Mechanisms of microbially enhanced Fe acquisition in red clover (Trifolium pratense L.)

Authors


Shao Jian Zheng. Fax: +86 5718643 3724; e-mail: sjzheng@zju.edu.cn

ABSTRACT

Soil microorganisms may play an important role in plant Fe uptake from soils with low Fe bioavailability, but there is little direct experimental evidence to date. We grew red clover, an Fe-efficient leguminous plant, in a calcareous soil to investigate the role of soil microbial activity in plant Fe uptake. Compared with plants grown in non-sterlie (NS) grown plants, growth and Fe content of the sterile(s) grown plants was significantly inhibited, but was improved by foliar application of Fe EDTA, indicating that soil microbial activity should play an important role in plant Fe acquisition. When soil solution was incubated with phenolic root exudates from Fe-deficient red clover, a few microbial species thrived while growth of the rest was inhibited, suggesting that the Fe-deficient (– Fe) root exudates selectively influenced the rhizosphere's microbial community. Eighty six per cent of the phenolic-tolerant microbes could produce siderophore [the Fe(III) chelator] under – Fe conditions, and 71% could secrete auxin-like compounds. Interestingly, the synthetic and microbial auxins (MAs) significantly enhanced the Ferric reduction system, suggesting that MAs, in addition to siderophores, are important to plant Fe uptake. Finally, plant growth and Fe uptake in sterilized soil were significantly increased by rhizobia inoculation. Root Fe–EDTA reductase activity in the – Fe plant was significantly enhanced by rhizobia infection, and the rhizobia could produce auxin but not siderophore under Fe-limiting conditions, suggesting that the contribution of nodulating rhizobia to plant Fe uptake can be at least partially attributed to stimulation of turbo reductase activity through nodule formation and auxin production in the rhizosphere. Based on these observations, we propose as a model that root exudates from – Fe plants selectively influence the rhizosphere microbial community, and the microbes in turn favour plant Fe acquisition by producing siderophores and auxins.

INTRODUCTION

Fe is an essential nutrient element for plant growth and development, and is involved in chlorophyll (Chl) and thylakoid syntheses and chloroplast development. Although the total Fe content of soils is much higher than plants require, its bioavailability is limited (Guerinot & Yi 1994) in most soils, particularly in calcareous soils which cover about 30% of the earth's surface (Vose 1982). As a result, Fe deficiency chlorosis is evident in about 30% of crops worldwide (Imsande 1998). Fortunately, many native plant species and genotypes, termed Fe-efficient plants, have developed strategies to avoid Fe deficiency-induced chlorosis. These strategies are classified as strategy I in non-graminaceous monocots and dicots, and strategy II in graminaceous monocots (Römheld & Marschner 1986). Because this study involves a dicot, we will not further discuss strategy II.

Strategy I plants respond to Fe deficiency stress by inducing a cell surface reduction system (‘turbo’) of soluble Fe(III)-chelating compounds in the root system, and by acidifying the extracellular medium. In most instances, these actions are accompanied by the release of reductants such as phenolic compounds (Curie & Briat 2003; Hell & Stephan 2003). Recently, however, two lines of evidence have shown that the physiological responses mentioned earlier are insufficient for strategy I plants to avoid Fe deficiency under Fe-limited conditions. Masalha et al. (2000) demonstrated that sunflower grown in sterilized (S) soil showed severe chlorosis, poor growth and lower tissue Fe content compared with non-sterilized (NS) soil-grown sunflower. Similarly, Rroco et al. (2003) reported that Fe acquisition and growth of rape were significantly reduced when the plant was grown in S soil, but the normal growth was quickly restored by the addition of Fe–ethylenediaminedihydroxyphenylacetic acid (EDDHA) to the S soil. It therefore appears that soil microbial activity plays a role in how strategy I plants avoid Fe deficiency.

Generally, siderophore production by soil microorganisms is seen as the microbial activity that is most supportive of Fe acquisition by plants, because siderophores are chelators with a high affinity for Fe(III) (Raymond, Mueller & Matzanke 1984). Many Fe(III)–siderophore chelates have been proven to be a source of available Fe for plants (e.g. Crowley, Reid & Szaniszlo 1988; Chen et al. 1998; Chen, Dick & Streeter 2000; Yehuda et al. 2000). However, because the rhizosphere interactions between soil microbes and plant roots are quite complex, the actual mechanism by which soil microbes contribute to plant Fe acquisition remains unknown.

Most Fe-efficient dicots secrete phenolic compounds under Fe-deficient (– Fe) conditions (Römheld & Maschner 1986; Römheld 1987; Susín et al. 1996). However, the function of phenolics in plant Fe acquisition is still uncertain. Generally, phenolic compounds would selectively affect the microorganisms’ growth, having both anti-microbial and growth-beneficial effects (Blum et al. 2000; Rauha et al. 2000; Yang & Crowley 2000; Dakora & Phillips 2002). Accordingly, phenolic compounds exuded by – Fe plants most likely inhibit the growth of some rhizosphere microorganisms while stimulating the growth of others. In this study, we investigate the ecological function of phenolic compounds exuded into the rhizosphere by Fe-efficient plants.

Microorganisms isolated from the rhizosphere and rhizoplane are generally more active in producing auxins than those from bulk soil (Sarwar et al. 1992). Auxin production is therefore considered a way in which microbes promote plant growth by stimulating enzymological reactions. Increased plant growth should reasonably require an increase in uptake of nutrient elements such as Fe. On the other hand, some studies (Landsberg 1984, 1996; Li, Zhu & Zhang 2000; Li & Li 2004) have proposed that auxin induces the Fe-deficiency response of plants. It is therefore possible that auxin produced in the rhizosphere by soil microorganisms plays a role in plant Fe acquisition, although there is as yet no direct evidence for this.

Legumes are associated with symbiotic N2 fixation in the nodules, a process in which Fe-containing proteins figure prominently. Nodulated legumes, thus, have an increased need for Fe as compared to non-nodulated plants (Guerinot 1991). However, few studies examine Fe-deficiency response in nodulated legumes. In particular, few studies examine the role of nodulating rhizobia in Fe acquisition in legumes, although this issue is particularly relevant to field conditions.

In the present research, we investigate the effects of soil microorganisms and nodulating rhizobia on Fe acquisition in red clover (Trifolium pratense L.), an Fe-efficient legume (Zheng et al. 2003). Red clover has small seeds, and hence, low seed Fe reserves, which means that the plant obtains most of its Fe from the soil. In addition, we propose a model of how soil microorganisms, especially nodulating rhizobia, contribute to the Fe acquisition of the plant.

MATERIALS AND METHODS

Soil cultivation experiment

Soil and rhizobia

A loess-derived loam soil with a pH of 7.6, and 2.47 mg Fe kg−1 soil of diethylenetriaminepentaacetic acid (DPTA)-extractable Fe (Lindsay & Norvell 1978) were used in this study. Air-dried soil was ground to pass through a 2 mm sieve, then the following chemical fertilizers were mixed into it: Ca(NO3)2 at 100 mg N kg−1 soil, K2SO4 at 100 mg K kg−1 soil and KH2PO4 at 150 mg P. Plant pots measuring 1.25 L in volume were then each filled with 1.3 kg soil.

Rhizobium leguminosarum b. Trifolii ACCC18002 was purchased from the Chinese Academy of Agricultural Science (Beijing, China), and was initially maintained on yeast extract mannitol agar (Vincent 1970). The inoculum was prepared by re-suspending cells from a 3-day-old culture in a sterile saline (0.85% NaCl) solution.

Plant cultivation

The soil surface in the pots was covered by a piece of aluminium foil. Because microorganisms are readily moved from surface to sub-surface soil by infiltrating water (Rroco et al. 2003), a 9-mm-diameter plastic pipe, with a rubber stopper on its top and several small holes on its bottom, was inserted into the soil for watering purposes. The pot surface was then further covered with a transparent polyethylene membrane, as shown in Fig. 1.

Figure 1.

The set-up used for soil-grown plants. The 1.25 L pot was filled with 1.3 kg soil, and the soil surface was covered by a piece of aluminium foil with four holes for plants. A 9-mm-diameter plastic pipe, with a rubber stopper on its top and several small holes on its bottom, was used for watering purposes. The pot surface was further covered with a transparent polyethylene membrane into which holes were cut when the shoot apexes reached it.

The experiment had five replications of the following four treatments: (1) NS soil as the control; (2) S, which means that soil, pots, foil and pipes were all sterilized; (3) S soil with Fe(III)–ethylenediaminetetraacetic acid (EDTA) (S + Fe); and (4) S soil inoculated with rhizobia (S + R). Soil was sterilized in the pots by first watering to 70% field water-holding capacity (FWHC), 28% water content (m/m), then autoclaving three times at 121 °C, 15 min each time at 2 d intervals.

Red clover seeds were surface sterilized by being immersed in 0.1% HgCl2 for 3 min, washed five times with sterile deionized water, immersed in 75% ethanol for 3 min, then shaken in 10% sodium hypochlorite for 20 min and finally rinsed three times with sterile deionized water. The seeds were then sown at a depth of 1 cm, with 2 seeds per hole and 4 holes per pot. Seed sowing was conducted under axenic conditions for all the S treatments. During plant growth, soil moisture was maintained at 28%, with sterile deionized water added using a sterile injector. The plants were grown in a growth chamber at a humidity of 70%, with a daily cycle of 26 °C, 14 h day and 23 °C, 10 h night. The daytime light intensity was 150 µmol photons m−2 s−1. After 16 d growth, when the shoots reached the polyethylene membrane, an opening for each plant was carefully cut into the membrane, and seedlings were thinned to 1 plant per hole. Cutting and thinning were conducted under axenic conditions.

For the Fe(III)–EDTA treatment, we sprayed 100 µm Fe(III)–EDTA solution on the plant every other day starting on day 20. On the last 7 d, we sprayed deionized water instead of the Fe solution to ensure that no un-absorbed Fe(III)–EDTA remained on the leaf surface. Other treatments were sprayed with deionized water following the same schedule, in order to keep growing conditions consistent across treatments.

For the rhizobia inoculation treatment, each seedling was inoculated with 2 mL of the prepared inoculum following the thinning on day 16. Inoculation was conducted under axenic conditions.

The plants were harvested on day 45, and the sub-surface soils were sampled and checked for sterility according to the method of Whilling, De Seoza & Terry (2001), using tryptic soy agar (TSA) plates through serial dilutions.

Analysis

At harvest, Chl was measured in the nearly expanded leaves according to the method of Moran (1982). The shoots were cut at the soil surface, then weighed. The roots were removed from the soil and carefully washed several times with deionized water to remove all adhering soil, then blotted dry with a paper towel. The shoots and roots were dried in an oven at 75 °C to a constant weight, and the dry weights (DWs) were recorded.

For the Fe concentration analysis, the dried root and shoot samples were wet-digested in concentrated HNO3 at 120 °C until no nitrogen oxide gas was emitted, then further digested with HNO3/HClO4 at 180 °C until the solution became transparent. Digestates were diluted with ultra-pure water, and their Fe concentration was analysed with a Shimadzu AA6800 flame atomic absorption spectrophotometer (FAAS) (Kyoto, Japan).

Microorganism cultivation experiment

Isolation of phenolic-tolerant soil microorganisms

One gram of the ground air-dried soil was added to 9 mL of sterile saline solution, and vortexed to make a soil microbial suspension. The suspension was then serially diluted (10−2, 10−3 and 10−4), and for each dilution 100 µL was spread on broth peptone agar plates. The agar was made with 5 g beef extract, 10 g peptone and 5 g NaCl L−1, with the solution pH adjusted from 7.2 to 7.4. These plates also contained phenolics secreted by – Fe roots at an equivalent of 0.6 µmol mL−1 gallic acid (the collection method of the phenolics in the root exudates is described later in ‘Solution cultivation experiment’). After 2 d of incubation at 30 °C, representative morphologically distinct bacterial colonies in the 10−4 dilution agar were randomly selected, harvested from the agar plates, purified with repetitious plate lineation and stored at 4 °C for later analysis.

Siderophore and auxin production by phenolic-tolerant microorganisms and rhizobia

To test whether the isolated phenolic-tolerant microorganisms and the rhizobia produce siderophores under Fe-limited conditions, these microorganisms were cultured in a – Fe solution of the following composition (in mg L−1): K2HPO4, 91; NH4Cl, 1000; MgSO4·7H2O, 63; CaCl2, 50; succinate, 1000; NaCl, 100; l-tyrosine, 100; and l-methionine, 100 (Hersman, Maurice & Sposito 1996). After 3 d of incubation at 30 °C, the siderophore concentration of the solution was determined using chrome azurol S (CAS) reagent according to the method of Schwyn & Neilands (1987).

To test the auxin production, these microorganisms were cultured in l-tryptophan (100 mg L−1) contained in the – Fe medium. After 4 d of incubation, the auxins in the solution were measured using Salkowski reagent according to the method of Sarwar et al. (1992). To further confirm indole-3-acetic acid (IAA) production by the rhizobia, we used high-performance liquid chromatography (HPLC) analysis. Briefly, the cultured solution was centrifuged at 10 000 g for 10 min at 4 °C, then 100 mL of the supernatant was partitioned with three volumes of acidic (pH 2.8–3.0) ethyl acetate. The ethyl acetate fractions were separated from the aqueous phase in a separatory funnel, reduced in volume and dissolved in 2 mL of methanol. Extracts (50 µL) were injected into a Beckman C18 (ODS) 5 µm (Beckman Coulter, Fullerton, CA, USA), 4.6 × 250 mm HPLC column heated to 30 °C, using four parts methanol to six parts 1% ethanol (v/v) as the mobile phase with a flow rate of 0.8 mL min−1. The peaks were recorded at 254 nm.

In addition, one microorganism capable of producing auxins was selected at random for culture in an l-tryptophan-containing medium, and the microbial auxins (MAs) were collected for a solution culture experiment. After 4 d of cultivation, the cultured solution was centrifuged at 10 000 g for 10 min at 4 °C, and the MA supernatant was collected and stored at 4 °C until an analysis was performed. The concentration of auxins in the supernatant was measured with Salkowski reagent, using synthetic auxin IAA as a calibration standard.

Solution cultivation experiment

Collection of phenolic root exudates

The solution cultivation experiment was carried out in the growth chamber under the same growing conditions described earlier. The plant culture method is described in detail in Zheng et al. (2003). Briefly, uniform 15-day-old red clover seedlings were transferred to 1 L pots (4 holes per seedling holder, and 3 seedlings per hole) filled with an aerated, complete nutrient solution. The nutrient solutions were replaced every other day. After 6 d, the plants were transferred to an otherwise identical growth solution having no Fe. The Fe-removing nutrient solutions were also replaced every other day, and the cultured nutrient solutions were gathered to collect phenolic root exudates through the resin adsorption method. Briefly, the gathered nutrient solution was passed through a column filled with SP825 Sepabeads resin (Mitsubishi Chemical Corporation, Tokyo, Japan). The phenolic compounds adsorbed onto the resin were eluted by 100% methanol. The eluant was evaporated to dryness at 40 °C in a rotary evaporator. The residues were redissolved in 5 mL of dimethyl sulphoxide (DMSO), and stored at 4 °C until an analysis was performed. The quantity of total phenolic compounds in DMSO solution was measured colorimetrically at 750 nm using Folin Ciocalteu's reagent (Singleton & Rossi 1965). The concentration of the total phenolic compounds was expressed as a molar equivalent of gallic acid.

Synthetic auxin (IAA) and MA addition experiment

Uniform 15-day-old red clover seedlings were transferred to 1 L pots filled with a complete nutrient solution. After 28 d growth period, the plants were separated into the following six growth solution treatments: (1) Fe sufficient (+Fe); (2) + Fe plus 2 × 10−7m synthetic auxin (IAA) (+Fe + IAA); (3) + Fe plus MA at an IAA-equivalent concentration (+Fe + MA); (4) – Fe; (5) – Fe plus 2 × 10−7m synthetic auxin (IAA) (–Fe + IAA); and (6) – Fe plus MA at an IAA-equivalent concentration (–Fe + MA). To eliminate the effects of other compounds (such as amino acids) in the MA supernatant on the root Fe(III) reductions that were later measured, both half of the + Fe and – Fe treatment pots were added with an un-cultured microbial growth medium at equal volumes of the MA supernatant used in + MA treatments. On day 2 after treatment, roots were excised for measurement of Fe(III) reductase activity.

Rhizobia inoculation experiment

Uniform 15-day-old seedlings were transplanted into 8 L containers (45 holes per seedling holder, and 1 seedling per hole) filled with the complete nutrient solution described previously, except that the concentration of Ca(NO3)2 was decreased to 0.3 mm. Half of the containers also received 100 mL rhizobia inoculum per container. After 12 d growth period, when the nodules became visible, the inoculated and un-inoculated plants were further divided into two groups, respectively, with one group transferred to the + Fe and the other group to the – Fe nutrient solution. Three days later, the roots were excised for the measurement of Fe(III)–EDTA reductase activity.

Measurement of Fe(III) reductase activity

Ferricyanide reduction was measured according to the method of Schmidt (1993), while Fe(III)–EDTA reduction was performed according to the method of Grusak (1995).

RESULTS

Effect of soil sterilization on plant growth and Fe acquisition

At the end of the experiment, although the red clover leaves in the S treatment were still green, the leaves remained small. Leaf Chl concentration in the S treatment was slightly but insignificantly lower than that in the control and the Fe–EDTA spraying treatments (data not shown). Compared with the NS treatment, the shoot and root DWs in the S treatment were decreased by 44 and 67%, respectively (Fig. 2). Foliar spraying of Fe–EDTA stimulated vigorous growth in plants grown in S soil, and their shoot and root DWs increased by 64 and 50%, respectively, as compared with the S treatment.

Figure 2.

Effects of soil sterilization and rhizobia inoculation on the dry weight (DW) of shoot (a) and root (b) of red clover. Treatments are non-sterilized (NS) soil, sterilized (S) soil, S soil with foliar Fe–ethylenediaminetetraacetic acid (EDTA) (100 µm) (S + Fe) and S soil inoculated with rhizobia (S + R). The whole plant cultivation period lasted 45 d. Error bars show SD (n = 5). Treatments labelled with the same letter are not statistically different at P = 0.05.

Compared with the plant grown in NS soil, the Fe concentration in the shoot and root of the S soil-grown plant was decreased by 46 and 35%, respectively. However, the Fe concentration was significantly increased by the Fe–EDTA spraying (Fig. 3). The growth and Fe concentration data both suggest that the insufficient Fe supply caused by minimizing microbial activity was a major factor contributing to the poor growth in the S soil.

Figure 3.

Fe concentration in the shoot (a) and root (b) of soil-grown red clover. Treatments are non-sterilized (NS) soil, sterilized (S) soil, S soil with foliar Fe–ethylenediaminetetraacetic acid (EDTA) (100 µm) (S + Fe) and S soil inoculated with rhizobia (S + R). The whole plant cultivation period lasted 45 d. Error bars show SD (n = 5). Treatments labelled with the same letter are not statistically different at P = 0.05. DW, dry weight.

Influence of – Fe root exudates on growth of soil microorganisms, and some in vitro characters of the phenolic-tolerant microorganisms

The growth of most soil microorganisms on agar was greatly inhibited by the addition of – Fe root exudates, with only a few microbial species growing well. Of the species that grew well in the presence of root exudates (termed phenolic-tolerant species), we selected the seven best growing species that clearly differed morphologically from each other, and expressed as A, B, C, D, E, F and G, respectively. The CAS assay showed that six of these seven species (86%) could produce siderophore under – Fe conditions, and the Salkowski reagent test showed that five of them (71%) could produce auxins (Table 1).

Table 1.  Auxin and siderophore production by phenolic-tolerant microorganisms
MicroorganismSiderophore productionAuxin production
  1. Siderophore was determined using CAS reagent, and auxins were determined using Salkowski reagent.

  2. +, denotes that the microorganism can produce auxins or siderophore.

  3. CAS, chrome azurol S.

A++
B++
C
D++
E++
F++
G+

Effect of nodulating rhizobia inoculation on plant growth and Fe acquisition

At harvest, we found no nodules on NS soil-grown roots, indicating that there were no nodulating rhizobia in the original calcareous soil. As shown in Fig. 2, when the S soil-grown plant was inoculated with nodulating rhizobia, the shoot and root DWs increased by 92 and 110%, respectively, as compared to non-inoculated S soil-grown plants. Plant Fe concentrations also increased by 132 and 150%, respectively (Fig. 3), which are also significantly higher than in the NS soil-grown plants, suggesting that rhizobia particularly favour Fe acquisition.

Effect of synthetic auxin (IAA) and MAs on Fe(III) reduction

Because many of the phenolic-tolerant microbes could produce auxins, we investigated the potential effect of auxin on plant Fe acquisition. As shown in Fig. 4a and b, both Fe(III)–EDTA and ferricyanide reductions of + Fe and – Fe roots were significantly increased by the exogenous IAA treatment. The exogenous MAs treatment showed a similar effect on Fe(III) reductase activity (Fig. 4c & d).

Figure 4.

Effect of indole-3-acetic acid (IAA) (a & b) and microbial auxin (MA) (c & d) addition to the nutrient solution on Fe(III)–ethylenediaminetetraacetic acid (EDTA) and ferricyanide reduction in red clover roots. Treatments are Fe-sufficient (+Fe) solution, + Fe plus 2 × 10−7m IAA (+Fe + IAA), + Fe plus MA at 2 × 10−7m IAA equivalent (+ Fe + MA), Fe-deficient (–Fe) nutrient solution, – Fe nutrient solution plus 2 × 10−7m IAA (–Fe + IAA) and – Fe nutrient solution plus MA at 2 × 10−7m IAA equivalent (–Fe + MA). Error bars show SD (n = 3). Treatments within each graph are different at P = 0.05. FW, fresh weight.

Effect of rhizobia infection on root Fe(III) reductase activity, and IAA and siderophore secretion by the rhizobia

Inoculation with rhizobia increased root Fe–EDTA reductase activity by 70% in the – Fe treatment, but had no effect on the + Fe control (Fig. 5). This suggests that the rhizobia stimulate Fe uptake in inoculated plants under S soil condition (Fig. 3).

Figure 5.

Reduction of Fe(III)–ethylenediaminetetraacetic acid (EDTA) by roots of Fe-sufficient (+Fe) and Fe-deficient (–Fe) plants both with (+R) and without rhizobia nodulation. Uniform 15-day-old seedlings were transplanted to + Fe solution except that the concentration of Ca(NO3)2 was decreased to 0.3 mm. Half of plants were inoculated with rhizobia. When the nodules became visible, the inoculated and un-inoculated plants were further divided into two groups, respectively, with one group transferred to the + Fe and the other group to the – Fe nutrient solution. Three days later, the roots were excised for the measurement of Fe(III)–EDTA reductase activities. Error bars show SD (n = 4). Treatments labelled with the same letter are not different at P = 0.05.

The Salkowski reagent reaction and HPLC profile of the rhizobia culture medium showed that rhizobia could exude IAA (Fig. 6). However, the CAS assay detected no siderophore production by rhizobia under Fe-limited conditions.

Figure 6.

High-performance liquid chromatography (HPLC) chromatograms of standard indole-3-acetic acid (IAA) solution (a), and rhizobia culture solution extract (b). The 4 d rhizobia cultured solution was centrifuged, and the supernatant was partitioned with acidic ethyl acetate. The ethyl acetate fractions were separated from the aqueous phase in a separatory funnel, reduced in volume and dissolved in methanol. The methanol solution was then analysed by HPLC. Column, 5 µm C18 (ODS) (Beckman Coulter, Fullerton, CA, USA); 0.8 mL min−1 flow; mobile phase, methanol:1% ethanol at 4:6; 50 µL injected sample; 254 nm detection.

DISCUSSION

Role of non-nodulating microorganisms in plant Fe acquisition

The activities of soil microorganisms in the rhizosphere have long been assumed to benefit plant Fe acquisition (Masalha et al. 2000; Siebner-Freibach, Hadar & Chen 2003). However, direct evidence supporting this assumption is very limited because S plant/soil systems can be easily contaminated by microorganisms, particularly when watering the plants (Rroco et al. 2003). In the present study, the microbial infection of the autoclaved soil was minimized by a set-up that allowed watering without contamination (Fig. 1). At the end of the experiment, no colonies were detected in the 10−6 dilution of suspended S soil, while colonies in the NS soil at the same dilution rate were too numerous to be counted. Therefore, the current set-up was quite successful in minimizing microorganism contamination.

Although there were no chlorosis symptoms in S soil-grown plants, their growth and plant Fe acquisition were seriously impaired (Figs 2 & 3) in comparison with other treatments. When Fe–EDTA was applied as a foliar fertilizer to S soil-grown plants, the growth and Fe content were greatly enhanced (Figs 2 & 3), indicating that the Fe deficiency caused by eliminating microbial activity was limiting the growth of S soil-grown plants. In other words, the soil microorganisms appear to play an essential role in plant Fe acquisition under Fe-limited conditions, even for Fe-efficient plants. Rroco et al. (2003) likewise reported significantly reduced Fe acquisition and growth of rape grown in S soil, with normal growth restored immediately by the addition of Fe–EDDHA to the S soil. It is worth repeating here that we found no nodules on the roots of NS soil-grown plants, so no nodulating rhizobia were involved in plant Fe acquisition in the NS treatment.

Role of nodulating rhizobia in plant Fe acquisition

In the natural soil environment, rhizobia-infected nodule formation in legumes is ubiquitous. In this study, we found that Fe acquisition by red clover grown in S soil is greatly enhanced by inoculation with rhizobia (Fig. 3), indicating that rhizobia have a beneficial effect on Fe uptake. The stimulation of growth of S soil-grown plants cannot be attributed to N supply by N2 fixation in the nodules, because fertilizer N exceeded plant requirements several-fold as calculated from the total N applied and the total N in the harvested issues (data not shown). In fact, N2 fixation in nodules is inhibited when N supply is sufficient (Van Raalte et al. 1974; Svenning, Junttila & Macduff 1996; Rastetter et al. 2001). Thus, the growth stimulation should be attributed primarily to the beneficial effect of rhizobia on plant Fe uptake.

Potential mechanisms of Fe acquisition enhancement by non-nodulating soil microorganisms and nodulating rhizobia

While the present research amplifies previous studies (Masalha et al. 2000; Rroco et al. 2003) that demonstrated the beneficial effect of microbial activity on plant Fe uptake, the exact mechanism whereby soil microorganisms contribute to plant Fe acquisition still remains unclear.

It is generally accepted that the composition and quantity of root exudates vary in relation to plant nutritional status, and the growth of bacteria and fungi that colonize the rhizosphere is influenced accordingly. For example, the microbial community in the barley rhizosphere varies with the plant's Fe nutritional status (Yang & Crowley 2000). Most Fe-efficient dicots secrete phenolic compounds under – Fe conditions (Römheld & Maschner 1986; Römheld 1987; Susín et al. 1996), red clover included (Zheng et al. 2000, 2003). Generally, plant phenolic compounds have anti-microbial activity with respect to many strains of microorganisms (Rauha et al. 2000), but there are some strains that benefit from phenolic compounds by using them as carbon sources for growth (Blum et al. 2000). Accordingly, when the root exudates of – Fe red clover were added to the culture medium, only a few strains of bacteria grew well, suggesting that the phenolics secreted by the –Fe red clover root may have selectively influenced the rhizosphere microbial community. Interestingly, we found that 86% (six out of seven) of the phenolic-tolerant microorganisms could secrete siderophores under Fe-limiting conditions, and 71% (five out of seven) could produce auxins (Table 1).

Since the 1980s, the production of siderophores has been proposed to be the major microbial activity that benefits plant Fe acquisition (Becker, Messens & Hedgees 1985; Crowley et al. 1988; Masalha et al. 2000; Yehuda et al. 2000). Siderophores have a high affinity for chelating Fe(III), and their Fe(III) chelates have been proven to serve as a source of Fe for plants, for example, Fe–aerobatin for soybean and oat (Chen et al. 1998, 2000), and Fe–rhizoferrin for tomato, barley and corn (Yehuda et al. 2000). The Fe(III)–siderophore chelate, in which the siderophore was secreted by a phenolic-tolerant microorganism, is an effective Fe source for red clover (unpublished results).

Strategy I plants must enzymatically reduce Fe(III) before their root cells can take it up (Chaney, Brown & Tiffin 1972). Bienfait (1985, 1988) proposed two different types of plasma membrane-bound Fe reduction systems: (1) the ‘standard’ reduction–oxidation (redox) system found in all plants, which is characterized by its ability to reduce only high-potential acceptors such as ferricyanide and Fe(III) chelates of the carboxylate type; and (2) the – Fe stress-inducible ‘turbo’ redox system that possesses the ability to reduce a wide variety of Fe(III) chelates as well as ferricyanide. Because auxin is proposed in some studies (Landsberg 1984, 1996; Li et al. 2000; Li & Li 2004) to be the signal substance-inducing plant – Fe responses, we investigated the effect of MAs on both standard and turbo redox systems. Interestingly, both Fe(III)–EDTA and ferricyanide reductase activities of + Fe and – Fe roots were significantly increased by the exogenous IAA treatment (Fig. 4), suggesting that the MAs may enhance plant Fe acquisition by increasing Fe reduction.

Rhizobia nodulation is ubiquitous in leguminous plants. Rhizobia inoculation clearly enhanced both growth and Fe uptake in red clover (Figs 2 & 3). However, CAS assay showed that there was no siderophore secretion by the rhizobia under Fe-limiting conditions, suggesting that some microbial activities other than siderophore production caused the growth enhancement. Interestingly, we found that Fe(III)–EDTA reduction was significantly increased in – Fe plants by infection of their roots with nodulating rhizobia (Fig. 5). Terry et al. (1991), who used split-root experiments, also showed that the Fe(III)–EDTA reduction of Bradyrhizobium-infected side of soybean roots was stronger than the non-inoculated side, and the reduction in the infected side was more active on the root below the nodule clusters. Therefore, the beneficial effect of rhizobia inoculation on Fe uptake by – Fe roots appears to act through an enhanced activity of the turbo redox system. The rhizobia employed in the present research can also synthesize and exude IAA (Fig. 6), indicating that auxin production by rhizosphere rhizobia may also contribute to increased Fe uptake (Figs 2 & 3) by increasing Fe(III) reductase activity (Fig. 4). We further note that phenolic compounds secreted by leguminous plant roots would enhance the growth of nodulating rhizobia, and hence, induce nodule formation (Crawford et al. 2000). For example, phenolics secreted by – Fe alfalfa were shown to significantly stimulate the growth of Rhizobium meliloti (Masaoka et al. 1997).

We synthesize the results of this and of previous studies by proposing the following model of soil microbial enhancement of Fe acquisition by red clover (Fig. 7). When the plant suffers from Fe-deficiency stress, phenolic compounds are exuded by the roots and accumulate in the rhizosphere. These compounds selectively influence the rhizosphere microbial community by their anti-microbial activity and their role as a carbon source. The new predominant microorganisms in the altered rhizosphere community increase the production of both siderophores, thereby increasing rhizosphere Fe bioavailability, and auxins, thereby increasing the Fe(III) reduction capability.

Figure 7.

Proposed model of microbially enhanced plant Fe acquisition. When the plant suffers from Fe-deficiency stress, phenolic compounds are exuded from the roots and accumulate in the rhizosphere. These compounds selectively alter the rhizosphere microbial community. Favoured microorganisms then produce siderophores which increase Fe bioavailability, and auxins which increase Fe(III) reduction capability. Nodulating rhizobia also stimulate Fe(III) reduction through nodule formation. The small circles with the letters mean the microbial communities.

ACKNOWLEDGMENTS

This work was supported by the Natural Science Foundation of China, grant no. 40271065, and by the Program for New Century Excellent Talents in University (NCET). Great thanks are given to Dr Robert Ewing of Iowa State University, USA, for his critical reviewing of the manuscript.

Ancillary