The expression and accumulation of members of the 1-aminocyclopropane-1-carboxylate (ACC) synthase (ACS) and ACC oxidase (ACO) gene families was examined in white clover roots grown in either Pi (phosphate) sufficient or Pi-deprived defined media. The accumulation of one ACO isoform, TR-ACO1, was positively influenced after only 1 h of exposure to low Pi, and this was maintained over a 7-day time-course. Up-regulation of TR-ACS1, TR-ACS2 and TR-ACS3 transcript abundance was also observed within 1 h of exposure to low Pi in different tissue regions of the roots, followed by a second increase in abundance of TR-ACS2 after 5–7 days of exposure. An increase in transcript abundance of TR-ACO1 and TR-ACO3, but not TR-ACO2, was observed after 1 h of exposure to low Pi, with a second increase in TR-ACO1 transcripts occurring after 2–5 days. These initial increases of the TR-ACS and TR-ACO transcript abundance occurred before the induction of Trifolium repens PHOSPHATE TRANSPORTER 1 (TR-PT1), and the addition of sodium phosphite did not up-regulate TR-ACS1 expression over 24 h. In situ hybridization revealed some overlap of TR-ACO mRNA accumulation, with TR-ACO1 and TR-ACO2 in the root tip regions, and TR-ACO1 and TR-ACO3 mRNA predominantly in the lateral root primordia. TR-ACO1p-driven GFP expression showed that activation of the TR-ACO1 promoter was initiated within 24 h of exposure to low Pi (as determined by GFP protein accumulation). These results suggest that the regulation of ethylene biosynthesis in white clover roots is biphasic in response to low Pi supply.

INTRODUCTION

The breeding of crop germplasm with improved Pi (phosphate) efficiency is a major research goal internationally, particularly given the limited natural sources of Pi and the relatively immobile nature of the element in soil [1,2]. As part of screens for such germplasm, the responses of plants which are starved of Pi are often investigated. In broad terms, two major classes of responses are observed: (i) the induction of the PSR (Pi-starvation-response) genes and (ii) the alteration of RSA (root system architecture) [1,3,4].

In the first response class, the PSR gene products are involved in a host of cellular responses including metabolic by-pass reactions within cells to reduce the need for phosphorylated hexoses, as well as in the machinery of mining inorganic Pi from the external media [5,6] and for Pi uptake [3,7]. The second response to Pi-deficiency is alteration to the RSA initiated by a shift in the shoot-root ratio, with sucrose proposed to act as both substrate and signal [810]. There then ensues, in those still comparatively few plant species investigated, changes including (in Arabidopsis, at least) a decrease in primary root growth and an increase in lateral root emergence (from existing primordia) [1117]. These changes in root architecture are believed to be regulated by plant hormones such as ethylene and auxin [1824].

In terms of the hormone regulation in root growth in response to Pi-supply, changes in the biosynthesis and/or the sensitivity to ethylene have been proposed to be a key determinant [1822,2528]. Further, ethylene is proposed to interact with another key hormone, auxin, to orchestrate these changes in root architecture [9,12,13,29,30]. However, information on the control of ethylene biosynthesis at the molecular level in response to Pi-supply is still mostly lacking particularly the enzymes ACS (ACC synthase) and ACO (ACC oxidase) which respectively catalyse the first committed and final step, in the ethylene biosynthetic pathway. Using differential library screening or direct transcriptome analysis, an up-regulation of ACO gene expression has been reported in bean [31,32] and Arabidopsis [4,33,34] in response to low Pi supply. For ACS, global transcriptome analysis of roots of Arabidopsis has revealed the up-regulation of AtACS6 after 7-day exposure to low Pi supply [33]. Lei et al. [35] confirmed the induction of AtACS6, as well as observing the up-regulation of AtACS2 and AtACS4 in roots exposed to low Pi, but details of the induction time-frame of these genes are not provided. Thus any further significant insight into the role of ethylene in regulating these signature growth changes in the root is dependent upon a more complete dissection of the ACS and ACO multi-gene families

In the present study, the legume white clover (Trifolium repens L.) has been used to investigate the influence of Pi-deficiency on the ethylene biosynthetic pathway, specifically the differential expression of the ACS and ACO gene families in root tissues. Members of both multi-gene families have been characterized previously in vegetative tissues of white clover and shown to display differential expression in response to developmental and environmental cues [3640]. Some research has been undertaken on Pi uptake by white clover and on responses to Pi supply [4146], and we have shown that low Pi-supply can promote both elongation of the primary root and the number of lateral roots; an effect that is stimulated further by ethylene [23]. Using the same hydroponic system, we have now dissected changes in transcript abundance and protein accumulation associated with the ethylene biosynthetic pathway. Our results show firstly that differential and tissue-specific expression of both the ACO and ACS gene families is influenced by Pi-supply, and occur both at the transcriptional and translational level. Further, the timing of such changes may support a dual role of the hormone in terms of regulating the responses observed in the roots of white clover to low Pi supply.

EXPERIMENTAL

Plant growth methods

A single genotype of white clover (T. repens L.) cv. Grasslands challenge, genotype 10F or a cultivar (comprising mixed genotypes), cv. Huia, both obtained from AgResearch Grasslands, were used for experiments. For vegetative propagation, stolons were excised at the base of the fourth node, proximal to the stolon apex, all fully-expanded leaves removed with the exception of 1–2 of the youngest expanded leaves, and then the third and fourth node buried into vermiculite (2–4 mm, Nuplex Industries). The cuttings were supplemented with a modified Hoagland's solution [47] comprising one-third strength macronutrients and full-strength micronutrients. For investigations requiring whole roots, the cuttings after root formation (typically 3–5 days), were maintained in a hydroponic media comprising half-strength modified Hoagland's media, contained in 50 ml-capacity tubes, to acclimatize for (typically) 2 days. After which, the Pi content of the media was adjusted to give a Pi-sufficient (1 mM Pi; P+) or a Pi-deficient (10 μM Pi; P−) media and the rooted cuttings exposed to these treatments, as appropriate. The media was replenished at 2-day intervals. For the transcriptional studies (qRT-PCR), in which the primary root was dissected further, the rooted stolons were transferred to half-strength Hoagland's media in 0.6-litre-capacity PVC pipes. After 14 days, the stolons were exposed to either Pi deficient or Pi sufficient treatment, as above, over the appropriate time course.

Determination of total leaf Pi content

Total Pi content was determined in harvested leaf tissue as described by Zhang and McManus [44] using acidified ammonium molybdate complexing and absorbance measurement at 595 nm.

Protein extraction methods

The soluble and cell-wall-associated proteins from white clover roots which were used for acid phosphatase (APase) activity assay were extracted following the protocol of Hay et al. [48]. To do this, root tissues were rinsed with water, blot dried and powdered in liquid nitrogen. To extract the soluble proteins, the powder was suspended in three volumes of 1 mM DTT and the slurry incubated on ice for 30 min. After centrifugation at 12,000 × g for 10 min at 4°C, the supernatant (designated the soluble fraction) was removed, the pellet extracted with fresh 1 mM DTT for a further three times, followed by a further seven washes with water, with centrifugation (12,000 × g for 10 min at 4°C) to collect the pellet after each wash. The pellet was then resuspended in one volume of 1 M NaCl, incubated at 37°C for 1 h, then the slurry was centrifuged at 12,000 × g. The supernatant (designated the cell-wall-enriched fraction) was removed and the pellet again resuspended in 1 M NaCl and incubated at 4°C for 16–18 h. After centrifugation, as before, the supernatant was pooled with the first 1 M NaCl supernatant and together designated the cell-wall-enriched fraction. For Western blot analyses, the tissue was homogenized in 100 mM Tris/HCl, pH 7.5, containing 10% glycerol and 1 mM DTT and the slurry centrifuged at 20,000 × g for 10 min. The supernatant was collected and protein was determined using a commercially available protein dye-based kit (Bio-Rad Laboratories) with BSA for a standard curve.

Acid phosphatase assay

APase activity was measured essentially as described in Zhang and McManus [44] using 8 mM ρ-nitrophenol phosphate (ρNPP) in 50 mM citrate buffer, pH 5.6 as substrate.

White clover transformation

Seeds of T. repens (L.) cv. Huia were subjected to Agrobacterium-mediated T-DNA transformation essentially as described by Voisey et al. [49]. Briefly, seeds were sterilized with 3.0% (v/v) sodium hypochlorite and then dark-imbibed in sterile water at room temperature overnight. After this, the imbibed seeds were further treated with 6% (v/v) hydrogen peroxide followed by five washes with sterile distilled water before dissection of the cotyledons. Three microlitres of a suspension of Agrobacterium tumefaciens, transformed with the TR-ACO1p::mGFP-ER construct was then pipetted on to each cotyledon maintained on filter paper moistened with CR7 media [MS Media, pH 5.7, containing 1 mg·l−1 6-benzylaminopurine (BAP, Sigma), 0.05 mg·l−1 1-naphthalene acetic acid (NAA, Sigma) in 0.8% (w/v) Phytoagar (Duchefa Biochemie)], and the treated tissues incubated at 23°C under light intensity of 120 μmol·m−2·s−1 for 16 h daily. Typically, after 3 days, the cotyledons were transferred on to CR7 selection media (as described previously, but augmented with 150 μg·ml−1 kanamycin, 300 μg·ml−1 cefotaxime) and then transferred to fresh CR7 media every 14 days. Regenerating cotyledons were then transferred to CR5 selection media [MS media, pH 5.7, containing 0.1 mg·l−1 BAP, 0.05 mg·l−1 NAA, 100 μg·ml−1 kanamycin and 300 μg·ml−1 cefotaxime in 0.8% (w/v) Phytoagar] with transfer to fresh CR5 selection media every 14 days. Any developing plantlets were transferred, as appropriate, on to hormone-free CR0 rooting media [MS media, pH 5.7, containing 50 μg·ml−1 kanamycin in 0.8% (w/v) Phytoagar] and after development of a well-established root system, these were exflasked and the plants maintained in a glasshouse with 75% relative humidity with day and night temperatures of 22 and 16°C respectively. Any putative transgenic plants were confirmed by genomic PCR and gene copy determined using Southern analysis (results not shown).

Confocal microscopy

Root samples for confocal microscopy were prepared immediately before examination, where fully excised roots were washed with distilled water then stained with 10 ng·μl−1 propidium iodide for approximately 10 s. Destaining was carried out with three washes of distilled water before mounting the root in 50% (v/v) glycerol on microscope slides. A Leica Confocal Microscope equipped with Leica Application Suite Advanced Fluorescence (LAS-AF) software was used to collect images of the root sections with dual scanning routinely conducted with the filter set for the excitation and emission of the green fluorescent protein at 477 and 510 nm respectively.

SDS/PAGE and western analysis

SDS/PAGE and western analysis were carried out essentially as described in Hunter et al. [36]. Prior to actual study, preliminary experiments were carried out to determine optimal protein required and ensure that immunoreactive bands were roughly proportional to the amount of extracted root proteins. Equal amount (10 μg) of protein was loaded per lane. The separated proteins were challenged either with a TR-ACO2 antibody [40] or with a TR-ACO1 antibody raised against a recombinant protein using the same method from the full-length TR-ACO1 cDNA isolated previously by Chen and McManus [38] or a commercially available anti-GFP antibody (Roche Applied Science). Antibody recognition was detected using the Super-Signal West Pico Chemiluminescent system (Pierce, Thermo Scientific). The relative quantification of the protein bands was done using the ImageJ software.

RNA isolation and real-time PCR (qRT-PCR)

Total RNA was isolated from root tissue harvested, and frozen in liquid nitrogen, at the appropriate time intervals using a hot borate method adapted by Hunter and Reid [50]. Genomic DNA-free RNA samples were prepared using DNase I recombinant RNase-free (Roche) treatment and 1 μg of total RNA was used to synthesize cDNA using Transcriptor First Strand cDNA synthesis kit (Roche) with an oligo (dT)15 primer and the product was diluted to 20-fold prior to use. β-Actin (TR-β-actin) and GAPDH (TR-GAPDH) were used as reference genes. qRT-PCR was performed using a LightCycler® 480 Real-Time PCR (Roche) and system series software 1.7. For each treatment two biological replicates with three technical replicates each were analysed. Typically, a 10 μl reaction consisted of 5 μl of 2 × LightCycler® 480 SYBR Green I Master Mix (Roche), 2.5 μl of diluted cDNA and 0.5 μl of 10 μM forward and reverse primers. The PCR was performed as follows: 95°C for 5 min, then (95°C 10 s; 60°C 10 s; 72°C 10 s) for 40 cycles, finally a 95°C melt. Fluorescence measurements were performed at 72°C for each cycle and continuously during final melting. Relative transcript abundance for the target genes was determined by comparative quantification to the geometric mean of the reference genes TR-GAPDH and TR-ACTIN, using the method of Pfaffl [51]. The primer pair for each gene is given as Supplementary Table S1. The list of accession numbers of genes examined in this study is provided as Supplementary Table S2.

In situ hybridization

Fresh root tissues were fixed with FAA and dehydrated in a standard butanol series before embedding in paraplast. Ten μm sections were then affixed on to poly-L-lysine coated slides and baked overnight at 42°C. DIG-labelled sense and antisense probes were synthesized from 3′UTR cDNA fragments of TR-ACO1 (364 bp, [38]), TR-ACO2 (270 bp, [38]) and TR-ACO3 (360 bp [38]) using SP6 and T7 RNA polymerases according to the manufacturers protocol (Roche). The probes were hydrolysed with sodium carbonate buffer, pH 10.2, to further reduce the probe size to approximately 100–150 bp. Ten micrometres root sections, affixed on to microscopic slides, were dewaxed, rehydrated in an ethanol series and then passed through pre-hybridization treatments according to the protocol of Alvarez et al. [52]. Hybridization was carried out overnight and washing and detection was undertaken according to the manufacturer's instructions (Roche).

Statistical analysis

Sample means with S.E.M. were calculated using SAS/STAT™ (SAS Institute) software. Means were compared using a Student's pairwise t test at a significance level of 1% (P≤0.01) or 5% (P≤0.05), as indicated.

RESULTS

Pi supply influences Pi levels and acid phosphatase activity

Experiments were conducted with rooted cuttings of white clover in 50 ml capacity containers in hydroponic media such that the concentration of Pi supplied could be manipulated. Two concentrations were supplied routinely, as 1 mM Pi (Pi-sufficient; Pi+) or 10 μM (Pi-deficient; Pi−). Under these conditions, the first significant decrease in the total leaf Pi content (measured in the first fully expanded leaf on the stolon) when compared with day 1 was observed after 8 days of the Pi− treatment (Supplementary Figure S1A). At day 8, this decrease also led to a significant difference between the Pi− and Pi+ treatments, with the lower Pi content in the leaves from plants grown in the Pi-deficient media.

In terms of the induction of acid phosphatase in the roots, enzyme activity was assayed in a total soluble fraction, as well as a cell-wall-enriched fraction. In general, higher rates [expressed on a per gram fresh weight (FW) basis] were observed in the soluble fraction when compared with the cell wall fraction, and a significantly higher activity was observed in the Pi− treatment from day 5 onwards with a rapid rise at day 8 (which coincides with the first significant decrease in Pi levels in the leaf tissue) (Supplementary Figure S1B). However, this difference was due more to a decrease in activity in the Pi+ treatment from that measured at day 1, as only at day 8 did the activity in the Pi− treatment increased significantly from the rate at 1 day. In contrast, the phosphatase activity in the cell-wall-enriched fraction increased in the Pi− treatment over the time course, with no change in the Pi+ treatment, such that a significantly higher activity in the Pi− treatment was observed from day 5 onwards when compared with both the 1 day and the Pi+ activity at the equivalent time points (Supplementary Figure S1C).

Accumulation of TR-ACO isoforms is dependent on Pi supply

At the translational level, a higher accumulation of the TR-ACO1 isoform is routinely observed in whole roots by 1 day of treatment which is maintained over the 7-day time-course examined (Figures 1A and 1B). In contrast, no change in accumulation was observed for the TR-ACO2 isoform at any of the time points assayed (Figures 1C and 1D).

Changes in TR-ACO protein isoform abundance in white clover roots over 7 days in response to Pi supply

Figure 1
Changes in TR-ACO protein isoform abundance in white clover roots over 7 days in response to Pi supply

Western analyses to detect TR-ACO1 (A and B) or TR-ACO2 (C and D) abundance as approximately 36 kDa protein in the roots of genotype 10F, cv. Grasslands Challenge collected at the hours or days, as indicated, after exposure to the Pi+ or Pi− treatment, as indicated. Aliquots (10 μg) of protein from each sampling point after Pi treatment were separated using SDS/PAGE, electroblotted on to PVDF membrane and challenged with an α-TR-ACO1 antibody (A) or an α-TR-ACO2 antibody (B). Coomassie staining of protein was used as a loading control with a major protein of approximately 55 kDa indicated in each panel, as appropriate. Relative quantification of the blots was done using Image J analysis programme.

Figure 1
Changes in TR-ACO protein isoform abundance in white clover roots over 7 days in response to Pi supply

Western analyses to detect TR-ACO1 (A and B) or TR-ACO2 (C and D) abundance as approximately 36 kDa protein in the roots of genotype 10F, cv. Grasslands Challenge collected at the hours or days, as indicated, after exposure to the Pi+ or Pi− treatment, as indicated. Aliquots (10 μg) of protein from each sampling point after Pi treatment were separated using SDS/PAGE, electroblotted on to PVDF membrane and challenged with an α-TR-ACO1 antibody (A) or an α-TR-ACO2 antibody (B). Coomassie staining of protein was used as a loading control with a major protein of approximately 55 kDa indicated in each panel, as appropriate. Relative quantification of the blots was done using Image J analysis programme.

With greater accumulation of the TR-ACO1 isoform in response to the low Pi treatment after 24 h, ethylene production was then measured over the 24-h time-course. Whole roots were used, but while detectable levels of evolved ethylene were measured, no significant differences between treatments were observed over the 24-h time-course (Supplementary Figure S2).

Pi supply influences the transcript abundance of the ACS and ACO multi-gene families

To examine the influence of Pi supply on the ethylene biosynthesis genes, the primary root was divided into three developmental regions: the elongation zone (EZ) comprising the distal zone from the root cap until the point at which the first lateral roots have clearly emerged, the visible lateral root zone (VL) where the emerged lateral roots occur and the mature root (MR) which comprises the root zone proximal to the subtending lateral roots. For the ACS gene family, a rapid and significant induction (after 1 h of treatment in Pi-deficient media) of TR-ACS1 transcript abundance was observed in the EZ (Figure 2A) and this higher transcript abundance was maintained over 24 h in the low Pi treatment. This significant difference between the two treatments was not observed in TR-ACS2 abundance, whereas no TR-ACS3 transcripts was detected in this tissue within the limits of sensitivity achieved using qRT-PCR. For the VL region, a rapid increase in TR-ACS3 transcript abundance instead was observed over the 1–12 h period (Figure 2B), whereas in the MR region, an increase in transcript abundance for both TR-ACS2 (1–12 h) and TR-ACS3 (1–6 h) was observed (Figure 2C). Changes in the transcript abundance of the TR-ACS gene family was also observed over the extended 7-day time-course (Supplementary Figure S3). In the EZ, TR-ACS1 was up-regulated but not the TR-ACS2 (Supplementary Figure S3A). In the VL, significant increases in TR-ACS2 transcript abundance tend to occur later (3–7 days) in the time course (Supplementary Figure S3B), and in the MR, up-regulation of TR-ACS2 and TR-ACS3 was noted at day 5 and day 3 respectively (Supplementary Figure S3C).

Changes in transcript abundance of the TR-ACS gene family in white clover roots over 24 h in response to Pi supply

Figure 2
Changes in transcript abundance of the TR-ACS gene family in white clover roots over 24 h in response to Pi supply

Relative transcript abundance of TR-ACS1, TR-ACS2 and TR-ACS3, as indicated in the EZ (A), VL (B) and MR (C) regions of the primary roots of nodal explants maintained hydroponically in half-strength Hoagland's media containing either 1 mM Pi (clear bars) or 10 μM Pi (shaded bars) for the time intervals indicated. Relative transcript abundance was determined by qRT-PCR and transcription was normalized against two internal reference genes, TR-β-actin and TR-GAPDH. Values are means ± S.E.M., for two biological replicates each of which was derived from means of three independent qRT-PCR reactions. * indicates significant differences between the 1 mM Pi and the 10 μM Pi treatments (P≤0.05).

Figure 2
Changes in transcript abundance of the TR-ACS gene family in white clover roots over 24 h in response to Pi supply

Relative transcript abundance of TR-ACS1, TR-ACS2 and TR-ACS3, as indicated in the EZ (A), VL (B) and MR (C) regions of the primary roots of nodal explants maintained hydroponically in half-strength Hoagland's media containing either 1 mM Pi (clear bars) or 10 μM Pi (shaded bars) for the time intervals indicated. Relative transcript abundance was determined by qRT-PCR and transcription was normalized against two internal reference genes, TR-β-actin and TR-GAPDH. Values are means ± S.E.M., for two biological replicates each of which was derived from means of three independent qRT-PCR reactions. * indicates significant differences between the 1 mM Pi and the 10 μM Pi treatments (P≤0.05).

In contrast with the TR-ACS gene family, no significant difference in the transcript abundance of the three genes examined comprising the TR-ACO gene family was observed in the EZ between treatments over the 24-h time-course (Figure 3A). In the VL region, a rapid and significant increase in transcript abundance of TR-ACO1 was observed after 1 and 6 h in the low Pi treatment, with a later induction (at 24 h) also occurring; a trend also observed for TR-ACO3 (at 12 and 24 h) whereas a decrease at 6 h was observed for TR-ACO2 (Figure 3B). Rapid and significant increases in transcript abundance were also observed at 1 and 6 h after low Pi treatment for both TR-ACO1 and TR-ACO3 in the MR (Figure 3C). No significant differences between treatments was observed for TR-ACO2 in the EZ and MR region (Figure 3). Over the longer time-course, an increase in transcript abundance of TR-ACO1 was positively influenced by the low Pi treatment over 2–5 days, particularly the EZ (Supplementary Figure S4). Of note too is that in the MR zone a decrease in transcript abundance was observed over 2–5 days (Supplementary Figure S4C).

Changes in transcript abundance of the TR-ACO gene family in white clover roots over 24 h in response to Pi supply

Figure 3
Changes in transcript abundance of the TR-ACO gene family in white clover roots over 24 h in response to Pi supply

Relative transcript abundance of TR-ACO1, TR-ACO2 and TR-ACO3, as indicated in the EZ (A), VL (B) and MR (C) regions of the primary roots of nodal explants maintained hydroponically in half-strength Hoagland's media containing either 1 mM Pi (clear bars) or 10 μM Pi (shaded bars) for the time intervals indicated. Relative transcript abundance was determined by qRT-PCR and transcription was normalized against two internal reference genes, TR-β-actin and TR-GAPDH. Values are means ± S.E.M., for two biological replicates each of which was derived from means of three independent qRT-PCR reactions. * indicates significant differences between the 1 mM Pi and the 10 μM Pi treatments (P≤0.05).

Figure 3
Changes in transcript abundance of the TR-ACO gene family in white clover roots over 24 h in response to Pi supply

Relative transcript abundance of TR-ACO1, TR-ACO2 and TR-ACO3, as indicated in the EZ (A), VL (B) and MR (C) regions of the primary roots of nodal explants maintained hydroponically in half-strength Hoagland's media containing either 1 mM Pi (clear bars) or 10 μM Pi (shaded bars) for the time intervals indicated. Relative transcript abundance was determined by qRT-PCR and transcription was normalized against two internal reference genes, TR-β-actin and TR-GAPDH. Values are means ± S.E.M., for two biological replicates each of which was derived from means of three independent qRT-PCR reactions. * indicates significant differences between the 1 mM Pi and the 10 μM Pi treatments (P≤0.05).

White clover roots display signature changes in transcript abundance of TR-PT1 in response to Pi-deprivation

To relate the timing of changes in TR-ACS and TR-ACO transcript abundance in response to low Pi supply to signature Pi-starvation changes well documented from other studies, the transcript abundance of T. repens PHOSPHATE TRANSPORTER 1 (TR-PT1), which is a high affinity Pi transporter, was examined over both the 24 h (Figure 4A) and 7-day time-course (Figure 4B). Here, in both the younger tissues of the root (EZ and VL) regions of the primary root, an increase in the transcript abundance of TR-PT1 was observed in response to low Pi from day 3 to day 7, whereas a more immediate increase (after 1 h) was observed in the MR only, which was then maintained over the whole time course.

Changes in transcript abundance of TR-PT1 in white clover roots over 7 days in response to Pi supply

Figure 4
Changes in transcript abundance of TR-PT1 in white clover roots over 7 days in response to Pi supply

Relative transcript abundance of TR-PT1 in the EZ (upper panel), VL (middle panel) and MR (lower panel) regions of the primary roots of nodal explants maintained hydroponically in half-strength Hoagland's media containing either 1 mM Pi (clear bars) or 10 μM Pi (shaded bars) for the time intervals indicated over a 24 h (A) or 7 days (B) time course. Relative transcript abundance was determined by qRT-PCR and transcription was normalized against two internal reference genes, TR-β-actin and TR-GAPDH. Values are means ± S.E.M., for two biological replicates each of which was derived from means of three independent qRT-PCR reactions. * indicates significant difference between the 1 mM Pi and the 10 μM Pi treatments (P≤0.05).

Figure 4
Changes in transcript abundance of TR-PT1 in white clover roots over 7 days in response to Pi supply

Relative transcript abundance of TR-PT1 in the EZ (upper panel), VL (middle panel) and MR (lower panel) regions of the primary roots of nodal explants maintained hydroponically in half-strength Hoagland's media containing either 1 mM Pi (clear bars) or 10 μM Pi (shaded bars) for the time intervals indicated over a 24 h (A) or 7 days (B) time course. Relative transcript abundance was determined by qRT-PCR and transcription was normalized against two internal reference genes, TR-β-actin and TR-GAPDH. Values are means ± S.E.M., for two biological replicates each of which was derived from means of three independent qRT-PCR reactions. * indicates significant difference between the 1 mM Pi and the 10 μM Pi treatments (P≤0.05).

Phosphite (Phi) addition does not influence the transcript abundance of TR-ACS1 in the EZ

To determine part of the mechanism by which low Pi may influence the transcript abundance of the ethylene biosynthesis genes, specifically within the first hours post-exposure to low Pi, Pi-sufficient roots were either treated with Phi or exposed to low Pi media and the abundance of TR-ACS1 in the EZ was determined (Figure 5). Here, a rapid (after 1 h) and significant increase in the transcript abundance of TR-ACS1 was again observed in response to the low Pi treatment, which was maintained over the 24-h time-course. In contrast, no significant increase in transcript abundance of TR-ACS1 at any time-point was observed in the Phi-treated roots.

Changes in transcript abundance of TR-ACS1 in white clover roots over 24 h in response to Pi or phosphite supply

Figure 5
Changes in transcript abundance of TR-ACS1 in white clover roots over 24 h in response to Pi or phosphite supply

Relative transcript abundance of TR-ACS1 in the EZ region of the primary roots of nodal explants maintained hydroponically in half-strength Hoagland's media containing either 1 mM Pi (clear bars), 10 μM Pi (dark bars) or 0.5 mM phosphite (shaded bars) for the time intervals indicated. Relative transcript abundance was determined by qRT-PCR and transcription was normalized against two internal reference genes, TR-β-actin and TR-GAPDH. Values are means ± S.E.M., for two biological replicates each of which was derived from means of three independent qRT-PCR reactions. * indicates significant differences between the 1 mM Pi and the 10 μM Pi treatments (P≤0.05).

Figure 5
Changes in transcript abundance of TR-ACS1 in white clover roots over 24 h in response to Pi or phosphite supply

Relative transcript abundance of TR-ACS1 in the EZ region of the primary roots of nodal explants maintained hydroponically in half-strength Hoagland's media containing either 1 mM Pi (clear bars), 10 μM Pi (dark bars) or 0.5 mM phosphite (shaded bars) for the time intervals indicated. Relative transcript abundance was determined by qRT-PCR and transcription was normalized against two internal reference genes, TR-β-actin and TR-GAPDH. Values are means ± S.E.M., for two biological replicates each of which was derived from means of three independent qRT-PCR reactions. * indicates significant differences between the 1 mM Pi and the 10 μM Pi treatments (P≤0.05).

Localization of TR-ACO gene expression in roots of white clover

With the differential induction of both the TR-ACS and TR-ACO gene families in different developmental regions of the roots, it was of interest to determine if the changes in the transcript abundance of these genes were confined to any specific tissues. Using in situ hybridization, the accumulation of transcripts from the TR-ACS genes proved to be too low to obtain reliable and consistent localization. Instead, localization of transcripts of the TR-ACO gene family was undertaken. In Pi-sufficient roots, localization of TR-ACO1 was broadly throughout the root, but accumulation was observed particularly in the lateral root primordia (Supplementary Figure S5A), which diminished with the outgrowth of the lateral root (Supplementary Figure S5C), and in the primary (Supplementary Figure S5E) and secondary root tips (results not shown). Accumulation of TR-ACO2 transcripts was again observed in the root tips (Supplementary Figure S6A), as well as high levels of mRNA accumulation in the vascular tissue (Supplementary Figure S6B), whereas the TR-ACO3 transcripts were again predominantly expressed in the lateral root primordia (Supplementary Figure S6E).

TR-ACO1 expression is limited to regions of cell division in roots and promoter activity is influenced by Pi supply

Although in situ hybridization showed a more widespread accumulation of transcripts from the TR-ACO gene family, albeit through a lower level of detection, the accumulation of both TR-ACO1 and TR-ACO3 mRNA was quite marked in the lateral root primordia. However, using in situ hybridization, it was not really possible to discern which specific cell type or complex tissue were the sites of any preferential increase in TR-ACO1 or TR-ACO3 transcript accumulation in response to low Pi. Thus to determine if the increase in transcript abundance that is observed in response to low Pi is driven by the localized activation of transcription, a TR-ACO1p::mGFP-ER construct was used in the white clover background. The TR-ACO1 promoter was selected for these experiments as the transcript abundance studies showed that this gene did display the most pronounced response to low Pi supply, particularly over the 7-day time-course. As whole roots were examined, the treatments were again undertaken in the 50 ml capacity containers. Data from a single transgenic line are shown, although three independent lines were examined. In addition, the transformants were assessed in terms of responses to the low Pi treatment to ensure that no secondary effects caused by the T-DNA insertion were evident. In common with the data shown in Supplementary Figure S1, a significant decrease in Pi level (measured in the first fully emerged leaf) was first observed in the low Pi treatment after 7 days, whereas a significant induction in acid phosphatase activity was again observed in both the total soluble activity and cell-wall-enriched after 7 days (Supplementary Figure S7). Again, acid phosphatase was higher in the soluble fraction when compared with the cell wall fraction. As we were assessing any potential T-DNA effects, the time course was continued up to 21 days, but with no change to the low Pi response.

In terms of localization of expression in the primary root, the GFP signal was observed in the cells proximal to the root cap and quiescent centre (qc), the procambium (pc) and the ground meristem (gm) (Figure 6A), as well as in the pericycle extending through the elongating zone of the root (Figure 6B) and in developing lateral root primordia in the mature zone of the root (Figure 6C). In a fully emerged lateral root, expression was also observed in the apical portion (Figure 6D) in a similar pattern to that observed in the primary root. To ensure that the GFP localization reflects transcription in that cell/tissue, the ER-targeted GFP construct (TR-ACO1p::GFP-ER) was used. Further, detection of the GFP signal in the root tips and in the lateral root primordia is broadly consistent with the TR-ACO1 transcript accumulation localized using in situ hybridization (Supplementary Figure S5).

Localization of TR-ACO1p-driven GFP accumulation in roots of Pi-sufficient white clover

Figure 6
Localization of TR-ACO1p-driven GFP accumulation in roots of Pi-sufficient white clover

GFP expression, determined using confocal microscopy, in roots of transgenic (transformed with TR-ACO1p::mGFP-ER) cv. Huia line TR2-1. (A) Apical region of the primary root; (B) pericycle tissue in the elongating zone of the primary root; (C) lateral root primordium (lrp) in the mature zone of the primary root; (D) apical region of a fully emerged lateral root. For GFP excitation and emission (AD), the filters were set at approximately 480 and 510 nm respectively. The bars represent 50 μm, with each image at the same magnification.

Figure 6
Localization of TR-ACO1p-driven GFP accumulation in roots of Pi-sufficient white clover

GFP expression, determined using confocal microscopy, in roots of transgenic (transformed with TR-ACO1p::mGFP-ER) cv. Huia line TR2-1. (A) Apical region of the primary root; (B) pericycle tissue in the elongating zone of the primary root; (C) lateral root primordium (lrp) in the mature zone of the primary root; (D) apical region of a fully emerged lateral root. For GFP excitation and emission (AD), the filters were set at approximately 480 and 510 nm respectively. The bars represent 50 μm, with each image at the same magnification.

To investigate changes in the intensity of GFP emission in response to Pi supply, roots of the TR-ACO1p::GFP-ER transformants, grown in either Pi-sufficient or low Pi media, were compared using confocal microscopy (Figure 7). Here, the intensity of GFP emission was highest after 24 h in roots of plants grown in the low Pi treatment, when compared with the Pi+ treatment in both the primary (Figures 7A and 7B) and lateral roots (Figures 7C and 7D). Again, GFP accumulation occurred further into the qc and root cap in the lateral root. These GFP emission observations are essentially qualitative in nature, and so to provide further evidence of any differences between treatments, the GFP accumulation in the roots of the transformants was compared using western analysis (Figure 7E). Again, a higher accumulation was observed by day 1 in the low Pi treatment.

Changes in TR-ACO1p-driven GFP accumulation in white clover roots in response to Pi supply

Figure 7
Changes in TR-ACO1p-driven GFP accumulation in white clover roots in response to Pi supply

GFP expression, determined using confocal microscopy, in the primary (A and B) or lateral (C and D) roots of the transgenic (transformed with TR-ACO1p::mGFP-ER) cv. Huia line TR2-1 maintained in Pi-sufficient (A and C) or Pi-deficient media (B and D). Roots were examined after 1 days of treatment. For GFP excitation and emission, the filters were set at approximately 480 and 510 nm respectively. The bars represent 50 μm, with each image at the same magnification. (E) Western analyses to detect GFP abundance (as a 27 kDa protein) in a total root extracts from the transgenic (transformed with TR-ACO1p::mGFP-ER) cv. Huia line TR2-1 subjected to Pi-sufficient or Pi-deficient media, as indicated, and examined after 1 days of treatment. Coomassie staining of protein was used as a loading control with a major protein of approximately 55 kDa indicated in each panel, as appropriate.

Figure 7
Changes in TR-ACO1p-driven GFP accumulation in white clover roots in response to Pi supply

GFP expression, determined using confocal microscopy, in the primary (A and B) or lateral (C and D) roots of the transgenic (transformed with TR-ACO1p::mGFP-ER) cv. Huia line TR2-1 maintained in Pi-sufficient (A and C) or Pi-deficient media (B and D). Roots were examined after 1 days of treatment. For GFP excitation and emission, the filters were set at approximately 480 and 510 nm respectively. The bars represent 50 μm, with each image at the same magnification. (E) Western analyses to detect GFP abundance (as a 27 kDa protein) in a total root extracts from the transgenic (transformed with TR-ACO1p::mGFP-ER) cv. Huia line TR2-1 subjected to Pi-sufficient or Pi-deficient media, as indicated, and examined after 1 days of treatment. Coomassie staining of protein was used as a loading control with a major protein of approximately 55 kDa indicated in each panel, as appropriate.

DISCUSSION

The present study has investigated the influence of Pi-deficiency on the differential transcript and protein isoform accumulation of members of the ACS and ACO gene families in the legume white clover (T. repens L.). It is generally accepted that the ACS mediates the rate-limiting and first committed step in the pathway, whereas the ACO catalyses the final step of ethylene synthesis [5354]. In terms of signalling events associated with responses of plants to Pi-deficiency, the role of ethylene has received particular attention, both in terms of a putative role in the observed changes to root growth [1822,25,26] and in terms of setting in train a series of characteristic responses associated with the nutrient deficiency [15,35]. We have shown previously that low Pi-supply can promote root growth in white clover, an effect that is stimulated further by ethylene [23] and an increase in transcript abundance of TR-ACO1, but not TR-ACO2, in whole roots within 1 h of exposure to low Pi [55]. Thus it was pertinent to examine the transcriptional and translational control of ethylene biosynthesis in response to Pi supply.

We first examined some of the signature changes that mark the onset of Pi-deficiency in most plants studied, to provide a time-frame of Pi responses and effect. For Pi content, the first significant decrease (in the first fully expanded leaf) occurred after 8 days in Pi-deficient media (Supplementary Figure S1) thus providing a measure for the extent of reserves of Pi in the cuttings, since it is known that white clover can store significant concentrations of Pi in the stolon tissue [42]. A second signature response to Pi deficiency is the induction of APase activity [5], which was observed at day 5 in the total soluble fraction and in the cell-wall-enriched fraction (Supplementary Figure S1). Changes in the soluble fraction are proposed to represent changes in cellular metabolism that accompany a decrease in internal Pi [56], whereas induction of the cell-wall-associated activity may represent secreted isoforms that are associated with mining Pi from the surrounding soil and/or recycling esterified forms that may leak from the compromised plant membrane [5].

We began our investigation on changes in the ethylene biosynthesis by considering the two major enzymes involved in the pathway, the TR-ACS and the TR-ACO. We aim to determine whether the previously reported transcriptional changes were manifest at the translational level [23]. Although antibodies are not available to the TR-ACS protein family, isoform-specific antibodies are available for TR-ACO1 and TR-ACO2. An increase in TR-ACO1 abundance was detected within 1 h of exposure to low Pi, which was maintained over a 7-day time course and with no difference in TR-ACO2 abundance in the whole root (Figure 1). However, no increase in ethylene production could be measured from whole roots, despite using laser-based photoacoustic detection (Supplementary Figure S2). Due to unavailability of antibodies specific for TR-ACS protein we were unable to determine if the increase in the TR-ACS transcript level resulted to an increase in TR-ACS protein. With no experimental evidence on the TR-ACS protein accumulation, we are then uncertain if this has contributed to unchanged ethylene level in the roots. Other plausible reason for inability to detect changes in ethylene production could be that either the induced isoforms were inactive or, perhaps more likely, that the biosynthesis is very tissue localized making it difficult to measure any changes in evolved ethylene from the whole roots used in the assay. In other studies where more rapid changes in ethylene evolution have been observed, Drew et al. [18] observed a decrease, first detectable after 1 day, in maize roots when grown in low Pi media, whereas Li et al. [28] observed a significant increase in the roots of Medicago falcata after 48 h. Although inconclusive in terms of ethylene evolution, our observations do indicate that the ethylene biosynthesis machinery is responsive to changes in the external Pi supply within only a few hours.

With the increase in TR-ACO1 accumulation, there was thus justification to dissect responses of both the ACS and ACO genes in response to changes in external Pi supply within the 1-day time-course. So we examined changes in transcript abundance of the TR-ACS and TR-ACO genes previously identified as being differentially expressed during vegetative development in white clover [3639]. Here, an increase in transcript abundance in the three broad root developmental zones was observed for both gene family in response to low Pi supply, including a very rapid increase (after 1 h; Figures 2 and 3) and then a second period of transcript abundance increase of TR-ACS2 (only after 3–7 days; Supplementary Figure S3) and TR-ACO1 (after 2–5 days; Supplementary Figure S4A) in the low Pi treatment. Thus the data support a biphasic control of transcript abundance of the ethylene biosynthesis gene family in response to exposure to the low Pi media. As well, this is mediated via differential expression of the gene family, thus showing exquisite control.

The rapid increase in TR-ACS and TR-ACO transcript abundance in response to low Pi could place these genes in the ‘local response’ category, in common with at least one ACO gene identified in the roots of Pi-starved Arabidopsis [4]. This suggests that the changes in the ethylene biosynthesis genes that are observed to occur rapidly may be in response to the changes in the external Pi concentration bathing the roots, rather than any change in internal Pi level (which takes several days using the hydroponic system). To resolve this further, we first monitored the timing of the increased transcript abundance of a signature PSR-associated gene coding for a member of the Pi transporter family, TR-PT1 (Figure 4). The induction of the PT1 genes that encode high affinity Pi transporters, is considered to be an example of a systemic response, where transcription is influenced primarily by the intracellular Pi content [4,5758]. The temporal separation between the initial increase in transcript abundance of the TR-ACS and TR-ACO1 and TR-ACO3 genes and the signature PSR gene, TR-PT1 further supports a separation of the signalling cues with respect to ethylene biosynthesis in the initial low Pi responses. To confirm this, we examined the effect of the Pi analogue, Phi on the early transcription of TR-ACS1 in the EZ root tissue (Figure 5). It is known from other studies that the rapid uptake of Phi can uncouple the intracellular regulation of PSR-associated gene transcription [5961]. However, no induction in transcription after only 1 h of exposure to Phi strongly supports the notion that the increase in transcript abundance of TR-ACS1 (as well as TR-ACO1 and TR-ACO3: results not shown) is specifically regulated by a change in the external concentration of Pi itself.

Changes in transcript abundance in response to changes in an environmental cue can occur through post-transcriptional modifications and/or via transcriptional activation of gene regulatory (promoter) elements. For example, the role of mi399b and mi399f in directing PHO2 transcript abundance as part of the regulation of Pi homoeostasis in roots of Arabidopsis is well established [62]. Thus it was of interest to determine whether changes in transcript abundance in response to low Pi were regulated by transcriptional activation. To do this, we assayed the TR-ACO1 promoter (as a TR-ACO1p::GFP construct), as this gene displayed both the rapid and longer term increase in transcript abundance in response to Pi supply. As a prelude to these GFP localization experiments, in situ hybridization was also used. Accumulation of the TR-ACO1 and TR-ACO3 transcripts was more markedly detected in the lateral root primordia (Supplementary Figures S5 and S6) and localization of GFP was determined in the primary and lateral root tips, lateral root primordia and in the pericycle (Figure 6). Using both GFP fluorescence and protein accumulation (detected by western analysis), we were able to confirm that the promoter was activated over 24 h of exposure to low Pi supply (Figure 7). This localization of expression of TR-ACO1, particularly in the lateral root primordia, and transient activation of the promoter in response to low Pi has some parallels with Arabidopsis where analysis of the RSA (root system architecture) responses has shown that the transient growth effects in response to low Pi are caused by activation of existing primordia, whereas the initiation of new ones is repressed [14].

In summary, we have shown that the transcriptional cues that regulate the transcript abundance of the TR-ACO and TR-ACS gene families mediate a biphasic pattern of induction. The initial increases are observed very rapidly (within 1 h) for specific genes in particular developmental regions of the root, and occur before the increase in abundance of the PSR-associated gene, TR-PT1, and well before any documented decrease in Pi levels and increase in APase activity. We contend, therefore, that these initial changes are regulated by Pi changes in the external media and any ethylene produced may therefore regulate the local changes to low Pi, including the induction of stress responses [4,63]. Critically, we have shown that these changes in transcript abundance (at least for TR-ACO1) result from the transcriptional activation of the promoter suggesting a very sensitive signalling pathway in the roots of plants to rapidly sense and respond to changes in external Pi. The second wave of induction of TR-ACS and TR-ACO transcript abundance may then be an important regulator of the systemic associated changes (as directed by a decrease in the intracellular Pi content) including the reported changes in RSA and the induction of the APases [35,64].

AUTHOR CONTRIBUTION

Michael McManus (deceased, July 2015) conceived the study and wrote the manuscript. Marissa Roldan and Afsana Islam conducted most of the experimental work with some assistance from Phuong Dinh and Susanna Leung. Marissa Roldan and Afsana Islam contributed to producing the manuscript and carried out necessary revisions in response to reviewers’ comments.

The present paper is dedicated to the memory of Professor Michael T. McManus (Massey University, Palmerston North, New Zealand) and Dr Nena Alvarez (deceased December 2010, Dagupan City, The Philippines). NA assisted in conducting in situ hybridization experiments.

FUNDING

This work was supported by the New Zealand Foundation for Research, Science and Technology [grant number C02X0406]; the Massey University Doctoral scholarships to M.R and A.I.; and the NZAid scholarship to P.T.Y.D.

Abbreviations

     
  • ACC

    1-aminocyclopropane-1-carboxylate

  •  
  • ACO

    ACC oxidase

  •  
  • ACS

    ACC synthase

  •  
  • BAP

    6-benzylaminopurine

  •  
  • EZ

    elongation zone

  •  
  • gm

    ground meristem

  •  
  • MR

    mature root

  •  
  • NAA

    1-napthalene acetic acid

  •  
  • pc

    procambium

  •  
  • PSR

    Pi-starvation-response

  •  
  • qc

    quiescent centre

  •  
  • TR-PT1

    Trifolium repens PHOSPHATE TRANSPORTER 1

  •  
  • VL

    visible lateral

References

References
1
Lynch
 
J.P.
 
Root phenes for enhanced soil exploration and phosphorus acquisition: tools for future crops
Plant Physiol
2011
, vol. 
156
 (pg. 
1041
-
1049
)
[PubMed]
2
Shen
 
J.
Yuan
 
L.
Zhang
 
J.
Li
 
H.
Bai
 
Z.
Chen
 
X.
Zhang
 
W.
Zhang
 
F.
 
Phosphorus dynamics: from soil to plant
Plant Physiol
2011
, vol. 
156
 (pg. 
997
-
1005
)
[PubMed]
3
Richardson
 
A.E.
Hocking
 
P.J.
Simpson
 
R.J.
George
 
T.S.
 
Plant mechanisms to optimise access to soil phosphorus
Crop. Pasture Sci.
2009
, vol. 
60
 (pg. 
124
-
143
)
4
Thibaud
 
M.-C.
Arrighi
 
J.-F.
Bayle
 
V.
Chiarenza
 
S.
Creff
 
A.
Bustos
 
R.
Paz-Ares
 
J.
Poirer
 
Y.
Nussaume
 
L.
 
Dissection of local and systematic transcriptional responses to phosphate starvation in Arabidopsis
Plant J.
2010
, vol. 
64
 (pg. 
775
-
789
)
[PubMed]
5
Tran
 
H.T.
Hurley
 
B.A.
Plaxton
 
W.C.
 
Feeding hungry plants: the role of purple acid phosphatases in phosphate nutrition
Plant Sci.
2010
, vol. 
179
 (pg. 
14
-
27
)
6
Plaxton
 
W.C.
Tran
 
H.T.
 
Metabolic adaptations of phosphate-starved plants
Plant Physiol.
2011
, vol. 
156
 (pg. 
1006
-
1015
)
[PubMed]
7
Raghothama
 
K.G.
Karthikeyen
 
A.S.
 
Phosphate acquisition
Plant Soil.
2005
, vol. 
274
 (pg. 
37
-
49
)
8
Hermans
 
C.
Hammond
 
J.P.
White
 
P.J.
Verbruggen
 
N.
 
How do plants respond to nutrient shortage by biomass allocation
Trends Plant Sci
2006
, vol. 
11
 (pg. 
610
-
617
)
[PubMed]
9
Jain
 
A.
Poling
 
M.D.
Karthikeyan
 
A.S.
Blakeslee
 
J.J.
Peer
 
W.A.
Titapiwatanakun
 
B.
Murphy
 
A.S.
Raghothama
 
K.G.
 
Differential effects of sucrose and auxin on localized phosphate deficiency-induced modulation of different traits of root system architecture in Arabidopsis
Plant Physiol.
2007
, vol. 
144
 (pg. 
232
-
247
)
[PubMed]
10
Hammond
 
J.P.
White
 
P.J.
 
Sugar signaling in root responses to low phosphorus availability
Plant Physiol.
2011
, vol. 
156
 (pg. 
1033
-
1040
)
[PubMed]
11
Williamson
 
L.C.
Ribrioux
 
S.
Fitter
 
A.H.
Leyser
 
H.M.O.
 
Phosphate availability regulates root system architecture in Arabidopsis
Plant Physiol.
2001
, vol. 
126
 (pg. 
875
-
882
)
[PubMed]
12
López-Bucio
 
J.
Hernandez-Abreu
 
E.
Sanchez-Calderon
 
L.
Nieto-Jacobo
 
M.F.
Simpson
 
J.
Herrera-Estrella
 
L.
 
Phosphate availability alters architecture and causes changes in hormone sensitivity in the Arabidopsis root system
Plant Physiol.
2002
, vol. 
129
 (pg. 
244
-
256
)
[PubMed]
13
Al-Ghazi
 
Y.
Muller
 
B.
Pinloche
 
S.
Tranbarger
 
T. J.
Nacry
 
P.
Rossignol
 
M.
Tardieu
 
F.
Doumas
 
P.
 
Temporal responses of Arabidopsis root architecture to phosphate starvation: evidence for the involvement of auxin signalling
Plant Cell Environ.
2003
, vol. 
26
 (pg. 
1053
-
1066
)
14
Nacry
 
P.
Canivenc
 
G.
Muller
 
B.
Azmi
 
A.
Van Onckelen
 
H.
Rossignol
 
M.
Doumas
 
P.
 
A role for auxin redistribution in the responses of the root system architecture to phosphate starvation in Arabidopsis
Plant Physiol.
2005
, vol. 
138
 (pg. 
2061
-
2074
)
[PubMed]
15
Sanchez-Calderon
 
L.
Lopez-Bucio
 
J.
Chacon-Lopez
 
A.
Cruz-Ramirez
 
A.
Nieto-Jacobo
 
F.
Dubrovsky
 
J.G.
Herrera-Estrella
 
L.
 
Phosphate starvation induces a determinate developmental program in the roots of Arabidopsis thaliana
Plant Cell Physiol.
2005
, vol. 
46
 (pg. 
174
-
184
)
[PubMed]
16
Fukaki
 
H.
Tasaka
 
M.
 
Hormone interactions during lateral root formation
Plant Mol. Biol.
2009
, vol. 
69
 (pg. 
437
-
449
)
[PubMed]
17
Peret
 
B.
Larrieu
 
A.
Bennett
 
M.B.
 
Lateral root emergence: a difficult birth
J. Exp. Bot.
2009
, vol. 
60
 (pg. 
3637
-
3643
)
[PubMed]
18
Drew
 
M.C.
He
 
C.-J.
Morgan
 
P.W.
 
Decreased ethylene biosynthesis, and induction of aerenchyma, by nitrogen- or phosphate-starvation in adventitious roots of Zea mays L
Plant Physiol.
1989
, vol. 
91
 (pg. 
266
-
271
)
[PubMed]
19
He
 
C.-J.
Morgan
 
P.W.
Drew
 
M.C.
 
Enhanced sensitivity to ethylene in nitrogen- or phosphate-starved roots of Zea mays L. during aerenchyma formation
Plant Physiol.
1992
, vol. 
98
 (pg. 
137
-
142
)
[PubMed]
20
Lynch
 
J.
Brown
 
K.M.
 
Ethylene and plant responses to nutritional stress
Physiol. Plant
1997
, vol. 
100
 (pg. 
613
-
619
)
21
Borch
 
K.
Bouma
 
T.J.
Lynch
 
J.P.
Brown
 
K.M.
 
Ethylene: a regulator of root architectural responses to soil phosphorus availability
Plant Cell Environ
1999
, vol. 
22
 (pg. 
425
-
431
)
22
Zhang
 
Y.J.
Lynch
 
J.P.
Brown
 
K.M.
 
Ethylene and phosphorus availability have interacting yet distinct effects on root hair development
J. Exp. Bot.
2003
, vol. 
54
 (pg. 
2351
-
2361
)
[PubMed]
23
Dinh
 
P.T.Y.
Roldan
 
M.
Leung
 
S
 
McManus
 
M.T.
 
Regulation of root growth by auxin and ethylene is influenced by phosphate supply in white clover (Trifolium repens L.)
Plant Growth Reg.
2012
, vol. 
66
 (pg. 
179
-
190
)
24
Niu
 
Y.F.
Chai
 
R.S.
Jin
 
G.L.
Wang
 
H.
Tang
 
C.X.
Zhang
 
Y.S.
 
Responses of root architecture development to low phosphorus availability: a review
Ann. Bot.
2013
, vol. 
112
 (pg. 
391
-
408
)
[PubMed]
25
Lynch
 
J.P.
Brown
 
K.M.
 
Topsoil foraging–an architectural adaptation of plants to low phosphorus availability
Plant Soil.
2001
, vol. 
237
 (pg. 
225
-
237
)
26
Ma
 
Z.
Baskin
 
T.I.
Brown
 
K.M.
Lynch
 
J.P.
 
Regulation of root elongation under phosphorus stress involves changes in ethylene responsiveness
Plant Physiol.
2003
, vol. 
131
 (pg. 
1381
-
1390
)
[PubMed]
27
Kim
 
H.-J.
Lynch
 
J.P.
Brown
 
K.M.
 
Ethylene insensitivity impedes a subset of responses to phosphorus deficiency in tomato and petunia
Plant Cell Environ.
2008
, vol. 
31
 (pg. 
1744
-
1755
)
[PubMed]
28
Li
 
Y.-S.
Mao
 
X.-T.
Tian
 
Q.-Y.
Li
 
L.-H.
Zhang
 
W.-H.
 
Phosphorus deficiency-induced reduction in root hydraulic conductivity in Medicago falcata is associated with ethylene production
Environ. Exp. Bot.
2009
, vol. 
67
 (pg. 
172
-
177
)
29
Perez-Torres
 
C.A.
Lopez-Bucio
 
J.
Cruz-Ramirez
 
A.
Ibarra-Laclette
 
E.
Dharmasiri
 
S.
Estelle
 
M.
Herrera-Estrella
 
L.
 
Phosphate availability alters lateral root development in Arabidopsis by modulating auxin sensitivity via a mechanism involving the TIR1 auxin receptor
Plant Cell
2008
, vol. 
20
 (pg. 
3258
-
3272
)
[PubMed]
30
Miura
 
K.M.
Lee
 
J.
Gong
 
Q.
Ma
 
S.
Jin
 
J.B.
Yoo
 
C.Y.
Miura
 
T.
Sato
 
A.
Bohnert
 
H.J.
Hasegawa
 
P.M.
 
SIZ1 regulation of phosphate starvation-induced root architecture remodelling involves the control of auxin accumulation
Plant Physiol.
2011
, vol. 
155
 (pg. 
1000
-
1012
)
[PubMed]
31
Graham
 
M.A.
Ramirez
 
M.
Valdes-Lopez
 
O.
Lara
 
M.
Tesfaye
 
M.
Vance
 
C.P.
Hernandez
 
G.
 
Identification of candidate phosphorus stress induced genes in Phaseolus vulgaris through clustering analysis across several plant species
Funct. Plant Biol.
2006
, vol. 
33
 (pg. 
789
-
797
)
32
Hernandez
 
G.
Ramirez
 
M.
Valdes-Lopez
 
O.
Tesfaye
 
M.
Graham
 
M.A.
Czechowski
 
T.
Schlereth
 
A.
Wandrey
 
M.
Erban
 
A.
Cheung
 
F.
, et al 
Phosphorus stress in common bean: root transcript and metabolic responses
Plant Physiol.
2007
, vol. 
144
 (pg. 
752
-
767
)
[PubMed]
33
Bustos
 
R.
Castrillo
 
G.
Linhares
 
F.
Puga
 
M.
Rubio
 
V.
Perez-Perez
 
J.
Solano
 
R.
Leyva
 
A.
Paz-Ares
 
J.
 
A central regulatory system largely controls transcriptional activation and repression responses to phosphate starvation in Arabidopsis
PLoS Genet.
2010
, vol. 
6
 pg. 
e1001102
 
[PubMed]
34
Chacon-Lopez
 
A.
Ibarra-Laclette
 
E.
Sanchez-Calderon
 
L.
Gutierrez-Alanis
 
D.
Herrera-Estella
 
L.
 
Global expression pattern comparison between low phosphorus insensitive 4 and WT Arabidopsis reveals an important role of reactive species and jasmonic acid in the root tip response to phosphate starvation
Plant Signal Behav.
2011
, vol. 
6
 (pg. 
382
-
392
)
[PubMed]
35
Lei
 
M.
Zhu
 
C.
Liu
 
Y.
Karthkeyan
 
A.S.
Bressan
 
R.A.
Raghothama
 
K.
Liu
 
D.
 
Ethylene signalling is involved in regulation of phosphate starvation-induced gene expression and production of acid phosphatases and anthocyanin in Arabidopsis
New Phytol.
2011
, vol. 
189
 (pg. 
1084
-
1095
)
[PubMed]
36
Hunter
 
D.A.
Yoo
 
S.D.
Butcher
 
S.M.
McManus
 
M.T.
 
Expression of 1-aminocyclopropane-1-carboxylate oxidase during leaf ontogeny in white clover
Plant Physiol.
1999
, vol. 
120
 (pg. 
131
-
141
)
[PubMed]
37
Murray
 
P.A.
McManus
 
M.T.
 
Developmental regulation of 1-aminocyclopropane-1-carboxylate synthase gene expression during leaf ontogeny in white clover
Physiol. Plant
2005
, vol. 
124
 (pg. 
107
-
120
)
38
Chen
 
B.C.
McManus
 
M.T.
 
Expression of 1-aminocyclopropane-1-carboxylate (ACC) oxidase genes during the development of vegetative tissues in white clover (Trifolium repens L.) is regulated by ontological cues
Plant Mol. Biol.
2006
, vol. 
60
 (pg. 
451
-
467
)
[PubMed]
39
Scott
 
R.W.
Yoo
 
S.D.
Hunter
 
D.A.
Gong
 
D.
Chen
 
C.-M.
Leung
 
S.
McManus
 
M.T.
 
Regulation of 1-aminocyclopropane-1-carboxylate oxidase gene expression during leaf ontogeny in white clover
Plant Growth Reg
2010
, vol. 
62
 (pg. 
31
-
41
)
40
Dorling
 
S.J.
Leung
 
S.
Anderson
 
C.W.N.
Albert
 
N.W.
McManus
 
M.T.
 
Changes in 1-aminocyclopropane-1-carboxlate (ACC) oxidase expression and enzyme activity in response to excess manganese in white clover (Trifolium repens L.)
Plant Physiol. Biochem.
2011
, vol. 
49
 (pg. 
1013
-
1019
)
[PubMed]
41
Dunlop
 
J.
Gardiner
 
S.
 
Phosphate uptake, proton extrusion and membrane electropotentials of phosphorus-deficient Trifolium repens L
J. Exp. Bot.
1993
, vol. 
44
 (pg. 
1801
-
1808
)
42
LÖtscher
 
M.
Hay
 
M.J.M.
 
Distribution of mineral nutrient from nodal roots of Trifolium repens: genotypic variation in intra-plant allocation of 32P and 45Ca
Physiol. Plant
1996
, vol. 
97
 (pg. 
269
-
276
)
43
Caradus
 
J.R.
Kennedy
 
L.D.
Dunn
 
A.
 
Genetic variation for the ratio of inorganic to total phosphorus in white clover leaves
J. Plant Nut.
1998
, vol. 
21
 (pg. 
2265
-
2272
)
44
Zhang
 
C.
McManus
 
M.T.
 
Identification and characterisation of two distinct acid phosphatases in cell walls of roots of white clover
Plant Physiol. Biochem.
2000
, vol. 
38
 (pg. 
259
-
270
)
45
Crush
 
J.R.
Boulesteix-Coutelier
 
A.R.L.
Ouyang
 
L.
 
Phosphate uptake by white clover (Trifolium repens L.) genotypes with contrasting root morphology
N.Z. J. Agric. Res.
2008
, vol. 
51
 (pg. 
279
-
285
)
46
Ma
 
X.F.
Wright
 
E.
Ge
 
Y.X.
Bell
 
J.
Xi
 
Y.J.
Bouton
 
J.H.
Wang
 
Z.Y.
 
Improving phosphorus acquisition of white clover (Trifolium repens L.) by transgenic expression of plant-derived phytase and acid phosphatase genes
Plant Sci.
2009
, vol. 
176
 (pg. 
479
-
488
)
[PubMed]
47
Gibeaut
 
D.M.
Hulet
 
J.
Cramer
 
G.R.
Seemann
 
J.R.
 
Maximal biomass of Arabidopsis thaliana using a simple, low-maintenance hydroponic method and favourable environmental conditions
Plant Physiol.
1997
, vol. 
115
 (pg. 
317
-
319
)
[PubMed]
48
Hay
 
A.S.
Watson
 
L.M.
Zhang
 
C.
McManus
 
M.T.
 
Identification of cell wall proteins in roots of phosphate-deprived white clover (Trifolium repens L.) plants
Plant Physiol. Biochem.
1998
, vol. 
36
 (pg. 
305
-
311
)
49
Voisey
 
C.R.
White
 
D.W.R.
Dudas
 
B.
 
Agrobacterium-mediated transformation of white clover using direct shoot organogenesis
Plant Cell Rep.
1994
, vol. 
13
 (pg. 
309
-
314
)
[PubMed]
50
Hunter
 
D.A.
Reid
 
M.R.
 
A simple and rapid method for isolating high quality RNA from flower petals
Acta Hortic.
2001
, vol. 
37
 (pg. 
33
-
37
)
51
Pfaffl
 
M.W.
 
A new mathematical model for relative quantification in real-time RTP-CR
Nucleic Acid Res.
2001
, vol. 
29
 (pg. 
2002
-
2007
)
52
Alvarez
 
N.G.
Meeking
 
R.J.
White
 
D.W.R.
 
The origin, initiation and development of axillary shoot meristems in Lotus japonicus
Ann. Bot.
2006
, vol. 
98
 (pg. 
953
-
963
)
[PubMed]
53
Harpez-Saad
 
S.
Yoon
 
G.M.
Mattoo
 
A.K.
Kieber
 
J.J.
 
The formation of ACC and competition between polyamines and ethylene for SAM
Annu. Plant Rev.
2012
, vol. 
44
 (pg. 
53
-
82
)
54
Yang
 
S.F.
Hoffman
 
N.E.
 
Ethylene biosynthesis and its regulation in higher plants
Ann. Rev. Plant Physiol.
1984
, vol. 
35
 (pg. 
155
-
189
)
55
Roldan
 
M.
Dinh
 
P.
Leung
 
S.
McManus
 
M.T.
 
Ethylene and the responses of plants to phosphate deficiency
AoB Plants
2013
, vol. 
5
 pg. 
plt013
 
56
Duff
 
S.M.G.
Sarath
 
G.
Plaxton
 
W.C.
 
The role of acid phosphatases in plant phosphorus metabolism
Physiol. Plant
1994
, vol. 
90
 (pg. 
791
-
800
)
57
Chiou
 
T.-J.
Liu
 
H.
Harrison
 
M.J.
 
The spatial expression patterns of a phosphate transporter (MtPT1) from Medicago truncatula indicate a role in phosphate transport at the root/soil interface
Plant J.
2001
, vol. 
25
 (pg. 
281
-
293
)
[PubMed]
58
Lui
 
J.
Versaw
 
W.K.
Pumplin
 
N.
Gomez
 
S.K.
Blaylock
 
L.A.
Harrison
 
M.J.
 
Closely related members of the Medicago truncatula PHT1 phosphate transporter gene family encode phosphate transporters with distinct biochemical activities
J. Biol. Chem.
2008
, vol. 
283
 (pg. 
24673
-
24681
)
[PubMed]
59
Ticconi
 
C.A.
Delatorre
 
C.A.
Abel
 
S.
 
Attenuation of phosphate starvation responses by phosphite in Arabidopsis
Plant Physiol.
2001
, vol. 
127
 (pg. 
963
-
972
)
[PubMed]
60
Varadarajan
 
D.K.
Karthikeyan
 
A.S.
Matilda
 
P.D.
Raghothama
 
K.G.
 
Phosphite, an analog of phosphate, suppresses the coordinated expression of genes under phosphate starvation
Plant Physiol.
2002
, vol. 
129
 (pg. 
1232
-
1240
)
[PubMed]
61
Berkowitz
 
O.
Jost
 
R.
Kollehn
 
D.O.
Fenske
 
R.
Finnegan
 
P.M.
O'Brien
 
P.A.
Hardy
 
G.E.
 
Lambers
 
H.
 
Acclimation responses of Arabidopsis thaliana to sustained phosphite treatments
J. Exp. Bot.
2013
, vol. 
64
 (pg. 
1731
-
1743
)
[PubMed]
62
Lin
 
S.-I.
Chiang
 
S.-F.
Lin
 
W.-Y.
Chen
 
J.-W.
Tseng
 
C.-Y.
Wu
 
P.-C.
Chiou
 
T.-J.
 
Regulatory network of microRNA399 and PHO2 by systemic signaling
Plant Physiol.
2008
, vol. 
147
 (pg. 
732
-
746
)
[PubMed]
63
Nagarajan
 
V.K.
Smith
 
A.P.
 
Ethylene's role in phosphate starvation signaling: more than just a root growth regulator
Plant Cell Physiol.
2012
, vol. 
53
 (pg. 
277
-
286
)
[PubMed]
64
Li
 
Y.-S.
Gao
 
Y.
Tian
 
Q.-Y.
Shi
 
F.-L.
Li
 
L.-H.
Zhang
 
W.-H.
 
Stimulation of root acid phosphatase by phosphorus deficiency is regulated by ethylene in Medicago falcata. Environ
Exp. Bot.
2011
, vol. 
71
 (pg. 
114
-
120
)

Author notes

2

These two authors contributed equally to this work.

3

Present address: AgResearch Limited, Private Bag 11008, Tennent Drive, Palmerston North, New Zealand.

4

Present address: Department of Molecular and Cellular Physiology, Ehime University, Toon, Ehime, Japan.

5

Present address: Department of Plant Pathology, Washington State University, Pullman, WA, U.S.A.

6

Present address: School of Physical and Chemical Sciences, Victoria University, Wellington, New Zealand.

7

Deceased.

This is an open access article published by Portland Press Limited on behalf of the Biochemical Society and distributed under the Creative Commons Attribution Licence 4.0 (CC BY).

Supplementary data