Arbuscular mycorrhizal (AM) fungi form symbiotic associations with most land plants and promote growth of the host through enhanced uptake of phosphate (Pi) (Smith & Read, 2008). It has been well documented that the high-affinity Pi transporters on the plasma membrane of extraradical hyphae play a main role in Pi uptake from soil (Harrison & Van Buuren, 1995; Maldonado-Mendoza et al., 2001). On the other hand, the mycorrhiza-specific plant Pi transporters localized on the periarbuscular membrane are responsible for the uptake of Pi released from arbuscules (Rausch et al., 2001; Maeda et al., 2006; Javot et al., 2007). Despite increasing knowledge of the membrane transport systems in the symbiotic associations, information about the mechanism of Pi translocation through AM fungal hyphae is quite limited. Evidence that AM fungi accumulate polyphosphate (polyP) in hyphae was first obtained more than three decades ago (Callow et al., 1978). Polyphosphate is a linear chain of three to thousands of Pi residues linked by high-energy phosphoanhydride bonds and has been found in nearly all classes of organisms (Kornberg et al., 1999). The compound has many functions in the cell, including acting as a Pi reservoir, an alternative energy source of ATP and a metal chelator (Kornberg et al., 1999). Although polyP is suggested to be involved in long-distance Pi translocation through hyphae in AM associations (Cox et al., 1980; Ezawa et al., 2002), the physiological roles and behavior of the compound in the fungi are largely unknown. It has been reported that the compound consists of only a small part of total cellular phosphorus (P) in AM fungi: the proportions of polyP to total P were estimated as 16% in Glomus mosseae (Capaccio & Callow, 1982) and 5–17% in Gigaspora margarita (Solaiman et al., 1999). These estimates suggest that Pi and/or other P compounds may play a more significant role in P storage/translocation in the fungi. On the other hand, Pi taken up by hyphae was converted to polyP quite rapidly (Ezawa et al., 2004; Viereck et al., 2004), and the rate of polyP accumulation was comparable to that of a polyP-hyperaccumulating bacterium found from activated sludge (Ezawa et al., 2004). These observations led us to hypothesize that the fungi could potentially accumulate much larger amounts of polyP than previously reported. In particular, Viereck et al. (2004) provided a comprehensive view of the relative dynamics of various (soluble) P compounds in an AM fungal mycelium using the in vivo31P-NMR (nuclear magnetic resonance) technique, and suggested that polyP might be the largest P storage in the fungi. However, NMR-invisible P compounds, such as long-chain polyP and structural P, might be present in the cell, and the absolute (potential) pool size of the cell for polyP has not been estimated so far. Therefore, further quantitative study on the dynamic of polyP in AM fungi with respect to total cellular P is required. In addition, it was predicted that the maximum pool size for polyP would be demonstrable in P-starved AM fungi, which could accumulate polyP as rapidly as a polyP-hyperaccumulator (Ezawa et al., 2004). In the present study, the dynamics of polyP, total P and Pi were investigated in an AM fungus grown under P-starvation conditions to clarify the significance of polyP in P storage/translocation in AM fungal associations.
Materials and Methods
Lotus japonicus L. cv Miyakojima MG-20 (National Bioresource Project Legume Base, http://www.legumebase.agr.miyazaki-u.ac.jp/index.jsp) were sown on moistened filter paper in a Petri dish and germinated at 25°C for 2 d in the dark. Three seedlings were transplanted to the mycorrhizal compartment (MC) of a dual mesh bag culture system in a 230 ml plastic pot (7.6 cm diameter) and inoculated with Glomus sp. HR1 (MAFF 520076) at 500 spores per pot. The dual mesh bag system consisted of two main compartments – a MC and a hyphal compartment (HC) – that were separated by a cone-shaped dual nylon mesh bag (37 μm pore size, Nippon Rikagaku Kikai, Tokyo, Japan) (Supporting Information, Fig. S1). The MC was defined as the region inside the inner mesh bag (31 ml in volume, 5.2 × 4.5 (W × H) cm), and the HC (159 ml in volume) was defined as the region outside of the outer mesh bag (6.8 × 5.9 (W × H) cm). The medium in these compartments was autoclaved river sand. In between the inner and outer mesh bags, autoclaved subsoil with a high-P absorption coefficient (2.8 g P2O5 kg−1 soil, pH 5.0) was layered as a P-diffusion barrier (10 mm in width). The pore size of the nylon mesh was small enough to prevent L. japonicus roots from passing through, but large enough to allow passage of AM fungal hyphae. The seedlings were grown in a growth chamber equipped with fluorescent light at a photon flux density of 150 μmol m−2 s−1 (16 h photoperiod, 25°C) and thinned to two plants per pot at 1 wk after sowing. The plants (whole pot) received deionized water (DIW) every other day for the first week, then low-P nutrient solution (4 mM NH4NO3, 1 mM K2SO4, 75 μM MgSO4, 2 mM CaCl2, 50 μM Fe-EDTA and 50 μM KH2PO4) from the second to the sixth week, and nonP nutrient solution (KH2PO4 was withheld from the low-P nutrient solution) for the seventh week in sufficient amount until the solution flowed out from the drain. At the beginning of eighth week, a 1 mM Pi (KH2PO4) solution was applied using a pipette to the HC gently in a sufficient amount until the solution flowed out from the drain, and then the Pi solution was washed out by applying DIW with a watering can in a sufficient amount 1 h after Pi application. Mycorrhizal roots and extraradical mycelium in the MC and mycelium in the HC were harvested separately after Pi application at 1 h intervals as follows. The MC (inner mesh bag) was taken off the pot and transferred to water, and then mycorrhizal roots and attaching mycelium (roots + mycelium) were collected after removing adhering sand particles by gentle shaking in the water. Detached mycelium in the water was further collected by the wet sieving and combined with the roots + mycelium fraction. Extraradical mycelium in the HC was collected by the wet sieving after removing the P-diffusion barrier (outer mesh bag). The samples were blotted with a paper towel, frozen in liquid nitrogen immediately and stored at −80°C.
In the case of mycelium from the HC, 5–30 mg (FW) of material was ground in an ice-cooled mortar and pestle with 10- to 20-fold volume (w/v) of extraction buffer (8 M urea : 50 mM Tris-HCl, pH 8.0) and transferred to a 1.5 ml tube. In the case of roots + mycelium from the MC, 0.7–2.0 g (FW) of material was ground in a mortar with liquid nitrogen and mixed with fivefold volume (w/v) of the extraction buffer, and then 500 μl of the slurry was transferred to a 1.5 ml tube.
For determination of total P in the HC samples, 100 μl of the slurry was transferred to a 14.7 ml Teflon vial (Savillex, Minneapolis, MN, USA), mixed with 2.5 ml of 1.8 M sulfuric acid, heated at 250°C for 2 h to evaporate water (until sulfuric acid was concentrated) and then digested at 250°C for 1 h by using 0.2–0.3 ml hydrogen peroxide as an oxidant. Pi concentrations in the digests were determined by the ascorbic acid method (Watanabe & Olsen, 1965). PolyP concentrations in the slurries prepared from the HC and MC samples were determined by the reverse reaction of polyphosphate kinase (PPK), as described by Ezawa et al. (2004). Free Pi concentrations in the slurries from the HC samples were determined by the ammonium molybdate method (Ohnishi et al., 1975) using 200 μl supernatant obtained after centrifuging the slurry at 14 000 g for 15 min. Ten microliters of the slurry from the HC and MC samples was taken for the determination of protein concentration using the DC Protein Assay Kit (Bio-Rad Laboratories, Tokyo, Japan) with bovine serum albumin as standard.
Experimental setup and data analysis
The time course analysis (from 0 to 9 h after Pi application) of total P and that of polyP in the HC were conducted separately using different batches of plant/fungal material. For these analyses, 22 pots were prepared as one batch, and mycelial samples collected from two pots grown in the same batch were combined as one sample (5–30 mg FW per sample). One set of time course experiments (from 0 to 9 h after Pi application) was conducted using one batch (without replication) and triplicated using three independent batches. For data analysis, average values were calculated from the data obtained from the three replicated experiments (n =3). For the time course analysis of polyP in roots + mycelium in the MC (from 0 to 10 h after Pi application), 33 pots were prepared as one batch, and three samples (0.7–2.0 g FW per sample) harvested from three pots were analyzed separately (n =3). For the simultaneous analysis of total P, polyP and Pi in the HC (from 0 to 6 h after Pi application), 22 pots were prepared as one batch, and mycelial samples collected from two pots grown in the same batch were combined as one sample. One set of time course experiments (from 0 to 6 h after Pi application) was conducted using one batch (without replication) and replicated five times using five independent batches. For data analysis, average values were calculated from the data obtained from the five replicated experiments (n =5).
Analysis of variance (ANOVA) with the Tukey–Kramer test as a post-hoc test or Student’s t-test was applied for data analysis using the StatView software (SAS Institute Inc., Cary, NC, USA).
The total P content of extraradical mycelium in the HC was 3.6 μmol mg−1 protein at time zero, which increased to 8.2 μmol mg−1 protein 5 h after Pi application and then decreased to 5.2 μmol mg−1 protein 9 h after Pi application (Fig. 1). The polyP content of extraradical mycelium in the HC increased from 0.5 to 7.1 μmol mg−1 protein from 0 to 6 h after Pi application and decreased to 2.8 μmol mg−1 protein by 9 h after Pi application (Fig. 2). The apparent accumulation rates of total P and polyP from 0 to 5 h after Pi application were 1.03 and 1.14 μmol mg−1 protein h−1, respectively, and were not significantly different (t-test, P <0.05). The apparent declining rates of total P and polyP from 5 to 9 h after Pi application were 0.94 and 1.06 μmol mg−1 protein h−1, respectively, and were also not different at P <0.05. The polyP content of mycorrhizal roots + extraradical mycelium in the MC was maintained within a range of 3.8 to 6.6 nmol mg−1 protein from 0 to 5 h after Pi application, increased to 44.3 nmol mg−1 protein from 6 to 9 h after Pi application and then decreased to 9.8 nmol mg−1 protein 10 h after Pi application (Fig. 3). To investigate whether Pi and/or unknown precursors of polyP were accumulated before polyP accumulation, the content of free Pi, total P and polyP of extraradical mycelium in the HC was measured simultaneously. Total P and polyP content increased synchronously from 2.9 to 6.7 μmol mg−1 protein and from 0.2 to 4.3 μmol mg−1 protein, respectively, from 0 to 5 h after Pi application (Fig. 4). By contrast, free Pi content remained constant in the range 200–500 nmol mg−1 protein during the experiment. It is noteworthy that the concentration of polyP reached 64% of total P 5 h after Pi application. Differences between amounts of total P and polyP were constant from 0 to 6 h after Pi application (Fig. 4).
The present study demonstrated that the AM fungus was capable of accumulating polyP > 60% of total cellular P, implying that the potential pool size of polyP in the cell was much larger than previously considered. Rapid and massive accumulation of polyP in microorganisms was first discovered in yeast Saccharomyces cerevisiae more than four decades ago and defined as ‘polyP overplus (or overcompensation)’ (reviewed in Harold, 1966); for example, the level of polyP in S. cerevisiae that was grown under P-deficient conditions increased 20-fold 2 h after resupply of Pi, up to 38 mg g−1 DW (corresponding to 400 μmol g−1 DW) (Trilisenko et al., 2002). In the present study, Glomus sp. HR1 was found to respond to Pi similarly: the amount of polyP in the fungus grown under P-deficient conditions increased 14-fold 5 h after Pi application, up to 7.1 μmol mg−1 protein, corresponding to 390–500 μmol g−1 DW (the values were estimated based on the following parameters: protein content, 11–15 mg g−1 FW; water content, 80%). Although this was achieved by the application of quite a high concentration of Pi that rarely occurs under natural conditions, the amount is comparable to that observed in the yeast ‘polyP overplus’ and much higher than previously reported in AM fungi: G. mosseae intraradical hyphae, 32 μmol g−1 DW (Capaccio & Callow, 1982); G. margarita extraradical hyphae, 32–48 μmol g−1 DW (Solaiman et al., 1999) (estimated on the assumption that water content was 80%). It has been considered that ‘polyP overplus’ would be the feature evolved in a wide range of microorganisms to prepare P deficiency, because Pi availability tends to be low in natural environments (Harold, 1966). Glomeromycotan fungi may have acquired the traits involved in ‘polyP overplus’ during the early evolution and could successfully provide a great competitive advantage in P acquisition for their host plants by facilitating a large P pool.
Our observations that total P and polyP increased and decreased synchronously without fluctuation in the amount of free Pi suggest that neither Pi nor an intermediary P metabolite such as short-chain polyP or organic P compound, which could be detected by the total P measurement but not by the PPK/luciferase method, was accumulated before polyP accumulation. This supports the idea that polyP formation in AM fungi is quite rapid (Ezawa et al., 2004) and thus contributes to maintaining cellular Pi at a low concentration for efficient uptake of Pi (Viereck et al., 2004), confirming that polyP plays a significant role in AM fungal P metabolism as a temporary but largest P storage.
Polyphosphate first increased in extraradical mycelium in the HC and later in mycorrhizal roots + mycelium in the MC. These observations were consistent with those reported by Viereck et al. (2004) and strongly suggest that polyP mediates long-distance P translocation through hyphae. In this context, our experimental system could provide further information about parameters for P translocation in AM symbiosis: the declining rates of polyP in mycelium in the HC from 5 to 9 h after Pi application (1.06 μmol mg−1 protein h−1, corresponding to 58–80 μmol g−1 DW h−1) can be regarded as the apparent P-translocation rate through hyphae towards the plants. In addition, the time lag of 5–6 h for the increases in polyP level in roots + mycelium in the MC can be regarded as the time required for P to pass the 10 mm diffusion barrier, that is, polyP (or Pi) moved towards the plant at 1.7–2.0 mm h−1. In the in vitro carrot root-organ–G. intraradices association, it took 14 h to translocate radiolabeled P from extraradical to intraradical hyphae (Nielsen et al., 2002), suggesting that the processes of P accumulation/translocation in vitro were slower than those in our open culture system. Differences in experimental systems may affect the energy status of the fungal partner through carbon supply from the plant partner, and this may be reflected in the differences in the rates of P accumulation/translocation (Olsson et al., 2002; Bücking & Shachar-Hill, 2005). It has been well documented that there are large inter- and intraspecific variations in the efficiency of P uptake by AM fungi (Jakobsen et al., 1992; Maldonado-Mendoza et al., 2001; Nielsen et al., 2002; Munkvold et al., 2004). In fact, the apparent polyP accumulation rate of Archaeospora leptoticha (Ezawa et al., 2004) is more than twice as rapid as that of Glomus sp. HR1. Accordingly, the parameters for polyP accumulation/translocation presented in our study could be applicable for the assessment of inter- and intraspecific variations in P delivery potential among AM fungi.
In the present experimental system, polyP formation (a net increase in polyP) was observed even after the Pi-washing (removal) process conducted 1 h after Pi application. The washing step shortened the duration of Pi uptake by hyphae and was thus essential to estimate the declining rate of polyP in mycelia. Two reasons can be proposed to explain the prolonged Pi uptake after the washing process: first, the amount of DIW used for the washing was insufficient, and thus Pi remained in the medium; and second, Pi was captured in hydrophilic (viscous) material such as extracellular polysaccharide gels around hyphae, which might be secreted by the fungus and removed by the wet sieving but not by watering. The former is unlikely, because an increase in DIW for the washing process did not shorten the duration of Pi uptake. The fact that adherence of organic material and fine sand particles to mycelia was observed frequently suggests that extracellular polysaccharide is present around hyphae, and thus the latter seems likely.
This study demonstrated the significance of polyP as the largest P storage and a mediator of long-distance P translocation in AM fungi. It remains uncertain, however, whether P is translocated as polyP without turning over, or whether it is translocated as Pi through dynamic regulation of polyP synthesis/hydrolysis (Ezawa et al., 2001). Further investigations are required to understand the whole picture of the P-delivery system in the symbiotic associations.
The PPK-over expressing E. coli was a kind gift from Prof. Arthur Kornberg (Stanford University, deceased). We thank Dr Derek B. Goto (Hokkaido University) for critical reading of the manuscript. This work was supported by the Grant-in-Aid for Scientific Research (19380040) from JSPS (to TE).