0094

As concentration and specific As uptake  There were effects of AM colonization and As addition on shoot and root As concentration, total As uptake and specific As uptake (Table 3b). Considerable amounts of As were taken up when As was added to the RHC (As treatments 2.5–0 and 2.5–2.5), as shown by shoot and root As concentration and specific As uptake. There was no detectable As in As treatment 0–0. Amounts of As taken up were relatively low when As was present in the HC only (As treatment 0–2.5). Importantly, transfer of As to the shoots was low, as most of the As remained in the roots. In As treatment 2.5–2.5 root As concentration and specific As uptake was significantly lower in AM compared with NM plants. Although not significant, there was also a trend towards lower root As concentration, total As uptake and specific As uptake in AM compared with NM plants in As treatment 2.5–0.

P : As ratio  The interaction between AM colonization and As treatment was significant for root P : As ratio only (Table 4). There were higher ratios in As treatment 0–2.5 than in As treatments 2.5–0 or 2.5–2.5. The AM plants had a significantly higher P : As ratios than NM plants in As treatment 2.5–2.5 and the trend was similar in As treatment 2.5–0.

Table 4.   Experiment 2: molar ratio of phosphate (P) : arsenic (As) per plant for nonmycorrhizal (NM) and arbuscular mycorrhizal (AM) barley (Hordeum vulgare) grown in compartmented pots with no As (0–0) or As added to the soil at the rate of 2.5 mg kg−1 soil in the RHC (2.5–0), HC (0–2.5) or both (2.5–2.5), and outputs of the corresponding ANOVAs
As treatment (mg As kg−1 soil)P : As molar ratios
ShootRoot
NMAMNMAM
  1. Data are presented as mean values (= 3). Means followed by the same letter within columns are not significantly different at the 5% level (see the Materials and Methods section). See text for full explanation.

  2. aBecause As was below the limit of detection in this treatment the ratio could not be calculated.

0–0aaaa
2.5–01713 ± 991802 ± 19122.0 ± 1.6 b31.8 ± 2.6 b
0–2.5aa610 ± 183 a224 ± 48 a
2.5–2.51949 ± 2331854 ± 7924.6 ± 2.4 bd43.6 ± 5.3 c
P > F
 As treatment (As)0.36550.2107
 Inoculation0.98030.4650
 As × inoculation0.51500.0104

Gene expression  Inoculation with G. intraradices resulted in a significant reduction in the expression of HvPht1;1 and HvPht1;2 (< 0.001 and = 0.045 respectively) and a very marked increase in the expression of HvPht1;8 (< 0.001; Fig. 5a–c). A very low level of expression of this gene in NM roots was found in some samples, but NRQ values were close to the detection limit, and therefore not well replicated (Fig. 5c). The effects of As treatment were not significant for the expression of any of the genes.

Figure 5.

 Normalized relative quantity (NRQ) of barley (Hordeum vulgare) inorganic orthophosphate (Pi)-transporter genes HvPht1;1 (a) and HvPht1;2 (b) in the direct uptake pathway and the HvPht1;8 gene (c) in the arbuscular mycorrhizal Pi-uptake pathway in roots from nonmycorrhizal (NM) and arbuscular mycorrhizal (AM) barley plants. The NRQ was determined using 18S-RNA as housekeeping gene and one of the biological replicates in the NM no arsenic (As) treatment (0–0) as calibrator. Data are presented as mean values ± SE (= 3).

Discussion

The growth data from Expt 1 were used to select an As concentration for Expt 2. The growth depression at the lower As addition amount (2.08 mg As kg−1 soil) was c. 10%. We expected this depression to increase with time and wanted the plants to grow longer in Expt 2, therefore the lower As addition amount was chosen for Expt 2. The low per cent colonization in Expt 1 was probably to the result of two factors; the plants were young (3 wk old), and only the younger portions of the roots were sampled. It is probable that the roots as a whole were more highly colonized. However, a low per cent colonization does not necessarily mean that the symbiosis was not functional, as it has become increasingly obvious that there is no strong correlation between per cent colonization and either increased nutrient uptake at the whole-plant level, or in the extent of any changes in growth (Graham & Abbott, 2000; Li et al., 2008; Smith et al., 2009). All inoculated barley plants in Expt 2 became well colonized by G. intraradices, with values that are within the range shown to have significant physiological effects on barley, including Pi uptake (Zhu et al., 2003; Grace et al., 2009b). There were no effects of As on the extent of colonization (Expt 2), in line with much previous research (Liu et al., 2005; Chen et al., 2007; Xu et al., 2008). Barley showed slight growth depression arising from colonization by G. intraradices. Calculation of uptake on the basis of root weight (specific uptake) allowed us to take into account effects of As on root growth and gain insights into potential uptake of As by the AM fungal hyphae. Arbuscular mycorrhizal (AM) symbiosis did not affect P concentration or specific P uptake in either experiment. However, the lower growth of shoots and roots following addition of As to the RHC (Expt 2) resulted in higher P concentrations, and higher specific P uptake in both NM and AM plants, so that reduction in root growth had no effect on total P uptake. This is an example of tissue concentration in small plants (see the Introduction section).

The significantly lower specific As uptake into AM roots compared with NM roots at both 2.08 and 4.16 mg As kg−1 soil (Expt 1) and in As treatment 2.5–2.5, and a similar trend in As treatment 2.5–0 (Expt 2), indicates a clear effect of AM symbiosis in reducing As uptake when As is available to the roots. This reduced uptake of As in AM plants was also reflected in the P : As ratios, which were higher in AM plants than in NM plants at the 2.08 and 4.16 mg As kg−1 soil additions (Expt 1) and in As treatments 2.5–0 and 2.5–2.5 (Expt 2). The presence of some As in roots of NM plants in As treatment 0–2.5, where As was present only in the HC, shows that the buffer zone was not entirely effective in preventing As movement in the sandy soil that was used. However, there was a trend towards a higher As concentration and specific As uptake in the AM than in NM roots. This result was also reflected in the P : As ratios. In addition, soil As concentrations for As treatment 0–2.5 were lower in the AM treatment than in the NM treatment at harvest (data not shown). Together, these findings indicate that As was taken up by G. intraradices, but was probably not extensively translocated along the hyphae or transported to plant cells and thence to the plant shoots. There are two possible explanations: the AM fungus may reduce As(V) to As(III) and export As(III) to the soil, as reported in ericoid mycorrhizal fungi (Sharples et al., 2000), or As is sequestered in fungal structures in the root. Both scenarios would mean that As(V) never reaches the interfacial apoplast and would not be available for uptake by the HvPht1;8 transporter.

This is the first study where the effect of long term exposure to As on the expression of Pi transporter genes has been carried out in soil-grown plants. In relation to our initial hypotheses we showed that colonization of barley by G. intraradices resulted in downregulation (Expt 2) of HvPht1;1 and HvPht1;2 encoding high-affinity P transporters mainly responsible for direct P uptake in barley. The lack of effect of colonization on expression of HvPht1;1 and HvPht1;2 in Expt 1 could possibly result from the age of the plants and the way the roots were sampled for RNA extractions. Previous reports have shown downregulation of genes encoding epidermal P transporters in AM plants (Liu et al., 1998; Burleigh & Harrison, 1999; Burleigh, 2001; Glassop et al., 2005). However, such downregulation is variable and has only occasionally been combined with attempts to quantify contributions of the direct epidermal and AM-uptake pathway using radioactive P (see Smith et al., 2009 and references therein). It has been speculated that reduced activity of the direct P-uptake pathway is part of an As tolerance in AM plants (Meharg & MacNair, 1992; Gonzalez-Chavez et al., 2002; Chen et al., 2007). We have now shown this to be the case. Furthermore, we showed that the functionality of the AM Pi-uptake pathway was maintained by induced expression of HvPht1;8 upon inoculation by G. intraradices (Expt 2). The low level of expression of the HvPht1;8 gene in Expt 1 and the lack of a significant upregulation, is most likely caused by the sampling differences, as mentioned earlier. Even so, we still saw an effect of AM on specific As uptake and P : As ratios. There was no effect of colonization on specific P uptake, so we conclude that the AM Pi-uptake pathway compensated for any loss of activity of the direct uptake pathway in P absorption. Lower activity of the direct Pi-uptake pathway would be expected to reduce As uptake as observed here and by Gonzalez-Chavez et al. (2002) using excised AM roots of the grass H. lanatus (on which the external mycelium was probably lacking or damaged). A study carried out on an Arabidopsis PHT1;1 mutant in agar by Catarecha et al. (2007) sheds little light on our findings because of the very severe As toxicity symptoms in the wild-type, which resulted in severe growth depressions (see also Zhao et al., 2009). In consequence, comparisons of As and Pi accumulation cannot easily be reconciled with data for short-term uptake or gene expression.

The addition of As had no effect on expression of HvPht1;8 or on development of arbuscules (Expt 2), suggesting that gene expression is associated with formation of arbuscules as shown by Isayenkov et al. (2004) for expression of MtPT4 in Medicago truncatula. This aspect of symbiotic function requires further investigation, particularly in relation to higher concentrations of As application. At the highest As addition concentration in Expt 1 we found the expression of HvPht1;1 to be higher (at the 10% significance level) than when no As was added, and in a previous experiment with NM barley grown at 20 mg P kg−1 soil and 0, 5 or 10 mg As kg−1 soil (Christophersen et al., 2009) we found that the increasing amount of As added to the soil resulted in an increasing level of expression of HvPht1;1 and HvPht1;2 (Fig. S2). However, this increase in the expression of HvPht1;1 and HvPht1;2 did not result in higher P uptake in plants that received 5 or 10 mg As kg−1soil compared with plants that did not receive As (Christophersen et al., 2009). Whether this increase in expression of HvPht1;1 and HvPht1;2 with increasing amounts of As also applies to the expression of HvPht1;8 and other Pi-transporter genes requires further investigation.

We found very low expression levels of HvPht1;8 in some of the NM roots, as also reported by Glassop et al. (2005). However, in-situ hybridization studies localized the HvPht1;8 transporter solely to cortical cells containing arbuscules (Glassop et al., 2005), which indicates that the main role of the HvPht1;8 transporter is in uptake of Pi from the arbuscular interface and therefore it is most relevant in AM plants.

In summary, we have shown that the decreased uptake of As in barley roots inoculated with G. intraradices can in part be explained by the downregulation of expression of the HvPht1;1 and HvPht1;2 genes encoding transporters mainly located in the epidermis and root hairs and therefore involved in direct uptake from soil. Operation of the AM Pi-uptake pathway compensates for loss of direct Pi uptake consequent to this downregulation. Furthermore, the AM Pi-uptake pathway is active in As-treated plants, but appears to transfer little or no As. The AM pathway thus provides a route for P uptake which does not simultaneously transport As and therefore provides protection against As uptake via the direct pathway and consequent toxicity. We anticipate that similar mechanisms will operate in AM responsive plants. Further work is required to investigate the mechanism(s) underlying lack of As transport via the AM pathway.

Acknowledgements

We thank the Australian Research Council for funding (Grant #06ARC_DP0662916), Dr Evelina Facelli for statistical advice, Rebecca Stonor and Dr Maria Manjarrez for technical assistance, and Dr Neil Shirley for providing standard material for use in QPCR. We thank the anonymous referees for helpful comments on a previous version of this manuscript.

Ancillary